From 1cc4dcd01a9646083319ca0338d932e473bdcdb1 Mon Sep 17 00:00:00 2001
From: "DEVSCOPE\\rquintino" <->
Date: Sat, 7 Apr 2018 00:07:14 +0100
Subject: [PATCH 01/15] -typical default for openml dbs -added ES_PUBLIC_URL

---
 openml_OS/config/BASE_CONFIG-BLANK.php | 9 +++++----
 openml_OS/views/frontend_main.php      | 2 +-
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/openml_OS/config/BASE_CONFIG-BLANK.php b/openml_OS/config/BASE_CONFIG-BLANK.php
index f8ce56674..3c7277997 100644
--- a/openml_OS/config/BASE_CONFIG-BLANK.php
+++ b/openml_OS/config/BASE_CONFIG-BLANK.php
@@ -73,7 +73,7 @@
  * DB_PASS_WRITING: The password that belongs to the DB_USER_WRITING
  * account.
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
-define( 'DB_NAME_EXPDB', 'FILL_IN' );
+define( 'DB_NAME_EXPDB', 'openml_expdb' );
 define( 'DB_HOST_EXPDB', 'FILL_IN' );
 define( 'DB_USER_EXPDB_READ', 'FILL_IN' );
 define( 'DB_PASS_EXPDB_READ', 'FILL_IN' );
@@ -92,7 +92,7 @@
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
  * DB_PASS: The password that belongs to the username
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
-define( 'DB_NAME_OPENML', 'FILL_IN' );
+define( 'DB_NAME_OPENML', 'openml' );
 define( 'DB_HOST_OPENML', 'FILL_IN' );
 define( 'DB_USER_OPENML', 'FILL_IN' );
 define( 'DB_PASS_OPENML', 'FILL_IN' );
@@ -115,8 +115,9 @@
  * ES_USERNAME and ES_PASSWORD can be set by ElasticSearch or a protected proxy
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
 define( 'ES_URL', 'FILL_IN' );
-define( 'ES_USERNAME', 'FILL_IN' );
-define( 'ES_PASSWORD', 'FILL_IN' );
+define( 'ES_PUBLIC_URL', 'FILL_IN' );
+define( 'ES_USERNAME', '' );
+define( 'ES_PASSWORD', '' );
 
 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
  * Configuration details for the Wiki server (Username, password)
diff --git a/openml_OS/views/frontend_main.php b/openml_OS/views/frontend_main.php
index 5739ff819..3797379e2 100755
--- a/openml_OS/views/frontend_main.php
+++ b/openml_OS/views/frontend_main.php
@@ -319,7 +319,7 @@
 
         </div>
         <script type="text/javascript">
-          var ES_URL = '<?php echo ES_URL; ?>';
+          var ES_URL = '<?php echo ES_PUBLIC_URL; ?>';
 
           function downloadJSAtOnload() {
           //var element3= document.createElement("script");

From d4bcd0c934e8c84f160663e091c2da9b5157d17c Mon Sep 17 00:00:00 2001
From: "DEVSCOPE\\rquintino" <->
Date: Sat, 7 Apr 2018 00:08:14 +0100
Subject: [PATCH 02/15] -added cron controller functions to initialize local
 environment

---
 openml_OS/controllers/Cron.php | 69 ++++++++++++++++++++++++++++++++++
 1 file changed, 69 insertions(+)

diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index 1d55e3817..477b8d577 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -75,21 +75,80 @@ public function update_tag($type, $id){
       $this->elasticsearch->update_tags($type, $id);
   }
 
+  // Creating main openml index
+  public function create_es_openml_index() {
+    echo "\r\nCreating main openml index...";
+    $params = [
+      'index' => 'openml',
+      'body' => [
+          'settings' => [
+              'number_of_shards' => 3,
+              'number_of_replicas' => 2
+              ]
+            ]
+    ];
+
+    $this->elasticsearch->client->indices()->create($params);
+    echo "\r\nDone!";
+    
+  }
+
+  // Create admin user
+  public function create_local_admin() {
+    echo "\r\nCreating local admin user...";
+  
+    $username="admin";
+    $email="admin";
+    $session_hash=md5(rand());
+    $additional_data = array(
+      'first_name' => "Admin",
+      'last_name'  => "",
+      'affiliation'=> "",
+      'country'    => "",
+      'bio'    	 => "",
+      'external_source' => null,
+      'external_id' => null,
+      'session_hash' => $session_hash
+    );
+  
+    $password=substr(md5(uniqid()), 0, 12);    
+    echo "\r\nPassword:\t" .$password;
+    echo "\r\nAPI key:\t" .$session_hash;
+
+    echo "\r\nSaving key in shared config...";
+    file_put_contents("/openmlconfig/api_key.txt",$session_hash);
+   
+    $this->load->Library('ion_auth');
+
+    # Add to admin group
+    $group = array('1');
+    $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
+    
+    echo "\r\nDone! ";
+    
+  }
+
+  
   // initialize all es indexes
   public function initialize_es_indices() {
+    echo "\r\nInitialize_es_indices...";
     foreach($this->es_indices as $index) {
       $this->elasticsearch->initialize_index($index);
     }
+    echo "\r\nDone!";
   }
 
   // builds all es indexes
   public function build_es_indices() {
+    echo "\r\nBuild_es_indices...";
     foreach($this->es_indices as $index) {
       $this->indexfrom($index, 1);
     }
+    echo "\r\nDone!";
   }
 
   function install_database() {
+    echo "\r\nInstall database...";
     // note that this one does not come from DATA folder, as they are stored in github
     $models = directory_map('data/sql/', 1);
     $manipulated_order = array('implementation.sql', 'algorithm_setup.sql', 'dataset.sql', 'study.sql', 'groups.sql', 'users.sql');
@@ -115,6 +174,16 @@ function install_database() {
         echo 'skipping ' . $modelname . ', as it is not empty... ' + "\n";
       }
     }
+    echo "\r\nDone!";
+  }
+
+  // Runs all local env init steps
+  public function init_local_env() {
+    $this->install_database();
+    $this->create_es_openml_index();
+    $this->initialize_es_indices();
+    $this->create_local_admin();
+    $this->build_es_indices();
   }
 
   function create_meta_dataset($id = false) {

From 918c5952648021b0b6926c784d58a2a34845059b Mon Sep 17 00:00:00 2001
From: "DEVSCOPE\\rquintino" <->
Date: Sat, 7 Apr 2018 00:08:36 +0100
Subject: [PATCH 03/15] add openml private db schema

---
 downloads/openml.sql | 313 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 313 insertions(+)
 create mode 100644 downloads/openml.sql

diff --git a/downloads/openml.sql b/downloads/openml.sql
new file mode 100644
index 000000000..9871bb920
--- /dev/null
+++ b/downloads/openml.sql
@@ -0,0 +1,313 @@
+-- phpMyAdmin SQL Dump
+-- version 4.5.4.1deb2ubuntu2
+-- http://www.phpmyadmin.net
+--
+-- Host: localhost
+-- Generation Time: Jun 16, 2017 at 02:32 PM
+-- Server version: 5.7.18-0ubuntu0.16.04.1
+-- PHP Version: 7.0.15-0ubuntu0.16.04.4
+
+SET SQL_MODE = "NO_AUTO_VALUE_ON_ZERO";
+SET time_zone = "+00:00";
+
+
+/*!40101 SET @OLD_CHARACTER_SET_CLIENT=@@CHARACTER_SET_CLIENT */;
+/*!40101 SET @OLD_CHARACTER_SET_RESULTS=@@CHARACTER_SET_RESULTS */;
+/*!40101 SET @OLD_COLLATION_CONNECTION=@@COLLATION_CONNECTION */;
+/*!40101 SET NAMES utf8mb4 */;
+
+--
+-- Database: `openml`
+--
+CREATE DATABASE IF NOT EXISTS `openml` DEFAULT CHARACTER SET latin1 COLLATE latin1_swedish_ci;
+USE `openml`;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `access`
+--
+
+CREATE TABLE `access` (
+  `user_id` bigint(20) NOT NULL,
+  `group` enum('true','false') NOT NULL,
+  `type` enum('data','flow','study') NOT NULL,
+  `entity_id` bigint(20) NOT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `category`
+--
+
+CREATE TABLE `category` (
+  `id` int(10) NOT NULL,
+  `activated` enum('y','n') NOT NULL DEFAULT 'y',
+  `deleted` enum('y','n') NOT NULL DEFAULT 'n',
+  `title` varchar(255) NOT NULL,
+  `description` text,
+  `order` int(10) NOT NULL DEFAULT '100'
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `file`
+--
+
+CREATE TABLE `file` (
+  `id` int(16) NOT NULL,
+  `creator` int(16) NOT NULL,
+  `creation_date` datetime NOT NULL,
+  `filepath` varchar(256) NOT NULL,
+  `filesize` int(64) NOT NULL,
+  `filename_original` varchar(256) NOT NULL,
+  `extension` varchar(16) NOT NULL,
+  `mime_type` varchar(32) NOT NULL,
+  `md5_hash` varchar(64) NOT NULL,
+  `type` enum('dataset','implementation','predictions','userimage','run_trace','run_uploaded_file','url','misc') NOT NULL,
+  `access_policy` enum('public','private','none','deleted') NOT NULL DEFAULT 'public'
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `groups`
+--
+
+CREATE TABLE `groups` (
+  `id` mediumint(8) UNSIGNED NOT NULL,
+  `name` varchar(20) NOT NULL,
+  `description` varchar(100) NOT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `login_attempts`
+--
+
+CREATE TABLE `login_attempts` (
+  `id` mediumint(8) UNSIGNED NOT NULL,
+  `ip_address` varbinary(16) NOT NULL,
+  `login` varchar(100) NOT NULL,
+  `time` int(11) UNSIGNED DEFAULT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `meta_dataset`
+--
+
+CREATE TABLE `meta_dataset` (
+  `id` int(8) NOT NULL,
+  `user_id` int(8) NOT NULL,
+  `request_date` datetime NOT NULL,
+  `type` enum('evaluations','qualities','inputs') NOT NULL DEFAULT 'evaluations',
+  `task_type` int(10) NOT NULL,
+  `datasets` text,
+  `tasks` text,
+  `flows` text,
+  `setups` text,
+  `functions` text,
+  `file_id` int(8) DEFAULT NULL,
+  `processed` datetime DEFAULT NULL,
+  `error_message` varchar(256) DEFAULT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `thread`
+--
+
+CREATE TABLE `thread` (
+  `id` int(10) NOT NULL,
+  `activated` enum('y','n') NOT NULL DEFAULT 'n',
+  `deleted` enum('y','n') NOT NULL DEFAULT 'n',
+  `title` varchar(255) NOT NULL,
+  `body` text NOT NULL,
+  `post_date` datetime NOT NULL,
+  `author_id` int(10) NOT NULL,
+  `category_id` int(10) NOT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `users`
+--
+
+CREATE TABLE `users` (
+  `id` mediumint(8) UNSIGNED NOT NULL,
+  `ip_address` varbinary(16) NOT NULL,
+  `username` varchar(100) NOT NULL,
+  `password` varchar(80) NOT NULL,
+  `salt` varchar(40) DEFAULT NULL,
+  `email` varchar(100) NOT NULL,
+  `activation_code` varchar(40) DEFAULT NULL,
+  `forgotten_password_code` varchar(40) DEFAULT NULL,
+  `forgotten_password_time` int(11) UNSIGNED DEFAULT NULL,
+  `remember_code` varchar(40) DEFAULT NULL,
+  `created_on` int(11) UNSIGNED NOT NULL,
+  `last_login` int(11) UNSIGNED DEFAULT NULL,
+  `active` tinyint(1) UNSIGNED DEFAULT NULL,
+  `first_name` varchar(50) DEFAULT NULL,
+  `last_name` varchar(50) DEFAULT NULL,
+  `affiliation` varchar(50) NOT NULL,
+  `country` varchar(50) NOT NULL,
+  `image` varchar(128) DEFAULT NULL,
+  `bio` text NOT NULL,
+  `core` enum('true','false') NOT NULL DEFAULT 'false',
+  `external_source` varchar(50) DEFAULT NULL,
+  `external_id` varchar(50) DEFAULT NULL,
+  `session_hash` varchar(40) DEFAULT NULL,
+  `session_hash_date` timestamp NULL DEFAULT CURRENT_TIMESTAMP,
+  `gamification_visibility` varchar(32) NOT NULL DEFAULT 'show'
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `users_groups`
+--
+
+CREATE TABLE `users_groups` (
+  `id` mediumint(8) UNSIGNED NOT NULL,
+  `user_id` mediumint(8) UNSIGNED NOT NULL,
+  `group_id` mediumint(8) UNSIGNED NOT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+-- --------------------------------------------------------
+
+--
+-- Table structure for table `vipercharts`
+--
+
+CREATE TABLE `vipercharts` (
+  `id` int(10) NOT NULL,
+  `run_id` int(10) NOT NULL,
+  `class` varchar(255) NOT NULL,
+  `viper_id` varchar(255) NOT NULL,
+  `activated` datetime NOT NULL,
+  `inactived` datetime DEFAULT NULL
+) ENGINE=InnoDB DEFAULT CHARSET=latin1;
+
+--
+-- Indexes for dumped tables
+--
+
+--
+-- Indexes for table `access`
+--
+ALTER TABLE `access`
+  ADD PRIMARY KEY (`user_id`,`group`,`type`,`entity_id`);
+
+--
+-- Indexes for table `category`
+--
+ALTER TABLE `category`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `file`
+--
+ALTER TABLE `file`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `groups`
+--
+ALTER TABLE `groups`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `login_attempts`
+--
+ALTER TABLE `login_attempts`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `meta_dataset`
+--
+ALTER TABLE `meta_dataset`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `thread`
+--
+ALTER TABLE `thread`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `users`
+--
+ALTER TABLE `users`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `users_groups`
+--
+ALTER TABLE `users_groups`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- Indexes for table `vipercharts`
+--
+ALTER TABLE `vipercharts`
+  ADD PRIMARY KEY (`id`);
+
+--
+-- AUTO_INCREMENT for dumped tables
+--
+
+--
+-- AUTO_INCREMENT for table `category`
+--
+ALTER TABLE `category`
+  MODIFY `id` int(10) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=2;
+--
+-- AUTO_INCREMENT for table `file`
+--
+ALTER TABLE `file`
+  MODIFY `id` int(16) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=5789610;
+--
+-- AUTO_INCREMENT for table `groups`
+--
+ALTER TABLE `groups`
+  MODIFY `id` mediumint(8) UNSIGNED NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=4;
+--
+-- AUTO_INCREMENT for table `login_attempts`
+--
+ALTER TABLE `login_attempts`
+  MODIFY `id` mediumint(8) UNSIGNED NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=23;
+--
+-- AUTO_INCREMENT for table `meta_dataset`
+--
+ALTER TABLE `meta_dataset`
+  MODIFY `id` int(8) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=64;
+--
+-- AUTO_INCREMENT for table `thread`
+--
+ALTER TABLE `thread`
+  MODIFY `id` int(10) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=3;
+--
+-- AUTO_INCREMENT for table `users`
+--
+ALTER TABLE `users`
+  MODIFY `id` mediumint(8) UNSIGNED NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=3375;
+--
+-- AUTO_INCREMENT for table `users_groups`
+--
+ALTER TABLE `users_groups`
+  MODIFY `id` mediumint(8) UNSIGNED NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=3418;
+--
+-- AUTO_INCREMENT for table `vipercharts`
+--
+ALTER TABLE `vipercharts`
+  MODIFY `id` int(10) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=3;
+/*!40101 SET CHARACTER_SET_CLIENT=@OLD_CHARACTER_SET_CLIENT */;
+/*!40101 SET CHARACTER_SET_RESULTS=@OLD_CHARACTER_SET_RESULTS */;
+/*!40101 SET COLLATION_CONNECTION=@OLD_COLLATION_CONNECTION */;

From 4f7cd5d05b6db23b86f1f77414a3e9266119deb2 Mon Sep 17 00:00:00 2001
From: janvanrijn <janvanrijn@gmail.com>
Date: Tue, 13 Oct 2020 18:31:24 +0200
Subject: [PATCH 04/15] recent updates website (#1064)

* Added inline swagger-php annotations to bootstrap methods.

* Re-arranged some of the api annotations to the bootstrap methods.

* Added swagger-php docs for schemas used as responses in the API.

* data edit api

* Moved annotations from the bootstrap to the respective methods responsible for handling the logic.
Refactored bootstrapping for tag/untag of several entities as well as attach/detach to clearly distinct between different handlers.
Fixed boolean types in annotations.

* Fixed issues with doctrine. Now all the annotations get compiled correctly.

* Fixed issues with schemas and properties with array types requiring a type/ref for items.

* Fixed issue with tasktype API description.

* Added bash/bat scripts for linux and windows respectively to generate openapi specifications from inline php annotations.

* Added server specs and info description to the annotations.

* Removed API key parameters from GET methods, replaced overall description and added servers for base paths.

* Repaired the code which got screwed up during merging develop into this feature branch.

* support edit parameters via xml

* Added python script for filtering out unused OpenAPI schemas.

Removed unreferenced OpenAPI schemas.

* Added swagger json and yaml file.

* support non-owner edtis and sensitive field edits

* fork and edit api

* template

* Update Api_data.php

* latest version

* tag_untag functions are now left out for tasks and runs. Instead within the classes two string constants are defined keeping the values for the super class providing the entity type and the entity special name.

* Added some documentation for the OpenAPI specification generation and documentation with annotations within the PHP REST API.

* type cast query

* Update body.php

* update hackathon date and link

* added info on metal challenge

Co-authored-by: mwever <wever@mail.uni-paderborn.de>
Co-authored-by: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Co-authored-by: Joaquin Vanschoren <joaquin.vanschoren@gmail.com>
Co-authored-by: Sahithya Ravi <44670788+sahithyaravi1493@users.noreply.github.com>
---
 .gitignore                                    |    2 +
 openapi/README.md                             |   62 +
 openapi/composer.json                         |   14 +
 openapi/composer.lock                         |  377 ++
 openapi/composer.phar                         |  Bin 0 -> 1811606 bytes
 openapi/generate_json_api.bat                 |    1 +
 openapi/generate_json_api.sh                  |    3 +
 openapi/generate_yaml_api.bat                 |    1 +
 openapi/generate_yaml_api.sh                  |    3 +
 openapi/swagger.json                          | 5647 +++++++++++++++++
 openapi/swagger.yaml                          | 3338 ++++++++++
 openapi/unusedSchemas.py                      |   55 +
 openml_OS/controllers/Api.php                 | 2130 ++++++-
 openml_OS/models/api/v1/Api_data.php          | 1007 ++-
 .../models/api/v1/Api_estimationprocedure.php |   50 +-
 openml_OS/models/api/v1/Api_evaluation.php    |  138 +-
 .../models/api/v1/Api_evaluationmeasure.php   |   33 +-
 openml_OS/models/api/v1/Api_flow.php          |  418 +-
 openml_OS/models/api/v1/Api_run.php           |  614 +-
 openml_OS/models/api/v1/Api_setup.php         |  313 +-
 openml_OS/models/api/v1/Api_study.php         |  362 +-
 openml_OS/models/api/v1/Api_task.php          |  462 +-
 openml_OS/models/api/v1/Api_tasktype.php      |  128 +
 openml_OS/models/api/v1/Api_user.php          |   32 +-
 .../pages/api_new/v1/xml/data-edit.tpl.php    |    3 +
 .../pages/api_new/v1/xml/data-fork.tpl.php    |    3 +
 openml_OS/views/pages/api_new/v1/xml/pre.php  |   17 +
 .../pages/api_new/v1/xsd/openml.data.edit.xsd |   89 +
 .../views/pages/frontend/contact/body.php     |    2 +-
 openml_OS/views/pages/frontend/home/body.php  |   11 +-
 30 files changed, 15191 insertions(+), 124 deletions(-)
 create mode 100644 openapi/README.md
 create mode 100644 openapi/composer.json
 create mode 100644 openapi/composer.lock
 create mode 100644 openapi/composer.phar
 create mode 100644 openapi/generate_json_api.bat
 create mode 100644 openapi/generate_json_api.sh
 create mode 100644 openapi/generate_yaml_api.bat
 create mode 100644 openapi/generate_yaml_api.sh
 create mode 100644 openapi/swagger.json
 create mode 100644 openapi/swagger.yaml
 create mode 100644 openapi/unusedSchemas.py
 create mode 100644 openml_OS/views/pages/api_new/v1/xml/data-edit.tpl.php
 create mode 100644 openml_OS/views/pages/api_new/v1/xml/data-fork.tpl.php
 create mode 100644 openml_OS/views/pages/api_new/v1/xsd/openml.data.edit.xsd

diff --git a/.gitignore b/.gitignore
index 36bf2307e..728ccc53f 100644
--- a/.gitignore
+++ b/.gitignore
@@ -3,3 +3,5 @@ data/*
 openml_OS/config/BASE_CONFIG.php
 openml_OS/third_party/OpenML/Java/old-evaluate.jar
 docs/site/
+openml_OS/vendor/
+openapi/vendor/
\ No newline at end of file
diff --git a/openapi/README.md b/openapi/README.md
new file mode 100644
index 000000000..85f6b44c7
--- /dev/null
+++ b/openapi/README.md
@@ -0,0 +1,62 @@
+# OpenAPI REST API Documentation
+
+In order to facilitate interfacing with OpenML's REST API, it is desribed by means of OpenAPI.
+However, in order to ensure that the specification is up-to-date with the current code base and the documentation does not need to be duplicated (once in the OpenAPI specification file and once in-line in the code base), in-line annotations are leveraged to facilitate the maintenance of the documentation. Via the composer package [zircote/swagger-php](https://zircote.github.io/swagger-php/), the annotations are compiled into an OpenAPI specification file (either yaml or json). Instructions on how to freshly generate the OpenAPI specification from the current code base as well as how to create/edit documentation for the REST API is given in the following.
+
+## Getting Started
+
+Prerequisite: In order to generate the up-to-date OpenAPI specification according to the current code base, you need to have the php-cli package installed on your device.
+
+For compiling in-line documentation into an OpenAPI specification file, the zircote/swagger-php package is used. This tool will allow you to search the entire code base and distill an OpenAPI specification from all found in-line comments.
+
+* Open a terminal and change into the `openapi/` directory.
+* Install the zircote/swagger-php package by running the command `php composer.phar install`
+
+Now you have everything in place to start compiling an OpenAPI specification file from the code base.
+
+## Generate OpenML's Current OpenAPI Specification
+
+For generating OpenML's up-to-date OpenAPI specification, depending on your operating system, you may choose one of the scripts contained in the `openapi/` directory.
+For all operating systems, there are two flavors of the script: one producing a yaml file and one a json file. The contents of both files will be the same except for the format itself.
+
+If you are a Linux of MacOS user, you may use the .sh-Scripts:
+
+```
+./generate_json_api.sh # This will compile the OpenAPI specification in JSON format.
+./generate_yaml_api.sh # This will compile the OpenAPI specification in YAML format.
+```
+
+As a Windows user, you may want to use the following batch scripts:
+
+```
+./generate_json_api.bat # This will compile the OpenAPI specification in JSON format.
+./generate_yaml_api.bat # This will compile the OpenAPI specification in YAML format.
+```
+
+## Instructions for the In-Line Documentation
+The inline annotations for PHP work similar as in Java, following the schema @<annotation name>(<annotation content>). In case of the OpenAPI annotations the most important components such as the annotations for get, post, or put methods have distinct annotations. However, all OpenAPI specific annotations are namespaced with an `OA\`. Thus the annotation for get, post, and put correspond to `@OA\Get(...)`, `@OA\Post(...)`, and `@OA\Put(...)` respectively. The information about the path, tags, description, parameters, etc. is then enclosed within the parantheses. While keys such as path, tags, description, and so on which only take a simple value (instead of a complex object) are directly referenced, parameters for instance are specified with the help of distinct annotations again, i.e. `@OA\Parameter(...)`.
+
+A more comprehensive description of what kind of annotations are supported by the OpenAPI specification and more details on the PHP annotations please refer to the official [OpenAPI specification website](https://swagger.io/specification/)  or the [zircote/swagger-php project page](https://zircote.github.io/swagger-php/).
+
+### Important Notice for Arrays
+
+Pay attention when specifying arrays within the PHP-Annotations. Brackets ([]) are not supported by the zircote/swagger-php compiler and need to be replaced by curly brackets ({}). It may seem to be a little awkward in the beginning, but eventually it works out.
+
+Example: Instead of specifying an object like
+```
+{
+	"a": [ 0, 1, 2 ]
+}
+```
+write
+```
+{
+	"a": { 0, 1, 2 }
+}
+```
+instead.
+
+
+### Unused Schemas
+
+Schemas describe return types of the REST API. Since they can be nested into each other it is not always easy to see which of them are no longer needed. In order to identify orphan schemas, you can run the `unusedSchemas.py` Python script. The script will load the OpenAPI specification located in the current directory and check which schemas are indeed used and which are obsolete. Finally, if there are indeed orphan schemas, it will give out a list of schema names which are not referenced by any other schema or REST method.
\ No newline at end of file
diff --git a/openapi/composer.json b/openapi/composer.json
new file mode 100644
index 000000000..2250e5b9a
--- /dev/null
+++ b/openapi/composer.json
@@ -0,0 +1,14 @@
+{
+	"description": "OpenML OpenAPI Specification",
+	"name": "openml/openapi",
+	"type": "project",
+	"homepage": "https://openml.org",
+	"license": "",
+	"support": {
+		"source": "https://github.com/openml/OpenML"
+	},
+	"require": {
+		"php": ">=5.2.4",
+		"zircote/swagger-php": "^3.0"
+	}
+}
diff --git a/openapi/composer.lock b/openapi/composer.lock
new file mode 100644
index 000000000..60b54f452
--- /dev/null
+++ b/openapi/composer.lock
@@ -0,0 +1,377 @@
+{
+    "_readme": [
+        "This file locks the dependencies of your project to a known state",
+        "Read more about it at https://getcomposer.org/doc/01-basic-usage.md#installing-dependencies",
+        "This file is @generated automatically"
+    ],
+    "content-hash": "57a385db00a99dcd2f118bc72879de3b",
+    "packages": [
+        {
+            "name": "doctrine/annotations",
+            "version": "v1.8.0",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/doctrine/annotations.git",
+                "reference": "904dca4eb10715b92569fbcd79e201d5c349b6bc"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/doctrine/annotations/zipball/904dca4eb10715b92569fbcd79e201d5c349b6bc",
+                "reference": "904dca4eb10715b92569fbcd79e201d5c349b6bc",
+                "shasum": ""
+            },
+            "require": {
+                "doctrine/lexer": "1.*",
+                "php": "^7.1"
+            },
+            "require-dev": {
+                "doctrine/cache": "1.*",
+                "phpunit/phpunit": "^7.5"
+            },
+            "type": "library",
+            "extra": {
+                "branch-alias": {
+                    "dev-master": "1.7.x-dev"
+                }
+            },
+            "autoload": {
+                "psr-4": {
+                    "Doctrine\\Common\\Annotations\\": "lib/Doctrine/Common/Annotations"
+                }
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "MIT"
+            ],
+            "authors": [
+                {
+                    "name": "Guilherme Blanco",
+                    "email": "guilhermeblanco@gmail.com"
+                },
+                {
+                    "name": "Roman Borschel",
+                    "email": "roman@code-factory.org"
+                },
+                {
+                    "name": "Benjamin Eberlei",
+                    "email": "kontakt@beberlei.de"
+                },
+                {
+                    "name": "Jonathan Wage",
+                    "email": "jonwage@gmail.com"
+                },
+                {
+                    "name": "Johannes Schmitt",
+                    "email": "schmittjoh@gmail.com"
+                }
+            ],
+            "description": "Docblock Annotations Parser",
+            "homepage": "http://www.doctrine-project.org",
+            "keywords": [
+                "annotations",
+                "docblock",
+                "parser"
+            ],
+            "time": "2019-10-01T18:55:10+00:00"
+        },
+        {
+            "name": "doctrine/lexer",
+            "version": "1.0.2",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/doctrine/lexer.git",
+                "reference": "1febd6c3ef84253d7c815bed85fc622ad207a9f8"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/doctrine/lexer/zipball/1febd6c3ef84253d7c815bed85fc622ad207a9f8",
+                "reference": "1febd6c3ef84253d7c815bed85fc622ad207a9f8",
+                "shasum": ""
+            },
+            "require": {
+                "php": ">=5.3.2"
+            },
+            "require-dev": {
+                "phpunit/phpunit": "^4.5"
+            },
+            "type": "library",
+            "extra": {
+                "branch-alias": {
+                    "dev-master": "1.0.x-dev"
+                }
+            },
+            "autoload": {
+                "psr-4": {
+                    "Doctrine\\Common\\Lexer\\": "lib/Doctrine/Common/Lexer"
+                }
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "MIT"
+            ],
+            "authors": [
+                {
+                    "name": "Roman Borschel",
+                    "email": "roman@code-factory.org"
+                },
+                {
+                    "name": "Guilherme Blanco",
+                    "email": "guilhermeblanco@gmail.com"
+                },
+                {
+                    "name": "Johannes Schmitt",
+                    "email": "schmittjoh@gmail.com"
+                }
+            ],
+            "description": "PHP Doctrine Lexer parser library that can be used in Top-Down, Recursive Descent Parsers.",
+            "homepage": "https://www.doctrine-project.org/projects/lexer.html",
+            "keywords": [
+                "annotations",
+                "docblock",
+                "lexer",
+                "parser",
+                "php"
+            ],
+            "time": "2019-06-08T11:03:04+00:00"
+        },
+        {
+            "name": "symfony/finder",
+            "version": "v4.3.5",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/symfony/finder.git",
+                "reference": "5e575faa95548d0586f6bedaeabec259714e44d1"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/symfony/finder/zipball/5e575faa95548d0586f6bedaeabec259714e44d1",
+                "reference": "5e575faa95548d0586f6bedaeabec259714e44d1",
+                "shasum": ""
+            },
+            "require": {
+                "php": "^7.1.3"
+            },
+            "type": "library",
+            "extra": {
+                "branch-alias": {
+                    "dev-master": "4.3-dev"
+                }
+            },
+            "autoload": {
+                "psr-4": {
+                    "Symfony\\Component\\Finder\\": ""
+                },
+                "exclude-from-classmap": [
+                    "/Tests/"
+                ]
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "MIT"
+            ],
+            "authors": [
+                {
+                    "name": "Fabien Potencier",
+                    "email": "fabien@symfony.com"
+                },
+                {
+                    "name": "Symfony Community",
+                    "homepage": "https://symfony.com/contributors"
+                }
+            ],
+            "description": "Symfony Finder Component",
+            "homepage": "https://symfony.com",
+            "time": "2019-09-16T11:29:48+00:00"
+        },
+        {
+            "name": "symfony/polyfill-ctype",
+            "version": "v1.12.0",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/symfony/polyfill-ctype.git",
+                "reference": "550ebaac289296ce228a706d0867afc34687e3f4"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/symfony/polyfill-ctype/zipball/550ebaac289296ce228a706d0867afc34687e3f4",
+                "reference": "550ebaac289296ce228a706d0867afc34687e3f4",
+                "shasum": ""
+            },
+            "require": {
+                "php": ">=5.3.3"
+            },
+            "suggest": {
+                "ext-ctype": "For best performance"
+            },
+            "type": "library",
+            "extra": {
+                "branch-alias": {
+                    "dev-master": "1.12-dev"
+                }
+            },
+            "autoload": {
+                "psr-4": {
+                    "Symfony\\Polyfill\\Ctype\\": ""
+                },
+                "files": [
+                    "bootstrap.php"
+                ]
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "MIT"
+            ],
+            "authors": [
+                {
+                    "name": "Gert de Pagter",
+                    "email": "BackEndTea@gmail.com"
+                },
+                {
+                    "name": "Symfony Community",
+                    "homepage": "https://symfony.com/contributors"
+                }
+            ],
+            "description": "Symfony polyfill for ctype functions",
+            "homepage": "https://symfony.com",
+            "keywords": [
+                "compatibility",
+                "ctype",
+                "polyfill",
+                "portable"
+            ],
+            "time": "2019-08-06T08:03:45+00:00"
+        },
+        {
+            "name": "symfony/yaml",
+            "version": "v4.3.5",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/symfony/yaml.git",
+                "reference": "41e16350a2a1c7383c4735aa2f9fce74cf3d1178"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/symfony/yaml/zipball/41e16350a2a1c7383c4735aa2f9fce74cf3d1178",
+                "reference": "41e16350a2a1c7383c4735aa2f9fce74cf3d1178",
+                "shasum": ""
+            },
+            "require": {
+                "php": "^7.1.3",
+                "symfony/polyfill-ctype": "~1.8"
+            },
+            "conflict": {
+                "symfony/console": "<3.4"
+            },
+            "require-dev": {
+                "symfony/console": "~3.4|~4.0"
+            },
+            "suggest": {
+                "symfony/console": "For validating YAML files using the lint command"
+            },
+            "type": "library",
+            "extra": {
+                "branch-alias": {
+                    "dev-master": "4.3-dev"
+                }
+            },
+            "autoload": {
+                "psr-4": {
+                    "Symfony\\Component\\Yaml\\": ""
+                },
+                "exclude-from-classmap": [
+                    "/Tests/"
+                ]
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "MIT"
+            ],
+            "authors": [
+                {
+                    "name": "Fabien Potencier",
+                    "email": "fabien@symfony.com"
+                },
+                {
+                    "name": "Symfony Community",
+                    "homepage": "https://symfony.com/contributors"
+                }
+            ],
+            "description": "Symfony Yaml Component",
+            "homepage": "https://symfony.com",
+            "time": "2019-09-11T15:41:19+00:00"
+        },
+        {
+            "name": "zircote/swagger-php",
+            "version": "3.0.2",
+            "source": {
+                "type": "git",
+                "url": "https://github.com/zircote/swagger-php.git",
+                "reference": "f10ab7f81d89dba97653a980cc90cf4b7b73f543"
+            },
+            "dist": {
+                "type": "zip",
+                "url": "https://api.github.com/repos/zircote/swagger-php/zipball/f10ab7f81d89dba97653a980cc90cf4b7b73f543",
+                "reference": "f10ab7f81d89dba97653a980cc90cf4b7b73f543",
+                "shasum": ""
+            },
+            "require": {
+                "doctrine/annotations": "*",
+                "php": ">=7.0",
+                "symfony/finder": ">=2.2",
+                "symfony/yaml": ">=3.3"
+            },
+            "require-dev": {
+                "phpunit/phpunit": ">=6.3",
+                "squizlabs/php_codesniffer": ">=3.3",
+                "zendframework/zend-form": "<2.8"
+            },
+            "bin": [
+                "bin/openapi"
+            ],
+            "type": "library",
+            "autoload": {
+                "psr-4": {
+                    "OpenApi\\": "src"
+                },
+                "files": [
+                    "src/functions.php"
+                ]
+            },
+            "notification-url": "https://packagist.org/downloads/",
+            "license": [
+                "Apache-2.0"
+            ],
+            "authors": [
+                {
+                    "name": "Robert Allen",
+                    "email": "zircote@gmail.com",
+                    "homepage": "http://www.zircote.com"
+                },
+                {
+                    "name": "Bob Fanger",
+                    "email": "bfanger@gmail.com",
+                    "homepage": "http://bfanger.nl"
+                }
+            ],
+            "description": "swagger-php - Generate interactive documentation for your RESTful API using phpdoc annotations",
+            "homepage": "https://github.com/zircote/swagger-php/",
+            "keywords": [
+                "api",
+                "json",
+                "rest",
+                "service discovery"
+            ],
+            "time": "2018-11-16T15:04:29+00:00"
+        }
+    ],
+    "packages-dev": [ ],
+    "aliases": [],
+    "minimum-stability": "stable",
+    "stability-flags": [],
+    "prefer-stable": false,
+    "prefer-lowest": false,
+    "platform": {
+        "php": ">=5.2.4"
+    },
+    "platform-dev": []
+}
diff --git a/openapi/composer.phar b/openapi/composer.phar
new file mode 100644
index 0000000000000000000000000000000000000000..89920d3980b4ac429ef7455d6df9c4fb7bc6c3be
GIT binary patch
literal 1811606
zcmdqK34EMabvAC+fLH>Ar9hUkkuxLP(l|TFjvYmoZGjgdDGRY2Wh9NPNk%iu%*eJ9
zJ8gkBEM=!a)0D6kQd%IPEQOX3c0vhDNx!C)odSQpHtbN^K>42M+;i@G-(^O!6Z(C>
zze{s8?|bjL=bU@)x#ymH?tP!Oy13YE6sIb+V!3uOw=lQRwRtQ5TU^(bTbCQ3t2A@7
zm1;Q$xKL`ea`oBVV10g}-YhrzP=bXD(>=L8m1;9LFjH>Kmuj`#=32@9d4c`WH&foi
zvi$eVdSj-NyRd%XK&4cx=Qdy2H>H1H&@5MH`lh5POK-0?a;>>?Zg#O+%}v)AmKv1<
zbFEydHj}GXrpvV^=&H@u(P*nuuWiUJRLdniJ6I_nVs%!va;>>ii%eAO)1_8<CWl{q
zJymbD>ht8PWvr5|VpmtOm>W1)ugv5IMh0`!rNw5Yb|BY4(<nY$YL%+Fa-&giKrH1(
zZm!i@X!aM22P&<(#i_pO`g{@pFH(yBs5G04<!15BbI<Oo%;pN^!!2+Cx+kl3h-;>h
zFD*>xdvd3r4yG!T2g)sZ(pRpPrmE%1>1rk8<#cI!t~@!lG&xhAEiG1C`JSHKwOy1l
zn>bi*NCTv$)F>CwvHD0G+mOqj)wi+lj5G82Ys>jjZNUI4nnp#cp$)l>J)5MZqq%Yw
zQg_X=;_RVDrByDBjc*$o9R=b`N{t$%*Ds>ne16U~nVT;yiD1#cp*#=Gm20`iVvSYM
z=}wCL1FALC?92C!>>QaK+Ph0z&uZ|Zg&A}*o7<3s3W13nWE)75Px&OtRjyUfZG`x9
zOZCMDM?)XFBGu&4uA^ODBXgxjfB$@GfqxY8Mrhvxe&#`2qkQ#ZrBTl1`2eaBD<A{T
zp5!-mbxlt09N0BJIk<QC$ndVA(L&Fr+}16R>T0}n)v9kkVAZO}u3EL~0jnN`|5_z!
z2ftUX$Nx7Q)5Ss4w}Hi0of=iNKX#OB<wmJhZ=lY?LTS~i!|&gJDPjGt;vMe;cK&OR
zX0TdnHg}g6Vrc#t6cG*K54!(0Z)lzfR55A>si185KOX055N>+;hreMp)~0H>o-Phd
z)fZdfirr2Gw?EJ!ApCgcl2JpDZ;PNYJy$tc4p2P&B!`0VC%1m(PYp$PI}{hfoCNjn
zc(|)i_}Y=xZ?^h<o$7BZFO+LD<=XU8i0O9}6X3RM-~9!{w5}bdL729N2s6NP`9mBQ
z!sq?ROHMQ_r?<sYugz8t1Q>4K;V=+B;flM9hM~7DhDI4fYot-XvOFF3&NXK{M1<cs
z^+g{uL~VP=iKQ9P@;>cg!f&3o@hOI)(2kaEmC}J4W}*PcXSX{t2*3O0fB2~3*w7Zo
z;{1XQlmO3-+S7y=-}Qp;8lJYAGIY2+9n_ya!O=nZfrk{2SpBv=v!hy{Dpdmvzfuf@
zU)}k#3B%Bqf}QpGP!gv-+EG9_{dZ5ib=msEwMy8b=bZ2A6W;!+4Ss~@+evb`*2JtB
zvhZsiZiE|u(f>jtp>2oma+7+f6y3@O;k%#t!+%>A$KJ&jjsHv_%AY^Xu|oKcfBycL
z3`b8pR-oQ<m8nWAB;+%yorJIb$1nY!A!*yAqviQ}%=&!y362cH|NWz%f8H>(mFTFO
zLIO^{f2u=4`2E_~USlZQax%JD8=G!a7Q&vr>5&cz;pcAu@!uPg)D)l%7afLyYNZqk
z<bi5@2;aH(zr8VSE08f*u<5ygn{Q~p5Psv#pI&Z+v^9cb<?5`P+XF1;ZglJrp1k?a
zvkXhyfigB%KNQq|=5t(q!gv4a&(E~_?fGbxT8p71HfYBZp7ENitA?S_PQS($4;&~r
zW9s!bOzxae2|s)I2m1|2TPjj=*tf=EARJzL&c7Rmwp3hPs#Z8u0$Dui@eT)J>y$ko
zFdR?L;KZ!~Ffso|yOHo;AN;JJS*0hXD{=PWT>dhpmGIEjm%hcSWK27(0_TQc3{Kf}
zdAQapH)c!I;#EBQfsSCpzkcLCKYwK~Er_1ktk*(hB<2(9%Mec2zT+K>3<at+>(z2`
zU||8v3R!coezz(h;XSjDyVsb>m~&kHovryQ%+$hS3kw{qUBD0Vob^!03*nc4e`wM0
zq=XWA6r*Jj1ed<Jtvp?!tKTG|XP@X05pI3mUv(RzGm;&WiAdHFBXw}NL}U*{H&QVP
zfBw~jziY_0tN__wtc^rWMV%b{17aT3z9W3k#Lw3m=HX+)yt~x6N=%bBQJ#r4@*mm=
z;SWFejsLJlo^^bUT!dALOmm55eyz<AKKZPt{h>8;!7(-CDUa#+mw<ykK={w!I{jY_
zHzww|ubt7I9%FG#6vh8$2Ydj&KM@n2{-$!5A->?45POlta1X0M2)Ez+QNPYOHlm%8
ziy@wSkRzJ#7i0hE4dq!YAlhklOt|k*84&)?O)vI?KV?8Ngga8NS1HF^);N*~PkH!D
z*P0wt6KWQYMjfmBd7}9fES}gMgb#Yp^&dYLG;|d(qY%yW9_-K%e&^zm=Np=pa5TDD
z6|?%3r#c*jCw*qEH)$*3h&aZ|LdyH#J(DBC_YD2Ud%;*EEUUSSCbsH4Tgf8aaqZLI
zZ>+7P?-k!LjIFJOAFl?N@aWG!^`D0B%oUP%NvQ>YX0BAJQSKGhB*M8L-mu}gu<dA+
z7UqcUD@q&T3m<gi(+%57Y&cmJ+V<3Eq&vT-V}|gxUwhoA4Oz<3%1}^eH%L<L>8>{k
z?|I`Z{ZKh~1>$m8E?hh?ba=YFz-S#YzD`9(_`7Sa-eYuaS}{g=>Ks1J7}~e!(uwfn
zr(Ab}p)IZe?Knbw&C;|{_|uad&xAky#pD0dP^IVO`k|V{@fyVe7kAj^pODq>=sJe*
z3D2APr6Ef9r>|CoYSj*uN6QQKMk{8tJ_|-Tt_a`tjhjDg$kM|zLT2RbF4am0oWrq0
z{Z_&k74G?4Lz9+M8cn$oqv$i*lY|GFtGx@H9+?5EgVRl;DuM2c${^wCAGy&pm{!;b
z-FAejBXpv$>blebyI2482P@%^Sei@CR(U?AQeDa)VPEg5FE&&uA1|VyxV_$NR;tkh
zl2m$x>p8;nHoSh=@T3-SQMKZZN-O$BBsV~+?0&*x?;BSnogLRy7UEcbsuqUu=$Gq0
zRFa-2A{p;I5XbQhHSmPj?*CsuHg-Do(@3c`7ss;oF^(U?AKmpZ?_$EAPD~RKB_mi&
zj+Caaf-VmoZZ$AsU3_y<rwGDpp8tgFjiSDexEk0*m`-TV$JL4ueqqNQHyNgmydl;b
z=j}VH$%J3K<FxM?mQEc$nndz?C57<Zho5!PN|1~lti>tW@mMEi!auy{;`XtFs9JG6
zNy)dB9l|?bdFfW8q?35Xq|%1PM%4N0AQE2oxle!1Fg*j9V-e(loQtQMgQW#-heQu!
zX6PT4Ji^iK-~2m6+DY0#8e!w`I#vBJ!V^Au(gzJmr-5<lHE|qoQzwk@lI~64G#s4>
zxhy$C|MQWq#|XE4;*BR6lGBqS8ai06wYF873v@`Z8kFB;<qaxi!dw63${!kr)X2)j
z5Md;)pP-LvbqGIs(`Ww7aHTpd6W7?{R1<-ysf3~b3+*t%x%WR{<BE~Gm=vR~i=3-t
zs&!D25+1tj>0^d8C4672xTikWtoom1cUF-Q{{20>J-Z0l$Mti&*dAFczjd<{5n<)}
z3w&hlDaks+dyCv!7~ad4{z%OqVd1&eR~We`XVqGqorMUc-a6GW!Z*EhZrtjfl2H#a
z4@3&I!3%WWCw%Bze*7(~cv?oq;k~f|_U{_>BV6^5fA(mrn#JKB?Akg?Z&ykQw_nlx
zO#528xR!Xx-@MjIQ($wKUqYq~u&w+X#S1I7Qe(;grgR?+&y8wJ_}7m(*+&^tPQI_D
zSVkJ<CU$M&@ut5A+r&nA`#Cq<Y@DR*UxZEnk~m1<eD$hxLU`NJKRVAarIa*<soe16
z`_Y)YNe*H4u4^iWq=Se&Izb}wqOU1ag#Uii1v?B&%E|S0EuNDY&EHmYLwNb~Ugi6F
zJrKuidW2|KWvY?vvdgv02p{v5I~zt%$}f%3>@K%3F^D_hAF}svRLvv&<l{cE+EArz
zxvvFNP+#!Bh-1C#JmKx{-+k0@<lA#)o(b(TQIvH7M!4ygT>Gsp-*7QRA-(iB>Vgu^
zOx@iyGCHUpourH*@hzQK2siz*;@yc3LLnwk8##KDx+H|3e(w9eXw-DjypihSfl4ig
z<tEi%!p*J!@@wV}3Zq+(F*K(r8p3rOFBmamQgH@ftEeIP@{G#4>6Te;Vf1$Gal((i
zVTa$$#I8$Rj~Pxb@;_WQP%O8o>Jq-}me)@kB~NQ#GYkPgR2d;W^$S<~5b)Zz)nQR8
z3yW0?Y5X^4b{Yi2Rabr1x1Soyo^ss<NY0@Xs%?aQm;UrC#z=}1m0wX}u=<0Nc!=6J
z!VkRmx*H6^c}Z400=I0o^)tOZS&ZR78X$j|nis+=e|g(??gL6}$jcn_a_wot-A}#g
zOhZZW#Vl4ED$9XIzg~wG;S+xM$1{fOtd-)5FR6csYV1nF6E=V5Z&r?ut_6IV>f#I~
zsKR@VsutnW`_BKQp<BK`us<YwJt>r@sg4l7c<eIoyrrhrl<=^mO<}uQZ6x72w^dFz
z0?$}hd^;IcigSCC_AB9auYR-_UdmXeNMkyWMsP$!o~UvoeDixA`(`66HKJ0O;LnGM
zUaeCR;ksu%+&d!O?f7FiDpC%p(;wyZf$*vM)4pniq^!0P!hWlkk*wk7MdJ94GDBEj
z|Lunxj<zyR;E4HxKiBRjeCua^|0jklWxS1`ZOFC2>SUnNC@s0aNz=bQ!tq7;rKevq
zWN1=utfAResa?g2FVH?F{LcHH<70DeHJTOa2qRojskMqj^9wEH<k)Ybe6F%Y_~`Fn
z=fgZ(+x1ZbW&CGMS~H58@GU>vRWmkMO6_iB_f^_c`_ri_er+iGS4i#H0(K2&E67_3
zWe~=(zxL`I{BA~Sofq~;graz{OFE36P_UOO_k<5Tdc~5F+E!O~F(*=W#r;h*e}fsA
zHj(h$N4@0~L(^V}4$W}wAUC^}VpxkP20f&ciomIpZZniAN5-@3P;vr|2>Vavk8t&M
zFSypQq+Hen7Kb6GDDOt^kw?O}9{sP+8LGBpR7EvfueTI?MAwHjmO(iDrkD5y!Z~eK
z#Bpa{LG6Hy)&L>MAax%C{6y``-tSAT9V14kuTrmSwX6Dez?Yp@_L9!EqeW4$<L|%%
zlg$!Nf9z#*CY01_$(mIdlk6MRXC=J%Q<wWGuq_gbH0jrz_HaiD;ln3?&j<3`n(?4t
z$jM{XY7sv0rJtKLPTI~VLH!svAJF-daLs(_Vnftc<#b@$0NsHNLuPS<raLugN_ggn
zKmTDv)poW_pc*MP6w@!2D8f^3JIn6_wDqSFn2_Ovyh0bseU|!Dgzw&bqTjVjOG*|N
zMF|a7s+HDK@F&^31PVswA^h&-&1V>UX+I^6$mO46(Q&Y%F5_zz8{xfszqQ4%rAI)3
z&8-f8SE~|!^6Dq7x2kE~(W<U>zuf<p+A6}o{^X}FS+2fYaYYOLk81mbm*4r{54HMv
z{EmCOu6<jv!G*Mho1ZBP!k+xPyA8$Kjwk{z`D8H5&LI5DUw__@pAGF0G-MfI;Z-k-
zD^wPQtDe#8&17mu;li$2fle(s&m;Ox-8kvdjvT`8f3~&C_(`dLR1Iia709C%JtC%e
zs=5-s{<UB9>3yk*B&sGs!00EjOsbbh_?mma?gMO{=rIR9j_F@U9cP3;SUdJ5<1A$#
zqH3B%j^zF58Hw8TFo%lpE$<pT%}{Mh_E}U-Q??cxT*pO^V;1ZWP@Sqkc<W7*CmY(7
zv54vyWl1RebJ0VR_)iaZxCn=De7ld8r$PWxH6O&9kvT4UL40pk+6arkxXzFAlou3L
zi}1B0GDLcD*B^V^j}7(s_{%~XJ&OtPdL@+bci#Mr1ry@A9dU{~6k&{cL^Z`o_>EWn
z+v^Nt%I-$>Em{>lp$xCYB*6Y9{PUmv<g<pT(}>{iU>wVHbiySpy!N)U3`@#-N7af9
zSIOYw7J98*9WB$FBu&GZc1aWA^ux~mYr~c@KS^wgX&C2?i09+#@DcWpue-zWbQ-5Z
zRf?i}bWKLM_0!MsVf;>-8^_daov|_E@FOZO!UymDy?-;xjtSox5x&zEAK_0fntGVw
zOSzCqiABjI@wLF!YPkuY@#rU=ZwNaXCSf>2XuWW?_5$JT2gm&qw9|kRRO%i^7$X;E
zt-8#F5502htBky5HN?A|PD5@{?I*nHQGfHu72_F&55&zO)wEYCkA(GCMJFpdjw0xP
z)E(5<w<}kKJvTk?s8Q2VUlCU1>AE<k$2{4MAi{Hw{Fh$=bTSVShG4gB<gpGB;TukT
zmbWvV^dLeMGlKu3G9rBCW54wg<8Ea%1>VVqE=LpLZGSNF$A;*<4%RfHX$-rOIQq5}
zGpBVmr-Tdt_?K@tl%3ebo-}r$T2qTyJGHhoaEi(K>5p;L-3j=G*~?yOs5+@8QE4WM
z$Fu2i4iDke?ppOe!_&#65)Z<58B|>ihleJW@UEd7e_+^pI>=11MXdc)m4Wcq>wfL`
zB$s0iB<)<Bitx<B|06Mkqffff=Uc1*&#qE{C#PCY_|Ydl$1gt4Sca*c2Qnn|uc}pq
zw{QHhHz}#mU8G7yN(kvTzRzygJ|q0hJHF==nL7<&;uwMNF3Bw((_S^`gp=hv?pbMn
zF{8vIxmg$Fgn#<H-oIHX5|9)jdA`bp@Wd0|;rCgV(?gFWW(2}i?@wL2?}aNQDUvzu
z<a;#2MEHQW|1ZB6nhN4a)ryy3%+t-ya3fFDNr3RS9Z&K>-%j?#&_yl8*nFYdI>J{J
zn?9!l>#4XC=~5)<)6e4wrC&H*f^r-ogm1q1vX+S{6>ru0zT$gz-X^TJ-n!c=wrg1D
zLtp<!ZJ6+Fzq#X4%hlHu5ALDL2Iz>o_=MNL?7`k@<$y1?dX0_=6`xgJlDdDULL+>{
z4?eS}ecS$sk;O@*ewg=g=X+oAJgc8-`=VB*%9FQqJIhrByQxxtuR1`u<(k7j@hTNC
zkh<P{u<Co&;1gbU>)Q5YOI2G3?ym4;I_ux6A|rgqhkkjtk-au4GOf=uG8+1G2u@eC
zO!$NTfBlsqSd)ohi0$r#vQ-`D*K6~HJIg1pvD(>Ei&jk}y$QocH6{GhJKyQg3&S4A
zds|4yG{8y^>~@t1zy78!uVSfG8mol4gZ18}hKTT#Z+!mYMrhl5m@8PXrBff_sr56x
z3zOZoIAIxS#h>YDBK*{UKi5Z$L3sv$d+G>59bUrf2(cK+oxV+%4TP7izh&M?hTY1j
zZ?gc2J_Y8Lk|i8|=>x7=rlvZ9tocjyBWn^~d^5KFF`D=<Tb)4hG(wCnN6V?xBQ=mf
z_=2A1AwzIFki>?t^bb!^w%jJLtoTX7_pfjy5KevU*M5ULo9!`d%OizH#=?(P;Dqy!
zKj?Q$R%f<7wlH&8kE62L2dVuayzRWM9~<H8vMaZ6N_rlvo)#{72t3|HCpzK@zqILN
zuP`jxI*s|3P2Z+Nknql{UgdY<vvq8|+KhDU2B<Upi}2oCU-3(8FD0x?XUbEH2ZE8i
zM-30*wWkbSZ8brIbkr0g<J?DEAcf&Mq9+U=z0nG%)WQy07v)ZYXj9>K%&w%9aOuU@
z`M~_s&>FI^{M`w?4Ij(vZ6ecgzSMUmwqbsxVkP|j9bfRKf~k73>Oq~4T;-Z1eET2%
z;!8&Jxu}B~XQtlptrnZmg<hP8<Ao#5A~DL7vTY+Zg^F+;|H)5WbE{$Gx_gBftqsBV
zh_a)Q@IT9|y*opdQ+&KGrHG3`dKVhl>um{XbCnrfHc+YIbYWllaCsH%2}Miz$1nNw
zgGLocoMBm-pRL!Hz;EX9#y`K<VIZt`zdsh&Y=c2v4s$)IBk#g6BOQd7-SS1BXhHXN
z87$Oi#YjwF#qJF;5%%5vCq74J1(<eKn&Cm{%hVbX-tp!)JkexR=tPsqNPHS`41S2R
zq=xWO|G3lp!JFIT*j8@Jk!+|~aFW!X>d~05D{X}D{i9oadf>(t;WcE=>i(HZjPPNb
zuJZ@Wo^g!0Vs7ZKP??Mq-umnl-fO}O)V-NdtcWVfnTme0TW>(XoA?MHU;V3x7(Rw_
z+9*hbpvY68GdLX(JtEE{k9If-AMxC0`y8Z|(Krs{9mn@Q_yeSo@GFo1$Q4E-7YEDJ
zc<FpKj_yQ7NBH0uk8~Tl4&;gAF}BWZ-lS*$Zk?nEFT3i6evB~+(uTYkKJTf2LZ?c?
ze>nUv6Gm8I9Fu(;M`V`KV|<+Yz=Yp?;+fw+R*d7gSqizmt+K~>jnYW?*BgGY%P{hQ
zX**_(EHg}d>NtVDh_#Wupwir>QY8HJf8;!3F4>kv>@mg0$}Fl<ZwTKwaG?(Yt%z35
zT^MK_Se(aF$hSaR->QfS|JP?fbeRc(8#Bw&s)+Z>alE!@Px||R+G%Jz=((?7+*LnN
znf5<P<tvp+!aMg)lnost$jefR8!TFtdFHc4IUBE0goOWm&xUgiVTXa^PudqVPTt?5
zwukV$FT45z!`GpI+Tx3K{EJkV2>ZYOkP{4X2g}<Q@mOog?IQ0|-;D4@FMs-GLwaF*
zadt);Bl!=NWWukGe)%1SearD;A8R51jB2p!A#M~9_MQL4DZ|{ssyJ5`w;RAa^*luW
z?O~lH3HSWP-G65ov9M{USRSLz!4bA=ls>}QJHG4}2pt>?!&WI(>jxrC{{;t|<DGEs
z(wF(T))~vvmBHk+F1iR`_3OX%F@VCd$RdJHQ4fId`j2jYvk7W>f*g*RhI|hV%Pa`^
zhTr+rw+$I0uFFW>A=AU*Ur?F|OE0+2uhp0DykR=G1BY;<UBuzEsx9G<fBEtwMpSWG
zqI8)Wq57i6M+q;u_1nL1s5(reu5uBm4q$oEys#uI&gNPtD8lc*>#ZJ92OlnhC?aY?
zQ*;O)xAn!}G@?3~<OHta+6*r(aW3Jr9`8sayyxUQZ!vVsllE-hGXN(wBJKB;G{U={
zGUb)4!-T1%jX4J~jdGtZTnP{S?80-5y!^7gInF>Vt<76{9EXHg%#J?KP%N*df?}{-
zmAf7K)o~#_@0};U%+M^a41y+-(n%UBB7Ez@{Pl)td5H_6u`+IOwBEZDV&DiQeD>E%
z?>1Z;Rwf=llbohvCLH<Nd*5n!*0skYV-FcH2M)?#l=S0qjEvL}zUwQGxzVt6khEbz
zBCtm@sl*WOd+(3#Ff^wxM~uht_D48E2p{y6)81zoIv8+Ai2T<aEVA^thdNAzANk6Y
zUT&B=I2a)&nx>e^ya7B=M1<F0IewGj>(KcDzHQ}MTtjJB61)b{BH|+4xVP$4X**cD
z09TBxn^d_8|9te^e=)K;=uv>mIgS+9$*S9g*MHz`zq?{QW32{m`4pVQ^EwqfVeXVK
zzs2w{Ue(4?+%-Hnv}bInM<+?bCx75)PqCUEgaSVm_aN9>%Kah{o7Mai_J8YBRYTRG
zubfL+yig8j;$|SZ+lja)bygz0_rg2hZ@4;;rMPs77omEuPM3sTZyK90R2|mGS*T+D
z{~c9J!p5GTtumY)`d@MKY-dE<=T(mg-}$;1`(;Q67e(<zIBrm}6K*^s=e@5EE=mkX
zjIooIAi}47>B~NQtAiy_WY9>2ZX>cj{t!3X2!H;RdwfDb2QS5@Gu+=%N6^rIQXkf+
zjU$}=+$}yv*r5}xIu&)cqE{W;L*o3tE`$l6_oQe0#Y=}ywCY9sgLs~%v=Nrym3y=)
zL$3Yc8=KQv^7|(`DG`2X_E$cz-GK&~gmf|HC!|T0HsOQseTWY#br>%WkNh{L0(Yqj
z5WeB-pSjZ5=`e9CD(94Vn&uE_r@bY7(WyJWXoxy^UgL0)cz7YYC(b>gnD0^+3IFQf
z-{Mb+bzo7^X}Bjs^-eXHghT&)gZFtehMmOs)j+}GM%J#S7K-rvRdZi4?lMfS!y#+2
zq~6_rxU|)3_N4VCsF{fTe+RqH6fWUg|MM(w)z-E{lH6i^^?*Y__~DP7@*$%lV}<Rg
z@PP!k36U8BdcF#Uu=esd{(&LQ7^W$tZqrA|^i5G*uTz|ay<IatIG17Hk~nczYCfJ*
z^gu;M*t_OluS6M3=_Io4ylXp6-Varm2*3Nt*=vmDm0}#9Yt#?X_ZHM-c$?ZK!Z$wh
zqxTrbjB%5q5qplf96ct#cR$f}G~v#leDOp>m$Ax7qJuTxJG&d3iH*1vuY;0N`3V2*
z=DN>0$_V=;QC@@tq;Sg0Guszya#>W2))#2^5WZ>0b8ANHrnXvb(XFCKj1%p1^>}15
z49(oU>&yLyw(W@XL<Kv0U64PSCj7w7pZ2S@;dZn|3Zq!!11I&GeSgm5TqA_9&+YP8
zf)1^`krdvTbr1JrKmLIWFE&x^UU>sv($<Bk7RkvXLQ51E;rFk6(37p9XB|gFu1hnU
z`K<b{g#Z4s&-oP{b_SB3uT#Zk{NDa_y;*;=RnhB3^jA!}71Sovgx$H{{HZaWF)%W*
zGLs|Iu>PG648nzbCj1I5!zN@R9^UIoXXjk6=m>xBS5Na(^X3(DAE1lz|5oLn@XCKa
zwA=X4m{<cc$+f#WjI{j|d`-%baK{z1UWThX(zZ;uW{mWJfL-<JtAdgKQuQ<lfAzpO
ze8RZMm~mjr<h12XX+a#9Demo0w+7sAvNNG;Yr;>y>W<$tWEnC1HpsT4jR4^*bQw!n
zs=w7ISY|lZNraB7SdZPR-X7t<{_ST@H}WzTZ3%1_HfIu)ovAA=!e<vhHMnwIJ8Fvw
zTu)G3gpYdC8~vHb497S@*T~#LAg!M%9>OP{((j$E46jk~z+2Ra5qG<S;L#tcs0g2U
z=)e7WgUm4~Y7x~hZgclV2k6!+I>KEa9{Q9?tG!|b==Lnm<3_0H0X6M9MM(JS>vM+<
zVTMPita_VJ9GGUt(Kh5il;cz9^L0cMe(1L5jfOO1MouGj`nJ7PsV+9+8&U64ABXV9
zpWWs2&NIBG6k1PpOp!xEPu_7&uMs+sWj3UDJf5PzCl|IC319O##}$`CkJ?MZf1kec
zd_#W03XscM4?V?P(kxMYR7FAf<b%)k<GH<t4jo=#upF^uOhfi3HQ)sGMG4<qzsny}
z%(Q+g1FK);hZKmUf|*q2Fu)65^wbX;uj@MS>XD3=50noR%R^6eSO^!MzRIsp+7Fh9
zl_D}J7vpijJGBD{-+kRL{T5iREm7OWGLw{#V&zJP2v2?8?0bx%jFp<|nu+3BIuQ_F
z`ORCsJDTA-YfZZ)w>Y@i#P^Xfvte4Wha_gd_BLVb8@Kv}dZynYU1YVzgpBfH(=^s_
zPyeIiy$>-K+mF%+D|UzH+MxZBN}cfQ8~^0NhIXK>)IC}ZE?z9M4PWx%vE-#R%|BG(
z65jTbfAEWlj8Jh>xIl)>Ia+f~YmSo5sqz!fK6|ii#I`qoR$I<SjxQ}_2zEldi*Vr;
zd7r16=_tj!i$o$b49l@rn)!1SFX0XMoc8!*B(_$X&mgv;-9^}X*i#k_X?wpVlt4mJ
zZ&wNlhyLV4e`LrqUD-J2a^nVWP#Kw)BghHFMeTROzx~u{&oRX9=RHp)`#den&nw%6
zA3XW(ej;4~rK3D|)Wj(vucy`)HR^E>cZP$o{N)e)$|zmISRJceQ*JY;4ya%WSKswo
zpV8CaG8#?AG28{pP=P%lksDO)3174E=O-JHr?efT%ecz<wiRV2vlJk<J*K`b;q>4i
z{E4-_b{P~t{JEfW4dL&6>F9wKAXun!VA&bJ->5JN|Nc>XZ#M)P3yy_)b!oOzt@h5-
zNk=TNm?4NCtlQ<sVn)C-jbpbWqc}bS)5VE~@bbzDV@AvcX>p_xjpB<G{7Of&xLD)Y
zMM}kL{SaI{9B3r&GY1@Q!Vf+6{G0D1?!^V1u`t|kRNRD*{_?}`G~5|M|1{mPgyvgy
z)+apjqQCeD!;~>$BJ-9mXS9%x{)+i^vk)GIO1S?QTfEnuF}l+zJEef#tcVG}aQpg6
zqcvk3rV%3!oXFLBwW1@;{q?thZ0Itk;551sCk(oW3D5JeI`xD%fA$)`X}Kcy7410E
zz__>O9bPKbX*%=>`yT!=pBIr~hST(B<(9op1wi<wr{3*@Z0qqmu?HC;<Sn$cXpy(}
zJf(~9k8i&6fhNZcZ_UtTjN@=CGfRzC`9|gz+$oDV4F{A4!k2vSz21+`uxmyGFpW5*
z6z;FI*9a${^x9__6P-w@EMV*pXaHe^zliB_UB42pZT^8jv{q~{RjFH)|Hekrwg<S8
zM0oJ-AH30s%5dyGS3(p%RC3r8^c^LL@ByFxbMF&oSxb|$V7UYidHqn+BYKHW9)y4J
z&Qs@%oHZRdizNK(XFFyHU$=SP9;=<P)bY%?+7~WXs(Povw{_=^@H20_<pqW$zbpwn
zW+VOcbe(<)U;nbj8w^Dlf8=2U9=E7gaCw?4>l#%t!qW4aeq3J=G##tQQ$9GKHzOI~
zxGt?#+_vbxrc}mRz1YxRryWiB<Ud~UrYIcR%i!iET_gT-q9-G^L>o-_lc#TbzA+gF
ziIzdQ3r8v>x9bF*A_!mj{Q1>}E3}i#;38p|zhgX~syq|E@ZFE`JYy>M)@K>CKCaDi
zw*~dcGU0EozUd|-Gc-fQ7aStV=-C(ic&lHBoSvX0sL7Kt<-;Z3I>@&3$_wG}$q(^6
z4WUz$Xj{&JRVy$-IL<;Gk5auPtbXW<uN(aVjw`W5t<-p>-&FmOD5#zdG9AsSjuL+G
zdk_5!!x31q_Bdv+ZB*ocNX_?^Bf{DZ@Aq3;<B?3-V<K<V$1yp$BzY(9BElyAuJYO|
zL_(PM;hwU+XX?NvyyNcc)*9u{ywB~0t-M85nebVE_@s)pGIqbU60_e#Pn9Hm-MhB?
zQ0AWd*xXQUkw@+V@lEUWMfkzL-1bXjY{z|UDLBy=_R#*%UimX?V&uLyA-4j@*}3^K
zPH=>;e&<_$bAPpELD*1qH>u_k?j5_xuNm(vUwfzAcVol0zVmp;7vV2^-~X8Vi!V3O
z;xarzRf6!BzxdwY-fwNmQY_xqbJgw;-ul01rmd|@@1t%o#4|Wop1ul4E%5bQ|C1Ku
z<H{f5YY*-7H%&a}ert0KoKCl5jXp+KuY})v$rJrq%8C1}QQut5k_@XQA-wzZxB3w8
zeYGUgTFjE%{74;ofEOIT@$alF@2iRQb9G#E-lw?^gui~r2mS8$_Tv!f5RT8_Mnv(1
z>hQOm`|=Q-R|${4?bNp!L&r0|y#v!`>C4%Q2A^=(WpCbTZC!jk-1tO^PHW;<Ym0>c
z_Wym=x48fQXwjLK-Q}4|iHhR6=kQz9h6(@t4Qp;RRxi0<8kWGyMU6_V$wS&C;hx@~
z?6W41M>RcSJ)b#UpGi8XU)8Zr_@+l3@$1_A#-T-xGLr^s$|2z`^-JfC!{Zs|5r<i8
z|L1E<g!v!6MqCv9x$1KKzpnjP?B7K#T~vc(jIVEUaUKjqUx149yJ&%9gz%<&i$3@}
zdf!@ZR47?~AvV@pMO!2M(Um{&!Kh~)M{5?uExM`Qev+wQAv8{=2*2@)`fIG?f;G~L
z#W^}Veb!lLJ%eb&#Ig;Kf5m?r+JWPs-VNq4Z0+DfVnm)tkeoCVKKqj=`#gv7<3Ns~
zS{yB7zAU2yqM(R>#P+@dylju~sr#?=aqr{tQ4{SAE;i%{OH2I_!r}M-dX>rRcyue#
z1Y_7c$~aRm`$G~*c%pVX;dk$TgAY3053P+f%9?d2H!V?o>(m?(_MY{dhZ%3j<6|Xw
z!%yE+F}LF{w95z|`k=wR*5)97AFGvMo6?XQrDW+<s)2+X|KkaMNwMv|pjRa%{_p8B
zg7C$|TYcnW=keh;0}~U`bAgLmP58=ti#|p-e7tQa{jo-9rP&DK$9G-jlVF3G@`}w%
z?9!vV)vqM{+M|#9pu^gT`e_6x;CnX`cY|o`<YbIAQbzdD3+H^C0C6YNkTwV;JQiI&
zBO~g2k9F7x-~XD6-eCF`#Bd^P<$2f|i+r@BqbG|FS5qP|v3#KiDl>$Ks?YPG+s(@%
zu{EBL2FIP;f6}2#`12oZd8x5=_KNYw@x5Jrbiz;VJ<q352I`O$lVjG?8R3iR+1Ei3
zdy{bE;D+ms&|pV?8N?Z1ng~Nlt9$SEdyXqdn(;x1Gt~AH-t&gT7a66QiW~L!nEXlG
za8Q>8gfIT5Py3|B4W0YsV)x01DAM}jbxvf2C-uyK(g<0dfoHnZJGICw!?Lt?-RjG)
zu*$%d8a+{E>C4iEf{Cnr=&6nb!t%*aJ=L&aNo80Tnhoqu`VG|cRnZ9-Zhfh@>%nNT
z8Y8HQy)&=b(RqqZiiGgTTh8+Cc(5IpMAB&`a|NiS?h)o*aOoMw#nbRx=b9us(6S2-
zhfMwVYIVXZfAr(0TlHec>NplBr>kS5;E19k>^|wuev={C7t5f;P>mx#D*@fNDH_7@
z5B=ALMp4kY*=S<Kd{S)(;om;4={MX@2iEpt0WqGAFg@(7bKdqfqazq9NjiAuE$Zmk
zD<g!Tdf>-SGXxzvnkRRtHlV|E!)H|2U#ayXtX#eEZbNrc#8GLnRj=an94!7^ZJqG-
ze|YW*Ry-dS_cqDuO)}jE-w4HiHBrnc3c}0Y_*TD2&Kjd01!QtIE5A<FobdH0U;8F&
zKM-HK{po6{*_?NGhP7*bO(zDB_uy#%n9@%84<G$i!<z3%I}Y~Apcp>EjVZ#fzV<yn
zTosI|H0{k6KHxY_*8d8A8HXg{_VQ`|4!hGbC{LVrdcyCxs)WCJ`QjQQJJ`+eEk~9O
ztH<IrdaVIP!Dp$v9xTlh?mzc#ZxNt3tEMV7e`{Ho^tbBNJ>Haq=~e2coTZ&ec<-t2
z_Y0}bTft^m@XsqwZczQ63A@T|Q&-nwvm8Cg9Ev+6H^m>yr9IKtWG5286Q%t}`X=<R
z@Y9*evU?|jFO();>II||;U`?$6!CX=BR+;tJCqZ}qN`Yqrt(bm%n#;7E$*&PaEatj
z1ti|O>rE%ZTTUTN{*!T%GYRh~*UED8zw27QT)-8)NW#u_!^Vdn{dINy+e$t0V5WR<
zcYUUuJ3lvDsy4x344E{!yR?8t=jy!9wuNl6Hf2i-(~E6hH5ReG)u=4Ann8hu#VK5D
zmz&j-Qn|^=X$h__PPYmPO-OOIAs2LSuDenP-Wu$14}{U(nyWN>w<P3%0%_4`QK=3j
zo4SrdkO{JyxM|zTsL&k=tJPR62aQE93dk73J?*ugoRt==BGxS2VhWNHE_7!KK4>uW
zJ$MY7&nyxVlfmZzda%&VehgaDZVeFgNk=@0p=oy)x?^3RkXr0r&_MJ&vqxs|y|{ui
zD7kLv3=9kO0q%MU+of6KFV|hIPnYnuhlTovTyjX~y63S+tWQ0j(6&Io9P;qthFo{6
z)Hnd;s5J0*bGlR;X*M=09<S$oh2#H~!|0oQo?@-c<_fBJnLX5#!+6d_Na;}^0ftCY
ztd_0$y<2AV6X32>3ZZG~?;jZ*njE-jeDAKk1KTFIU9@|IoXH4MrQVP$@ymRL9{Zy!
zU&q~G^T+|s=g!~aR1WH%6v2kBo*wDHZkeCus#Xru@}XR0{0b-%yvMMuHF284C#((~
z#@&ofj0frF-YpP#A@5wd-kC}x&nCKZ{J-ug`7UB>jw3y4v3E<2d!|)T%lz5`XZwZN
zA3!;MXr|DEf1t(TX3=BC(}708jQNn6_oeD^O_Im84bgjuFUBFt_Euc6UQ(MUX*h<k
zfGbFC0w^~~VXT^#=q;&BScZUd=bx{(J-0QN7bBm~_2=@`raZ8EEhe^z`@VcJ-xs}H
zSuc59?3Do7pqpW0-NLN*F-$5gc)An?%S62swG6cjFm1?*4T*IWYo2o^$j2W4BxF>@
zL|w>FPHr0>ot(^X$mO9ZbX5wIliP=P4NXq=;9-E1_2XSXULUuyy<*Y>gd1{R%VK;e
z>0L0f3;c>FwEXzcdtHyjkwV!{7)mGFrWE4Uq}#k#*W@sTST}FpJhXRvm%4w&V$Ren
zv2w^=aKQN_<(b^nlK1_(x+bDF<4WdwZW9V{@PZ`1j*|pMYLr`xjhb4at}f6*?c$PS
zSpscc@;^!%Y|(=Iwkqaza^%eE6f^U5e!z=Dn^`3`8$}bBn7+YMtzN@zT~(WAV6`w5
ztUoyGXk$2u*DRZ*fc=aJLZ=5!u*bV{vvnx!^c*H8k7{2Yvg_TL-=8ZrbKPDiXz+my
zhf`$2DIZ?o2nJhtg4T7W*whtal){39f2C{f+aFO4Q%p=$i$Y*5K(ZP4*ucdP!O2;s
z+mq7334Bhs2TP4f_plJr9}O7Kl2h}(^K)x*_?N=7CAS75&M578UDH=VVjBsjr2$2r
z4iQq{nw||4HDCaR88~`2$;(b%aQ6RyyTA|OWrQCqcg$js)i!}U2YbG>u&fCYCk+L;
z#{Zu-BYr-0ccMe7xyXS!as=Dx=I{wdv=|C|NB2tH#bg0LzZH6Nr=Ol%UFJuR3o*Rj
z0`$nBtztn{<rtsF)wrz!)ZA%(kxJj*-%61dwCp@CI4u~kWulG`CF$u&h`!+0h%jVQ
zniT1@5Ds(q(vR_@>ak8<h0lGoWsVk9r=RUeF!l7Ud}{XH1W81O%Fp)xweLlpQM+hx
zbRFgbNzrQ93~^z?scMfDzE=LrY7^b?YS{H{N$<<Sh1(QV2n6j0kqWcigb0X}|GeRz
z@Na3n?H2{_3cCMNxi;oj!^MQZfvqnicSBSoS%nqhwIVr}%b5__(bz9E2@?ty9jkqj
zc2MP<ax{50ex!>N_=k5UGIWR2+Mb@iJjZV!4z^3;SNmqb0Iv9g11Wp1lj?p?k_d1m
z6HOK?s)iapY0plheDxxZi&8GYh-o)0sz|6Zkw~045Rwp&2e5kzuI*J-tcV~LX9<J_
z9lWM8&3*Z@JD=J+3x)1A3zpySMskQHe{CV33q*T_qj?30P$%DW6!HXF@ONw_3%Z<Y
zncYb)3Ms*0A{?)eVc{Is3a3JcN*xi*uJlOG#)^g>3I1l974dnM4G6dQo!yscMs|ek
zi%8oqeu+`Qxm*^z(s@nL(n90<L@A*xAdRaUO}IpX%h7bWA(4R6WXl;|)`3QSaY5Xy
zh(Y&E3d^5y_T>@GwovOhhmjDGtFh={r^8p3CQj<qtBbAj2(_Q4rU$D>F**H_jdN>6
zzI3z9>PmAGzg&4X8x3=6ujd(aF5;=^_2`kTQuBE<LcPZNG<pT|9H^K2x22hxL2ets
zkv;}*5ya6%+bF`{37YW4+5%7PPwCUB1A?I;1u;9Re<m`5a$9|}=N$k0iB<|GNnh7h
zPT8zc5FOEK4@F(%hj{PPT<&rh$|Ep<?kMNrg5_zE@_pfiDn)cY?GtrJP=E-;p-m!K
zW8RfmB@KMBDfc@TsxU#88@(<-hH(u?1R4Y@Jj0y0La?HSrdb8C<Ia##y%6k%q^Z1<
zSUoU=%tx5*W1)<AWv8u}4iCV)Xid)*^4Ho1#(<BU;Hyln!oJJ<Cz|`$_w@H4rD0U;
z_)F?JDtvs;J=}N3ezgQiDMXRLc+|#-3q`TDKQ}*fP63Byudd7#@@QP#302B2;TN}E
z2mE?3Bal^Mfld~@!ppLU`a;pI02SS6kv))@OMML0O;-$s{igB(4Hm;bD8@&$J2HYS
z4g;>j8CkX}G5t`+$-0B81JMYCW(-I8h4aLcm{M@tF<>w9;Q}qf^GKy=bIv&vkK(}!
zQHTtIW-y1kiKj+3`#1+zyG-5$Sy_Xii3*gFw7AH#cUPt4An#?MYDEfGn9q`~Y2l=Z
zLcOI>yuP*(nk8RfC>BZ(l==POS^knxFn;{D<m`zcpOK*bV`E@hN&6b@L>(_cS<gbo
zD;ZAF#xX~wn<${BW3r!2v;{R6d~1UMe@8lqAAtZJzLB{FO}b%VD1(-$dSE7OOnNHJ
z3OYKXI*GU_C~(HFO=F(!pe>qr*=~=qZ4Hj%r8G)wx4~~MBoyEJFBFYiTcxTHnW%hF
zC*p_07)auS*mH$HvSRjljHx3YiDw<@xp)WIj!d&OTOO=0EFl6dL21Ziklu?U+Kb=1
zIpZ}=r0bpxTz!2-_dngh7tsdD>@kqs-<wB!$HsrVyGL*wZS&ek0qz%v3f4<KDnBJc
zS)5<$UWoG_IfSU7Ya4FkMQ)aj8rhC7OTWS^6(<;K(ZSht3mKUReOOf-vjvzfs%Ff<
z(ggA0u>?l@*0|uKcp{Z1wz8w0<$cco(>0+`0kZ8Wn>I4S`AC%zJLYiKysztPF%aPE
zvB>I}msKoqx914LJ&5R+kLaP`BhA+0R8Mnj@n{8_i)DBVONb;v@IK*sk|Zp?WR;KT
z1_DTQWL*zb>b+YIAxFVoHCyP)Z^nK5_~PAxEtfdhpBtE;svKCXFE(>d>ax~f<N@0a
zxi$1`IVk(`__HgYJA`HCEI$L4tJHE+b@(HA9`35>;!sWUHiTpPauamb*5oO6zh4@&
z+1LbzRwTJYd}2)wH>e2}WD>vO^39du**37;KZKiIa#Q6T?9oi$=Ato{@8T4O|FdA2
zN6g3Rfn)nKgSvD|De3ZqqPsa)Kjh`8l0`LfPcbeul3^|QES3I#T$h9dl{P~o6DwAn
z*0op>W{*}8Kq|2!x!01BG1O*PUw2k~?oKSWs$nOp`1Dmc8PcRBNN)EywQ&Cpqo2@Q
zw_P2X4nyN28V;wem#htFAEp+3t+YOy3p18jZrX>@u)W|!br4(KxL>ENe0VpPw%<^4
zoO9R3(dZv<pcjs|s<>tuLt=NS1{NC<PgeQS21OXL!7pm2T^?YU$CC1b+bLQ&!j2t5
zZ@rTgrX{Xokt7szOPxUysb@yGErM<3UGy58C$g-h?nso<h!h;-Hb-I`jdp|+L9(4a
z+&BiG?J92SYVnF|Nozn7m@qex{3csr&B)kj@7aFi6h;XvUGyx|`XFqqb}__ZVL0NQ
z*Tn3G9HfVfj9N?FP@b<N^D6NkhDl|W%M0~+>^D*PV(=4T!AK%xHd4Zr#4=)0AVwll
zvL5jBrAQP(8`yc8WETcr<Q*?T_-jM1_Y4v5isU3AzMOBNnVZKo4mo5!IQf*KSGlHr
z`UV3lTGr-sGj;4G)atDbw3gv3kf`F(u#QX0N*RyD-l>cdKjblXmdPO!wGH-LP{YRB
zWOJ@Ei#b?34R5Gf4B>DKRg~M6*1^m-(evf%xt-K_%bU7{{K>^ys%cC^@v@*rw$(Rs
z3^;2N7X<^ZArTNf`)ecp%>b2*7DNk<5zW~dXk?z84|mZsC~_P&us?-8Z!DV<;>3KL
zH#`|l%yNa$nbagkAe34e*fj8d%=oec_9?mg=rZKK5puR*GXpa0J~hQ^*tK_@bBPKY
zx-^7tPtL{@FB%sdq-Owu<Aq~Up30wilue}+(N{JnxS^HIF?23~TlExlT$%ZixfJ;h
zCEQ*F@i9$IoCqyV?r43~i}khQyOe3GhRqPCqwRJA#PYjld$(X^5i?yFsfe-AqQaeO
z-#n+8Y@@2px8KB1Ha~)%JXR8X7j~c`DkJ7yqt&g(5+}ZZeUUkgkmV@0aT%R6=Kd>N
zUkm{p&EZ3GOe2{XO?;gbX%WJxxpOg3J#y6Hf|q4x-i^ES;CvL|7YE+1c0}OrI>N=>
z$nZ3>!hWiCkp<omWFsC|d<ug!9zc)@AP8B+5@UjiH=;qh1S^+$3W@uB-pT^G`egLI
zkj#hQD5ryAS+|B$Gp1vF?cQTddp=uo8>MrChR{w^C-UIwhTKNX=88}91C@moPU)#M
zt5R+_h{PiMfSY-Nu#^%X3<f1Q<T)P;Zo>Xz?T48&sMnO)PhFmXX4$t2%5iHcc#TjC
zlrDH5jP=#wb8&y1c@=ewz0<{d9s%jd0tHB$!ll>}4!7`)RJHt<HCRFF%nRLRk00xb
z9|&~KAB~b!{+_lkTiX&yKfb({W6Rr1ag7f2JKG%x?{<sf*_7d?qu<WB72KKqDQ~1B
z`Pc9awwWrk6;4)2iL=$scDxc5j_~-MHS|rmw-(2u8$2lEw%NMwgC@j@2{l621uh6r
z9!aUX44S{tL2GQvjW!7WN7Tj(a(R961BSiec5_x}(Pe02Eg#57HYs|yB)7t)r1UWV
zqCK`CJiD&QWD0Q_Xe8th77-p4VLTn(91Nd=gyJ+KBm{<*88uRbb_#Pr6uRaEm)q@(
zw^CzPn?EI%CZ<2CCi$YInBgik6$P)*sI=;cW+2ba>oizaLF^RtNsOfIQSq$_q63%V
zt7PJL!AlSfI=XGI^H{pWXe3kiH+lrKgwmJj9Chfz_+)Vcq4-jxTSU!TE-YZ!kRh7q
z`~ISIvlwfACsTsOL9BdVuHZUP7RD;gtuY%Au#1Ulja$4UZkGz0U^*Dkt!XNExV_zW
zQwI9n4R^jZ2%4#dkrpJX>@ydkZHb;<?rO6;)*w*H69p0GBS&(Gn8qK<0BMU2#F(D8
zE#rKRGvEOFo&RX-sL!_!!FxSyi;Kw(?5!+RqRVCkF;m)6n2dCH?$fk2d79oeSx2O<
zgQzjcR86uo4s=7UzaJeQtbB|r(X-C#;kVGZd1c-_MizbzN1hiR_}w|zPOJ!Z3|UTr
z%5b{CgSe-Q2W|t<RZbmm!Q+)|l|diQ+lYVSIMY<%0lH%6(7?8#(TtApiqN(<WVPc@
zAlS&^`RYnbNyg}4zKd-xx$Qe-=h>3Ny>*8H^`*_LIhGO8Y8P3;q11^IBspI4*3T3E
zZ0y8DP}dC{LjZG1qU@jL4C=O_sBlnMHV!lL1zEjKT4|tG8N~YaM6+C-?e8~Xa5Ni9
zgsQ>uLLnW#baSCP>HP-e+hA!^Za6nCl1ygzV7ALA#LS%&w=bO5DmVv5VVo?elCgn5
z8C8Jz?X+@smoVJxk%bH_;>ND~$^yNZE~nWx;l@X%m`)i4uv!=e`Qbf-yDr)`G&wRb
zzBBI$i@DR_EK5~Vem-VHW6TB378aWb=!ff(4LPVtJo%MAgH^*8V*I70Y?Hdlv7wQH
z(Sh;3qZs%hQex=p3K<bYCVml(GuRv(9~d7VoE+XWJnkEtpwSwWeRVg`z~{)$k;#jP
zM#qNt?wK6kmfM1v$c@fu=bek*)p{dTA-{9y#k;MB%P}jgwt<ZEA#%Jhe+?qclT^2I
zL)_cRGD7n+<r!4K^tQU)YdL;nqTy&yQJfw+k_GIhiEvo^QR2AKq3uP%MsV(@Itht@
zM4Xu}L~3jU$Ukv13yhu2C;VNJlaLJ%>bMR2*3YC}L1(CvuGVF#7*>Rb&}4KVJf*VO
z?bvvBq^l$z0b+*1v((hH9)<lAmW>Fag*7oTYT%2@3QYX9grSm#9izQlWcS<G;nYMA
zFH&t5hEqsj|5idYL$ALja@)oNjXpw<+$HjsAw?w&C;n8|4DK2j8{0iFlIR`l*X6oM
z%Ao@?IF_ajPg3tmX?B>=*GTn*ks7xZa!N}WpQyRRQ_`@DEE=-T>S>7KRqfn6I!w!R
z@lYfXPbV~HP<P5>2n(`4!N|x&6!n9nNzT(lG2uH+FOJyhG<ATvRyg-;H8+K<luyh2
zbPn-ZP}tCH4vY+5<W<%NwdneQZ>^9Inr7Uri?t#$VNz0_TS{-#cqb;IU~F9!^Dn)K
zf_hPl3D+~(*yzxX;j!_d(aG_F(H%qMk~3u<uxIwZxfW(~BA>$gw5wC=aaF{$<qPxp
zJe*Z3CjUiU;a!GIr==@u&S6k{8@TJ@izh{7b14rrkNH+tGzCrA)GUZG;JrHpC6}9o
zI?W>!E|?JwRzGVv#@&BYZFIEz^mZ(v*RVuqfxO{QAjK5mh>3^d+_R+pnO2u?)Sbf7
zH2|hTwjCkh*&#j&rUly9n6y(muzLt*WZUHSy?ddvzAw~CHbrm(1F~3Y^tQA|Cn&if
z>!1l1X&|Xs6hIHPao{)K24}3?U4$!TSr<gqAs0rU=-#nw?}Y=q#`fVby<v;odFi$M
z{(VUqx=kqSN9%Z33Ow11<HZhz<xn+wf@8<KbW10}1jVvf(9Ubx4w)WX?(OL*4^$Si
zDYx|e+&LT1-MG>B`7(>#$PsM2%=O+(dJLl#nGF4#luwBMjus+1=Rvqx8?OmJDALq;
zP{osDgV_&VLD1f33uoLmo1Z-M8aj~eXb9Y~<7vIE%LP62$#sE~)Q$|1edH&ld?U|s
zkjok+J;MC3JsDx`?@h_LVZ-empjU#ry<3tISH^Nydeug>FX19+Yv6yxgFukReQ%b2
z(1^sW-3ni3WANb{MXP!w?P9Fe>J1hqA7Sh2Qw5tWNG0Xop^FS6<c*-h!~_H96029<
zEYwKOw2gt}1s$HlwryzCcbOKr`R7+*q+J5k<`k}(*qAXH(*@lnYtxC16hA_0$e`U-
z=DA6W+`4Db8)_Q$#+J726jXFeym%MnhOS2Y)Isit+3ywLfztX|Yz&=KXFnWW{0CRw
zIPcNzjUv-cmT^d+!_8W7mu1MvFRtg2lN6mXh6L)!!csv*^zXB%a_l_>)gyOI2r`Z=
z&XWGd3!#*ziIvhMxYO?4O4gR#S!bSi?s=q7?i|Xn@e;{oE3OXuD9}guO!W?3>G+dw
zL{QV+T;r86uIS)53UF&Q)={+_p13jtPkj6mu?RiDQ<ZT8r8?lr$`>!stt-wmF2^an
zdt<lceYvXW;mDC`BBokv%@$;(7LP-AuceZ$ZQ|Mstlgvwr6-L5cDgB_%Oyvri0>N?
z8wux|9|34SR__+=#P*XOgPKyvUAz&?o#v2$VSqQqDAe%4uHnnnpNli$JuF1Z!kCJz
z72QDfBqfjSn^!nkt@EC%{(jueI1}7&&?8BUQFsQ57FrDNJ>hBtmwXd6iWF;_3iJhK
zBdi6C$dr=d__Rx?V|>Ii646mS@4S-4M_X7$#)P^N4YvrXjQe9udhsW`p+!cYB$e3G
zg|S<)9B%Dx_wvDlSaG*e6{agv#zr>qy2$8!8JD|3`SlCB(Q!5^trLtiysdmNY8#Ei
zwsg6R1rnKR`$U9DM`l0~$6TeE2(7)QvWIEiWX6Z(V=V{S%e0X8&ZRjg`3q8Bz@pj7
z;OX(__L%6DgEpbL_(66*1uzB|&yFF(NIrwQNl%$IfYgGA=3eC~&ER=<f}*5=n|zYF
zN)j!O9Xk%B;?jd8Q0Q#F=$7_>wkG-52IU-HNH#ICnQ7GJg^7tRw4(cXTSaj{{uDcV
z^&&QXx##aKuG0|N_-A@G^H>!F>$l@aFI9nDK|q!rzxHxuJb5bn@;uqhQ!B14(CQHz
z9<RY0@t2_2Hnq!f2bcK}j)PDeQ;8556mC%ZUTxP=oL7Ye5Fej@t}U(?O1U{72uvm_
zNR2t?o^)~q^4)*k=7Yx?N*K(o!~a;kZ~ayPL4i?hBqR&7<i5*``(uSk^5O#e7Bq9H
zM%6<OCK<k;g5|);q-!rP6t?!`e^h~qiN2m|&)9J0(e=fi^|(`GL$rkFoi#|P#5of;
zEu8M!p(;|IYGJet4xWoE6i|~OzfTw?n?V&ZSvoDz8$F6+AQU7Icd(I+lscV|;rvu_
z6Pf6O>j6xNa+M;a<d0|x{%*ZhMJ@`1n0N@3jd1TNLgV#Dq{(hD+v2kOf^K*1#n6V0
zs)Nsk5S!T)GpSvu#v?8V%SwC(i@1owI41&T5$16O4atM<XI2)WYbmMe;%H;V5lk{o
zm`b@}6Y|;)rz46#`7BLSMR8n1a$~H>m`sc;ZyFO#cE$wvEs}M#v*K<+iOGxo8FQO_
zJ?ZHT4<pUxh?_uzLsn`^jBUgwR3#Y_hs>jcKYQdr4i|9!KG{m=T}-|Fmq9|}Aca^~
zcBMHu60p&qG0#Q1JN{&mnaR7{B|OeVqxoa^aGV~&jdA(BB6ExkBTA+z$E(;Ap%jY4
z7>jhJQme)${zmQ>@;;XkIVlcJT$CCaZ1QT?Si5j)qIWJump2TEM!A`Tl=4w9o6_|{
z*jj}H6_?XhszD^%X%50D2u$M8CC)(E5j56xPjOHz<KZwOnGHDvhqvDSc99})8WllQ
zc0=y;sIBNnc?R5iI?0oB*zA0)P*<K&QHx2nu%gkZ??mA<DucC*+6eqHsN=bh`qITS
zyAPEs2j+O22}6W|efPA6NPX`{tQbNrX1Pr)-K9u-&(EEa>*Wn5lzll8lI#}p0v{RE
zI3^@#gsq3H#h`zz6)1qi4Ps3`_TVU<>D%Q8igW*{Ai9q$svZ)LrZVVe4JwXnl<j1Y
zMu}sB!2!ET+JKn)oZc-XgDtsrJUq!psn&#O=b<8)vhe*cXf0M?L8o9R+zJVe{#OF`
zL7>2hK}X~Yr*~JR>_L=qCZij#*RSucun_+LAUl99X`|V=u0_BAZ@LfeJBxSBh(5W&
zeW+B${U7o-uz8YzwM4e{6@T(7=gD|+ZPQ2y#8n7{i`SV))f_*3>s8V<;x65bddJxj
z=`~J<$wO7aiBz*iX4JqEt)=}Wp^?rd@yN^e6M@~tfc6&~(@be|cFKk9<r#&KbC)<W
zVP*2TM;V&tO`pDYzu|WrS})tGa7uWC+%#@_m~7Q2-OmV%wS+?g(^&q?Q8m&4)b>DW
zArD6_@y+oP{g)Dy&vp1HhfiZdCs5zUmgZ;cwI#QWS3{mUzBW)pYKwxz2M2~beJdgL
ztpcS(ybO0hYj<hEpYrCOg8XX}F_Ej=d5sv83+Wk&*ptD+NRy<ftSC`OhEQ16<}^NI
zvNL?T>Et<cdjHTCww1V+g^y56gH?IC-q=#x5TSyb90u-O8xodQl*%bV8*YOW@6=zg
z>~DjBj=A_v_IC_WH0owol#+6j0u>8Wj9c%>{G`vxQSej-4H|MIR8rr@7UX)B;o5B7
z3PL>IjkNpQ#5WOf^Dm%q`eU*!^l(oBd$#fiK2lykv<tp~{3O|hJTj>Z$d)~VB(NiM
zb3MqYMwN;jLXw;tznTmV1t-spVVyKsU#!mXEN_nY2r>o8UEF{Jn^-qg5dG(^4Wtnl
zHmu1VnycXSnVtX5A<S7u=5wpgn-fZMM;LH5_-DO};Sk-PpnFT+lm)vd?5OmO^;qTL
zg)MhN+vkMNDVP!x8S;-2Cr5cGJ`*2^te~IlzSWhQhJ?sYMLNhTBJPIly5T+B_f8I7
zIzF_A$!~Dk9JxqIU*Nh9q9&NbUGbt5VGzl+5O9?~*I;c)T`6fKlUKhovtqD@_9p}p
z!<;zaV8u8NErt-*4cExHh)ts+*PglIs9)~;;%+QcT){{}>`QRSnl^ahuM#l|Vhcp}
zOiV;>cTbK2y9`q9nhi3b3Ox@g2&|J>K~!OP%iqN69KtAO7Y$_D#|Z99b#=nA-lpgL
zLf_xd;<^b&$zQ#qy}({GNsFNLs+O>*MAl6z68{PWn7{TIZfOayx(iwiYf9Gmyh`lQ
z=WpkF77k>SNO+}+k#&sMsswHZ5X(fORzHz8?$;zSrQt}H39ft?E*VeHb0wL`VBL(Y
zfU`GlWPm}w%WwomD=h?K*n1taSb5qVw{%EiMyXLQgn6!z{#kuz(Hv|!ALfAWRzL!_
zPx42~wZ(Zxd%zDrp24BOdlNz51u!Tw#Vkz1pyI+FtTb&}QCsDU6SY_`3fONs3g6Ny
zJQXeofi=G!C32-!Zq3@}n#e{=0`7(!VqDW#<(A-nIRg)A=Y)6cmExm7B-Bq*44%W!
z(mNuTZwnV~a0am|)>sA1;&P37jIC4?=m~B&N^DU$gQX9yT6Ar?v93!WKXpVlt{v~#
zQPfEd(~|7I%+P>9EjHvTt=xQRiSrfqg-SS_%^9!WXe=(^h6w4d?sB7nhv(-6%AJ?6
zQ7zl|P)i{%_vMuYyZF~Af&x$phxYEGVISu$Lo$hTNp5A7jJa}kVRo@9p_ZtQBO<_0
zMd~z_o0rwC=wb7m5LE+3hhj!(VhZ6=v8iw(^y^zZ9_~j{U%qFeDK|z%+Q=?RbqF=w
z7lGgs6ArffCYtLC6Zx$@g^5G!dnP8%U|_Z}Q9Hu_HG1%?ZyjF79!^Z0S=f4hRO-m4
zo}*0;INE(c-yo)}j?J&xcljE|fcLLkUyxXE^t)%>`ZY%;5O1D9G&NSjKR}U*e20c9
zSKI8dAuc;TXZE<uTQ+ZXp-Fq0me1vgjBlRU>KZ}ufVT~~Gw@6It><Yz)R~Drlc6Ms
z6w_S;^Am|G-S))REq&{K<6F1Hn|Gqc5lK{pOY9Cy8YAe6ty{P%qX^{eo!b~=am>T+
zNsv>JmQpv0atnbX5@Tkl*_<AZw?`sVsUQj+Y4RjO<nU_zzAblVp-Hjde3qlWZ(faL
zy8bQup=%?XxdXOknE{ZSXs%xmd?TCbWNg`2>b+*5_p-@}!{^QRP8^<{?ccADHugRv
zcskJwZF4)*(zg50ku(jp<`(b0^tg7K(h}QEwU*f!^%^6m2}R{Uz!lena`$Wni|f|s
z_Dx*7X@5*;5&1_ehvQ0VyBpMxz^n5nI<~apmIpmLUL%W+#RUhrGoG0>n)qQmCGNuS
z{XLryu+43jLi}_6da8z)FWp7(6)<h>-`8Dnn=-LBp{&VGK7Kl6%|r_+5o;zI@?W_?
zj{9=P&c1w2gs)g>y0OV(Gb87Cb`2;GD&T=DPAfzzwG1XVT;_rqe+Ji@a|_i~<zpV`
zDa&9M%RmixH}&g>$5?XTMuebQxL?X(3~ST*!&r>;>FCH{vlslZl3%>Y_n9Uq`9yF4
zo`HUgX8VX44hYALYg;@RHwo((z?2A0`r<Fw8pot5f`plLtCJ;MBnA6KTAUlLEFZXj
z>gBtqEBFW|WX&kdAcCm;7C62{+2Lc9sQ^o?s=OV8=Y&(G4HvgrHn`(FR3SPyh_CvT
z5jaS=tKIl+=v;XalfJu~P0xC7^5x`Xh*qh)QLndxcjX%6saycNz_l+Y?jo-*2@2DF
zvhQ4)@0|F;V))_5L^uRmA*61Ti`zYZG@Xa7gH_{}COymaX@8XATrB>#+ii<J^QhZz
z<O{`AU!sN(zJM0EMK8mCvR4;2-tB+**Abt<7_5hLxXC*mKoE-&%@UX)3YQ<zr7zr&
zZc=_^&qceA6p_JF6g~?2`banTh$FeUxyjKO^^DAKo75Si=1PlQ2s7vv{jLHf;64Cj
zN`@GZOj+|+A>zh%qJo&@&O7fsJCD+(c`Uqg(_PmdmF($~3DZ4DEc~L|e(*IbI_54)
zhSP0O$s8JmO4i}|%I02#lD)Bk?r>B>9eX37FE`StPnXLxbYpp&UVRbBi8X$OkaCog
zm5h0f<oWenr8(rZy+B2>eGW&KVi4&j>tVq#3fX4Va)SvSlB~i7iVHM|&QhVEN(pau
zpCoU@6|yhCU;N-at*fY>x;+U_>GlArU#7)`>*{srwo#Br$ufCF+LzIznZ{WB^HU>E
z+@OMnA@+^d{{C%47hbf3Q?d$CNQtkj(s;!b`x@>8wHcz5YhRu0I5KoCOEJov7waxz
z@jeRIY(i4CFTq&oBqOY3>MN3wyO%4JRN{vIR1buFLnG0OyNPRv%f+F?(`9_JvtC=_
zJzU#M(32TPPDQc^wi?~|aAfB@x;pkfdn0D`E*C8bd7CIq^z`TU)~ZXn!af*ye0(8?
zn@|1UQ%i_Q2wCu6#1JNK41*x%#U^&;>|Xp#y~2S4ADpn|X&98!u3URYH-pW_KNb!m
zn$j$9ie7tW_adoVdlq1fGMFl|u=Z^K0>x$<<uW!L5KhUJ4gwF&W?-9;c$20S59g8D
zBj8Mc3-B@5t0swQQUf!wI<Ss(WYmtp4r9P6)`YSe`{9uE1I(@EyKT$f6w$xOOrA7f
z-1Xx3v-~i0n`ghhL8xp|gWT-j?&GmuCNW8=!GnCndlyBt4=pI^9jxJIiLh+JU~pw&
z0R6s$#k8<v!TBvYgf<toE{&UyY6YECX>Kb|EgsNIX&upt&daad(m2EH35{7)_F1?;
znX)Re(&|&d{<i7<acnA4ju?W$rYfKOUzy095pC5M@r9SHktMU?JicU%;gX>muw;BC
zyM3L_DVHBf|4EvS`%iFTY<7kWgANxB5kp<b-`~@$>*6$fQ6Gu_k{aR3;eB9wxVfi1
zO_SNcx1b|usIb~?mn7E|@HSS@w<eJ*F;l>*`kalrZJI>vZZ=kf1+*ff|10{UGSH3*
z?Mlaq1bw*z=|RO>{Cv}V>F^l7Mcu(H;>(KsPMq6!oY;oDyvIckvD9#{dbiBiXDYKx
z1vq?g##&VvpJo|e4;J&Z=8zzF`+-4OftTeraJgU0%fSnI-p^T;xm;SqelEX~jc;o|
zr+2=0CbzS{(hp?7=94-MYiw)N=~b4f_$;x}>5NHttB$-;${->zM9{>Wa6F=xbK-n8
zT(@UZWYhy&6pa#=3=z5*zswK0aA{zgs#~6Mq7=f0wH`y~SkJ*oq1lW^7%udU^lrgX
z6?~D&t-44z<JfH4$mrdIBLfWtLn&-fGJ^LqC8L=DKN7Ms6wNVjXr+eO<WNASjA5y$
z@fyN9<VaRC*(1}|?WVSF6~7WSZqAj?_<v}mk{r3NCDfO=m0fmZ0>_!>ocsR}Q*))}
zoMS5QXzMhIq}j{7I?hIaWVPLsFrkCl8PPNmWq3GMT0z*0=|=Og_OL+NNl1g($*Wh(
zxFsJBzx(g7ZSAy+ZCRbYi!D}iAgD!G#D%&Brm)^u5^@^b%w33MrTPrKXCwy0YR=-5
zAlN^pMHT^c$@Lj~=tbvCM4In$kG7SY(~XL(qY8QXbUhXZ&TevPC1rio&noT223i4o
zOhbFeyEe1<7HqHjv-voHi3}B~xf$O#!&_X5I^NaOC6jUnz2(C=kpu%SGW2qQ^`T5R
z<P!49b%T7o@Vq(0x0j|#oQ^nU7O_O4=!i>=&bg03FdQt|H`oF+IyX<FyYYOBuZ4I_
z3A?uZG7{pIOW4Vj4@e*w0AySiSL=u^MI|a&$YE6^yS`xcN(6p@0ayK3D^rb91F!qK
z#;{_VYqb`dxO5T56m3dOwXfbdP!x*hk)j!P1HO)8ggH?jul^ca6WdD?kUEyEIcP1k
zDp6y!4ALu%ZR!0QFxt_x_JyM|*_4clQcgH9gfpV8W&+trb@4!@X5&+eWS4LSGjU|b
zsnfA&8?V6_x36KiCRbF&EA}ze*((Ecxe9K8WZMyMNgC60I2-~6pgFt$Y>7i_h9PE6
z1`+&gAu0eX;!1H~&>DM7VITU4s>92Z><GP1uZvM;5xWVzfG@#HE;VM5*o*%Dy(8nu
zAsg6*{Sj{eYdHpdtbLx#LB!9|QO>DPgK(sbQ|#AdwWm2L9orkoX(_OVHjwNB+k%F{
z4kAvO#Kc;2{ftw>;;^Er{H(R-UiJ51Jh1Dcp~=yqXJ0ftI<yTXwo6&#1^Ld4Rf*Fn
zv;F+$3S-OHR2DMZLIeSAFbuDfY}*po<?b^rCP=U*9qHrNnJv(nJMJd@r(+I4ZILEP
zUzRRsU%m@VgFQG`2ojKbL+LuP`437lM-t|G2H6z&bc0TG;haVFaPyWjr1qbRDZ|5s
z?`gQb?uhO((p8@34cYkikH}y0?a<wv3WOBU=?uQfMrQ=RH1lg3$L`^CkF?8_79YWh
zz3}?Q!7H;I0(hcg@nY8KQeRj5nOR5UzTRHP!}(B={W<(Z?emMX^*VkSA^TLZ_wTnk
z^0zmcJHG|%4xJOslXEkV9K}Hg$y^y*7nxXHN(9ykvbM6K67a$gcC5?-s7tegTK-WX
zoFw}D2lwvYJ+Oy^O!Wg=Y~F}@A6WQU3*vwX>TMOoFL}7JA*)jTg*#A!wxqOpV%#4N
z-*9QPQk-sPEQAwefB(qn(B#1A;7;89HaWO^o2-~b7#!uY!R)vY5)wAYE&jyi5??X#
zT1ce>Aa&tHIwevKl2x_aOYhVa^h?uD((UkFyGC3+aUYR|gU{^M^p+HL?v<E=#h>6*
z4Z|NqeA%fK{?G>TCOwm|VAgxH{PFh*!DPP_ieqSfH5x@1!={?LJ3vCNxH2~_Fx?fM
zUrx=kRYE>ynbZkgVg=2%^UY1PsXc5rC}m>`T7jNZi$Tri2$OjQw|huJXY@c#HTJqO
zJ+Z@Qa3Tq>aS#<hXypS0;^NObS5-%?cNe>7>W6B)Y-P84<lLclMG>tK*JE3}1d?Iu
zUc)%nl(+T>2Aq;>0S;r101_R?3}nvc1Tv7WGb`%jA_!h!MKI!A7$qpE1k@@qL%G`I
zN_^UFXNddn^UhO^K*<NyPWpI__)IQ_0%eb<E$7`_DE^wwV=5cBOT+^S>?dRzV<{p_
zItJy8G)9BaJqdZ@!$VGd{(L4oSSC$M5a{?@JDk{Nne!G;E5K8HM3LVMb1lp^(-%J#
zxP&@{=oQeZC|F##^cgx#vg;hv8EpG)(C-1D9bbso6(BbSTL{*24|1dIVbFqng~b<g
z$EZcjY@Yq6-Q~_mW~_VT%8(ji#tjDqc#97u85*8ujkCLc9K!(-IQwu&Ank-3q2!0X
z7@sZKk1k>oYLV=cxEVs$yIj7pq}eGkP0^ap28;_9?&1nLdMWdG!y(XlSk{oiy3u+6
zdJ@W|xZz$U#iFxjWD+CP@pk=lcwmTaqDP`UX5?X);4e!NFyf0YdDkN>VL>E_g!2lO
zH7WW%*aNq9{FO~aPW)F^>7dz41Ji=b1i?8D-%Hvj_%JK84VWUrbTwXQ2K~hUL;#cS
z<Y7payA-3*S&|rN5vLRZ#)EGVJ;>^AOpQ3BB;V!`HIC~Y6`B*4rVtvAp4=8Bw=ra`
zPasXKMR?%W7s(Kp)70(qCE9r|T#%oK*oxc>LD^^y7wnE=5e*x`h-xj3xszh}ibANy
zsU8{NcaLjwB?9)=%ZRAt`?3U!wEW<%qHf;IWGCjj7c!dNl*2tmMpm<jWQZ|LYJ`s^
zja18s*vX;UQcglU>hDh{R*MT2oEJc}sWLkYa<PI`-h=QDTMG79y@>Bq2Vj+huNF@;
zO`rJ{#~}1*K&7~M`@zv+H(3goj3w6;ZB*16^-cOT+DbMP8tu}W=c1g;Oqf1YS_3MV
zlVB5}1&W}>v1CMhpNJhf?<W0FV!gO~tXe@X3Ie2w5*@Q5@p2EmqT#*E#dC)5$5?Xb
zs5Q8s$Fe984r9+<UAQuk0Zycv@vK-FarTiTt%?9M{`qXlX|;h6bNIMYcMte1kti8X
zGq6T-lf0#2Vw@o}gc&7neI^b9M99gHwkTR_xGBq87kh*#fb~*rQVV*IQ<1C{!?JRD
zi+-9oLP3<ays&17HmE7K0%DPwso;4@F%cia#xfh|blN05IO^P6m<nr6gRj^hHR?Ao
z1noXFTd6S$je*vcuh$S)j)HLPdw#Q*868j*|2^6Ot=ruQWae9lPMX@7jCX>}H5nbw
z3&>4K_DAGN1X2BIEk!Zg-|djflw;VUZ4ayjA|%unOhw+i37DdroIqhMi5@3(w$dDF
z)DJJAJxVZ49Ec262IuNNv{Bwl{kN>PE>$?Guykm^t6xnx@km*EXU<|tkuZJSN|}_B
zqZ*wI#0i%+=-Vi584d|WmEPiuE=N3$jUkR|ZUM)F9@vE(DK#)Pz%kI}xN|P^y}h-1
zFK)@GFgzh!<21p&TTsK+?I-{u4?s(pL%{hq9m5FA`1J=E#DO?W=l%TwORhF-gKlP-
zAVnHk;(+qN{eff9D9;SL`=uOd%|dc1F7t=qZipgRSCz{P7f(0s7959e%wG?|L^$ap
zWHX|j_zMgi_ZT5AxpBm$&2=>qr{&jCb!{v*5HWhAv{&pAPLGMmrs)UeU2<8z-!S@y
z+@Adu*6vr~Mkpx@^WTa}YTFs*cwZCBsk7UUgCU9r(iaOK#doYj9gV$8ja*@9T?$zz
zN(miOrqJC<jM=)w8odXnp*W#(1+7E{oN6U{pgmn7^}lV(k}A-iDkm*bX`LH%5f}1B
zCtB2my-E{=|2y^_d9(Qw+x#BwXc=A_R?8D3^*RiG%-tJti<}8w?a=0x7V+aJQ7l?2
zP7Ltgndo=>f{w@)ZjFxM8aIYXR)RPx<Ktp+c8`CPwEdpuc8TG@hU$)^|2@`Psv^5K
zPJs&iViyRUPzXDWQ$((8`<A$8gJ#C*0WMcIaGEW`dkj?)z1^Oqpc!W(mTP1so3qU5
zi1b(83h!yh^F0uQgaO7kL@wGrv}YV2&>DQ!z>XpKNPc5fZqp0P4DQ`CHa<Eqyl0%n
z1cB^O;+&ns=5|i8xj1+cK0<Y|VDq=cCQFTv4oLyr3slCtV6MU(P^~wxQQu}pM`H+|
z#`02Rt6>|i{$6g|*JXVr*-q^N2m7Nsj=2qsNeJ6^$a%@y;7wqTgq1=bY%sgNN8*P?
z<pQZ9XtOTXl+Gotp`@%)nsIQ^D_79=diU(zGlaLJc3m%0@8NyWv^8^!3*?KTB{{L8
ztKnr^W#0j1t7C&Byy?2>kjt(z_A~K6gl4gOj_qvrJ0z*8ug+l!)xe)KnyC*RS7r}p
zhDr^BWo!1uR<1LnWdN51-jBB_Kmv=?!A=`Wt#qR=!p10s#FE8h2pu%P2oWde{@L#%
zQ@9~r4E|oQ+`#=mT;;~w8kB*fv3z(}S^@3^4&u3~6%(k$@?0`b#CdjTwIGQOiTO)&
zF5*q=C}R9nw$t3|RwNKjKD*r(j1riN_qtJ+p4YTN3EU@TpbCcL=XLB{N<VTsmp<k*
z>@LjAL=<qjEvzCF38H{ld1MnOgK>1u<?bVU4q{|j1KQ1R&Wqgex$v1n7gCU2ytN+#
zBXDbMA}J(ngc@Cn<jmM6U%LvCQE@we>~#&dva=mShmg*Ib>%^-fKd)!=)USPY^x(q
zVh}7Cu1Nx~weC+w73>U9aOsr9;S`a(w*<KcVY(Vc;KXJl947%t#OgrCv5>J5i;Atd
zm}Q+5MxZ+=X4IK0iovU7+)-)@F>K~o$cq#wsZ%r_*<%D(-64|V+IBk&TpaVrxab>4
zup5>pNsTj^S)*Lw<CSrc6H6!SR=ILXe~9D~*z9UJeQY}e@c=eGTX&5qQCMgu<66@o
zifzNWic?&+RpJ<8zhgFY*Gwcjmuq01{+v6foqMr!_1rTx$~N%WQL#F`Td+80KC6|0
z>L=R4(!fV2qG^_mNkGR59t&(74;bfRG&CW$Te!-adpQd_d#M-Uvr2$~cL#LF;5{X0
zYfzclS7NwBb;1hI+`5(@(!3=wVngf=B!amPe|o1YKJFF0TN`48HA<j3xIsTc$300u
z=;8O6f6~PFDI~hjko#I*>Jm+{xTJ{zK)I-iFB0O%#-@DG#m~2l7{#u=@(sFOopGeU
zAk4_Hq6-${ME;Bq{JUIQl7US!-kY?xX<N!LrB1a$gm4+{*~uC{>LM8(8iRu8C<_k=
z*vC@jbU%C5h88Y&zyz7aLS!Qy^`0DvNhL)WOO$B@!>P68J%q$_)q8Os{Htw2ezn`M
zb3`CwUTaZZIdh5fpvxmY4-2`2=#Dqh5nnv9#q3<Mj8K|uOz^nnh+pnl{3*Z-PYWmA
zu;POJoDt(#nW-ZNDz1rtHdChsxO%b7Fk4WxQ9l$JK-7&4oedKnw{d^7Bl*wG*4kSj
zhV}{o+t}3^bUWv3tpQpVYZbJLN8GB1Ch@B~vYw}w=w(q#?>;T;E-b+~@%MfhN4^B`
zxxkVcC~W|4g*VZG76iSW$L6mZHEdmCnITa}+zy11)vNcb%Z_SRb?`%hW!w}Sl=dp8
z&8UQ=L`=hiyFa16g|*Edj_8#251%=u9`pv-ZI#P_i^QMoXq$?p_rs2IaU9cWcAFFd
zmoc5P`eYr|<(&)wJ+!o1Jsukyc*p70SpYc(#>rU%1F>N*vp~5a%Sy2d1*~`Hp;zb?
zO4`4NX?E{n%Ft)%P296us$N)cU>X^mD>VxF&NT+h)#_;d5ah-?=3ES<kcZH7iF&|}
za-7yUsAQvoQ<V#;i}D9cXuo#if>_O3n;{!S)7`oo!ggb`D~~XBxrUD}<d;BY{Sg0}
zUMk6xd0bsrYnAwD3SXrHq->)x>MW09<Uo}_#o#s>-$5%By(7x8jZl!2rIv-DowXAE
z^kq0}Kp0_BbiE*D*Nel^Rvu}4iMos*Rt|$Axl*WFK3L{j;S7<C<SF2n<|~*YfwsyF
zs~{^UV!FLyIws4+d_zP)fL*5=C&qoaeTk&4HDW>;_i$nYBL@dFs}{6wFS9J#Bd9cY
z;leyLT;K)VWDqn6>r{KS8snH$Sp-(|z8SR@*fcfe+N8UUCh`;3Hn(t|<vj#jbdnWO
zu}|H)sDtq%PRYTcS<?nQY7D^5hW*S4ySIP|YZ*x%P5VOQC446dCl14B8D<iaovt*d
z7ptYlC8Y-M*YXpV%mA_($Vs4J#iDaUOFSSY5e|AwJ<j<8LZ=#%;Z5&gSEYup8p@rG
z#v&Xh{+}3k$j*ImjN<^*sv4llZ@v1}-+J}$;eUepTd%%Re?=0e<di}E)2Jjf;;W=5
zfzU4z=f$Q7;&r334=b984qU^g6g-!fkqydNSnsaHa>E{j`yh{sc<UGEpB>|Nrrfrn
zBLY3-_AW`~YDxd$B?Guw5~%DjD{G=(pz+De$siz1jgZrg4v4Je!E;Hbv!wYe<Z`bg
z@?ZC-<ADUxW(A307^(L3K5b+S@^&8=Rkp}h{J7SNg45o^yD#x+y>AQoOAVn4^hnA8
zGzSxDEH4CJV}|1ihbtNJ4NoFY6LYUzXJki!2C5YV!<!l!%1a3?kac2lkQkx^MH6|s
z*MxarRlnkp>d4V@EoYwt2`I6%ATvlp<FzcfIm%sc<JT~nsYl0xccJB**sP4*^VIi$
z^^8R3Ac3hOjZ96AoO5^ue8_Vj@-0xz7(t4E^bT<@Ey;Y{X2<eTiKrIW#2qX70#O}o
zmv*dT?@=S}zyy5KeL>!&e1gM27FWjqf^k3?M!(zSp8}(^D$x_JbrMN#Yut))Fy8rz
zJ5yi{K?1J-i8??vYz^s##G@sNv<8YKYE7(`vTj`$r;8V!`<uLktxg1Lo{wD-wsNxC
z^3Acwg}fuXBSObKCiQ9Fo3Rmm#BApfFLWws9{PX{IE-sHfrDHYA&`3a$-2ssPYXpS
ztn7*|CWOO(+y}m8lgAwrw^;3R<~gY1LRrgIaE^KUFlcO77EHh{C)jrnC!UGyW%s8`
z%t(~V-e_heskt2nB2i<kJP!vnYv<R`ZzH-(%&kuSn+rbt(<w;k?L7UUa5Mv{C38qp
zj~$UjpL%6HCtf1I1Yc&R&@$^rue0lW$7Wh^#I`_V6QaDfsVekU^J8UXW6JAXT}}lf
zo5YUrx7|L~=x6>wwLVqi70&qs_>bt6dzIL0WJ=r;nsL*-?l8K_);R4uxBqCP0z`#8
z?kL1LaJIoipTVm{8+&m$GpB{K@KBb-f+cDPECa}qMp7dShm_?Cazz=GT@GmGdV5_<
zJ((oSm!Z%M1Ie8lyNF+qk_}OJ7!|*G)V3FfJrggZS!96{Z!d^lD5bP7!e`}m3yK(B
zn;~G4Jg_c*G<aApAjlmmxl_o~b9MMBvCMMwUHas#0+y^!wd7}J<Mt8p6-aATPUyjL
z((jJL5u?*_XpYBtaaKo;rmW-SZapVoN|w)`*zs)pI$(?zvzdi_*a6mk93gT5I@v4f
zYSNzdjlJ&gA0ET$$<cx5K*om4m86f6={~x%?BWx{O=R4YS+^KB)H-mdhgEglTJOqz
z5PqKkr&GERs4wGkPXxBOr7uTad~Mf7?0h>e92=0)YaIC<m_fTxBoMR2^TF_79Z5pZ
zd>sl0i-(+=7EdMXA>)kZfNH!>ZG_UqiDzmZ{`tcVYt>lv;X?iDB=hdq@|u6?UBF^+
za|Kbt*TRPZjvr@$j%2vWuGz1rxzV6{3PQgyO@X!%*h+Zm?RAxdhRFvupTUOj!=;s=
z@}(jBKAf`7!G-nrfJ~JdeGp}$8F!CZtfgFmc(}b%%E{p~EKsmylCxf63q30#skc{5
zpfEy`63ruzPc{NvOa|s^B{)e$pReH#*oAnXgAu|HAC-7MClX<xkZrmPmr=r$;L8?p
zIi2e25%;K%<Y=PKk4}fZ4#)sPH972pmbpUY%X(uaOxZjVG6FSF%PQOF4w)DLvd6(g
z-CHxXZFqd|=$c%)c90+6s!*biHbgMciJUBWi-Q|yHWPT0C`d^B%E%<PkjO~dhe}KM
zI0%n($QdMDo>0RU9E^&16MeY>Xp|fT;$7+ZYDatyF*$a%=^SvhtX(i6VpdKKs^Es8
zm|NyOryzG)rc@u`(7oD@_XJ^&dHDqFc8S7{R+qa=Pr+@5S6qfT`p}##nV*xX8L1C|
zA=g|~kw-5~4l*9N2a9i{zzD#M;ln6?jl{TOBIM=6%k)U`-e*j>lOcoRCvx!!B<L9I
zoH*^;&RGQL(!?8+UHC?D47&tk3;DS^Zryj+hdWg2Ds=TT+}B1HOiz;7C8PxPFXVBt
z!pPpSp;4xT<DKlyFodGWgjg=TW90Hqew#{NL1s~}rWF0P&k=i)V(pjoXjsk$yDd5Y
zxIKr`e}WnZ0D{lsL=E6;2G>s1j!xiP25uGW`X(x%3wSPxDDRLN^Hp5W#Her!hoJ7D
zbfSA4?-F8V$MWmO!K&mwcP!CG-3tGV(>bOaA082kq;;nWQ3X=t<dC9?3x!A)@z<I1
zh>EO8CQ)%x=8s)K?M0K)x@oE|S`_O8BEW^ahhamR27{1K?4s4Ei{0)7+*)@SIH>`4
z5t#;&T|J4gQ5Y_7eD4%bWpN)PdCThb9InEIcbY*O)R_@H?o4bvccTlV90QkeQFY)n
zS<tr&ax?om!SPh3HMKZ>Rk_t$C#p==;bk9y&$~FaY#H1sy`nO$`c>HVWVA0SX0*i%
z*LtT)&HuyRyRNl$WNV}UxeA3vmgJ5FKBRM?ZD(P?HVHNauyX)65lDbyArVQK;B=nn
zyN!K&`%3ovjzd+=lMvX>>b1X>?gW^#W*tV28a2+Ndsy0HijAlQO0pTIHY_4U2B}V9
z_SQKTt_QdBXlXI2D+X@D?uN-R<wTi?qT)MaHr1%*2-oS&bM951D`kKzILf%8hc)=W
zv`{1N1cvCt(>9Ye$W5SJ5gNOuRX^aFBSO1|4lz@sACN?g26Q=)bD?88Eg?~<R1kCj
zuwc=(lV?-H_P_0(0zUY`|Mpw`0e>RkT7L4j)q01xmAgzP>+>L(8g)1&Xk0Y5!rqIz
zHOvL-O?&Z=nGkqHi#D~KN1!U@v{}vu-9#$RR5*4_E>Xm(@cJJ!nk>BG&_P`Z{LfM|
z<7YWMX_^${9Afx?fn(DzvK{U-sA(&>88EIDE_f%R0*7>JNaWjr^|W|nF~Sngw%qZ|
zTKt0|v8;fKAEj!z=|BLQ_GuoP)i;9ZYZSe7DSV*!ukqvA<-k!C6|iW`%&vUcx{#;p
z{eXpLVL4|CfbVFMNSQEKL1Yyh;b_Fq2yJue)a}v3Xv@qqrp0beoly~mrs^=g_a>02
zy2})#O2)2I7&Y~aj3Kp3z~5Sz!YF0u%2G=SUQL{983HDgh2=TuG~5RUXB-SG&bU-q
z!q0Ye#N%V~vuALH;qT@i!Cy}yW}Kd%Hd!yc+v|=m`H+Tf_!MakaXx^C=>m>IQX3C3
z<{hX9`50!=VN1P#S{KN1O)t$1hA4lb*ojjBLQ`mBsHId910U~;+a{_H|5K#<Z%}xQ
zLMz1OTl^h0iRVANb?ci2Od!E8{9)NNWIN>)hk~n;WdrpWH0IIT;cszTfZG#IUAl`8
zp=ZFpHe{;1^;Exr`ine(TtogvJb2i5C(Y>;%0All31cGatv8K>J2O27XH%q-h>Jmz
zz%boQt-Be8Dqmgc^fNOd=IR=_0Qa1UClcEZ)3SgJOb<<~A(%iiTE&)joafjq)EYCC
zA;cXUKtxsm)C6?lmBP)~N%#$+0+=iqqgvp3xBz7sxzjM7Ob<wlcDA4RL$(h;nZIP7
z&3Vb-2M{yYDOaB20S%f~0nS^R=cKPeCMwNf3!?H|je`PAt4*PUaAwIq7ucZ$rLm!A
z!i1+kC1gpv!V5ODn+~`9q#5@HVVT9&;-1(lLI+uLj{6l^pp>kA*SW-*D&B*x<i@G_
zrgjG))N5mLacgybX@`mUSmt^gn;)?#RK?DA9;uiKF(lIlcSDLvb1!S+qrL}}O%I{q
zv})c8am9Lv9SsTlNRiCEMfRTC?RPz3nfQ&z7TN1eKNLUu2<pnK;U6`0e}G`1POJZl
z8SE`%eqeHC7@!y^0h0zoYnUt``1lZaNxtO>@t3>&&mEW^#32~f5&y)QuyI~6__c?0
z{`B&z9Ej_X+t@ngPBz#BT$S-kdIURS&dG{0Siw1z@_vEM;MzguuhFQ<iXGr6L~Jo8
z2?ftk>bNt6FoUVWlN_DWR8oiG3nBN1z4LabNwZ7xk*@&bD5QYTg+;Lsz3yOk55fBs
zsdVI-$|aY>796f{*dxg@z{x_V%s|Wl+S}D#8mM4=!K_BO^g2y)%&6{o<M|Z!8bCfh
zyJ|url5juRRBlO5jkL=U%ijsc+q0{1VCGL<7`Q0Zy=V^lCE)fAx^mBuBDXyl28Wr5
zCh8#%z*laZL76Z%!%YI;Ixy{8U0BY)_nnD%gZm6u%==7;SJC^<MjTu|OeYiuqhr$z
z_etkN_U6M(%6+Ngxw`hg>y%KpCAU}QWJ+5rEN=1*N~N%>NHSJ670EKuEQ`$&Z1bqW
z#^1kLZ0x={J!X00#qKu;x4yac4Ia-eP*0$T{1IHmk@Bs206Rc>nt6n%tj-{9^|zC$
zCT*&e$)_J4)b+46Y&Q>K-8t^O52b{(GZH_H0?901k0roSlBu1PQcQhy6N1S*g^6D&
zlARyH@i{?*>G4*D9olp1;M~Nn1O69e9nhBh2Hx4YLM3h|h2zvDnRc~PASv%Gn{P7F
zsk;#eD1>;*)d#t6EEx@N(o^f~@cca|gR{ic>hIZ$nU$O4-i-i@vQTiKYEq$l6EGaM
zNOzJit4}8z1{0Iq_`;FHmAuua8CUtstB>a-TvM_E9ch+ZtDck>yZxy1G0i+=S(xXV
zCvY4QeM-C_ro~?fK68@bvx3RBfxugh2M7U45h4eI5&12OAuRW+7blM^GH@n*m^`Nl
z4%-UEeIYMVCT4Syi3N7vTd7chd68R{GA;aN1Y!t1Vlou^=)xu3Cl294*zUy|KE17e
z9f-!mXPg8CA!7?wQZC#lvbuzGL>3j}4-G=?!D$P5Osetr#_)NMiP19iZ#OZZoZv}`
zv~Eg65ywoj)?m^UJF`MIE6H2a0v=~urr42D`>7d|`tQ>LOR!K6v8`%_RKGwSflPeL
znI*_*!{O3Lwu@~z$n-j|C{Cu`gP*x#t;+F02nD{w!6`FQRRbhSm8*aePfBKT&zbDE
zY;KeLyLxI-oZf^sIF+Mwlxunx!8ifYAV<n~?e9~~^Lum?G}iCYhikguql==3Ja(3&
z*g(*rc@Z^o;@-Wuh{#XXu|t);`zp;DvAEcGEz7p+;q|J0eFZN(WjEMZ(J82=!xj5+
z9oN3x@+b!dsTv4kpE=n4<<#2k6#q6enn0%AxLBmq0_ZH+%@gQ~CZluaG|m!4gra<*
zo<VB6F-?A=kgm+EBPFJwH;?LJ=%>%r4h0W-c+#289u3^lR+nXte&T2&WsT-*FX5g2
zi@P{wy;XmqH!*XzqEnUWJn>iPO`VK`3zgm>*SY*EAeP>GB~7Nh?)WAAjaia8Xwu7X
zy0S4UZjr05xMEki3&dwteytThFIlW!wv+F(Vn-9_VHWDF#zL%E8MB|-{I0h6RpUUB
zbvR7C&74!&f`Uo&i8a((140d}p)l3t1n!AkV*&PHhrv0Nz$x2vE^C+AkEB~y$Iqz6
zmSl?TJ^F;PjGFDCs+RmA%g3vai||Z$E9Rojp;gD>rx=bx0A_q(zgP<wLY`c{FA#of
zE6ZX_4e1u!bwPKYwslTEH7Ge7H~|FGy#P4aO~V4*hi-9EGA8lQ>`Gk1DCIpT45}mO
zs529-abo#DJ!cdF{fOz_Oy)@98Co^$(0pz6_ZIz!7i-0epo$35Dm%dq-i%J_F_1kD
zVln-zPJok%0`a;E)@2`tqpyeuk+5uip4{)RT4ILFzxFySzn*Tf4cMDgN5OxYuq$4J
z{3E#j`5)rr_pj*8ck9Ez$<7STOSU92SUDxkEi+OrHJ<z5_efAmk6)?i=bznuUDAzV
zSRl3u1y*%Z4s_zSp;Y-3xYiV@a>CUV%<T15)0ou87c7@~4C5kmfTG$<e?&tgL)~tZ
zbJ4$`f&+zv74Mjd!w|TF((6u)lqi27C{*rCUOr84s5~_NYM~22SdA!B{4&7}qls~4
zt-88Q16+F~kmAfjAV<A@+{7?Uip10%rFfkDt?sIB3gwl|p6`m?I0sR@o49>Dh$cc5
z8j_0m*4++DFa&vkd1R!A4M>#w_u8Jg4H+O8P(B{}kAv;7o{P`>ZMokl!TpazabH3P
zdj-9DWphcHH-^`)ofK-5U|O~KWBMzW?}>apM4<KnJ`5Cn#7mPAi{`X5IzKzpgf|G|
zB3lG%B4NgM*Pj(DK$NJngbb;Y7e2SW-T@3WP*>R#<ri`Fpv1$;pv^o`^yC!E6nq|n
zZeb`{<w7sY3f}1WxkIF!G+~W+++;S>a70*o>gZ9se+SMdf$$4+N|Me~!c{O(IR1sa
z3CZyx6-oB5K<M4Qe5Nk0<%}YvTa5Iq6Ryt(aB(2G2qOdNd@?qKWX)CtD_I=bd2wTW
z3Dte#;y(QjJhO0-8;*O0-(+*=;{Ie7Ch(|4l2i=)=r}U<j)>SW-FAG6aNPIYjG_nD
z2-$`ekv_(@Sl7&yz<)~irsziPM_04v#ghx?HSk!>q=#o3)tdzPsXsL_1c*;ltz>wE
zG4%L5o&;2KKtL6LQv@?#voygQZhq-ptUo}IC|>%#%f4~Q#1PEzvUtNsd}1;|aU#T}
zE@HWP;+kcMgLM1n$bUh<H!SHDyugCXWwOLw^`NBxY~d+f2ssC{VPs5Q=Lk!nBN+`~
z=2y|c<m{LL*<h*=M@wZ8Gc)9evNn-j->4}IYo31ytrfhC(U{qP8}&c_scYPu|C785
z{^ev5LqI>*z}*&iR!IR`>$>qv+%<m3t(%iA!_qR45`HQ*>0zDR73~q93DUq1I#Z_)
zB#Vs<u`_U@D`x&M+@g0YzQNvLYWM|%U?O3R`fk2Dpf#GWzg9D%O_O4Ogm^Mb#P-5g
z%sb9ryN?^O!9G@3Dx`cB+|X(;`a8HjpB9%M68SXhyCK$IB08{TVMk)1?`w;ImnPOD
zgFqVcK$gobyp<->anJ1c8!Yslc8w+TfA3MwWmyhdE|W}o?is6w<mE(`E$agE?7+f+
z+`}jYGaNA3E<DI}8=m_oa>rUy`^vW4=g07{ftA@3)}XGEFgRA#Bu-;vFR6o->(0L2
zODrR~7M?)*5S3=C>)MpGLY@`PD&ngO4{{^;9L#v)iF&rB2<ITb9IjH8f_w4T@{Nh_
ztv=mHsDR^p*M36Bx=ZXwb>SQ>kiD1IOwWPhrWv5zMOnN1unGFs7<;8jqcqa|szLPn
zEYAQ{oZ1sa*Wrlhj~qCtW30;RqzCk>Y*7NxxHY-{D=<$p$AdciKb8avZa3qOon;)a
z38@G(?8w}x`lOa|Eqeb|=V+2?uoxU^N=(8uIvjQ-l>)-trHMp%S~IQf8_S=<WqSZe
z5xDK5+^f|D#{GI_LMKNw6ai7!G~CaEApT-;F{x5x@U77TPO}lx3v`VrBH}SgG2~x2
z;GS)>nb`4)w_;Ww#NSX8OlPSPQ<9wb+LFjXfKWA#TcRgT-w4ue1d}1_g4K86I#4VG
zlsB^iPr~3vWeMD3fR;c<-+Cgdq)5>7e7qOaJ;F%Pt~Jj4SPxk8klGvkxfSm1v{Cgp
zg`bb^24(klC&yJf1*`4zWH5x}=bNIUom7Fg-L0wLyd_Tp;T?71c825};%mf2D6f{c
z@t3+uLtdLWEpff~koyw2Ogky|Q$rS47P%f`Uy#>C&G)17;fLYMz<yMMB-xeCd?Do6
zL#=^aH?TBNvp^n?I!U<G+!C}H4BmvA=LW&6X4IrzX}VdhtZt(2!_v;`%A#0oo6Tfe
zJXi1vEFTUanly&NT~j&cumOo-tU_#)fVNdZGwXRAp-r)qaa%Qgz`+sZC;`Od<qAYa
zrON3hwW2A2rkhpy>M_<KY(|4q4{!joOaYEOxLI6N%Gv2$DuIo5N{Xi$<cAYzikNn5
zKrsj?O~n=bOGU{7*n=8)>|o7$#c>mQmloctTsX+TmJFSMVz&J&1JRqQ)C4`+MW0U2
z^eK=YDu+njVW$|yQaX`hCtunzvSdIta8cl8k*o-sW6j8K_XbH0Kg}#&PD5*~RfI6_
z+ORu%L$3yH4Z3;MYawS7iRl5BgaSvc^5uj&m{%{PWalJfFO^nItaUW#fv+ha1}I|d
znF0@!fktn-XIWwm6d^dATx)R?M8VOvS$acskD=c}Y?jr#0txwT8a!Q-&uade(*(29
z(AN5&K-_WDXEM@rU!JwDW5S{)3vy1z<TqE#*7Pm+jM+4De2lOzsB0_k@c;?VkG2c^
z!+dS+$j%aOQ&fosV?etx>hDs|1uQ@=uE_k%umv7KDNL7R%?Kns1@qY1ZmG#R&_Lck
zgg)HCncm6gh*VUrkc9>kEr59FR@g-{L8#^eIlOh2u4}~m1d43b)x#VZP&i|E8lpM+
z<KaMR*w{4ep4Pjs)1!Y2zL$Ji!(<Yg9k|2}o?#$ycQSNq!2vvOG50*<dca+xnpthv
zCh|I@w)54&n;A*^;OdO{F#&|N)ZqIRFNFa-m-G0LD@A4mFf#~%hlDs70$?|q>!OdB
zpL3Y$PN)D>gX^snzz7%$3zeokEL*`PlEg86CGNB#OOtbXGtC&5|B#nqLzz$hN{1_O
zx0287sJ69y8-I;kFbwq0PFm~(jc#eyHQ!}C(Y>p|D)L71-1&Mqx-qhP3Iw{oLkdTN
zSGhCMW*LSs{FQm?;IV~mb!w^5Mdq(8f9NNU)KTYV2nHPVqb1xR(v@JiBv~G}8!P4<
zoGs?kF>j<I@8e_h&IkiXPUWGvv@b3!oL<I<^Mm1WJvW^sBO6ve7Ki#6MR@c4F_xnS
zdcE?jWmFcHYID65hh$?=XBw#T6r^BNq^Mui0B=gOVzF!+9$$7TSO+DYsOU;Orhq0E
z-yd{xo~`-<brP_}k9Z!DU=Dcd8;*Mb$Cq^o{6?KcKeO|-m6V6$H@4$w7#TV60T)Pl
z?cJ*_h#JI_>6J29sy=<c`IdEz1-!&roW(n1#b#vz<i)b{^_k+Zxoi|a!lg?S3N|{1
zMcBp}I5t(dZV(fs>?C=7ue9htB?(uQT<yg0*7earNaOoJQ7c>6@z1^Y`A5l;BPiM#
z7W$vW!rL=t*-Ef@L8Q#wyytihnMDA4YQaodUa^S?EZH@BqXfd04V*PDZxHye^4A=|
z8QRJgwuKc?EaiXDR|U`#78TkN#h5cxBI_PTQk27)MA-^R4sKgU2G5OU;&@_RGCJUu
zXI?;A0_qUZ1%ygvR&Vf1h28i(U<Mu`PMuc?P&|Vocq6`j;W1Gqd1z|@dDyg&!{lYe
z4?MKFw3UzFG>Is7-ydv?B8m|ZJ|wsG6?o$Kq(8tB6S*Sua80#C$ur%rO=DNLz^x`P
zb$6a25DyRKNV9kx81Ll^R_?{|ljMuk<Z&ySk0{Zo;xOb@GWngh7`5J2YU|J=InheJ
zyXU(UN@>)9n<)-81e+L@4APzU6NK|CP_ndV3;dyN0(-=c=@hwjx;J?YiEj)w@F3s>
zUsl~%cvswbJd7KTKe9WG2VsZtyX~wz^&Q0@#a+Y0xLZ6&(fZ>W!3Yq+8?|COKMvsM
zu4@HODLV2B_YZ}Pt(Y`xHWjM`riyBxHqwkmtV9n;oJL&ZmPZ9LtETBq9iSk_PIa7x
z<xEnYst6$?uTrMRVMBN&>S>_rTJsSADu3xYY3L7WH(FdodeWW!t&NRPS$%))#rDq9
z`Z}SikV+0I0U3%HR_(I5nFP!91imDhwjL5mCMM*G%T{ku04bbO*sH_BeRh(V#9DNQ
zi;K#@229p5&|PdVRoQe}dY_@(VX_ueN^1i^66Q^P&EJE@`t5T`{89oU)oH^M#psgk
zv;A(Lt8|+XnR7@#BHLo#BKaoo;MAaX<f^p#o~x~rkOUk8CXiSg5zUYXln~6mwR(L#
z*yD2|W(vUJd7tk|dF}|gv1W;d+MW<$=pAs<8X561kOhLI-F)Iln}+)^loPEPx&iRc
z1WoOZ<Ss*rLzoj9w`QbzEuObVZv=~&Km$q<udV_$lM7eWdRc~XN)qC+MGlra<;Fa`
z7H`+)M{oP{^S5e{p(%`h<1#`tO8;~lIoBy(<D5Nu4R;}9?q+aZ(~^x;P0+)HgY2ab
z?Kz;_IYUCzJ8~{n<c02p$F$pLvPjL|i@yX7Ff(<*=X-gwG8hYWu*Zqdhj<OZbqOoL
zoa+1bbNSV_{Mkl|5o%uQ^UO7ZNPj9O<nPk~#4U(Z^2fnDDQQ7y$loU{pi!Sc8ubp}
zK%v1u_MdEQZ?CQIgL#8gJqP6zFNkFE%}0GafLDOM5u!1MXb<p!%^)(tGAg3@{&?WA
z%?dbRl_k&$e?#qIg`?$Tj)`A59`WxJkWF`6UU4M?61brSn@sXXN~(k#TEOv(A1c!R
z2`L*ywAR=Q1;$aQSJqh&HBiZZFnqLxeB%72HaHX=4fH@L=nCy217TfOBDTR_4-xg_
z7E(gb&mjdbHvTz`FUZf=fFkR`f_7(4aq;ZLdrE2JC-zF_WS2EhYUVETyB({tn_a1%
zp(t&hN_~wO@&e*RqM23`4f-kO=7?1!6B|CG8{OX#CH&s(AW8ySFYJ`WOE9*DOXfV|
zZrX&KkN&7*{Qv};?AFznvHtqchT<2h%)Q}<!shS~0T;+KGI(O*RBb^YhsN~uL8Xk|
z(R>hW@L6WfQt_M<1kJ=0amR%YlNyL&VwjeWV|<Pz^Hf(4^KpsH$w#=Bm$z17ec9jK
z+W2;Ld1rt5`O1V#`S1joi`uZljB>A4&$ev(MI4O+X85Hm0Bs;GWXP!mBqWU!%_O{m
z2zq!N$pD3J)vJex5akHF!&T&rqKlSvmW(c^F`xQ*c#1wjQ%ahL6<l-XWwS*WlV-`+
z#IP%7$7Xh&nHQGkB!!5W*B03wm$KGZifL8cHqWV;RlULgZ{@Kf>w~om%>#x+ZXOR^
z_R<R=z7j%sC>GM{G`n_D>Q^zXigK3b<jIZ+bvf0Ed?Evk#N53FD=+s3<WkW*<L%Sn
zYxNrna)hZa<xOBv#hOdJ3Ia-|`;u;`BHflk;F(SUqhS~`)i?ofhOjW$Tn&3R<r$ec
zeDz!4KhemVHy_DoyxPMjxVzo8*=VHP{Op7`K0XFD4pE0`zX$=_)fBZJgZN1vox$Q+
z1a9HOx2l0a3RZ!?jUCTHW(hJ>_(FJ(>~028m7Vbw*!z%n_K$z={`2<U==I;e`7iKN
zUBvd0&#(UrL`vN<;T$?@<gUIL*nf+5sJTBSn^tz3*d7U;(ItdA5a9A0k)|0W@I(M;
z5a7>kv~b96e;;bKTPr6~Q6ew)5M0iaUc!vffihOQ8<k5%RJzYW_a$Mqs^ci^4G_f=
z4%2}278X)?pnwGkk7-S;N$6!TbC6jI8v;Ib+8q&yKxGL>R)B)%byIv=!&t;$z7_5u
z9ykOChJ#L^%)zu3agNl1-w7YGS$w0UsYuckgH6A*0X!XI@giS2-2&GEcPDG18d4=$
zoCJ=iz45gKRh82lJEDMJXdA|nrJkH^njw>?-M()=BExMjF#gFIdHuhM4|HbX1cNmb
z#=$@$q@)BI`Q|x>O3o)%1wf1-wYd|fEzbJ(r@(b=ldK^wAor!N<1zfDmOV#(L#PwM
zp<h|Ad=T~xf28=JC%`|DJrkbXxO7Z!c3B+182UrN1p}5+Lb2}7G84V`VS~a|d)8*0
zpU*7Jl(S(H$$VINh3J$_SES_kSh;Z&VKkvi(;VQr3>cWBm}YHXk?}oZTvY22Wba{Z
z^Vw!-f<vCi62>oLiD0R3G6VlMOXS-w3fhe^%;c(`U0@k;-dVS6_lQrSId&H3Po!|{
z-pU$0pis@~r~U2K&84j+{EDY~5u6RqxVh+9=dL1|rycb)YW1_j{<yb40wG|^TF$23
zfq&<)(-_0+YxMB0lr>3PHMZBDuD#evQ<&Zoo3Kc=Y9*IdRp<*V1!DX3A2ZN=j2n0F
zOFb2xyl7UW3i)bUns!Y!fD~QDvEB}loO@pVkh-GRefshVsAKk|o8m6#l&^T5)jY}C
z3KF)II&<~^N$V}l&K>b46_r-5WUh33N0>KoBp{687$(A_ok%JjkuCpi6Qq_rNJx*C
zu195KinVNR9Hi+;pdzvGlJEu0fcnK2GwH)Wrql3zETK(NxV&Ba+}O{#_9+n2^z^FH
z%&U}BRlQ*F=?BnjUZtVNU7ZOmFft95cII({s5oZrQ_injSdk%D8hjv^NEa(XxU;=x
z{)+st+8<SBAYiCW!8_fXmpaBvi>TB9*{4GQv@$~!G{uG{9fYWXfHK?@Odnb;$y{s$
z0j0d$*j0ucS&o@;aL0{kc-YjdC=Nmf!`2g2qX2JRc@PA#G9ofX9-D`NPmzt#LnqLM
z7EkiBdMCTP4#~Eh=t%f*GELh_Vr=S!@LIYC72JBYy0shXnyKXgkG9t~Uer-gAv?t^
z1wa>YK*-cZ%Y6S3SHo2!a2h6j7oa6K2v8+2jF+FEE97yBxI2Q9nJ!6YmC?Th%035c
z(R@@sNbZp0hO3-kx}Cvn1oNVelwfH7z}zhIs2CYRWg2qJYv_gkH&~5t8iT=z290LW
zh@BBc<w6{Tz5<nmK2P3a+4!dnR3h4Yr~)ouDB)r9iBEX|XnInUiCj^j?et9w<)fEa
zJ`^f<BjOh)Bmo|XFp5D9+b67VtEG_a+`+G|WzJa^%wR8QS;z((Xot@_y|c!H2M_35
z1oIJiZPcV}L(K`@Dm?<B3{`6diU-qOP#wIYHxxYi#4^Pjq&>lH&6XIv34`*^PV~%7
z;@w2p;-3=m-rUK_u+(>?Jy;d5Vi*D3+W^ZG_ao&4cvjytUQV@>xd<E6wd9@xv*AyZ
zd*~M`f`N91t^SC?Y(deVYpZ+*qq>N^VmmEwp$*lznrk7UK{}IHUXq}!$y&clv?+x5
z$@iNYI<L?z#HvY?EnI8kPBbBxjREyhr*(?SjNwFE4l~GHM=#G{r0DE@2bD*zm*lOn
zQTY~Y8+$Ln1g>r9mw!3A>9$}m8+FR)O+sbjA($H;VF~H6Altz_1-YkWo|ozN_z;QG
z)Ar8#;}jCzoJtkxkxeazKP6OW#a^oGuLk&$x<E;QU~Blj9s&p9@-1Q*eXfgyW(T%S
zb%bnNXKuyhw{%5@X8=TxY*!ovMeEK$o(^DXCd-fY!&-bgqOT_UG;q*v$~stpfW#?*
z?%gFjaR{nPuV8cIQ>W*95|&|;%~PQW`~bR}Q=nk<A`=q453to@#0ch6+{l>*EYgBq
z6!9A1k;Wo`7_8(d8v}sO!tqCixpM%?-6&A#_4(2Z*k=$>rie$YMwa$roG?idL*!##
z3q@ZI7Ez6ZCGr+}P2X_tk>f)o9(R^|heom%Ns|c88Kg<LPtym{fTQ#nF}6rh>mCGx
z7g(iyxsAV0TW0bq1rV{E!1S?Xno8ja)Tq(c1A(SSYTQdBX^UlOtM|~`haP}fCtc)=
zdO7Spri~EuL`7e?ZGr388yn_*@c~xWUkR1KQZWa#pg7w+Va;gYBVhfQn5r>Luw7^<
z-7{Ckd1*P+dB4WKZJHuacB~vJAh6!7@2AigPz);XmXHO)_4`mTbl;+g9BD~_^R*U1
z-#XnYo*%`C$pe#&eb|xd6_FWwa3;zlXcBN~YK`RI$NP}q_wj}TU8C$iU0g&9ARn*T
zfPEcJ7bv*Lb%~Z3!ACTia+qB~QxxrDg`M9o-Qp2sy5xjE+u7M9P5%Az+7RC4SF)j{
zco=)GnN?q?IL$Ipt0KcquXjG|-fUDVhF%W4TiVwqy-+4kWLwwxL1!`&08~xIP5Y9R
zPw5*eh)}miwuZ0h^P)$CejyGjlD%HCu1FTLtWGQ0MaaCGP#&sO`}~kh$F-n~<B7V$
z8_ebh)r$-b6xM-(WMH*}aj^`OKo;;hvf7>2;FQy;^-#xSoSA?*@5tNqdewoq@X<QI
zap<rsF&rp0&0IC+2~Whodk2MBksD;(VMRrz`^U}2I*Wf~=*$BT4ESMb>&4oOr+LqN
z*LO5xvZs(jL?>8-_ZhU*ad8NoVK51NB9=DSp!k@<zl>_I$SOnuiiYulo$rSSaI1Lq
z>G^|&%me+7avRv5tJ>P>8P$57qcJvno?7w!)s1zIz1`_(1qEDCP(^B-4t=m}Ul!ip
zU-%DBzbSje3!Z!&1EO+0)$kAL@d!Opk?m{%0$vu{gTn>rtGkHz8Z98k3x4G&R1EiD
z^f!On{*TsKw;2a%VgxNi-4sk3MYp#)-1-aK7M{rz>#r>pQ?(bB@LU!V0{Ip&EXrYG
z=THYJ^4#X-^ZA{-f+QvARS#QFmQ_!bOkYS?gedMoaJJxEk<X>uzFmVyV{O)<#Vdup
zO|xl9r>(FnOI?Hl-pTgN^1>4*FDz%Az=?-19KDj+!ktVdVbGjXQVh*FLcE7S4{(Wx
zNHcp4h5(%9n{9qYW|HfXi<D4@b+0vCRGp&frT6!rtgWx^@1wT&?b_efNL@xN5R5<z
z__Zdod450<GBb_R)du*RJ?UDbxH>;sETHO&It9MVZUj0MAl<hlcR~0lWq;}6FF0bw
z7O9BCtqyPoDc;#<oHvRS*iy~Hr()0T8fdW#MdBMBx@$hNUptM!84J^(9JqMkg-uh-
z+lWc8lI}3OQ}rAq$ErH5nYxUui4ScOGj*GSKK1z^r!=`brO@c04sWk`pM{=zrH#gy
ztm{ozzR|Iy^G_P$iMadWm7GlL5ZSis4@A$bUxU;=l0VUomZJv?B`=LG>cQ|09Kzw{
z6g-k>k>t?$8(t_A_7gV%oP+@;#Q-o&l6&xxOPajU{k)?f528cQ#d{-+!~+A?&)r%w
z^Dw8wR__9eJ=l+ex;By_Tq{G;4UD0kI|t7AaM+1RI`n(^>@ai}|9^vkI9yNzi%F(m
z3^*?fMoZ=!=u_yMzz}N0o@G#k3LlPpmxx!x=@O@9P48s~;|k~rKTbu+xDZ$*Km~oG
zv3Cb`hRjv#A9N_9P#Mq>yCtMvdJ#K@l^GJ^uB6GBE9*`wWhrchT_8h{8A)et1MSl8
zP!C+fvR!!uq;?KXOq<ZOE^6D`>kB*U!dTK)01ke>wnNa1J_EtfFfNjsgEo*A-o_uo
z`HpID&wG71E@Q>)?|G}=T<bS+(KMfT+o(iSYmLXS>x|OiK%!>#@;`xafO6WH!oHMv
zYd7i$ukMdVuoV+LpCs#+gQc*wCCk}<RJ~y8s*8)Q9e`s2_k&PP8Hs`PItPM`4h^lh
zP*8J~Scs0HIAd&}o+QY~F7QjLK~K>WMQg<yp7XriZBjUJnmJ99J1uYvC9Ht@UK5t0
z%8<AN984%H8hIaT=wYnvncie2P}?Qt0kPc|^8|VjQ}%1Misf9gl~e^tWbsZPzwA3!
z^ac!srBehsi`w#1?GW`Kx<_4kd<4z^e2u)OI5+W_ddoeV^ByDtkt<W^myIbLE5e^!
zJ$TdiVHN`8i0mjK;{|q8CI^FcE?EV_#V>1;QH87|abb^zYPkfSz0w&3Mn>Hq;H7`k
zJ%Z{;VGD>~2ckm3P0+I3&)Ux|hy-^MLTcR$fuEjaiF~JIU#Je~BWOgx^OCcMKn`J>
zCDki`1I8P=6Gibgf9%6|gKVf>Cet1P4nO@V2)>*2?mxlEfD9{ha{{6lpu5}<89k)_
z+6V#@Xs$JtpS$(ZN$cL<zo^5b5GG^fnGna!nDg>ew*HiA4*qhx299o0_v{~Flwr};
z2U<KGfiSnIm=rR}2{nBFn4ph)OT&!#I>C<m8Wmb(<dGl1a2#L!eo0DpI%q>-P6X$E
z!e_XJ;+cmg`XNQjr4Y41aSBl$&N?c0kw-qUp;a-Fyf!yEk5U~2xe=yx!mY)bT|pcJ
z9!^bm(RYMuSYIVYy)IRc-3Lz?IWXi6&8V7F6$#I&`e(AUS5L`+U5`{186Lpq3aNru
z^|{!<;UL<&g!v(;8mU&}&NG>ixRd7SILsv(<|joTrgg)ou)V6)Lz9YJl6zB(g6}QF
z5+2UWz`~Jo&XPx9kj9Tc?K7v>Dgr8>yj)*@xwSScQo1zQu4(~)i0*LECufEmW7s7P
z#ySVJSpnE=X2OA6yAt~*sJMWByhJCl9_bO|Ez|FigDPhk)CULeAr7DoU?_5w&O60(
zswALdNaW@Paadn~F|pn1yjdWDLp;uS;a&Ia0De6ScjxcS-$6%LU?@0Z)02>-Mf98s
z)o$4LzX|H=(*JvK{#qz~B}SJ^<4<L4%Qc%(s58M^9N?Cq1f{7I+!na%ygC$Y%I~Js
zI%ti$htyIU1}H@~^{I*}31TrZJgJ-`aiX_{3=2htzf4~M+o`z&2D`9Yb$3RsZA4#3
zu^A{wF-Iwg+&ltj4tXY0R91z;<VYDfjhB5d2CKXuk6~pTqsS%iqk(feJyGeFK*TQV
z_tCl~y;Z8+Qf_fhu5YlK`cDDPfIEt0Tj1JNiyb#ZHHWa|!D>NyR{*AZb;_iABflT^
zVP-}oZ-RQ%!tSfWbrKRL&A`hULwgfkoQR6N+vgb83eG0OE%T4S)rO<MmQ?(!ha>pd
zU8I`_YBLU|im_TL=rl<MPP=_kbUvY?^kMDpoqM0x{w6<E_0RwrIKqqnf{gQKpX$_o
zQKyQZDxRCl(3B7?JSrZ&EJC58CM0q8@Bq56WGIO}zh2*Ij3v(P)gz=g@dFh#cW@7L
zAvKn%CVR^x(M9S%UIR4FWrg-U<61ZYa$V<jv;->wQ2`xnv`i){9l7__z7-UE-6dZ7
zMSv;w$n)dY@n{}OWmR`nHiJ<_TM-_q?$`ncL@RLh82PVvFnSG3XMP>1sF8T1f7}@Y
zLU1b-O#t=Qi;D>G@W+jc*$c1Hfrka>r=ZCwX`Vxe$GyT*v2Nv+I5=2JRSn>>%I@35
zB~`*>1qrv_V?nWB-W%6j55B!(IR+Tng%TJcX9hHOs{6)G(K~9Mivt?$F!~FtatUuN
zdQ-Tk)XvYcKr>`_<V13*z%rNPecoz7yvLqUBR8Tu9q^wTe`^+O7~nI5I7ZBe?*&Zb
zJS^$QLQvM_Rq!%$Z5@&`tJ|~+w%Wlig0N+5_)>%Bl(((_mkK<ceQW$*m6~VBV;#}x
z!MsJ|s_^hOS5S?@T$hn$uixo034sUfsA=q%vVD<St+qu7-Yd)&oj@eSIEBSQ5W7kr
zjqAFwW2%$iGM^<=7;(o+2@emDutnlJZ>@9WL;d77zDqpc6@}yuf_kLADhVBea$7h6
z?F+_3gdeEoGAa)Qxvc@!(t_uYP;Dd8HQHQIYJ<?xzXVsxl0e8zIP53Zk|Q8uE}5I*
zxSv;P94my_XxWLWpzb|#jA+fV+0*D_79f7hUW<xXYXH95R*IXGFfPsxLzu!Zyp)!S
zmC?F*vb44yjx|r09yK@K;gsD=sZ-j)WLJSgKX(Ig9TJ%PRe(sYp8y~BBkL>Rtlq9T
z0q7R4F2q(djsZKso%1iOC>SA8*l%<U@*)N^i~IUdw|fz?_O+VU7%>a8+&s)`$e%Gf
zX7UjH79U1q8iWvFjgnmqs0cvlx-BsC5DW&r)!~e6ILSHEgOKJzIBLYFHp0CB4F1dF
z0A=k+{?gAddv9M~lT5M@QnZG&s_leQVH`<`ds|ZYuz{q2rLF!K#PHObo^J9XbnWLI
zp-1*PKb42bRI7<Yq3~11JNuz!1Ye)RPSr*E*4nrWM4qRhCWps&!&#Pk!p<6k+x;bP
zB=RL(G}1#+*Sravd+uAB;ivi*d=XUqxv%XZtzTD-O|@z6Yk4%yw_O;~kP=j^QN>lA
zvdTDamR#@h(SCSY90&VD54uKO>JRKg>fJx0Ju25v{%xV7;*4-}FP}pzxQ^U=gc&D}
zXNP#EdjH`*3?*{Q!ky$n?ejZ#?tCqj$hXv5%g*G!3?Bx<K{U3Te1B~P{r&yU{X6$E
z{Q*^_ucT@C9JQ@Px$k33TbXmxI(GcIbnGV<)__S}7hw}@i<$ZP(<y-KO=X4~o2xHi
zUJT2B@$&iO)h&SF&hLn7y@P*`a?`cXopC9_edNgK_B$|xK701+x%XU&eXX;@bKBv7
zPiKeoS}OpSy{@5C1Q@=M=Y_SFs6*fNx{yn9OOS_|uLlSF-~&(w3p5V<!Kj|b+KVR}
z`%hP2tZpr>ll0kh996VdP5?Q#`30#30-|FA3Uw}lj)&|Y{mmQTV*mDFVcdT7U<}{=
z9#r8EXWz~~`d965V6Xmmt1<t#Z*HNHH2K%utv^qx2DK6KTatDn*uycAB<)1S-5y%w
zc5gNns7*vNA0nw0>YU^N6JBJ1!{rsZhHTFamPg5iVebN2%9&ie(0Ar#+P`rWftvg-
z(F^m?*i)25fif}67ALFi_N`rthr&d0*2UAK9~CBNeX0FE?aD3_bMMs?E>MdchCUS6
z8TnQGYs?Fg@Erj*OhKmiHlIkCh2A=g;EXU!?4{qOv(ycUdMC`DN`X@qKIdosk_JY~
zm$-&!E%1=uB2eJNTLcmqU?&C5xijQXn#j+t)?s!&`;VCexI&{0q0L6_v5ML7U9`M@
z;MuVA!ffnbteB6!{b)1N?hA90+qPmxeD~4j!_Et{$=%Ks^O9CIWIXU|JUdC+_>g1)
zUa0UyJwU>C5jLknHA{pKk5B08Pp3gUnM+|E=Y2eHcj^@jK%gbud-knWFxq)m8AIuI
z`LS4PtHld-HBcZ)Waoh}rHCaB#Sh4Z&V%^;vRHbw=!~12#KYRg(GeCxzn%hvt!!NO
zY>m#L=Ii#W)@ZNi)twFAYvvX(kHFP&hkl4*-5>;bIS_}i53y$Mn-5QRyRoXl4Y0=o
z8FoSBkpBIk+v>kwtS!J5c>%!s`vNXyc%t=>=GzO&$G55h=&dHK5H&qXWoGB+jC5WD
zzL-=QY+?1(yd{l4ywg8#k$R0$dK}*1Dv}5)ou3^K!D`aN1v*^&`~3a+dxXtN_ZS&?
z=L=Yy=V>djH`nZ=ZDTd$^D_YMlJ-xHy3R+JEKtCU_X($-P#1U#Yl#zqIFa@gIK@(6
zYL(yQZu5d4mE-ZtxFbHs=P)II9d|`(4zhBx(gt=FX6Fw0Fd{B|3K+``-w{dOfieqB
zxTeK^!g=Rr8~wn=Sq;K>pcW7!-M{W(ZQ*=0TmWCGhv6PXk2_!7`{J|S(P$wjYG6wH
z#HG>TIxu$zy#(gu4eWIS>&PCU?%YWi%*H7=Q+is!p+<A$nHMA+X!;|-1Wpq4m7tzh
z%CnX4!P-&H8;JC7^+l8+t7JC7X$UuH^NeTbPf>ip@e05MJ1WeQ>HZUqi19&Dx-#C&
z`TrdE<&TB`vOg+0Fnb?)Fc&2vsD=H)<3s@Y%a3ht_-o%cBHzR%)rmEHmPHwQi7LZu
zUdvRfjBrt^?~SX1;ZD;VAljvb@1)wt5d@`i)P1sw2X(ZN$t9gleTllz5*Id0T73Z>
zr}4X5vk3#`<QzK`T%RU?P-kpvl}5CWdC+DBDcTDBG}h+qw2&OF5taf0o@}Dvsv;-u
zVu1K;*v!Cr^gH7P{K4P9>9!x<yZ80iU&kY3%y~w3nqvKO{?d;O0^%H<0RdYIZKyHw
zEIM5XM+l?(GX#~D^F`A`rLdgreLY)wPlGvy7p~mYdO;ft=U4RS$z=c<C~^;+u#kwb
zV$K1lSw4xr;#j=g^`D6~?-a(^+F54+Z6dsD)x#QO@)S(KnD<GieT>RojFDFaD)I$+
z<;WuSwf69pDy$&tLEVP`!?1yH2>U&+huK;E5!u7)D1Xh=Rty;;V@Wkcp||`Ks2a+8
zqte$zO`&lc(9}WmgKP5uRr4-uV3M+3gI9uzRf_x&;Ozl?+DJ1zo)qr)Z|^Fbq`%%R
zhk8OJ(p$t?E}sW(W7Evl=reYgdy9-JZ#zXOXQlPzJdn`o%PgszlFW6-W9Z?_(S1dP
zXCfep08WaPP6{ZMN5H&*ZIgC&Md_>yN{{sFF+IpM#JY3|*OP{(<BoJu#c3n2##9_n
zjwJvc$N5E-T(!lBPOHeOiqZ-y4JM{>BL)sh=41Cz`6QY~phNz|Z70Fm$`}tF?>%b2
zzejeN51UqH?L(E`rmvNyeL6oqgG(M_pI8Y;(~k&0+#c1KTGpPyEC%E*WbTqUyXr(L
z=$l?~(z$0s$)$H{qfKm5mgi{&Ey(WT?zOIHHSmN`&_;K~kx=brGn$7BnT$59_BCom
zY5EqX4WOUcr25C(D5$ApsWc0gpSlDHTA|viDZnM8iupoFkc`MXu$w1UtbGcM7CG&{
zlL8I{%tE$NUwR2>tGUXDYze2A?X5d<Pve)*&4XI~j&WBh$<r%Nw(+#ZX%7$sc03rU
zr!TgveXhY%rIYgS#BFgf?4pba%r2N+y)|))jx4I^#*{?*;`VrtV7+L{TSIQ<?qL_P
zZN@#0YQHp_4vSyPK?Uyke2RT~iEDakbNQTv?63tBF0T0VetQ8|SW|xU*_DrC)z7ZR
z2fY}uNJl*>dOSR*WJ{=eVzM-kn7S;04GO+8;Qt^G6Vt<K@VsA1+OLxH<uQ<gFd^4l
z$Xl}^!IsH$H9d<B8w3;FyJQuQZjd<4@=%;z;8Yhg*nn}-z}fhyI)hHS5Ny_8mPSWC
z*~M=dp?xUEQ7oBInxzkn6@$KdOdU<bio-@+I$}O)>yx!!U;$e9FEmNK7|6}9r_jfy
zC5t2rr>2ualVu^53V1&QY$Pd&WI!bdj4w25aZ5spwLfVQ-vvA(Da4>a&y#S5%e(c{
znLF<#gH*NTI5?i=-OIr(=}e!S-5yM+u1!{!nY7&d>i+@YGfcZIOXkxK!#l;P(z}9w
z+pY!GI3=D!(LwGy9anb;Ci_??0QNPHM@u#fo8kbay#TP#UVPu-{P&#UX%)o<;h+bR
zpzzf)EsD=uyr$zOH?n(P{AvkXg+};q@f{rbtl(AERQ$@!PlrUz^911Jhhs&9rXB*J
z2a@(RJ*djU3GpGt@Xf>2=#_JQ-t6C$uabb?%hRJl|B}fF&rr3gKgMr`#GPIS6bB=)
zVyj5Wh>vJeRte_=uwcAFBD0j_vEyJ!_R8Q^2-R@whAi0aLsneSM)9L7IV>76u&MPP
z{!_u)q+F3Ny;0Wz{VRXIxVW*ogLJh^C~+s+Ep5iD9*#v&Ma>`2)A$Lq5rm0^-SQ3^
zGxgT+80_AR%*G!tF0O6wFKult{Z#vx&NIFVBhJ+QA^9orU5sf{yOK^rgZ%Z5Z#F$~
zADUS8YH%XdM*9^5RqXLO#hy`pM)?KNeg4IF{T$}La&>_rv|^oa-Y7J%m)uRV)k5V{
z66PQaJF2b+=#K8^OwrQt86e7bL2DM+TVdPxPdg|wE{#SORkIJ4w&8AG68-%LbU@?t
zA=1Fi!4x2r=ho~$YJY;9IJD)K441L|^_WO472HV5WL=+dGLj%8NoNm|h^wg+5{x!x
zUm#_yeXQJNk>$rVfW(SuMX&*2_zhsb#d-h?8!5YWJlkxD$XW7O{xBLGMX(?u%_B_<
zi+M+n1kPRb;rNA@rlz$wFC{ptyzLIoN00dqGBIalnW^I&sCy4uI6C+k%mpq}%g=#W
z-taZAH++k`NP4It7z-*27LCK-Sl;BdNiO)N^pIh61uV9UTAPX8Q9+gAgo^kC2#Bgc
z6EPO<-a?{AOpvN)ZEi)yMlkXqS$TaQ_MQ%$jFi&yHqtpb)+-Y&$OEn30gD777n?va
z$8d{__~VW3wVj{#Uu<kWhbs|aYFY=m@M8xPKK=gX+Uky?jJwCDllR~C#x9vZm4|Zc
zLh6-iNBDvI&{^lEx9qaX3wxOqv>)y6^?6U1k#<}B8imAP^??F^FSiFks&Mx{1(9m?
zu-*@P`)YQoc$Fsi?-IQhpw>3Of5S$40Enu?@>=1sN^weGki``BU{|;HpKUx}WhJ94
zwIB{sDHO=HHF=7QS1tpiQA>jfQp@jDyHfKP=M-0}_E;t<Q;HW;i>;nytyME8GrR9=
za9n|$hn-PCIG|>z-_HX}aye3AlK~Xc<1$9t-HK&IT-czaEKX2cUSF%7Ba#GaBnbKn
z_}(S1A-ub^*y<T5fbj{c9?0%wx?fNRYJr53+k|%b49*ElJI|o`K+H{Rh$&>gp`*g#
z;vN7n9PK>-*EoE%w{&*4f{44ltw9SxY{$w!Qi6vL>IYP?0lLhPHC$lW*Mfeuz)XG$
z7_<UM^m|QH5;G?-bn`M1OHzzcsH{PxcxUuxJUEkJd<ICM@@N~^F5ui3{dX4&KSC)n
zq(ox2+L<3$p6)Miym+$qRPc_vd#jYjwdSY8%v-#_WL7gXFQ7DBtnJW-AFTHP2-2WP
z)_3aMG4sQ7lyimROGE|%HJ~Y%YPXL|N;^|E8%C$HDGAGO^+I5N-1x_9FZNL#24F~4
zEV)YUIDHKvAc+p8OC9U&T8>sgWYNr4P=vzjqY3wL4~b_e9EF%$@sGXHXMgfOkuI8#
z_E9`liIis0Br(P(!LdKy>DC<sIzsTOaA5M6T;%c_N~cLQQk}I>aMQ3yiDJ{wlx?$g
zEqsS$Er`WnQ}bD}APD6dG7TSSJG+29PdousRPTVd0Sr_zfdz%sXj)?d(hg97(y!I1
z(1vm|A<rt1SA|@jLwQ&w>@(Qu7;=gEn(%PJU5xb}fOZI?DCw8s6~cUwhvg(?D=<eY
zw~`A(J$jMdpvUxk*tvsaIOz89?*RWE;NKSizV(g&Z|{IRTd&yN+qJtQD2FBCEJODx
z*i1zR0(O{cR?%2=Kz`9zeo+Z{gk9yZxU0D&&_4;hgs@om9(HgO3EQW;XZ*C#qaHl=
z?C)@7a+&9&K5Z}lIz{89;jncXznG@i3|A87t9ke>peNUjzRRY3zKVGe0-@zL{5HMD
zCmYas4Zuf(`qn-y<8(Ds^9*Oru%betfHP7eit53tH0vA;aTVbZEAx@F=h*pn#w6i*
zXbT2u2kkynP0?PRW_{cfT{+Y*x7Nkvk4c|Fw@Xe)t}h)9;nUaQc5DdRd}RN8=O!ru
z5+2BKya5UdTB9?-Z{^kmE}1hXfq3Qv>b@7uJGf#A0slmpkWNb+R{1;&j`HE)sCB`i
z!rID*f(@p|rrGgudzOq*#+>cgG^xrErdS*ZxI^rfjGDo?Y!_#e1?nz3$Kbn(k_GV7
z1WIJ*hqPzb^0vl+J>!VYruR3`&tzF--wbl0n`CMD%D<@rv#d7+o3S@`Fz>Q8Z?n>4
zKrs<2vKw#+2v!7I<JvJXS!|TZHutWx#TTL(E)9vrhoa)p(iM?GFpdJ6aYtm&=GS3b
zOEM0&8!V|Kya6sxWISwR5*8NdEmnVrzK)Z#=W!IBH9O;P>?QjyQ-Ttq)|nxbImr0-
zGK{8ok#IHkN;z<i&n5LPyIhp_5l%rDv}vXmA3U;K0FIsnWisSM!oSA3vE((kA3R77
zOgceflq7yt<3%~)PzqNUBj1IMLWpV%I9;{keCCVz$Yx7!#TL!|LEJTcZNyNSO_pup
zGOQafyLXMD>hom^J3j+J5+um+;;uLZ76%XxiR8#Bc`rB-2y~r+=Rk_Wg7VA2LY&3X
zpQiTAW%SMNX(bXk@UWZQyJtpl?avL6I<a}W^AgUTL0kCCZLNGSr%saynYPT@BOM>)
zwL>H}A&}+@Pm#DK+n=<4IYbz~a@?Otf*9ugO=l;2r{iTl_u2F7u-|Nl4nEQFlzm7@
z4Na7Vtj{|T_+KrpudT4yJZPC9u0hw%u=OM33pgkp1l}@XutN^^x~i~0Mb>a-m;p>o
zLUctQr*1uvygrsYUniyWwh1_kDMYj+BS8LJ##bv)HnVKl;$mlX*h2kAnFbYp;7phb
z*tqvxb5HST#zNbPO=J5O7rP^qnIU@t7><nAXEUZ3me9c3>_z8b7EBV#Ne>}%&q~^3
z&JzmjS=_cR{>hfj!S(z6caryFcj}`3d;1uk4v3abznmM4CVo`-K(y6bB~ynKlq!zt
zqiJqb%sL%aji4lB))Qt?(U07tzVc!pRXAlt5Bl;Ed^#n~s@?K(!BfO{9zsaoZN2x&
z+%buWCUUV|ml%D;r6MvFu2CWrx)xOfw<BOYDG9wFG_(=Op#}|an_xt0f0$9xoU&q0
z8foa8FqqI4zlv~^;GUT=!9LqY?t3Ah&(&k0%DX^%nC{!qr<gS`aOOaUx0gXKH;zi)
z6wl4g46qL2y;vM@3z?#NaP!<_`R<(H^)12@_xPVDhziNQD;MM9Iubduys2%(UQ!A@
zH-xu<KAxFl_UtFhn*d3sYl|d^qJ7A<PNu0Kfmd^UR^8YsRb|`v6jgH)_f(?ibh2ud
z+*vN+8rj~np9+xTY`mG8rAb4cK;_B3k$=P`1(F;qyG>HjC(^9&uvQyRxsExc3smJ4
zd-X&hQnFl!0cE|Wha)LcluHv~SIfClN|H2&!5$o41u5SaG+WaMxz6O8BEQn5F%f6U
zbW$yz2;Uhkzjq1Eu-Dhb%{5KWEAa~sU~E@u=kTk6`^uF4s;^6pYs~dKBf2~h>V4W{
zlzpDmZ?Nvep8!ph4iQC8Zvu0Zzs*dtER>AIwWtAVi@6;(uX!dl<!+TOFL2t^0nWL=
zWHP1QS`_c3<>hTK3KmyABxqn%kSY!9R$WMAeRybo5VO$g_5lKc%+X-5aDZAQlIfMg
zvpski9?aSkaqPb{J82Pgp8Qw~p-_VES(%5H^S@|eU>*->Il9EMsuVKG20RD7I6Xf_
zjEH>({9|n~0v4`9zy&Zv^N8H*(Mybdx;F$9t2L7QZE+DfgtQZI0k>!iEQLI@(9pVx
zYO2CLCz;;5Hoh7q!CMJWGifuuUw5|2Vx{9OmdMZrwf{ub6xIX;goO{hI8;KW8a$T_
zecx_xyii^Xs*iFPT<We=z9E5jCmnEqqR3J+#U$E1V}`{deV&=Aa_>z>%GjLbswvzn
z@1a{NRU9EpREQ^c6I0!82ozXO$$C8y;<^z|G0M0GGZGx6{*(*?c5Lg8aDwcPSl~Uy
zn-7Eo9Jq4RD#R+!yboFeri*d@z~Q2b0?IJQ@uOZBzJ*~O;Hr&pH*tL+s&tg#vjx0X
z4NmSc@{~qV08|0lRrE4}!YbEPGV`NO0+aWby@FJ&Tm+Gw#$kvUQ4_Zz(`f)o;W%Kz
zqTmXTcs40EsYV^&2cA4*D}w?FpJtV9a>6)hs*j7)Oou>bR<1d+l5rHV#j&2f86q(f
z1UsD$eqk>dS)?w`JP+x(<#}MkL`_oGOj-~;0gTCW@O(KKCog66fI>_+D>x^cn4e8}
zCwdwc_H>80^eb_+OZGDPU{TnHz0tTW?zFm5D%GiPPn>%46}B#{C_KhGDIz9)nfj$-
zzH)x@@U!0Z9U(XUa;(28mS8Q3f%#>GfeT9^)-!hBmgz#<<O*0JXEDEob_FQ`Scbz!
z#U|^)u+!5WiFndtIa1V&lOqzc1gbhgw?+)EOlgCALafHBj%C}{=#4P*Jm3)N8ub<N
zy8yEW^Yic&rDN1p9Nq=H!P)MzlT^uj4UMRUUY5<k6DGREfq2E?xYJG(AG%se(iA$}
z%x`_##Y_6L(LRjZNYM!i1<T}qbx|FAC{@C-4@l<IqxXci&{tyNCO9;EljP8hA5H87
z%3%6oay08J(L^d9A}!4;@e5v)d4O(*`RM$l;1a{5kR|bl!FjJuGY=T)H+1Ro+gCkf
z{Qiyhd-&17G66PJ419W{#@TrL3ELL%sRR+`d}`_5>Z&kq<*Jkly8fv|6qJh$D0>ro
z5Fj1TI8>HTT@syO3R233%QN$(mY(0nm8D@Sje2P3J`d^Dz;8d|m8I7s2GQe6T2fSO
zOnJ4*w5u*MI#pkBh3Q<XI?(%;S0r#by#saTGs+7|K3wpLqCA8Hc^%4BO_J-e{VEiO
z8d_3B(^c)V0sw<x0M!t-pu-`RTV^RH{0?x~eJFqCuVcUW?P4q?i!-2$AkxCz06}1v
zlV2ahDf&&L{zIpSB}l#-9z~rA(8!pO5RyXq;^6q<0Wx#`<LLOIFkVg&v2+X}92mW5
zyo75?o23OD^y$iGoVBDNI6g=DBcTR#uRcFM(JFaV1IuwoV*#(KLUdB|gtZSicX&aw
zWgvR9^K__|3*?hI(Fl|xwQ##ioUBoFT1JoP!ao+LEdbw~f8goU*lji6y}s4hU26XP
z`qsbfr~c4ax5WQvZ}0Bj-u>p@-o3rO&jBmxG3gn?v)3W?R|#7MDzIzyiRbC_2mVv&
zw6Cf7Ug)%F<k<!!DrCR=#4@Ms!yOdqEqCiL@ym;)=d1ZnOBz)73evQ1LeD;DXm_(s
zu=uvTC){5^8ef(C<b9K_BkV;9#6sAXT=ueAk|Gg)TL0u`+-ZKjKi_=)*+TlqoOs@G
zEP-j_*>6ToH=IWg{~3ucF)fd(nqa487cg=Ht3wqG5N{avc*jZX?L}uOc_Lt<)iT5$
zLnD1anxTah7T`7#t{}S`9-v`xm#{{%BAkiETDpqNk`OUsta*T?F+NNE_A3^zL*{Y>
zySQ3M0Nr??m+qs52mJfdt%UmttBU~!Nd1Ll=W!MYVUV1>meyEN7`w@hpf*rL39Lyp
zs&SymL0oIYv@f25o9~$P@o%+d&KD{)0UAUpG4foop4;k|GGxyUlOqYmgrp#XA&Twg
z*Lc!^_Weco=qP-tToO<vRo6!L3V4xpMITBMbjF`3dZceFCZBxIH6(vxu{r!v@x6RW
zCfdsx&Qnejq<Mi74E0PJteJU*YRWs7DDOG_3!`||@%pEu+JjF=k2t^hjrNy$<lXq;
zqmh!cB@c<63mwa~2r(pk8ka=PojrB>NeWmfiOK(tkZ)k}5#->Qo%X<ULZ~QSX$_=)
zU4m)l5ZIFeu)$9!7varO{D5h=i7D1$W;q<H`3SDRqmC}Z87@7f*hjOXmqAn~xT`v+
z*I~xH^#q#X6lmiLRMMB+CH;*E#Ii0_H*5#RX(<a3^$Kzvc;>Le(r@)hWVg=MvvHKn
z8OisP?&nmSnD4;1a{_72e-g>8Vt=JdcvtsYIgIIOt9$?0n2DBp78Js$?CgB)Ih^$+
zV^UFdJ#!P?wO(_MTslxnGxIR0z?-=FK%M+If<#nzmXjS(gl0%+Xd7s!U$Z;x<wka+
zEK%SHMHHIj<BApX70(5K%7RP@TfU~Cp-pdw^oy{vb<_K2u-#gg%+TIcOw!(Rp7yIu
zwN!+ldIxtBB8){~palm`Gi_*(p!kGLU`n5vAkt%{1#JToJy!HkYnYY|X)39OR(%uv
zV&g&!3L(%7`$&|D7|@i2s%~?dpHV%9>7^QnTrZ~IW475LkP~8|=s8VeDIz)`ra>n7
zrA%g5#i)2_mGKWYAOW9_EzozB$TaYKo7sf`Ln;if2S<F6uJY!fEOzvvpeb6=s~{?k
z0M*<y09dH}<ez**M#&v&cNsjEc!8%p^1?qG-)t(gzT2$L-`+d>WBpJ3=f$6U_5b((
z{6DjeTdzMu9RxO#yY$igXW!g<G`GM!Id(<j7m+6d+hp+Rj$TnrSS4SCx+M&UjTds{
z5e&s=^zx<-M6%cU2gY(iLf5-oGH_-)gQ|eg06eaD)hJY$GY_KBd@WopW`9$)25q7<
z^=}6!{o3{^v-&;w6~F&iN3$6-0`8HCN{RbN5NSYM+D5uz0peQdE9(A$suzS{D{30*
z)JC1w@bJW1$`+UC0^<j^p`RGURl@`BEBzaoU!@Wf;$q?NiThcJ67BYkOi}Ev!M-!e
zvrDu$*pci~Wr9`9DyL3sjEED3{Sl(2j5bw9$VKuK41zTAvEG4sT@f7M%(xB0%o?>)
zOPT+HMg~IY!B5S1J|&sn5QqnQH$tJ9l)!*a#~!d-bb$pQ354n(T@*NI|J=-Lp=amG
zyzYm!Bb`o)f5}0EjsxCq^e2l$5!)W4(wHsv$zW)>XO_<h9qloQ9}zsdpfI6iR1RY!
zk|)kxsm_(p)L)97+~<~a^cDB*+%u>E93Z)c=3nT19kmSb8iJUCl(wUKNS#tMFKuR4
zNYe3BH&Ia2SE>GaST>K3fmEMW>&>CYgzZYz6)fNAzSjhvsZSd9Z8_uuilCSQnUQT(
zvZvNwB@bB=a;6aoL?@!JMNtYpq$dW?K1kEy_&Lcf=dM!7;mH7OR&pj%gRgz9fgDs9
z3sOC*LC1@<W`;u2wRIyP_Z_dBv%{x<cW>_6r;HAo-Gy_TcJ_J}n3s0uk{XsFhp_CO
zs(XCBg@q*n6Gr?st`vzF!}?{;IYf1hJFqUS;0f31%LuEmp-~`%u%rgLfVj{-BXfsc
zZh${2jcrh>QqE0b&cMjCmU{&|BM%qE9qQNMn3Jjf1sNHx4A>o4k=f2+Si@??PbVXF
zSPzJV$#>!M$@=c29$X)pVuszrM%p^+Ic%ie26)+9Nsp#PwpA0&D_de=e%h?Lqk$Z|
zJZJL9Kles|i_FR1_5wH&Xv-1j%%`^CM|<0!-MaM+nyL;qEH!S$b4kRZhl7Kd=Emk=
zZxceN<dsU~3{Xuu0p1%71D(`j3{puS#l?n3N}ZrGgj`!W2?5Cr%n%ZhRcNPuNTY=3
ziQ45lBHy9)6+A+6AVV7>*Pq0%WV?(iWpxp&6j<D!vi2WT7QuQ6*Le+prxAtmQlSN>
zQv+<FtA^^D;YG~V(`k!-H)8Ycnfp-pvQC_GY=AL0L~Y0!atXwXmSw;jlq5uaK`?V-
zS#eH8e$m5E?g3b-ETukAwniQ#5~=_PMj_i_qVoj!(7t6TLi3h?lsGsx0I*)Cg*pNU
zy;lDX;Rs!!*r>JYrGI~GBBX>!MtIPJ#s*WW9RB2l<E?Q8D_}^#gA%wSp*}g9<Ye-O
zS9_c+u5%kRT;Yz0tXIhM5E}=YP{a;dw6jop{aZi$NCW{tH{MEmD`BOT$6peyDKIR`
zN!Eeh76#L5!HVaqG>TLE-%7=8=`trJgBf)eD04^qLlM;D=1yixr7>nyU^lfF_2>#5
zUKW_?fKHvoqire{N;-v>x*Qf@gr7#w1LMC%`zS}>gduOidHF8pUdaIP5~7PffI{{x
z@3}7=8uaXsA<N-RXj_dSZO{m+*X-dbO}+enQ=`3NfEv{Gz(u5zs5R-jYVUQJ!7zr4
z=@{ZIIEU_#Z2`Fxc#25a^htNLeSXMTGF0C-lV)x~n1mu4K8(WoK61d&>e$%*=R;h5
z4^heO)|{4alaT;VjfM~j=TvVwOm%DZDa@`r`x{#;t6Ld7ur|Wz$?vlHxykIB<bC(G
zY`nDFy{Bc@Cv8Z@yL4^9i~K(&XL5q~k1#q+7>Hg`44J<n074}>MlZ%H;yI5Yo1D-R
zMIU%H;f^DAIDG@C7ryy))L7U<#k_BXosiA=?HmrYuzJXF;*`L6IANwaWa^!CyM~C#
z)%~@nNXWIi|6=ur{q?mMtJ|4P&~JGP`5SkDs1moG_nb@(i<K71N$_~bC|Ji8l|Ztb
zPE-*fA#Z-5Ua4#u<(+UCwV824FL3dPrpVe8L|Z39<)mPOD^zBSJ-d-?GA{>YQIqKL
z*;$nwChd9JUfSWnAn}^^k?xuw*V(=PtYo+uowIJKWEzzLZ`bbJt~HTmg;I$3P-A&K
zd_i==DrE#P8>J&)QSmeKmBie6`5c3n!zimh6Xh3C-J~4XFohbl7$5k|*j*JQ4dWeS
z>^qjgDbypqUxH!2-SPH~10W$Hb$WvAnegD%=cvu>3(5ss1Ne;0EB9XuXV`CAv=m?~
zZ92oBIH*FX*@v}4m(O#SIa74-lqOwC4!)?sO1_teAj+}{SqjEMZEuQ8DU4gHrt)RE
zLLHRN?d{QCBv|D`3qvaLoh1#)sEQwauY+2|z#8|I4H`K>WZ4{?o{%(QT_i4$?uecZ
zX>Zo+edwav40X#sHHL>Az9sOc1MGeqg=*kZ%U-lVX*foF1i6=ZhFj36nfnKMH@kH+
ziKQOfnT1bm$;D#12BZuKlKaGC>ilRRjaZdBtaP_<NGZqQN5T)wiE}}4+oCWH7me`w
z(*&&q({<Mc6`jomD`EnJf22GKIrZgMzpYL+HVv{y*~u7N5i&v$+)59UDbqP&>QQ75
z5(%=>UCA%WnfkK$B1sAfz&D71V3<%6tku7CktV85#PN=WUTp2EA1VUdq=Jt$9k9dd
z0$4!BTI18vt@+wA^UbAEir@@7JzT*^5?A#pb(exPPHra9Bo})8te0Yf)IP%H2^VFP
z8Vt*V_$##Py>e=;L3#vlS&TiF07ItQVx*F!HO++(TP+g(Aq~LoQ=%=#)Gk9P9rQYy
zx=Q~WwD{NfM#<ncK5$RFt3Ig;BsPNLF~0&SgGfahs}jZwmpw#>${bdf{IN84*EYv;
zRcS<M)+QQDUbrtSJ#QudJk=@SB#|u_A(TK`+}^zrf`IWFb^_`voX(2>>vaNt(JiX`
zH7!}mgfLSpCh}x}q$YG@6bMR`geIqUTBtptR2DNo;<Axyfw%F&Veh=nL-%TV8-J+;
z4V`CsJ}R3Q($q2wxROkv?us}bhdS+~*iQ|CoPE*G!{+KVODk@W14(*s)Z4?$CM3Qo
zhE{yjZAGdJ))Bx7ky>ElxJ-g-ae1VoV0-|A20dDGWlWk)5Dp{;?l9~!NkHxl113Gi
zbXKLLUT%L7Lv3<Ct<Fx`NANe%LfMRlG@~ETi#fgw*#ugTmn&v;{emYMtf`n)`RXy2
zhI2Np56Cf7fMaD9V)HUm+^j6tUV@&X8x6Ns2!U0q91c1oDKP|(>Z=h@OZ>!7l0c!-
z&BQ!3dUBdYXy-iTibyo$0C-F#Yn_<0N*m`YO9_N}R4x$|?if^MnmHg2ikPgfBvS5?
z<$@U6E<IEZF*-j!1`{*|2yFx=J8}#~>KFN%C#rk|J2L3L{`OyQ`F|dg<S6F|#Ws3}
z;Rgm=fNTt5hj`$GiBte|6y5+xD+?wK(L{i!R&G0k*@h?As9eQ#2JpINrKPlz3Z^z)
zd8SZu1lq6mIeizhPfpKDQAutz<h9S0V8O6|Ny(*TsS~O%C9S4F1)kLRD8!^XA{_MA
zn72$5v~L;|*Pyge+S5s<YF8VO9cuqcPN5laAV(IRLrIe8T_aCql1>6h%CG>QH-rY@
zh`mMZ3#>F`$o!u7*#Sc}W1*UIU>iC{^dl}_HCNIy;h#8baCLOMT<sn(#}fgz5d*eR
zIwZquME-=l3%qnNKP}-L)9On43u;NpG=O_ySI9cuC;5PAfxnEP;|~xt^f9A}GbBox
zjT7%Z(jV&1<X}pzVv4D#3Z_{|_(!NZIn!ZE!jOBqHyD7I0`HUohfPP8HG+_c%+6{I
z=5~jcQfbSkX<*4j<Ck!XcIB6=MJnA~s91ny3UX*vg{YGW#eK=l%y(T{zeJl*f^#DJ
z73|LsI!q$?$h~n}RzM_848gcn83K5%1xaQd0s-5SJQ8X1**qFLnI!%d<Fn=~!(COq
z9@pp0PwDkvu~JWy_!B}$S+iObJP<j7N=WvM_!gu_mH3F-Y+1+fMkWQ~J7X5Y`Y?0v
zo~8!hi8SCMd<1Zc_C!=VvVX;Ok*WI-%RoPv^-9XzNx5#52&E*ead@CB7rA9n<rKC$
z#l2;dL{NA(R}CWB7=sbk0f2Znf}JaN&XX)<gp}kw-!#sls#cV97SgTs8-!O=(_nJp
zQDs><JGDe@fj)jJ4^x<|h`lSJz6ZVbyhsK$#&IlJ8WrSXpGn`KPojj7PfT*npbxW>
z28u$9sX?Nugp=kfQu0)1KnF3~@}1M-4kdwzY1@({lFB3b6SIZWM!I2}@*H)SY;4^o
zsPA<chk}gEBS2<_9TYLs-=;TrNhP+@X21irxLW<k@)0T+my$uMT5_C;wd1asJU`Af
zN%qfjO%E@(Lelh%{?;W#&vFcP4uv;URA-txlj==%lXNe$J8Q|PwE=h6U&~H=`s*pm
zy;)P#Ra**TOw~z?QVY(Seuy=1OU|xQ)kiM9_!pE>wv}5i0|Dh7lLY(=q9)NVnV@8L
z;4_=X@-#K-bjB=9m-Zwj>hzWJb<z1Xn8<M~ig;hjJYvqKqb>O&UK38r_1p{npeA0`
zYLcKiYP+3I1j{m2;ny!&2O=m6eMy5@l?zcPa3cj^=zmigN>ALM)A@yF(HfH^Sj;u9
zRw`;9g1sF=?J7mz8kwMvt$@_KYEq4#bq#YF0RqA%f2cCY7UqI%RCbnQ+$KCtDI_O4
zG?qPwKa_t%?Lf*l$39>>nOxXd)pEEl>qq#Hz?cVh(04)FGjO5B6Uh(2!N^?cwYpf7
zit*sHX<uSL{JT_e#knE-bG32>#E%Fug2NlV=&^9;M>Fay(GWOwNo(isV0H!?^7OJz
zZCv!d*5%0503uQ7+NoVpaB@`%NJYM}+a2vw0OtzA6f-`)hd6*4yU8vJ1%BSdB_+e1
z;I5?}!_ZNfno8!#7l%Diy7uw8qu~HjZJyaN+9J+93?(Quh*ahI<b<~qsdQwvgLo2n
zCy0(PvzuwgeWS(g;fyRBrwyhlBoT0jH_6_m#_ZM^qs(k6_NA}^;gRPgbG!7L5pw?o
z`wvril*@iIBb23*_yWbya(2p2u1FJ-NkC8DB0Kf$L3OgNi?xus2%dRdnNAb-tMR4s
zwP}2w$~mzMFglD*sOaSh^k!gR(37=mK$?u9-}%ngje2H!6YY4+RT!set?6t<lX)N6
z4VgQ>y8D$xcd(crn}5T+N77=4M(!P?Skj_Q*baFqHTSm1moTZ<c^Kh^FJ0M7SO)c_
zg!wQ=V~ip_y~l$g+}KcU2pO0rG+%}#a%*q_cAp+8N>LjODu*j<W&+QkegWhyXS8f2
zsjESmzZ!HzU;;8&NJ>EUd?^K+K3sM%zI7WMISUvOeLUq=$u{54h3Wbck@jUfr<q_7
zNNzJ3(mE;yj;*fLjH0nzBtYYo;9|pqXWbNQO+Q8Wg0eD&HPSCmd!2z3nqax4>B)Rh
zy7GEY@wnJLEZl<G($n~@Wv+(Tp=LJo!D)Hcu1ZfQLJ%_33yr3jTWFh145=JrfUJ(2
zKH)wA7kyJqjj)&o<w0J^);V>zg8Iw8RX)ItsT8U<`{@X}irRJu6>B=ZGf~Rr8%dn5
z<swoq9&pb7Pl-cFDe;cCMqofSIU$@W6$O&#h7_Rn2j_=xI!WA@DCNw47QA?wcM?QC
zMJ2mDz~>8Lu~+Mm;z(5UU+O44<JK@X(7UKX*WokJT@aW}oi5W_rfF5|OOG1yIbsd@
z!4prgZhzmLfvk~?pmd)%o%%E*ZVvt?pIp(nFbs8Fvibl-<5}{zqkV~ltOjp>HN{EJ
z_j|>?ZNsFaBqEr0Ci*$Ut$fUFZrTF409~&t9F_p>?bJ)ik|71ZIOf`vWmfJ|5+`~S
z8L?X4dQTb3wa?RuYD=ueCL>PQJ`p~y`rv9Y4K0}1sBL_4aQUDlWzVpCC5n7ioK?Df
zfLD-yqll%N5pYN<ropTsK{FJSqVVA856KDMRVR!jIj4ycWTNGzn^T_&h$1KgnvE@i
zyyK!_rC8@X6l_vp6a=pRB2yQAQ^^dK_L%A!V$$Yo8$&FB7ZM)1p%mp*yS?@ylTkJP
zCdJJ37|KFX4J)K<zrxBMfCuhnvW-*(UiC~cQmBxllG}T!HOPoFP>F(D@Mk!HV0@;&
zGq`Wi9!FG6I{-Y(z1Dsi0SA656;WZ4B4;((kScSd?W$yZWq`8HlPuFkw-YHw$st#K
zqq5Ss`MYpPz*#&s3h>a3m~i4SQZCAWf`oU!8wN|43+L!AnA@u*@+}Qt17ZyN&3B#Q
z0OtuYV)P|ddpU*!q%*tYD9@3rr1Y?r;Il<c$7cS@#4w1b9S(lKl)d|fOt?6)a^n!k
z@e9hnVR?se*Wk{!X*uU62PP)?E(&&l5gwf%o-EGHG%fNIE)1Q%8TJ-oMKNyZQuDP5
zI%SWZ+AT*ovpg$2xAW|bvj!Lc=n_o*;3A48nLgnx%o7C}-I%y|;uWX$oPiS1zl#~)
zMl&@m3xdbEt9jv3JFwGwgKiEFA!Ts6n8m1biVknrMu<f_J03z>(5-;^O{x_09CXL%
z>a>N0_206pO6_7Wd?Qmo=^mf-@b4I9MUjm*{&~gRg??d|bo8cA6CxE&yBr&vL~FLs
zsXK<UBvU^3gKA4)*bIymQj-xw;LBNWHgyAqyA@*=*%TV(K2gnwFhP3O8g(9fwaT%(
zU3~3`DH){sJ+J2~o1PxxYzKNMCw=Y|Ig;wM2Rl8QhP+#C|E1JeIWs%oX}>62sE3}s
zhcApqERC-un#OxuEzsoDoH0?=2RPJxZQ5Wo0qX_ep*NWr*{nQ|^>HfAepsewLzXZ}
z4-9+e36AHuR1q9W%CZtRx}TQ@7G~KE#N>dUz<e0}PDN*ddrq%@;T!9qjXJh-1WUbc
z3l0~qEVwwiY{Hu)n0Jzdea@sH!2vW(PNR8Y!i!`qq%Z4GB3`No9TJ^z-}$hq(iKYE
zD;2Dm7b&}O9Vn%4E-#_u3GNtS%|aS^9fbPeWiA>7H~zs_`C>dbSIycX@J(tDR;{+D
z9-n~ui66jC*mt0ens}fjf=yVAJo>Ae5S2L&t=8P?oO}Kuj6l>_R2i8pSmi2KUAWJW
zZMI1Uak6^Vye6IS88JZDwmPSSw;emdL4shiym+ch5G*;obnhX_Ob_o+di5ieIhr~k
z*#$o%!AIP9J352guVPHOfkZ&pIDwZ#DpoTFAG(=gFYPXeSDMbc!h3c)v(DA46Y~zM
zd_bBi_y45r_2WCSs%SO^UmPq|WVdNoN0e>%&_p{olm)g_($ST<ded62=JzSa{5amh
zx@6oMQ`+dD0W`5IFwfeB%2<R)(|R-CtMU!CaVa-p<U}obwNSD|rkfHcWDnHK%UbBj
zAp>o-`?gMsg*oYj9|Z;ONUed}8g63n<fqdmcw0D3xj2bFAgp%>!}7vN0M~mgU}XTN
zpI0FGPqFX5-$F`7yi)gut6}@TzOl|(56tN!^}4xI_~*yss1!X_{?|eJ8V=G`(zbfK
zDkU#^`|_?Uy?V^W7jC)`%uBUe)lRgT!9JoG7#P`%Z=2t6LBxU<rc413Z~6wJQ9bKE
zh_EZvX{4PhJa<NF*F2#^vKr*w*kfV281~TB7?>k%Tx<+)AlUiD5bm!Z;|7MF0STk<
zJ{P1LKNBMXYPqQGfzOk1Q*N;l#*(El>nmE82*F}4XMOLKzd+S${~MvobFJQZM>{AJ
zqo?Vi)Y-X8vaY19D$xczx77}Q;I6t;c>-5;4qMfZ=@m;Zf^&rXQ~4y9TU8hCZ43!V
zE};=FC)`JR7!vM|w;qILsA=ez5--1^yY37XTx#J}V@}u=P}6k7Zb?!Q_{_iIMtK~!
zrOI*?+hsxuX`pGpLf-*LlNQpUD1?#H(*u&(aarTtc#16frPw(5IrS8pVZm1t4w+`G
zQ_iz;%Ea8)!HQz}%82+pLnL17SQBEm(G2G|E;auh<`L>&6yRbAUKpnm)g;%E%4@rR
zJd&@G9So`EVsM8f$OFs597&uQ-m9LDe1~hT^|eif?hcOEq+Y6of|;<S{8e~YRo3dF
zMHxR}B3L02Fff@g+N89`o*sakUxd7MkpX@QJPi**dH6}7HL?yU30pRH*+=3*plmFQ
z?x4O^^$G}!QOtmXGX<ewVTCd;aCE8=G?D)fCyfw&Gj7tl5E3+Gh&M@V(2*@0!T+D@
zO|RhNeJBb-{0zFiaKc2B`rqVCuLO9dBOcO#ec7#KMA{@9-v7>&mfwT^Hf34eM_muP
ztXS{ZP(Q2mE8durlI(r_;;YB}pl<Tj!(8n(|7v<)c-yPSTxRP`jp2-c>e%gowI(0N
zMf|1MUgca`zxw0{FMk?5Jx2CEyn(Ptza>d7UusMU6r@=cYMnFxFfQuisCF&aeKND1
z-w!+D;S2QFtg_?bB`G?>SjOixrr@wkNK158he1gk68A!8%3AD4c!BV-R529dYLA!{
z?vAv{QF<m_!tJJkYFoU*TRmht0M`pKocd@u)<Nrx%ZZ#Y?f_g|<djfF^|;$dTPWJ-
zL<eq3`{4B`Of#$xN&2P|ra8Hzl-<UcW<?3$;gmInNnEO218xh^_33qXeMkNeYp=M0
z0yO!A=Wfuysm7FHs(#h&WwgTKX_b!p+X?#R@D47(U4&;djc#7Wtt{gn=nq{~{ih~7
z%eTFGQk1Za=wQa!3MlEJc8bq(1I8D7F&0I>84NFR5c!YmnI%L8U5X9dBMR2wx^K^J
z<f>v~L8+#Bsf}P<YY=k(z>%UXdNUie3(ekX>~eZXlnH8QFYfK)dd5d(Oa0TrB-J;C
zx55B40A3HNn+j-#WKsuZfFRQ9<}1)kE|m{bjGnUY)oL_=*`Z8F0cE{qv#Dk+*N8$2
z_}>IA1y32=s(*pv-(%LU-q{;vuL!Kc6sYsRriGvpxpaeydsqEWF)jQ%YUG`~h!g%4
zc(^I#NXmFYw`l}ceE!1zRNA#{EEqSYO9^9RY5QxA#n_srtAVFj*EZTi<>KZ}$@+rW
z2ymPWBASr(RJX0L&dbvMbO9xm7ysA9k&CpxnxI}U2aF3+sRs(xvc}TTM?`oj?C5*&
zIdDf=I})e1wx&a(k8JmdPAp}d1_o6>MM(sP;t=psLDSJW#D4}IA$$owepCiq61U*`
z>v-thjtV2Y=FUVjvAyz=49Lm!DP6cDw!5%nf-Q14K!hg2Q`KBw6Rq_}G{#T{gBUp}
zn`uSh+Z>kEyIh8*oAzAjb3~kz#4ZTF;2)*|EqNjF3nk(oAEAoLAA9{jA#he$4orXy
zqGm>EtH~VN89CEE;)1w~Qa|}rtYt#4%ezsk4Vq0&<9hWd7qE5{p9~<^8pXe@9z_()
zgxLsj`JPjj)kj2zzZ$+GB8zsI`UPMp<+vy4FYI5?9A0JV5<S~yD_FV0`FcMkE2C^g
z&EfsYI8j`KyxBTtU|RM$4wy#Y!p%+cOl^P$43Vj+*jG+B=pSf#UA5l;j}vvWPKeN2
zF3Oea)i+IfFZ2JC@-_KM-h$(a1Q?AZe<{L{La_*{d{k*IpwU#en|d?Cbb4}z>$Lv2
zy29g?idYuA8*?kfPRi?l<uu5yQWhAy;FsEI(IC!{(DF#7?>d(pOBB-bwQ4Vr!;!gl
zt}tM#LVM~i>WTeR9Alu-wJS!`gaydKga+Rq6ISJ2A%!|O+^!APP}l)6i0U)KHJalx
zV=<sJ3bu0WGu@RHwn^IS!6chxTWhL5D9bWn%ra6}Yrb>su~XFkc_}+p^YZ*Jqkswm
z7}Oq(-A&gP|HRhFA2jIc@9B({di%LuVZ~0Xa-}BqXilF1L*x0R0HFa;fLZKz*C&aY
zBt|`=pM40OPan1*Nf;z|f1DFPwpAqGVtWw*OO>R@Xa{l_Kb1yF<mf7Mc{S?lzY%W^
zTDlGp$p958jIHU-2`j!`T7tp432lv?W*Uh{RFzg)-J_H$;7(0oS*<<b@V^F^w+Kn6
zo#)zliLPgGNiK{3Yiy5lx!~@}T^$6spH-n;|4qh6W5w0C7exGK>7a`Dm`muA`z!eH
zt85Id5Qt6XH*LHQUraQ$UCV5w84><{Gd`*DZjw1qm@bRqPA?vpBg@TuC4g2^`q!BD
zbj!Dm*{|Gm*r6Z|gl*Ta)<x$s0aZ*X0|1b>3RW-jBvNYfFER>>r7*p_%EXzBrO<3i
zl6O0p3)`sZkhK<OB%Jd;9iEi+5Rwt3ikRVLl6A6l%nxj$Oy&iW=L9Y>PriqN5^*rO
z+9#EKnVb|<(CA!D&6F|90gR^C^br{m{)ssgY_dKwAE-_h%zDNMQ&NQR<;K2Pw}<T`
zLq{=*yl~3LC44T-w#tAe14!W*^$)Ytue_yoA@GhCqqzdhX*^IU(o`~lGJD(S2QQe(
zCh~sh+eS&bsn8SLib`Z^<qw08C;G@N1}~PqaEmyA+WbO1UrAI7Y++Kn{stLpwfCDE
zx|dBWwx<~e_vE(b2I}APXU=qLEt3XTo+9-1a@_6hZECSMjly!{@JF+&PiG`wi&^Si
zs@26+&uda7k{O6SJtNi+L}*vUBEg^>H-2s;5hVg%N_whxm`P!}xQJx=`)e<@cb3-I
z_m`iq+|Iw=*xs4=9-qJ5TtT}1%C9H9#>W_W=jC?g$5*`G+Su6H-&|V$Zs}=QM(Glk
zw^ljF&8>}ZSC@B6hr?u+ww9l*y{cZyr1$*%<<7?X#?s3E%FE}QrPIi@AXSU|6{;P9
zlufERGCc2Fuu!zZ>sG@c-l%(Ajy|GWF=S-uQsGnTSBmEazeX|%?G3EBptmD-#!jz@
zrPg_Rn4ae%xcnnTTO-E0*wxxLHehS%r-{x0r3M&gB7$2Yx~wc4k|Z(Q?!fG$>#5n7
z1P-Fk2K_cNydc2tj${+)cOc`Jo9L%A{>w-VghF$aAZWvdsUrAmy8bd2;mLh6R0Hs+
zGn<8MnG!G0>1bIcn5xJtO32)t8_ru*EN~=XULkk{N)Yd;66276p(@pS#Az0g8GB3>
zs$+U+@9M_R3^`R?mnBE$<{`bpeik`vfnR!Gn|_qf711@jYj7k_^&k<`?bt|1Ni{5-
z(QNjt`g1s<)8J^#NgLGq7-hWSF*2)t%u>mK#T!)zJx^ql;((UwAZByYUcg7Fo>%|W
zyFq40I2fsDpfZ^1Cj7?UPHvLL#riy4YhdbtR{Ibc@|wi6ef&o}|Id*W5d{~uf20v3
zH=nHh%Ccm29ZPbCF^h!-hyy`9t`O!D5Hj9^7&8o{%t_P>5I{vgY`_p%<=j~~fnZjJ
z{I7qN>b0O3<3VL7APv|XwQnTQ@r7ezPHY3n!Lu>5KmN8EupemfcnCXvDAhS-|Efx@
zf%dX#;UkoA<W$g?vZQH;?*)6aN+s6^!RdC&Jm{hS(mWgqOe0qET~IjEPDVP7NDZai
zSzPv=aH5{nB0OZk#C{vjpHP9=`#gdbkJgTj0_W26PVJP*We^wI$LT`u8*yy|$$=6X
z$x#48TW@j=Q0~A*VCSz`Kp5IM8=}zwk9`Yxrz2udaA1uyzj9}xxk0{{!xP1(1d$9|
zfHwrAc3MILGn9z}nI1vGs*;0X!r0*(Sh~#6r6a42OOOdR&^{n-9SNjuH5v|4Zo`e)
zM>AS$>t)dhI>8wwI-e`0x)5ok53}lkMjd%FYxCMF17MlbO|mUi^W_OEAMyrX#mgt(
z{5Q%A+d{A<DF|r!1c|}b0=CIyNY4)tBV7yJ=su-*koNP#fC`cyF}+84f_1mpfu%4>
zKdi*s$-HjnvNAvXY`z2wLWfBtoex~abvl$X6AI?kb!rCKi4o2qC|)IZ=}EfCS^l;B
zUXI2V`17fm>i_#Vzd(_Ry(JV9s663PQ!;r|;+`Ta2z!zLB*Ka1HRUTE8=ViaD`8g!
zs@HPJZbAV;BE)o=_1*tybkTA4gOk!KJxV4%A-Z|IzYBj=$8n6@y%8XN!`wVEWMEq|
z<6lMGJ@*l`hhQx_P@+Z7VGag>=lU2YAULHY4aS2voqqNkjsDAY!=!}FN_8V;h9>wg
zRb=Q2MG`mtX;K=&?f#y|nkr$rdi{)WR!y=JpxgBq_)qdSCk5L;l3Wpr0Qsm)PAB)t
z-Y8g?^p;UENZA7LH}w7!mqA>quCA7T(gFNHH0la*@86%t{5(b7NHVH^%n$>GDnUnu
zLltSaR=5;}LC`<p%*1Is#91hmAy))SrSxtcjV@1*1}Gbi4uzb!=>E}i|G^~go>e??
zz+$4ora3;vLkKymH0yvcxzTf@{2mZRy$T`ejS+ksz0gy*K~%`I^CUDGCk;-RHpS41
zaNu#0&ofSE>;45K6GRJ%vz3{fSx-7FRtZWb3dxQ{92D0?3MSeMV4c7g8Ne$0Fpo1`
z3N9(w8CfHf!&8j)eV(|Gut4#i(_|I+kQgU|Y@9=QR9U@3@eHR!FHHPYa>NB?CL=vV
zTP6aPI`X-ACEFc_lBtt;l~?w1(nicE-qDf7Z~{El<?Zymvw(pbR*_Dm*OH`_b76pQ
zkZVlDP;7R%GZq)OSC^oZ-hZ+5d=-vplbSqvxxT)$`s0r5?dZIRGmA+N*(r_pq(`yJ
zf>iR03vxOW=nq^4YdT5%<n*z?L|Q}iy9_{SO@Ge`Y4&utF0}MD<h6z+TF~aZaMe==
z9V+YkYw_8G50m$ndrcQS7`Ii!^IrUoVtS!mdU?p=vZ&#6hAA;S(V1;@39Zd(?(KG`
zcciWedoVBNO;*WUIEye$E#G2oV{dIEY=5qEudUY^?!^g&aNKJ!Q@H`A;`9D;S(dC-
zab{1ttQdjy^9OKJ{!tTIgf4cO2fV^YCem!}*~atLO<0X^Na|3`IrW$y43ClKM%!$x
zuLRo{P8ORs2fg+@i-V$3!b$~Ud9pZL9|1p1x9u!XE^#MYsAWF5b7TZw#V*l+${j0{
z;gwIZ6Bi%CY&?Vj3jHe$JkBi*2ff|_{C&yZu=;fgdt);07Ik3~$B5K6fCfqz+r^;%
z7)TH0z$`U(9MTE2$+6DQrm4}pLcb0~P9TGdeH1c*^(`uFUeooQw%&kwAsoCUEoLgg
zpvj!%y>NQkTat>obp-pky!Q=AYO+TXX`MQd{6)qTwu4Fd;Mg1O54%(f5lD03xsO2^
zr-H{#%5v)jnr}dpFz>6>7@H{Vdp2hA;i2$?tr|YGMU<|<{Ex2g`N?2-$UM;zA_8yy
zzIJ#5b$<tmm=m}=g2E;4ZMTEW&R5R9VhrnG;vl1b>kJ7mCmIs%7*venJ*2}_vEOP7
zOGrZAs0M-U5IMhd<Vm&!d~RnrV<;xEcz)<KR{IF8&ihPes7#xl3#AVeQtTEd^AUrD
zlvaYO=2+=vxpyVff($x1JDS6VQ-h1YR)%Yidk_X|j7p))i?AOIta>@@kxZYA$7hQR
z3se+yNsEh4?!iNZv_rY}qpA;}1}c>vz~r1lzK(ZsaVcB57VyT?a<E(xc`ElJACa%X
z_mp2=8E@b#2Md-TiJuk|f`wy<^w54<>W-0s7TGmBV+5E0H!^{C(n1S?sbd(YCqAGk
zZ^bi;x#=&)m4W|m0bN&6vI1OZX2;}36WP$I@keDZy=cl>BCCEYG&~ug07bqpK9p`m
zpd#eSV4@k~#d=`&Pa1Rk+pAlzR<}ea_O||d_YS$s#ahX8z_1(y^^{zuRu9t>9)Vvt
zA%B<OPFQ>BrG0v)tAm}(qpccPKd0JIsmy1N&c9*d^s-mrdIL~pyG2?}`?QgA7O;qq
zfj5i>T_?PnWCf)gaZ$}uc`kFFug}IZ?xg>cw9ad{OJnf7sNV#@<CagdnsJ>lVsdW)
zgbJ|w-h;CN`nl><<z|29wb+RC5moS4X2*4yIpcSLC0RTL#atIwNW^AsWv7ZLr4M#P
zNv+qwOWpFMQ9Kxa!#BbQicy(<4ViH8j`u~^)~NsS&)qxCuUpNd*MEF|=g+x6ad%~h
zk%CLvM>Xcku?gB_<Q(Q)Q$2=|X%3lUkVf*rWTYym?Pp8PyFB<vGYE#7oGUVXqELq<
zo%Omia;5k-r@3@|lO)DbwiUd!@SRYht!{0vZM+Z~oLxlOj6-=az4bglW2?vPmtZc$
z`61k4ww<+4N3{n2by@?DUX5MJLlS@6CqhW&u5d~P)lE1)B|?0b9~I2BV=iwVE29c`
z&I~pNWzTwRG}bn@U740Dv^o^cXv!nx1c2yh^kTkNmkW#6QVBf(W3GY+${#k_g^M9`
z%E+=u^yYA$K}ErZPe;v9N1uY_%h281Gj!A#HsV%pt*);wZLjXrj}PP27J;hu{U#YP
zL1<_Z)YapUe>MV6fBEOF@caGyir@9QyZAJAY&e2`^cv1dR9ZMp5Es5LJA<9pF_!Sj
z^$rm>H(Hz<zW&*lOz=gI8t~taPKtY<j-1iT;MVN5kt(^C?Ha;fC=UXKhn|VBu5W6y
z3x32F{>UDfKSLn{+0<DoJhG4E(i<U_GZJ*+^a5-0EP>*@`vhmX9;_@j=Tzt6!qYv*
zR;h9m^e<0fYA~}WujA*c_h>#k9GqP?t{4X6rpSJ~%Bd!ZKRhpmsc`K=d$J+lY0qXU
z3jivkKu$9$oB2j~f_Ac91>&Q3q)I@LSo%mFj>Po16G5`zr(R7bXXr=L_(ZZeQU2WT
z4N#k>-KY<cIBqn8fWt`jESHqwYs;ta2`riZ{fwARYF-O#CYE=PP>7Uykx23x|1WjM
ztnYSh1LJRRuh-soTUai6(D`KCT#8KRldoC}C!08UmH6KP;I?vA$%sr2SEAtSCI*Mh
zVsK$e_I6zpU><X%W^OJJ%^Cc=ra2n?Dg^yxwgoabg@JE}ya={*G{amZ{Qq_h6~X|v
zVQp}M%%R=Uo7=R1z`;i;3HsYhC6K+k=lP`q266~gF2VrKVONYG*U2nVw3weG_d2gc
zzDOPF6wc?z&<oGQ14Du~<p~VI@&FynPO!Uu&xc^c$gL$WvU%@pz)U~4YY!egSXq6w
z`SS61t3S;&`Oo9kr)w{2c(T5>T*Fg&IP-jMZF%k2r5BHn-~4v+ru+2ki#v~(zF&Q^
zw6w9jbo|v4-!C73hu>G1`j1zh{rdCv7o9slKOcN||M#7hFOO!PyjXtM`|;Nw7nZu6
zd+qM{tM7kYx^w?*d$_T4(0bngedERblb7qCf4Tnd{_)-I<u88vcJ=wIXTv-1e){e6
zPVa~AO!t20@X6}&yY<6gKRf&R`QLy1;<q2p?`^z$_xal&e!u_iPlLM)t%YYlwKvXx
z__}kycd`A|tMT@orB~lBefILJAMY&9oc!E>eERgPH-2`q@zu%Zch8RBd^>)#^YZuQ
zlg`fJ!EgP~FP=W@ef9IbZDh?@Io#Y?zxb{5?uYMQ9u0qbcYgP1X0Y(<{U4rOJQ;oQ
zbo|}t7ww<M-5-aaU5?Hk|NLy@{^i~C_R&|JCr_Rp9^89%f3>x~F&rPA55M^F`&Zp}
zziynrnfa=B@%MM{MxDFgpDn!UJ^SroWoz{7_PcMt=&pb9<nz{L|MB4$tLMLs7rr_?
ze6h6l?#^)i*Y%T)VXJfb>*t*p{h6=2gEtq?zw7>Zb~yZgedoJppKU$6`|a}i#;f1I
z|NQ<J2j5;CY>&?FzPjlCb~N7nzT0{6+4$XWm!GY6pMCr7>C>Z`oySWDKO7=k<>B(1
zuYbFD+Wu_m^yJ0C+h-e}Z~ky_@b2rKcgtTs`}W}TFYf$$`9tUY`Rdz?`^PWV&Q=Ha
ze*LNS^~~AxS6>{y`mMj+`s`xk@XpGYXUku=zBuT2?i{>bKRms^{n^DAYk%*qJig!i
zd311p>Fbx{t?u^bm#wGUgD+RUyf^dJc=@Z(4%d!<K3x0q&DU>#eY^47v(MUJj@lb9
z?_F*#_qTrf;^N847xxy9zdBgyzdRfM{fqw3FYZ45>g%T$&p&T}J~Ld{dHLnz@4sJr
zH~RT_sr%c;(;wfyded7wUV7erb*FoNczkq!_~P;UufKL)tX?3RVdv@BPwp&_zyIvZ
zZ`Utct1}ndtLIDSTc@9$ZT;5$>hV_p`-P|9y*ylAx>#LW`u^cVJqD{URtkq8o|ONO
zy?5P;>&VhY|ML`SHLgrqN<c`y6y=f|0g@;qfdR>OftpGn1t|fEGZPTL>~Y3<hWfdm
zuAk(5-@L?SU6Pb6_o>=rySi<1tymE=X3UuPn1Ky=!xjshSN7QU;*+h9yUFHW_Gah5
z{i1We^zzM@Uz{zSEuG9yJ^Sv(#^JYj7XRG+>FtluM1SbdUw!xJ$J6VVCqJCd&z{Zx
zefYx{=SyopZhW`>^m_35n;-Vye)sdEtFN}c{b}Ry<ap|G@yqAu`_H=AtY2Pj%r5r#
zFJJud?5lU*UHtIv{<G!nALlRr{$Tyf;j=p%ub=FnkG|b_^kl93)!)B5oFBe8+g_jA
z|6%mQ7i)tz_wHZ4-5mA59FAtMW`8`te}D67_s?4=KfUeswjMkiKE3zm(c6R7w<}MF
zo4Eb*?>~1BUv%F6IQMMobiXrqv2pz51-#B1Kd(H!8vOXvqem;F?=Lt0{N-r=a`gJ$
zSF4}@bUFXk+vB+(zZl(lbM|KK-V0RTnD1SD_xIORXZ@vTk5|_(9$da1e*5{;cYk~L
zZt#5W$@=M&)2o-?y;)x$y?cArd;7fCAN~Ew#^!@>hcDlJ_5AG7VC{!5`%f;OO`Uyr
z`Q)J2ySVdhe`$Vj|LO9h!TECc(aXd2@82!|y#4v+gWl@gxBa!BA0N-3fBVzb{*QP1
z`_DJF=J(&+nLj-G+tlFY{g<Er?ako&!TfjYOOKx4Kltjas~7h+zI{DfUwbpZvUJ|P
z`tgg)vy}&5ezn>=@2y;Xxw3fp?N@)ldNaRvx;>cM{Pw#STb)0@{qpGO(c2&Y{{3Qi
zeRh8FXs&yG|7d>y=u52bo#E4!pFY1nTm9ki^ONJtFGfqBZ#+7D3PyRp{O#1}WPbHs
zZ{@7}=4}0R@#iNmdzUNsetdFx@#CX6&+nYRdAI!X`m1lBJQ}<?`}yAAHjWlAj}EW?
zHh=iz)q~C9!EAR5Y~b<tPu|WxefCA?!RxuzpC5g{|F_Tkqq&XM-h<BR+qo}S`=duc
zEPsFX({k^-?W^mHCrj6-pRfG%{de<^mjC?u)aH}Rum1dKWBJbf^4i(ou7_tw!^@YQ
z7ZYr<WUH9`gb0Mx>ahZlN$|aMJdn!}G(iz*sVZaf1!|WcG5R0!*Vi{zo<q^uTU>jx
zz6Trl{#Os6bbZsPcurm0S`>5U{W)@hRdJKyPY)3mhdF(S46uUhWw)&SXxCo6JYISY
zwL%pQ?Kvt&cN52oXg}AY$B3>+{sWKhsCxu=3u=*3*B}~#Rlrk&dxtp4!UKrmMYxSD
zJ_536s(kUF+<FIBD0YZ2=9AH^hEfRy?5~Y`V(L386t<SIMEF~hqn?>!4ig}OW7^jh
zkGW<T!TWoS8##GA8+Dg$?eS7Ho&du{nr1;MJ_HR>V-%l^U{a7s-Pnx)78@2F@D!0L
zQJazeqDipSoWxy%JOkeOiMm~k5J5|%ZkenJ19F4G#rbG9l;ps9AHke1leh*sosZ*!
zkfj&&Oha?U;&}QiJYUQl*?=}(@HLi<vBYs8bo`%wMmk%-3WLlFHq($&iVU*{f?7~?
z0g{-b&k)j*^m6FM*=g*Ojb~kMAUFs&)u7uRSYj_NOpBP(LX)_zDBCE(cGT&eW!<`?
zCAzDGJY8FkjuLhA6m8xY7+PHf!4^!1Z@1kC<o*KyVqQ%Ag$*HwO>?H~|473Q#ZkkZ
zC5sl|bfdHw#jJ9;(o6!v<IHAYwVwej^H%XR)vyO+s9@aDOB)mYLijp*wv#jjq`n|%
zp-~yPTA2Z^o0HS2xQJq`MYxgdF1OimBtu3A2{8u`)p5w^1|3`lhl3eBJa@NoZsY;l
za2yLev{om^5)&6ftGNyaYpcqr5TiW4x-&ROa(uBr)9cSF&l%1fk4{gL)kT&eT;QN{
z_4P6is^?gyahRkBjsSLmYkDru<ia{^ZaHH@J|)P>9a&;kRhnAj8^)!qr`|zPuDl4&
zhnGCWk~sG5JQdIY)6c9=r^%$l85U3g71CJ-i}G8kpw&bY!__fmN-Oudwdz5gx$}*@
ztk9Kj|Hk05b&o%U<?2ogL(0a--6V@=hnNX5hy!FRTxeXveA$?7Y*`4lVKhQWajM4e
zAx9Ah#Q3EOtblPF(i=mC)c{|#8S#P}ag>ojF=C4;L|7P~{tP3B#5mjCHvDkalBXi@
za2GKCshBS`ieIMWUS`T+{!h3~Zi0hiSZRVBI*W?YG-oxYS(4O@8~KhkMy*oDi#uCE
zYS<cRt+NMZqCy=`Z{V_6IS>Ox4Hp)Y$E|4_Tm0YCG)~Y4fc!DbZvFcV6xVUkF$=%V
zT)01MxgCUmsr4pM;?;3<{Q(Oe<V>rX^4p;j5_l}+Mg>W#W>Fy<T%@!#%RVHLc@ie3
z;!P0##IHPJ#}jG({-ue^pteZBZfqdMb-_Xx1(mR#PX>A69a!NQ<YeMb2~Cnv&nbmE
zq^AgDidSHY5mNsSb3>Ta{d`DEdQ1jNhkOhWA?-ljeBq(b*z*wzfe?3aBf3i-f*`2&
zTaW`BO8xJN1gM`G!M2VV+0FC^Ne|Ikh^WHIjKpbG(La224&^AU&aDX4t7iy9SZ5Ph
zrtg~?&zR8845K-|B8GhOLC*DkH0Bf&bxJGsU{%0W9lnC^-K>F0_2FD`HTi8<PHB!~
z#fF4idW^1*kOs+x(H@{e;oI&Eu+u=S%mAYbrzM{L^WYfzl@t(+pO6l5lKT`%4*5nl
zEPjH*i?oMX9<(!<;fgY?PK7KJ^XbpmgqB^kMiPk7#}pzem;uZ^G*qUeag!WY1WE`=
z@OT$sfdu|fiA3c<7G|v7<6?Emz)0XIaVfxM*w<9;-zx9D2@2jc&2gmNoE>zLW4K$#
zexxvQp)GL|c5!55({f9m1EIgFb@g%K>8{-HlPokUCo^M~d@6*l*G~>t6UO&gTIAtt
z(Lu70c(?&iU;Ehs5{#!pCmG^495ar%e2y@yGA7|KJ{K$T5llW=8I2|uSfeI+*@x^0
zG7Hg|3&E3g4)|J#XoD-@=rwyyH78{sRycE$<v7s}8k5&~_KX6ZlwTwmDizB`aTR;-
zW`RX{_rXZl50P+%TCI=daFt@6$i{SW+yi$N_Xm5U{+=Dq&@V6xVR3eFo5&c73^Ae(
zxdphuO6NF4rhhfwsnj|hhGJD!Mk8=B0V!qpcr3KGuwZ9rn5T>~=1AX~<DUX{i+>4O
zpzv_W<0}#e+VswfH0w|s{#84F|NEI-u<kDVg<sd3-C_R(hW_qGzkgCp&o8%H?!NBe
zpu|QW2dZ)54^!GhryCGp@QuhBL}obK@1R!IG&fl`{-Pu6nj6cyovgPl<XRtqu8OTc
z!o^8G-5aL4)jcJ8;cUQ;hI?m}{8@(@Ig;Urz7+KG!2$NzKk4owZx?nc_Rp@WvEldD
z=$f2^YgGR=0u&dcUHNwbnu)&gOtSYFX0me^Iv1N#0kS}G@$}H{;$YALF~d0|r^^cK
z9bdfP30WU0eqZ0Quh%4PW19)478_Te>e)I!e7DPgt@RF&jlm<a*}n8i6doI4>VLpx
z<sUnEa@@bdS7VNxBwW+;)cl6BC|h8Ro%9C)hn_jD)i`}FY$o|UxSG^*LoN{bBqR-m
zI(Q)zs^`>k;KJ`;`SFj!2@__}Ri7={!bi|&kd1*YB1t6s@Oy`&pxz(2nibt{sqmR%
z`iyJ}C-6|l9nP8TMRrQUc!3&OJS97tX{__aEZIH)ai>oQE{~JLgU$fn<<D-jIV1#K
zMl5+E$?xaN@o@hI_y#ml_?we61Q9Cm7W!z%LS-D5o_=c%ZQj{Y0&6yeA5*~jNOXSF
z0SgxIY4BDKhr6;bvr4-xbJXlP%sNux1haG)+J|m&%LwiPX=yqPa%ip@A)`b>M2m0W
zpM>g#D}XWAM@#TVrC9SR`O#5A!FvE?J?o+H<~b0RJ%`Ir*mdXTu0#6?`le39Zt0S&
zhgc|{{@p<zF(O-ybPH6%P4aAuv4a7|A1#-K@qYOo$W^SQgggVH;U^@>8#jaOJu>AT
z4(W={MhK^|vzoK&Y5h0OKf6Q*T6+i1QFKt%v%13B_xRd!0k`Hm_q|uuC)kXH*Vt?S
z=IJLMdEcVU4-=XuE&I08P#x2q!vIyIpZ3oBgRr2dH;=EI$9ay#10~T;I%+^hfQPgX
zkSPr|%0x6x%^uD)7-34TFmtFZi>!M8;dlTzBQdjlBV<Zl74z3lu;oL}9hgCp@GF;_
zmX!>D6I0vlA~h9>hZ4K_xjX11rj1*N7>jPm-C7t9E0L~70F)!w9h0IcLoX^{k+TN|
zgP<&~7?FEo)6LrGh`j1}x2)kcr7|LAw(-axPD1@rWd+U5Hu9{$?ZaGBJt~%B4Ldi`
z{YQh>x}wY#B^4eS5Dwa_TLXH7rsQXKIusJg({CjizF7k`<kcZo9zDns3z`(%am^64
z{1Jh4E~zH)8!G|9YOq{y8NL;Iy9EJH!4n?%QfZW(OMRAzDUj0DOk*4TsomxZ_O(BT
z3=uAb+9u~ygh0{-MoidqWzESbZol#@egF9+brGOM`62{fCx^_$BChGmcBy3cR>k-M
zAL2!*D_9Z#4|1)xg$q>xDil#%4n>i|&}g-nZ#$~2h~LFf5@O-X2C?kp>SR&xEQ|)w
z?gGI+2EISX_XFZQ91v->JGY5TH?3oYy<$E}x^(jV4or-jC<g$TD79x4G2t{Pn^jU(
zB)#BCbzR^<D8M(JEQg~mZ63CzC*~vDI@}fSA7H_8HZAmwY#x6@GY0#Wn)L>#kUDv^
zY)tQFkp^!ch6aGiwMpFa-h|&ze6v#>V-USc*T$8Q)TbEOj$ReDp)yjOR|A@xgePqN
zWR3-c)Fv3~reo`cN~;wV=To$eTsu^TiyQh3#2>f=9KBxtq=?rls&FW{hKiTGOiNB7
zM21@9&$2==;knjNaiA#|OqrJ!04TJ)4hL$}7Cfv-;uzRE)Pj|M#1D$h=6Ro7M5K$<
z(bJ>G;OU}Jc-D$CBUCeie-Uh*F`IzV$n{_2sJVo%d#C3o%vz(KgEV~lC+xD03>-aB
zHI^2X)9L{Wg;MsY1P+KIE}g?Kdo$uK$Gikr;!7doD%!1DTiwzclwcRhPUTNI=Q6lq
zAruLyQi=kd)?YUK?<m1@GXe!?De5ODuzxjrDKVS^7r1dpOtKIus)R`W_)O@%3`U5=
zC<T=-9K61fAmO4fiz84sol);!UJ3gYn@{;G{u{K+e7}60;z|yASt~Q8OH?Tu9eTpl
zw``^U^WH2;&;F1p-O}N1nJPgg#|2E4BIxQgz<25NSe;QK*<6c{E+><cqC9DU6CqC(
zi}oQyoShR`hOA})a|@ic&;&dYmoMreH=O(&F-cat5J^l5sCZ1vCp5%znk_O!L@kK|
zx`@un6U~J^IW-8Qte8~Idm9QcTR`fQ05Aksk{eiC4!mh_F;W*vqyhqzdS!lZlFtw;
zJot<;VPl|%Hkr-_f?}iVm+BI10V@xoD=Av#eRzXO)KhE8PBRnL;NOO-nl3`d!5zyS
zAn*&NN1wjN>T9h;-zzHO<ji~g(hNt22G`P8DLlTd07`b*CR9;nPm*Hx4Ayq~l_OS`
zKT(VDI|tuto^ewp-L2s&rby$UZu3e=Zspj9NR->`%B^s8?1k@Gy8coO6yb*HDsEjI
z9ib=^ll7@H1>w{}Mx<3>u@)8zz6D^RZg)}70`^OFS5iu2sNcZ|CFS4)B;Oatn2v3b
z64dyboEZ++hf={>zhJJ=*Gx&(7eZ?}@=Mr4%;1u@&I+#@hgJp&5PBTl+!C9eC%mO0
zXkEj_LiU!F5%-|m#tGrmd+N)QF_+Ec1@)(h*K8l?cGRthg%a!<;)zhMX}riFT@eX(
zP!CEWu;UK06pEb>+>%KETyt@RxKwQu_%-cgNtEGCx0kV9zf(>sbKIoL$B>&{B^>oI
z7R5InCEqEvdKcLnTNWox+En@;2quIgGo_A%p-Sb9z?40~hbU5kqu}Xn4!gUbac9kZ
z_6`0c3CurTpvq0y3Dy~c1=GLvkU*RWLOQ#HFw8(oSh$mqJ4{lrvJQ+7AiJPQMHn&?
zlA?%-Cj5)Ng;i3;!EM7$d+mi9EhofZ+wNcBl!B?L){^#<QS1~vIdpC$yic84wIo4z
zNF-vyT4bIT6yl^jC0McwlLmMgyaYM<Txjc@B^Y0k*~nZIbd_kjJ0Q~3DwUiag57?I
zBQ6`tf(FJ`^X1U(Y%RD&eNQt0;wuz10;3opG)b3t;?|eIlzf`f%5X>Kh>Lak>J`}z
z{~$r)e=bd{02|9^un#pW&02*~%53M=ZBhX@Zh!v`T=Zd!M9}I!GWm%^Q-~2DhU?QE
z`=y{YC?&v2+I#Z6`86W7ABg6(%jZC+9dYB93JaF(Qp`Wnbotu|WBkTRnnY?}jGZ+X
z*4o+H>Iy^ure9(4<YXK5NzJk3>bEl)@dZHlU`thj*j-tcxf*3npD`dLj_S(RNxTRw
z7u_8)h>E5eK)8XKS5+X_3@jyAi!zoSNK`py9b2N3K`%6Tb5J8H_cR-dJ;9bzE~m^X
zyfYQG$}6D55Pl;hr=RNW5X_Ej%CbG9f+hP?^qIHJV=ECkUQNlKUF)75=@ALr7x95E
zO!y5k9MI)d;nY*9HI(+I%2jfT(CKOC9dcj86S>Feq>HSwl%UmE@GW8bcmZE-VdXt%
zbE@wAF|6EzHA2ft1X(0qqPIXuBsx`b-e1Yb=49W2)_uTSfu~KXHZMkpJ{eLk3y@$H
z>d+;;IqNm-ha6MsVn;+}UkVSbL0883lLZBmZdzjT*nqE`h{=4_I~*P7R`TNcITUFL
zo0t=S8@bU4q?|0{uAeg|l|aZP`WNvwZ5?6R8VD-LF95UgbOU8&xA&ec{y>h#Ro=iP
zLD;7Y3%%j@;3xen6Ri4CyUm754`t@XBme{wpkMqCcuP3g4ZaeAC)N(4g2?i4-3@$-
z=~rH8f!(J8crA@nH%SnVbY)Cxvfio&j5te~o`$R<D`?yaUK69#JB|B|ufxhZIBQ~E
z1DYH<`k>?XouqRU6JrNU$^n0j`KPR}YQ9B~<{#t;fHJsl*&7?CGXX4N86SyRCk#MR
zX3;$fy~H1hU=vr#CoulUuqG7;Npl~eAh5@D;d{73t%&hu!j^Wa!&E)2<8WroD2i(v
zp~p{3<M4p);yygWecS`8evpKd2V-aK1Vpx1uQjz2;ZKQk#g!tQcY<+w-a)o`op><i
zUTGdWO9ROuCUVSqnC+^9l7Z$-4zC)dLQG(+f;i5Hb0#ZH#&&hpA_)*mHIWrn;2p+m
z1r%T#LuMP3VW{**JL^X>JaFVD`s;*Lq9zR?li(RXB{)c>x0$B6l>#a<@bJC<PSLEc
zr}X=X_zV*X%aNJG_ayg=F-KHf*0^~jU*O~m-mGwIcYyQ91RaIsI>=;!<K4Mvqpg;k
zF+KdsmI$vO*j7(gMZk!Cwb5BRKi`u;#tAjK1mN2Zan>0WRnxiH$7k6$cN=p+U@~0e
z{%s;}W@bh_AlZV5Z!a0J2B(yUGCD<3#3G!6R>+pNA!>GNC52V>G>8I{2Zihny9-Oe
z>l-4<qeiZq1v_zcT!yMha+NHrf|bzkLf;|%mzW1Gde{U0u)<|37PbYKf;TsTK_hbt
z<`T2x1banLYQN(mh~I_h38!FcArO>;!=re?1&gIhwvHVP(s&P?8bVz9aLP{`PTdN5
zP6jF~k|Nop<J?rM$wA1o7H;RVk7^^D3DrX`^QXfk6X_O;aMrRw@iRGArjQIL@{{E;
z6JJI`Ws$vlpXqug4RPq*_rf?+Ep1lTy||o60aVqjLI#kG?SPTw1&JbFSzn^$%wz64
zeYcL@lq0HNHAF=NvHZmc6N4GmcF$W=O~5!p{yeIpquQyy!14E?2F~psn{*^2DGepa
z<u3^s@)Gq3+@~!BDu}E~#V;~%12mjfaMonc?>5Q}&Q+{?RgIMWAL${v3NbQ~M!>b1
z@+u`w$^BPZo#eCw=>~vI+$52kGjmdt82FXAWc&=G8bKEni4`supMwW9;Z#P<*puc1
z1t$C2rWM8_x6#8D3kyTAxgm-U@#txkgT*hvk>d~oM4abw@96jh|34D;oFKrwq4wC0
zanzBjty2pe%bcLeng*u|-`-7<Z<01(MSWZ-F19HCV{#1$qW&u9HfU=&^B3w>b)1qi
zP7VzcrfR*goRw)0_C%9R+YjMU9Cnui_JPIl2}+9F-{ghoaxRVFAVX<6J*y_UE<Ai<
zFL~SzwfF7rW4=I5jYGKzO1;B4g2%Z6?kT%!?$EHu9z<LBku&F)-%E#9a@YlY7}b@`
zy9~&QpM0THqe3c-!1qV*KfQ48BogPFNhn`@C@FQT3rFj0vv*RjokX_ocC;^6;J_M0
zhlfVOqK$2s6ginugL=qyOD8fJ1C?>(J5M*Iby;b#y!1e|_*2p;g*Tv@!W;yziOsMI
z{a=RC$$U}-h{~>Gm4Cw}|GGP1<q=sRo@3-GBOVik!%AP-LNlNX!m#8mVX4-WatQsA
zkcBPyiJWFV+3s6E=*D~2su)m3O}eOuv;cG5SavtE?do{4Djb@Mv24ngs#BMN6WTPw
ze!(`Lgbmmn5+0CoFxw2W^>3(q&@$&OY1AS-sL9-*KN(t&lZ`NJbsZ{UlXD~rpE`g|
z8-+F^{@|)UGKIae6NuBR{@^ea&sIS4yuA!8RGX0x5YUZDuJp7S+mh8V+9mNQnI_AL
zKeK&>Ph~r2C>81R9_G{$`H78HIZy4~QALCoO{@y3q)QMsV%f67P$4Llf1hko?#%M)
z(LEo*siXjgK%k`LsPzMm0kxIW9irUvLDIi2t82i`oll3aV%`Evr<(|U2=$i9Vunf~
z#4YJdfvJ3j>sf>`lt0=eGwC9FN?ogxhT)nmW8uT#Tt;7&Q1V3;jC_70gxslNcZ8if
zW`l6Y($9<eRuJ|CrdCc!vIJmKZbxjG1iCH4PE`&DrFdANbAbw$C?0O9FUj$Zq5)=V
zS5?0>zHBAk2%Di5rt&_a<bqihPme6}(h#0SSO8S8t3q#yi6s{rp-9h3pB1m*%I}@w
z=np(uGo;3F&;#@a6~yO`0H}z$lK9`cI6r4FN@Y^ylu0&_G%!*dmkeYbiT1%gD#`~c
zk=b=bvF{kL6XZ(UEbE61dB$qT5o~w9q}ycig7ys@X|<4JO{K++pvSIq*u4W}2>z0+
z9wd?P`5O2oxlvH9ctGr<c|IJp=U_Abl0@vn05UH3iz#g83W+FF&5v?}nU7-TBdp2c
z>0dBCfPM_B!KE$6#|x6Y1=kAuz{xu9GB>6xKxKW{e*;kE#5pWpfi@8p1r!8C+pLwn
zq~x?Jr>nbk8+oWHVm~gp9>VUniR#GubEiqXRum>-f$bND(KR4_-=L%gwcw9Y)Tn@6
zAYYGB=}XnCq2T%bKAx5#)AZL6(<Qe8ZrQ@eNZ;^A!pcL&T#_6##v}+T&PxR((FYRk
z&R&XP22CtLV>i1-Madds9Rg6QHi=6~?H+U>Tijdhuv8$;3e!11QysGenJ_I;A;o~k
zGKN9$ae)TCb4s`9=H{_ObGEsL>kVDHs|FH!Rz`jUy=ac)1jNE}k8dgafuK=uBu;o0
zjbM5H!2ST|2+c($;nD<$y&j-cg_QZMd>MAwFT7N=IdgIFx;J1JwvT%Iw|f44i+a6+
zFw?k|aPfqLvLCf~`B#BY%P2xKTIdUkH$`;Lz4={2SJ6+Pt@BN8o;*Q*ojn!Q%p0*U
zX3sDwy#bWb^^jKMAb}&bz-n`3I0Z&+EZGH>_EUx`t?wbS#yVU*6qMO7_MU<xXt5z4
zU9U2$hZUSzNLk^*=YMIau7)U^K9dd+=woN08AOd|pcyx?S3CEx%^)7-JSnBDei_tQ
zxR-G$(d|ZSeQUMRIS0-!yMS3$ZSQ056^J*C760Cur_UmcT-Wz|bw*yD$Qff0Hn{It
zHz+X~*{#msD@Z0e#k?+um=jYOpYv3+*E8WAL#&7b%&NejYT}YqTrzies(IZ#fxC=<
z9b9+h$tj8{K{4c`eW+mgFDWTz3+cYK9_cWWKV|jYdL%6>P>;PzkcH#4EqH)}D1;K)
zkf9-GFl~x8kxnygRJs`}Qj<4SZ%+_oY-KVRM5UVo!=OqXY#2J}h~UcIlPq+gNcA9u
zJF|)cKJz`qAVktr2s%k1Ghsb1-Jvz+GjZ!r6kFtKlNp!OF;!vqBL+p55`FwwUgoqv
zlDdV>|NiIyhLD?ngZ9Wen;!LX_!ECyWi{^&qec?*hr)5}2W>o2a_zSe+?zXh%$0%e
z(Z4)Y0+lcn4~?wB#A!J|LPsKM<O799ze*wgr-<e06R0JG6~P##a?UxW{~c|@dk-ni
z@j5K`r^EmL=l_HM`}))2LoPc0W8BAZ1Jah=fd3{_EI&;aI+Zb)iGfm>DFNXc`jiP&
znwS3H|NJLt55$C_3>5ZdTVuF-Mx^1RU>iho;~Hp)M~{y1w+;oH6P4$*rUbvI8(Ts`
zY!YXPzk)8rErj3chQiMr3_OU~($%=rxQ8KAd8!&s1&hq!Ia5GBE(ZwC%@Wvfy<YbE
z7egW=*$9|WASz)W<#vU|`rllv|NiIy%jGIG1|BreE=(%wdXZ8g?KjMUaTD@$9Muq>
zh+VmWeZ~~dCL<^YgM5_IDMVWKPWfEg0@L`2{HBN`(cBuwgnB^BM`uEIP}{bRPu4aP
zIYVGl7@iD-9I(K1>Yz48B#c7zD6Xqg(5KEqQ-%aw)^V!X2iO%_^#%tQC!N9f9n_kE
z&2i?=fPCX7lcR^A+4ApVr6XJ*u@3_fp$*ohxg9dKjU(cQOjL)zfiHek)`*?rP;z#(
zB|j#sB(Z)n3BF{F69n!^z$w9RfGguvNdwHpe&MehIA<~t8joX}&?>z!UWn)vgb0kA
zZoPUV7!b{*H(bUgIjEuP`EXFD<+gy7i>V|56BEJ1ttc#w($Se4@@%ZtKiYiutNdI4
zWEz`vi*?ME=;W!HwfIkIYUPY><A+s>A#E=v3sE@hUl=?d=)?~>v|wi%G@%=6AhMFl
zJh?klMo8%=HkNTpuM3KH)R9eb^QX{KM2?bUGfo^p5DJFgx!*&Kk1mVwMtQXRr{x#&
z(0ivBr(51!#(#any8*_37`}MS#{5j#vK-@)4(y1K>VYXdvUYp!LP^`+Z6w807Z#w+
z>nxCzU<X)qQTK7QGyp(Ay&=lI5{z*(hDbrogI}ZKy8>70E?t!{4(CXU8g*FC+g5{X
z6S89GMB0IvK@=hUBWSkPREN^^jbc|BIfQRySeaI#J>&ximr!^hG?f(v7?B!6H%B_J
zyn74vx=P?griE)B0o_YTWYvp9JUbu*rbzc0N)>UXt>%<l-bG|4$~Kh78R6D>3CZ&(
z5syuglm0{m4p4o!R<sYxZFCLTZKHA2V*Uj|Q1G*Zs09Lywqc$`7@lj98AX$?M4UO-
z|F^04ucSOtaooVxwDRgi-{?87n?6vR{)Xzf%P2i2_w=Ir@NoyBCgNl;DnR8FvYRwt
zzC0MlpA&rW@?f~sIp?;&`7`7ByUQ1+sK&2_d-ZRfbpXg4+HUd!(N8XL2UoXL(OFrq
zH}d~#22ru&k5K)2OVWjR^*>TSq*!)Wi#QMZID<Z1dDlHaEDJ!(KUQUTgsZQIm@Z43
z=yiFAL*df)jN<HoSx=3XA6B>bme!Y7_Et7G*EhGYLtzRR-@SOg@nY-g%JSaQ)5Ygc
zR@iF3+{$h+PZ}W`xhe|~-J`-+eL-FUKbFu93DmgS2dA)5Hjy?Ful|xkMW~VI_s4lq
z6le3b!Y?wl=tV=+)XD<I`n!Ja@v=MEhdMx&<nq@`{@*43US)kdZBnQoD<FgcUrU*J
z_%#|*1R+#%(sG)ZDC4iV<3hH{P^wne7H^bkvtE=en<<R&V29~ul}AG<&8dAj`XUMp
zCx2kQMrc9+QgtYZ@u1Owxa1*aE^U+#tgNG}4~8$<wfL|3I>qkM5kv%hoZ32c>p&4n
zi<SB0WR$?}Z!R!|Dr2YE838%+y|b+#L@<qBx!Gz}(mFSFp}R4YR*un18OOrH#^%c2
z*7oA|i><w-XUjn!o8rYTslS*Kv)a#jft@fs?j9<qj~-@z0d8(QwBjutPKfyV%_i2+
z`R4SIa$8dg?5(ljf+$`RenCA<w1=GYNv$?&GBOyUIu!nI9WSyxno9woGrAzNb(gh*
zy3{|z(Zv%J5YUjtKWQ46C^Lsa_Hn5XPqRht4&dpwrR;k8UQj9to9bW7J-J?_r(fF#
zj4!8cRVhx!6M(H~LY3Ghe5%jyF>(c4LljC3rW&&@IZo;OV=-34FL9$VVZSlFKHX=&
zD;Uq&TRpa68oIJEnR&p2U_c7M84-j@U{ylz<3KR(s4COAX^+(geEQR7CL;-{0#I}j
zvy)u6E&)13z0Xoj8(SwiSfc4$YKk1-?g&)RmV~b60?6JRg9d<6kDy6EivWNm>{dMO
zgs##Xx>(Nht3n4>Rjo9>S~|3W3@^6Th#T930F9f`Uf?*7Y#?}h@!znRG|}T~f;EY8
zlg^b%u!S+ok9L|A482kFmB!bT4rG6oSW1Cj)rT<E^zWLC82H0W4U@<-Ut7WZ1{Cz9
zeiZ^=gYIi<7FUa>Nt!l;`hN2%?hq36z2{@UN;@yM^|>SBCP6oIatkV}cY2M+9Ejc7
za|Nlt$T@dd$!sA7>E4NL-~!60{k3jTkdkLan-IY&e-awGB#?)t)s*2v_KYUGS)df>
zkwkZKWyvC@%q%hUDyb}_BUkq(?;f(+vQ!#3?gIa(6y1D{ThX+}W~xptd_Ak5Nq$xo
z5;~4a0UkcXi8$*pEJ67c5NJVkN@aRl^dgm@p8mm`M$IiIXLEwDPO=~*-`=K<`WUQ`
zG1E8^Uu$rS-GzFu)j+kc0Gv3K0-`;C1gUdOPntm)Ey+z;D5NJr8Dc0uzzW>Q7<ueX
zI^jF{pK<~x5?lZ3rSTNRusHseFFFd-XQfg12e>fo_{O#KehaJTy*o`Zmd*M9-(_Tx
z0Wt)n|K-a#;f{5LhlMou*SPEqdLkit!N~Xx=Y13wtpvgE+I>`@FK{a3EOGw%CV?!{
zSIt*Dp3nxl#9SuZ<5>%e490<7^t_BoXrZ~k8`ERj#JbcS#QoKy6lS4C@()!n$}_<I
ztRdJUIWRC->mkMhmp~rhLkU8m5gG`EbhMk;NrV#I>oMyWhXgAI%qb1AFeuj3uYSRK
zIWUH8&NN$3GP((Qlw4;?s|U51O7e=&N~9;}IID_cT;?3!5wk+8swJm$+ih!u>g$(s
z6VDJq_h`(E=YXUy@-#OZ-`mPt-`cJ;^Gte#iGO1k*p=Kq3(Tyz11EfuWtP7vIfKHt
zCG;Vjhyzfs7Z&1pMJEwIRu1(3zS1RgIP69veRl(AQXsyQ3h+0%JjK<c7nLmzjxI=%
z@9{~&hxtF?ik&hipPfel62`dsDsm=n^1fNW(S=s@znMYfCn8wSCpNy(-?o2^3~5O?
z!S1GUGS1odWYE7j$4xDg?Z(?K>=XuzB6u&*r0Et&Ja(-n2Jx11h1G|SgNXOIVb1z(
z9`w9<9=N^q<N}wwM4*NZspoF7Uv}do0h!5+oFv0FVd0i0f3vW#y0y2sxw-gb<5$H(
ze6zl>y}JH<aShJR4MqA;J(Jhaj<1mAM%F9d)FNcHuF5H%iLp#KGH|$<XBZg$)H59I
zp_H;0Z-QnHxX;!GikKbJRe1&%KNnRyZeMz5)RYlu@#pPGJ-BltTV?wqmeAktL($`?
zGQj%w)a@x@wiR2VD_-<&OK2$U=-`3Wph)uEJd#Wfgvs)jwpWE0)Ey&l!?}ETN!~7Z
zP#+j9Ph`H8qGl&zD<tKfH!qM<d)f{6-nN?jno<KiwR?(n?DJ+9>ERobBT(pMXGt8G
zox(b`C)8YI$SP@Obgbrzf6=&uY2}>=wWA|?695VCl|!3o|2W|##LAgMgAv~qovipk
zt#!PJW>y$eHFTMTS`hA>)yxWOx6>rW#VO?#y=jt5)!*hTH^9up>w|ldg8-EqVaV-|
zLEx+(nY4sAi-QgHDAn&CEa^uDVkO@8P^B17nE*M^Xe9&FNBk5AyhIXW(&9EP2}TQv
zh-1gI5?rKO<u>=c3Mpe%tq&Zjfs$<NXe;ddI!Qvto207q`z^e!G}9|cW2J@XQnWs?
zwNP4AWi6)dBwavjv^G*1zPN-2yDcH3Hhybzi9~mrp|RK{_Djn9WTmYMpq?X1YSrVa
zdw>$ZF_4FQ6i~3o48Qr+Gf8{!aaOnI)<~fgy0m&;fedz%?`$L3z_43FkuD+_g<&H|
zmchgd-RbgWP4cg~K3bq#>Gn6L$T(c2Qh$cy0m2v%N#g=DWv?>+z^Nr1#-nyo4uBS^
zx(DMa8cx&F_pcJs)Y^g79mTm@fe!x;G0x!Z($2V)l7bxQLN<Y2zY^zGAdTD^5;Ns0
z62>DiV%Ye3{t_gALXvED3eN|}{v#6&KgHJ(do3PV3?^E&Wo91^7ZSSdBA$T73n=ih
zs0gNGm-j?sL56Wrw5ouHg@zuxh(Ke{SOs{VP~;9EK9b(7C{JOv{?wQ!j_g>8tg4|c
z`P9<bx}8z0$!xSQztY}I(}V0v5bKi*cz#dBBGBTW^0v_wzv#Le!PvJc2$a4V%SmS?
z6$4(knm3}iJD421AC}VMGR1C4E@fD(nv5@IUaYR~uC52I>*J+b)Z1uJyC=*X*=D}U
zzl06=0k)|stJN7i^)vqt0U)MH?D`+%&iF@d%@^mqaun<MQ7uiq>NA>-iJhqf4jV92
z1~Cnph5)qQ5eyIrN&p;g9qt?;MfH#d1m42{kvs&saSNIqa$xLMQFSf*YwEVV9=Qu!
z=f-rA3f0yOSAX+Cc9@kz;}wm9<HNfRNIc%sVmd6CPICD%bS^@h*oQ6<;U+_dWZR;U
zH(R8S`zU3hmq23x@z|<GYFbl6dm8F}q`>Ep*Ruk;aMkq7r7ipdn?$u5P|jGeB(C`p
zfMP$5?j`bFFfIt=(kVbwYf65A!@+fXaKTHgZ}I@>wxM;58*sq{%!kdpa&e>fi7bhk
ziJ+)!n0p{LInLcXz}eXAhTa@Kgjs5SznVLuCv+IlATF*CGM=jZ1;6*!>WRG=2t7{@
zRW&_u&rPx#y!kZUv|+CzrZ+t1E+3*g1@7bqf#LuF7ejBf73Sa@8wN7pz+7jx7<pJV
zRKUQ&#c<R=ZM*h7pqu3;nps{!dd<?}_R8{tTvv>4&Lsm-_f_d~b>FLK=T!$-gYU+1
zw%+#O1dE?EaD3A!w2~K~f90i<OhP*++LXMh*Z^7NI@SsH8N7N>fAmh&$XlIxR<I+Y
zV$jDK)ZrvBEca!9savJtfpD@w>7jPfOoN&Lg*F1%?Uno4T>rv^Yq$q5eTB>8yvv~I
z0i@3a11%nsV-Nd+!1P-p_39X5H@c(}@0Uxo@UD8c9&mV{G{qsZ0&(B4-v3ORq!W@G
zf_xHE5@{HyBLmA**H&(xmQBYtCI@Oz)5rC~r@f!MIU(P0Avx)vAEl&xc)|Lo`IFQ0
zY8-e)4-SfwJUwIz_AVuJ6K9!I2XG-}bJ423G^7J?RdKzN<h@6^4}GSwN^fJ%&i~PF
zSIqSv)c|(R4=!?y+{8jy$8~XI=|c1tT%*<1+u3HJ0kq|wHix--z}^vV-$2O`_ZvVw
zOhTlq@>#<_bE(462M*QaEKK81OdD<oBB+sNu5`$`j0rZ!hA1gGtV%HwM7QwfWS3P>
z)y87wN90J|%rtE*ZU3HQ^^gKF=o~n8^~O?{SSZ>UTMQq*MulI!x}GeX+9=BBk*!i{
zNj*T(GJJ#av5LCtiL=;hd%}g?1a4S>R`nR|mHmqzE%i0P4Okff|A}l(#;r?BR^wc|
za<4pj2i0mA<wz`-jpR!88~BYxE`&`sW%1fZNRH-ZS(3d@1!t6L=y%cQDGE6t<}&~`
z#cvDI2VsO!rZS+XrW7=iYf$;mk8vbD19oa^n;{PHIsw#h5>z~5r_*>1hUn{|b&7d$
zx`Zq^JA2haJ%BWDVtv6YFuXW8UbqP$5FzTy4`!{md)Vsz+0$$L+1Npe>{o{2si_|k
zpgHZ};3Z`Vu5X9Z3ELmk)f>KTv_88Hm{1T52Ri}(_}L}E3*SS2=Xv)4)jAFgpbrE5
zvjV`tvJtlpIBd{KPq7m=6O=l+NS*tlooo!wPZzV@GYtHJ`Rj0#&}A7SHc54w7|KX=
zwX$kxVTD4}KNP^`+QCpuD#d|t2VLBW0zpe@f-Q0j8d9m_TQj0Z@|`J?w`3DglT7Nj
z!TaT&%;etH_z~EZ6ipGQVT}WoX2{2w2KoWi%SrfBrtW$)eOuHYI~OM-=y+F!X)uuF
zdpBE?CF3V8y~oL?i_>Sx!WfRk9ms6+82gIr=+=25BvpfzgjNkZX|I7|Hf*Sb^`GFP
z*`D50hm#|4si`uHKYV`p(>oLQr6v9FpoxT6d}5K`>-w0<2lS-AX75$gTNNG)+HBgb
zpGY6lla)7y{A>uW;>4q%aW;@2V-7v&ZQTA?oQzU00w2ctdjuz7MqK6oEA%DnGR!WH
ziyg8zFCi3Aj8A=NZiAH21ZB&Xzy2)VAkV!di3S^Hci<kkNJzdssbwQ>S}ri9p(O2K
zSSr^b-|z@G7F`_m&gAuQVPR?g*|WvxIEv9!@qx{%qu{?vGnh{x8i938I*^={2t!G0
zMs%cvM+z7EB+n3e3K&=2<+`wN03ps?Kp|eIDrv5XB1EJ=mi|2^7)n!!Qm2{in1kwR
z%u>&<g;ISC<&kY{#V4YzI*emMM((?zfU(Of;BQk+TZ2H_l|b2QRyG4_Gvk4?E05r2
zLBY_d7^JQx$AK|W-F?wasOY!UBw0W6RSvF^B2X0@bElVUbT2KNuwtTaReDohBwb5{
zBa@Xi52-fb>iTP0X?LGqK0`RKZrdD+$g!q4THt0G(j#g%8ubJ{XbNtuA7aj1azb)g
z_mR6k{0llbwlt}!3R~FE?ndx0NSd@$69A{3nhcDklg@DXtiw4(OSF~KmNi`j9i%;R
zY3<bu@%f63%btAieu651#f8R79U6ac+V5C|1zM_9*$TZZC%S%Bk7YlHV?qtZ4Kpu|
z-;0@8)la20>RlOW=@I@4QdDAVt8EyTTfbTpqpxa6`^v0BjD>wtBij@G$P^!Cf>PCp
z2C4FCDT@GcVX2dVRG4P_A^gZzO`!5ne5Fc|#IG2lHYbIs4PqrK9`0HxwdA>L4-h2c
zZ;L$GQdJUVBAmN7M%F`NnxJxP4snuKRDX?p8{W_GT>}OX!w`&24Ann`!*N*)v_$~}
z7$~qGMb&hH8x9Fto=riVnf6rDR&eV}CTQ|@nZCh-b=+)jU3UjBSxh(#%!P^pTvKl}
zcxpdH&#35O2L`m=juI+di9p%zu!028nU;9T-Y5=?Uc&J|$;Ha%6|+?n<pX`M^>Oqt
zzBK$87KAM0l**$NGpDNN=Gu{ft$KpY#`G<Dj2(yLO!`u1bSzO(in5>pYoN5%{K>+G
zB!FrEftJ*PJd&8yL|!3%xiu~3Au7g)G;nk3psOUbZyD8U9RfBu{NbfSFJxNSB{*<^
zg{Kt^;~IAO{weZcU|^XUNx-HK`^W+qM=pDZz1OdWdlw0c3XSdTbc(`X$<(>9yd1&k
zud)w*h#NoxMG%vUQ&grH;+&@J+o1m8562{dCpY88$#m@SqNubdQ5-CpDSJ-pwu6P#
zeA9OrJtgt<6^7H%LZztmZHkIaOV`xAw5e#^^mfC`r_zmU>WBmcYRPCy7LTE*<>uqZ
z2bji0tV_Hlh~OA)fGx2=4!pX-cN2J&`W}@;x5@TfiP|kMWpSUv=0?7S<;XR)wo2l*
z7o#)oN1uY^XV?!xlJyAeZ1%6LH=my-n&4C=pn!Z~m|+l7Lm@7NJVf2L7;DHJnreP8
z*J#ljb<%q)^bh4%($<oIw|Kk|7rP8-_;lFV37J0E-Q*C)E0q4aJ9SU0;aKHHpOlmZ
zlcHb4CJrvDbAhvS8R-p$UnC0#Wv!6IYRbRHo2{srEzO5iOL6hI`xp+`{s5JP=t>Gt
zH<Yk;*H8=|c^!}utcHVaw8&H8`n34TM=0ST1v+Hij%-L4=pbbE2^g#N1N=LkLP+yN
zY?fUhp0;RR$&n8!1_Jgj6^vqvhU)iV@vWwh3&|c`z~X@>6y(h;^fjXzvfC&(o=||I
zh(OH2+cw<JCX>cRmn^h++QBwdFenfxmxS1G{-7#fBMc(I-5k!aQ3Noii10J@vm=x2
z;+z7I@$K<ronIi!et10jq;xpuTG*!zJrG+tOxcWplVcHs32oD;!sQhUFBjKdtn58s
zf4;)E^ltkSR|FC)5FNCgENoE;gvLK6nBiu3bOHBy2VTju_Rrk`6o_|-5gafdl@qqK
z0#O>TP(qoUy<+mwzsO7?pR7M!SV&G_$jMWzQ=rQbE5T>Y8R|hXw$919=|pX-_4U_Z
zuiy~FHo_4BVfaG!n?K+_I6_GDtT0AtH{isDPnHgPDEnYeK(<bh9^3_pf=I%S4Mafh
zN~~1Vg4!_G7;fS{k23DmeEN~}#-Pc4rPPZv28<;32Ly|vc6J0%cs4^shw>BzhemMq
zLD+`#(7)q=2A72zml;RUvwd+|O<6Jw4Oz=3LA}+8fGR_5vj@_^Ig%EW$h$RObD6AM
zD=V#}ng$IhiQ>ZTZSOIfi_QC-v#WQGnfD$8sOK>Lef>x%*P<*5xOXol>oC3ar%*&$
zRv0Hq!Tc6M@fp^1|F)WeagjSxGnhmDSL-8}tnhU{U0M2eZ*gtSd)-h5gq~|WKFWBV
zPX%b!$n9?UNEoRs=5R}?vJ!q{5NOEoo}Y}Qa;vbdNMRiOb_O7B%z(P~L;DkYBdGxm
z9bf2D&jV=}fLA><G7Gok|JpEBrQmrL$^x1=3q4lj>_DtCB0!Zx5{$LtabbCZn{HT7
z5QX-o8J|CfsdTKyzdTi#L{)E+hp=UkA0(hCLo`OmxTj@V+UaIt6e*8FHlS>Mgtowl
zsJ<S8hj&H{H#gueY5c?IU?EwgO~^jCou*C>#H<k<BJly2hEmxqybjVA!O?64grCLb
zl2sr!@Q#!7b+kwrDjtF3E$L~4Y3-i}2Vi}q>n>#(r7IgrOe{sX_841mFm8Z(xj?7q
zRL7bp4mRH2!&@G4m7Sh&A>U7Bsi_PGb5L&px1r-vXd$`?Ztx$ZK9JCz8u?~%RTMbV
z`Ur&~Q?)%G)ZT0qiWeyl1mw~kuGT0>H?=fj)*>l@1|Y8?0^+#*WoG*aY}9%RG+U<R
z3jtZZFI5UdK<M{FsBTiSO3C`P^Nx3M%XkFW2p(Wqp$^g!6^8c?RXNXtqEhfykhyiC
zafGdvk?O~ZT!oonE65zffcY$4u!i=+Y)anR3bw;cg}aSEg%$o&*ufYfWe0#j4t3#i
z&Nl+H_YcS&lGfON#**;R<N`Z^Q)je=5l%a}OwZfLQmV$Ngt_Mf$|L79;%X?DXJGkQ
zPR&bu$ucIMF%6rF3gRrWJW}4bFu_nxrgDJ04b7Uy@mtNTaVV3KMT+H<uO7#~eYl03
zrQ^})e7G<>3mwj>Y^Farnmz0v%+5VP&MVP4a|*o?DvoD8GVcnaOkd_n6xbCKH%<{y
z4TTiTk!rjpbjn~uo>Qi$95<_|_;dgT7sy#o<}RL{<PLSKe`Kn(t+>chu_+;AFC-hB
zHXhvZMh1|zut+-YO+^7C1ri(bUkbT&y#~s3>XAvRvYaBccP11H4--}_7HU$I`2B^*
z_(l<W)vL3zP4INuQLl8z%7PkF-SKFt3Hr3SmB}tcI)Rg}Z(2V&?9Zz{&cPrbm;{E%
zAl?^%{7dMd$hsmS#%Rb?Q>Ll~&urFHO`C&04HGKd1r+*Xe8F!2!J)I$;gKA8iylxu
z5$`AdsNO`ff3yoI1iJcYOy>0UZ1E+LVE>z-FbLGEO2^+Oawp9R7=mj>hhOj!Qo}C1
zjH*^rW5g7oLQ_+O;AX5)mI`e((ui6mObn^+&ht~ypZ{6iupnZ03%?5^GiJev`mX=+
zM2`}L3c_~cYsvI{#1`RKhGX68wqPGL@6FtwnS&y)FCJ|!K3{qYBX4zaYcFXP8f>nt
ztt@V>>@DM<1V5jjyZ1%=-h6xhL1S)i;lY;+bNA7f|G1=f|9Igk>6|N+H68GprXGF2
z0X^A!KTJ^lwx_{eaYt9TIM3H8>l+Aw{agS@bQ2^+K&J&8DuT<X9PwX7yD;^G=<q-;
zlF@CeiiW9IYmo%jpisYsJpqdSrKmrExY#-5r*}?G!@c-fH+A0vF!gM=^glEdhy}Sy
zLVM5$ghT(T>k43KAFVPJ<G%zDWUDZU^*?Iy;aqaWh7t>baI&FzsKkYJRR&`GmjEKN
z2P2+V=?U+MKLayPVJp!r!L(*$kX7ot0&k_R>-q!q1-{ZM{0<-`SS{?P$>1*ZTLE>U
z+c7X#@K#i9kqnJ=6Ia1rvDYVqzoP$&IaG8%W*!;qw+um}p(@xjE;Jd``3Pm=&bk}}
zX+fT}3cq1y()x=GFQJyg8^T?9R5vbILt&f>j8E7XD7_@?TBRoq$q&d_b25_>+vZ$K
z568@=G+xDgD*Kx#Dhd&lsXxkAao(PsS+n`-)F0oRJob)Sn?qjY5NZw;l!vhg^7U2%
z%M??YD9MqC#uAF4NI)G}++KaTvbV9g{d5a>#C-!Up|FofIF<k(h9r#ZmuRB)GK>Do
zv5_`|%CO=^bG<(|hl;zHK5~<!$d+8dEU(m3ms-I%8;#UcG!a%IZ=1w7@4$rjv-kSC
zE!j=+wIrV78cn2`T<r5OBa=N&hVrSx(-eD@v74d9)Yz6`8<djdIgd76TRM>bmg_cq
zXYuhbxVXg|{x-Ycle;+&cnB<QbKF1e{x&O**#m-noziQ*&=2WA8>a6=f4|MjyV#VX
zqNXKZ#fB*BSJMt}VzbfcM5cT1^Upt@pTnv&rWzSk-aya96<)%xGepXzA`EN#toMjr
z11oY+q@#-+*6gr1Jm?G#5vb+;Bf2WS+`f-^9RA|<>)t!*R7RabneM^J3()1C^)2f&
zO)F-B4ez!l)i-TY%-T)gAybhju8O_qWOmp;?SMuFU?Vamz!~@xgtqp3qx}o0KBG1Y
z|025GKJE`kw4vo2>kZhEoRX2zt4{$L^tg4SO}b5+A9P3>DV7d!=gR0fePa7y%%Es-
z(%ElUPXTYj6mZZQKwy=Qb)wwXr{_4+hn;HtLq>Kl*;KPLI5<XF6DJ~>2qQ0>s71~%
z{8iX&Gk7`$()0hpkgQ=Zn{GO=u0+CQM6*V0h`yx+kg4VfcVX+`?%g2-7XzhxWe4BD
zKc$+u_dp<7a(Zt#JlUhBb2LCUiu9w)#YiuK9P};`!H-&Vc(9M?>~asLM0mmtDpn6b
zPaYp63_p2HRHw2ac?G_?a(3Ap^v|^JWf+0}fbQW_5Ld(vp{&xO_izh580+BT0AZZG
zN8nW^*_gIT-~o3cWh-iw6$}RD|F}ae69DiqYrAyT=%?5^UC;1M37pox+)Gt}GqjIj
zI>aK?izXD5iWH+}3>8Ez942xpK%z*bD}*UngUTU-d1UGM>?JqV(#|ToP}L_}`-vef
z63xxfE*8kFTE)qrag8Z#(Gc-B7oAQ~sYTqhmhGs|d6@u}3lwf+^a$Z3<r!vpCa>v{
zioUAi*W$NA?LVof1Q#S*bF^4X4lgb78pTBe3+@w1Q~-=SroYJxi@!(xx4e_PA*`Ye
zQtKV?n3;f%wes5m4ZLX#YK7a<eX=9Zu^cj2cSx37z%|LK0L8({&ASRpnh-B6<=k(C
zLNF3gENDUiH6pzXf<!XyGrG5#zsm2#KZrkk5-rs93b-dt6uznHIb$C=T0)N%pK^dP
zpH6^Lbu(MRz48b<n2s$RYR4Q>QQ_{XX$U3ztY-1GqrjU|x{z1w)kom0qTjka7TW^l
z@6!xD%1ZblJ>NDdAFkbcg+{XE3RU{XDPRw`a3I|lNfqTb;WX<s(#nO&m%pk8Eqocb
ze&~Y!Y>7Q+4K?>75>8!SQJN$j7Uvq$*9M{~Sp?{j=F3Ze*l++Q6p*ye;x0-1!j{Vt
zZVT@s1a^d;ClUfA2}aAHU_<AP>!3pTv88Y){e@AWUig)^J}wBzqa)m<hC?tBD>VzX
z2WdgSK>w;(O1)l`9)D4;>?lP=*R|i4KOx=weog@|6Ui9TVjoqi4TSKfqa4Vn5NYX<
z++5+GqRN1-T4c1~%W{1x;G3gh3wI6S9oIl_0*a;~?gP?Cw1ueV`tPDO?9YRV=nsMo
z7(Oi;i+M{<DH}<fyB!!Ci`fu=ip_v(0&w%3@g#8*XI(eew^lay7~O#Y7@?XsYHPLl
zC(=Fy8B%nyxhMJblA?wWZS_S1(ws;rsYum2<H=!iF|^T3r`Z+6%5DF|aAK8LKcQ@6
zbKG9e|KN;VnKAl$AhWSKySpn#S#$_L_qMPAvj=Gt8<OOLcT|K){5Woc@oggQ2e;))
zqo$yaE9u%Az7Wc=-z{X0$z{ENvg-J`O9c65&~2q`e(^<d8c6y(A<|W<0w^UET`rp2
z3l0NR8Jx$hn0?_R^&hw~zM$+MTEh|Ur$IcU)tvpW-Qj1gow@d(XI|a8&A-1{Xzjc`
zJ$iK;fAzlEzx~baZ}50}mQ_<K2E;?@syh@lVBBeesas2O%7N)=a=n|xg@H{R_(8I2
z=3EFMH-4!Y^tW4@VLDA80}0GP*tv&l)~alX2=>SGD!zSd3hDBUA=bLkL@|B@Ol#kt
z!J#!b&x<cLskV7q|LSe|RXDT*M&jf{m@Ah_S2zK`Ox%KFr@ML@h@({ziXocAv5L4(
znSLq4uUkb(1bpKJwQpL@|M;)g?)203XDh${uiLHJU#D;X2WUYzEcP_EajVI<Odnl9
zF*AMg0*qo~bN%t^8eaHbq^OXZ$rV!rC^j?g+Sc7!03oZ`+X*wJ6%TiUP69Ml$9N(i
zd$jr-x1~OV>r!E_AT>b675OCE4nnZOdVt9Q-&&*j8dflv9{I%hJzxhmnOF@Np~*B-
zuh;_UUEcgQ(^z?j!Y<64<`{PoOguR;x_3lr)3Am`FPiRm<K8y60uzq7SYF*)THFkG
z!qSjYt0|c}mO4xv=DNB`=@1YM;OMCQagV>|xafTKK`055rT`Vu10Yi5+`HR=Z&JMy
zs$l`80-Pa>%hyzBRsEyfiR|27sjNrdwHzZ5oKf5~rOPI*Dfcwsb7@>4i!wz!suI|M
z7=`5bAg5Vc=V~UIP*4H_D>0?$3A9w6Kp-r$hp`pSYcAIk;s>2c`KEE(Ikq*HTxgd1
zE?*J7x^~1lPf#9q-=S+YiS3qp@yCM$NY9WqTthRp^wExf_lh>MY+B?@Q%SQAWbwsR
zr$Re+mr6=C-Kwl?B~zA-s%YTp2#T=IWv2(P0P@68K$MmF?pj*3mgiCvX1B?s6nV^n
ziC^$zHEv130XN3oLW^5mms^ND+_KZxPKgSASPHWZ+Vlknbs)#BIuvTCcg!lMOY+gW
zHrJo29<?^Mtfozw4#iZ#ve^L<ydv98N;45<vOzBJGYG$;t`3K-D`}>sPvemjSc}(3
zn$ZHIKwTG#CveNn<Q3|$Q1J-nr_v|%-U3v5&#V~$94fl5f?vaCg%L6{LWno)Nqj1p
zC{Ae5xk_ypb(6i<G~CrDewvgh>>^9Edx`@i^-;i0yq5t9A=6t7fs7_~R^F$1#)ES(
ztgMMv)u%axdH$Ts1YsP{Ps=AM^|(xoI#8<2>>&&PWFH{`Xf?aP>`u34p!NPneI`I-
zI^k{Eqe(p*QSuYPVX#SNxqSouGbKg*t>DQTx+Up8`7mHq`>FhM+7vCRmj26sK^dR@
zb*Iz*`PG8_M-_GiCvN{Hin)B50>=jxo5=m)qK82nlU)4_ybzTL@D)@B%t^9e2&64?
z?v+o#z^4$Gv}XnxVGFZY$}SQ^+4-UV3`0xCLDFrjz6e$HGY?!<BMhxx51EAejVSFy
zsD~l~K9Bm2VpDZ}$n-#S`cHA&fWnV^a5N(+{L%H2IJps88St{KSQ4KlML_8>o<iVM
zEb^iX*m5XLum@Dyr%<i8VT|Jh<|>o%_FinRB^yB(6(I?xQBxTflWPxy+UFZjH~7ik
z7LIj2TLJtWE~>fuo$D|vH=({jREe^OkRz3uOB{!lK@(~qOKYaWL>fe`xSW{Kx`h+}
z1DtZC_y0nRb)#-D;&>ux8E)gtGpNHjbczy{eO&lAIGJhOm`45u!p-ayY2^QrfF$DF
zQp#s;aqWUr9g9E$o@O9+4A!3Fjy}T~$~hzs-z3tuBb}J*xR5k`E0h=fUJG6?G!_pi
z-{Ag0JqZqubN+*{eyF~{i&K%4d&U}q68R0<Nm7PSCgc9IGAx)8bwLt`{%?Y8TuVFw
z^VNS~cUo0crQ|}G3>#JdDQ*H5O>kifcP4ioIHax*P+9$v)(<Uy7n+wv{o$cOg}Nsa
zp|EkB6OS9F;>)okZ46j=bQC8u$sqYhvHz;6tm(z?2vh@8l@lp_r3fp17{)2C3U=#r
ztc2D=dM|3A4(o}<yMIHWSozmu<Q^0fjlw&m#Pv|1x`T{?LK8_h+QO%22cBe$h0=^8
zex6cVLr!OCb>sB;s?()Xpvze~yQbt79x>3TpE})@@R&xy2?6H?S1<mJrHJ2SLkWu(
zZAJAcw!vc?*F<FYMe@C}>9WcimIae>Qy_M4r~cq7#JZ}~DX#0--G)vLF661|l5{!A
zRx|qyEZ(W^K-g%iCjH5_F4|DU+J6+Q!a?<uom{!b<y34P{QkZQztv8ph6B`3q4+_Z
zMNny41>gy<$@;NS6UbUXfjXVdRwe5pz&RD@dd6pe8M7!^&v8boVAd>)>fzu?uFJ-u
zIAMzHf|gl`h5u@%s|3FJBSgXdhlId?r6H^F?_WLT=l1T*Zv>M@5C_vkp9&GB8;@bb
z4S&J{gH0s)NXxsp;Ysz71XC$3iM3~9m0JvR!-{wS%LuY)Y?a#I_(kk`%yKK4W$R+0
zx8JeS6`jV#t|VU~*i4mw+vUx-Ub@KKY`(mCD>q=lP=Esbn^dK~pUJ+EeLQwI=aL#E
zs^Yz#^MC{RH*wV_-J-f7vltW{n_n<~1j_{V`v)DAR-$T(T}sDWY5QPC(<o#~wN;Js
z^FHpDWNW^`qsCuxnB5ti^vDQ6aA#=8jHKfm=Mb*l;`1FoH+44r^ubgTg;BzL$|pYE
z5zvJ1SMOR`#`oBHj_HrJ>Ab(Va`pS|-)gN?eeiV<8pu}1_ElV?`b{Tmbo0h3X`hqr
z9ezgaGcgnYN%h?t30KiF6Bm3B*1(>e^2GR&DbNxqYE8<2OYkS!T~T4uFSaiBpF@<D
zpsysM1ZqN?^EH{h!%R38*7I@|odP5{CyEk^xN!`-?TE!TpqGlPOMvGyY%apH2i+2x
zbw^Tza>||T$Gz9^MDj5P4Bz_h-V;`0Z$)!3{gKLU`=J%o37!B}IOLINy2sr~q%AS_
z21DF8hswg0lB{ncI`yqjhbW-0P66*-REJ8%Sj+JVtOUm^>ijX_c<(oYL-q@wm@sfo
zN*2SP)xqXMFmbZvvOewg)-StfzwMsg@{cGnd;lybUpMg2?kyDrySETj(?|HqKIExm
zHw+M$(E36&5T3=S;u@8yfMBMW%=K6epHC`DgN7C>o%tN&1Q`7>;#FB!RHRkvDR~1H
z+9<x?yzCG2?70DrZIrc3lJ3qN+_Tg$h0L9_>ATH7wjBg#F1KEE3>DS&65LhFo3Nt@
zdd@BW2<Qp3qngg{J(%f(J=yo*@SSTfTA8dbCae%5DybZRPz!aUiXpSXWtn~D_#!bt
z`E{fUd9mbQ$R2e44T(fba=?5>BHTPwAMBxk`QmIyj<nfjAy=#&M$wPau5wy|W%(RJ
ze+sWuXmEGFB1x#x=$12f{{w0{0dcM7!7(nG6_zgdX_w|b_r)Ab0eP~S$djTwyEvIY
zVK!I=Rws$tq#5|Gwm~2`H<qSy9j<-{5(f6{TWl$J4MZ6~(FrrU^oD&pYVwn2g;QUd
zho|78;k)|!my74TdP~l5G8_95jyxof)cI5Yby7VuICb9R8?D83(+BHvIYjl%s==*=
zKS11v+E@A%@xoGhf$FNIV~^S4r7gwkdSMW<Tk)v945cUQPV&jptgzDAQviEX0<3WO
za#u_4ue^g7nGW^3Zo~84=l!kWN%%=tuK%cmE?L>ER%r4ZHWU7dN(?r$W&&zo$%|ee
zkK9p*L<KoPc+x*&Xru|bwI}~fH5#+>=PZ8eL;keC<c&A=r+ZWQWEMT}!^W=s`yb&i
z{Vo6A!z+EiH#>z-&@cbZkCXrDZ~VK(0d8|Nyq(%5c(eL%KCu6vNuK(jc!z0}|KW3%
zP;ewnBj~Efe~lVNkz=+^a`MN*X4-D}=z7^M=`r^tyJKOrq$WY?;G~M^xE>x~jCihT
zvwMW(U({w{wz~!FM4NJ**5kcRlxN;U1O#Flk{o7b_lzaG_jvlYd)R8eJM8XX99cq<
zHuZ-yyulIu9CW3e=pJm`11iz#QiUt1*qqm!oF)w*>d3`QAhCGrp5-3w@qFpeC>|Yw
zoUlEVRRz1YyH-(VRJYWQL^2cHPo~RA7Db`_7`WGf4OqRd$DTJ5q0apqrr$sTIB0wH
z0SGt1$5I|fluSOQlvV9LwxRYFe(f}^LQIoMIaAEX1*b#7Jnu4TfcW1cLvtXIQ{3`4
zxCQ|*>qHi>fogmpsj**k-Mg$ZSn;`-65WI2e&ZH&S%j;XSjKW}&8W@*yaMNCXGeoR
zXdML;Q7#4{AjbZ5gtlC*))t)L+DK@&a-$|IEujo*gSUwyw1_jP;GuratbS>wRJGut
zn`Ee~dNx{ENH<c`Wdvx{|3s18mCdcy_2;T}kPjBNBJZLQnHQv;)n&s1%0bt<8zt;L
zw=e06Z8gVDi=qiN=$~j=s8p%)&jJXhM<7NlQzAg+%5c?zhQmJUf>(oIdlN=tPuDXp
z1)%+q?5{I`2@huEWft4*^!^VJwZ_g(gVZ>>T-+pT4hp~)W&<Tr)p_hpz5M`Adl;pE
znM%dz!h&x)*0bc^F%t$#<M;i+Tb$wqUSm(s^)YA;Z)`{*<mvv&AwSSJ_y)rsT#3?q
zaEwC$@L`1>F$@N4KTJ!CEy=m)!^YD0%cM}|-+K?9`<NsWEG#UqJbLlO$c>7#C@~E4
zVdy4gi)z05JA!ZGMZ$rKF9CiuxK<l7?Zv)A_3cA*z=`HB-*RB2OC!O&(L(to=()Ya
z#lg`9Lv}%lz^MX~rzp?L6LmrA=mBiX4a>4T#LMQ<Nq@hCv%6$L>6AH&P9Zuv!`kO^
z)#_&&zWAXaN?S|nlf6pjs_R=^un^tX0a(D9dL$0YIA}>1GfiZu&R{|JM4zIOBCUB%
zYvW3C=9vpH=%JQ%C$N>!C;J6(C?<zW(L-F1blyEIRK{}?%vKEsbR0wtn6O~Es%NV-
z)xh4jkIUG15hZr#bZ-bx>N{k)=!}Ez@H=QOZnA{luuTk4M%ex{!7R5@5u7k{xr?!b
zfHMvs7__@IR2*^&MjL@8c+*)4rwAx}4;s(A5&m2t9d~<8HNjTpCUji#(E1eR*;rb(
zfgg}mpR!N`;1nw)$d~j{TBhqRZd#Qqzmcc^^yM=z>jmP%B@uiAvP@s#U`-R*L3qj>
zAroGAr*Jdt9!eJ?SD-n>w~)!uhslbP7XuXbx7_9|w<)qD*-}Zrd#lThuN$Aw-JhGU
zLPs;~B%iJpEbpTd#6b^c`n6mgk1`O$24Vgj)y(GQ&J4LILvMe@p~@48dPGKey9Pc%
znn53zS@8CnK_9BZDXI_>yQp+LkZ!W6)2VlH&U60Y-*iAiuj!v1p`WBwa*~5%8N4L;
zdTp4x@+a8y%#X{m@E0Ggt?d0E9DMKl#m(ocftEx<zzb5R;KS{*LbH$yNLI)~XQawE
zCQMiC;Y?#2t_<N89n@ZZ0}Tkt6u4}IM1>BhDL0VP0!|5bb<mv|n_;DW7g&wzmx177
zCjZJyna=9-Cwtqga#|$WV6ufJWtm5%+IE+sc-U+U<}H#2gc*$cVWW1aD!TG*7U?VG
zVF`*3ae(!te@>Bwnx8oK%OgqwwZNfdXMHh;YZ|Wc;{>Q(EmzO>tDsp~8#kx=62Xxz
zN5X8Lw<yy=3UVF_*)h2F*>AW`sR}yF9uC!DKkjgQ;_tP`JeVBJiV`(FU*B7OjvT?o
zrERF6PDpx1DGWBzHub7&%`=Q$eG(7L^@iU=$oH=hF5r5UOeYre2Ro=FXVi*c|9)@l
z#m34e&jXnaSY@z*WP9&ur(E<7wYT&MhGrCWE{NnxqL`R$st6(yg5Ax5#fq{P;*}S<
z&JHiII~Z;sd@^=xuyLAXSllEc3^7$mHa(){Yq#}cd42E2>awDWRZof*rr;Zg28I{7
z2&;Y3*xmh)U#8L3Z(xLc@-4Q)EXa3OtbzH5f4ZkRfsoF1eKqU8G)`9f(20NsCmn&J
z9#csip`Zi(q?p(6dMI`$A4y$s5gD`m*{0MuJ^Z|N+B-m8OExfQk4X8X#n1N&erN0T
zE}SepP`0-x^rlULxzt5%O-V3$>x;7!&87=2NU}1T-v{B)YjCQ_awE$S1M(ro4Znyt
zp{tw@iL27B+<czLthn8mj7yWljs#bZC)VW`IRvpf0|!XBYvOJNN_F?sAjRr>mG)rh
z0X06(V_=oG9&wCl7^iJ_Q*h^I@iD#={NkyG45$XFO!;f$4cwCOtngx$D*VgGe)7w=
zt0l;KFFQ@Y)U#aZL&r&5u#>2{SM(v}KD0+U&2V^edMZb+g`T1$vN@Rd)19C=nLWZi
zzRej3bp#3R$YUEKAE%gs@_5V{snmns`3S_}^NJdAE4<_`D5!LiwJf4=HT#gu^28d@
z->?AwC_NvBWzcT7Wbnn4US|lppg|JhHZwpBGO!A@n1oH)RV&o{%Vs$ACoZ|k7Wfbc
zqire?DD58hHXT?XJ;N4}qw2^q{emc&i!vhNL$sr*fumczVqCJ;#Ur6!hJqDOvR8tJ
zG{ePjsgW&<WX5&=BFIZ4%Q$wmT#!QC&&0z8W{2XRP>*NeVgmNmIV5aO-hds>%7J0~
z55tW!&1a90XxHt$t^AHJ5Tir4bstlo21=8!bIdl{$P#zN>xj_LWtKoAt8?uM!`f5F
zmflGh14l$O8oR-+G*XTYm?h}nX#|j0A&3Xbzuqm{_r67Xyi(xw6u}0$kOimz0bwn8
zdRSoE^3=iG@Q=I0gX{r!%N;J2E;Mm9U8TMkB_gpBiv+FguGwze2XgiwoS+Sz1n`p9
z)VnE8hNM@D60l2g<;2J8J~<x1ugGn<yIFC8o_I&F`^7nwh>{@=|Mp(@x=84N(;)T<
zR*?%BI0<a6_2_DBDvzr$cU)c#cfkv~s_##ViU_nIB#z=N9!8GdA(s%g<caW6q?uGu
zGzO6(doIR>oxjWSAT)-v8yp4s2)(l{SSnrIghTGaQ~8b0>~KEU1!WixU*iS*M{oCD
zqtPB}HskdAG}Z)<F&_JpZd*+`?M^-({nc1OcJMW#*~v(^r+BH%`*fXD`uJD_lr^b|
zAg|_Ts=F|!hM0oSdq^KSX$*)0aly3VNpma~EydH$H7_Y(0sn&zH9c5Cg=+pzIf<U4
z$Bi?O<aaMtS9oAXPMy?pi{FwMyfFt^eJ*glMZ!>+tfa>qLO@BD`<d3y$KlO}9P9-;
zp=$vKxe=KjC7BXW%r(NUGl5o;<Pm+zg$9lSN0*@LkQQHuXRV<QLx#bN>YClb>ER62
z_{Jg*_Jb!(hPkp<p|JeEz(^ab*;b6S6(en8L>?0(?SPT?m70O=+^$g|pec=A^aAU=
zd7+^3;X*5%uts>O-B{AKg1ZOTD50pL%H0oJBT#j^J}7bVmQ-zV9(Y8CSDkZQi#L$N
zXUCmOZVx`=5i~iUM)&F@)Ba;t3>w!#8|RMBSU>Y^TXB@ewCr5lWH2VlE5)c(%D|zJ
z{dCdOSg<ZM*o9OjwTLrVtRGzDW{r`kyqWN%{1KV6df4;lA%-ntEbmm?yR2ZRCWi7C
z=sMU!j2w?bG9qs$BaHB=<fD>B1@?49?1C@_#K_Ea_yW_IR$VEuAq>I0oLcvk9r|d@
z_oztqFult!^kEW2TfDA31x59XuAr6mHSx#d4Q8a{t*D|!y#?uAa)k&{T>yrkEzlf~
z-PlkN?Ll*?feop1+Of_grX%Ic5cH5@*pu}cpcRVZAm`bQ5PV^csOr84f%0lc1V3FX
zb^Oa074$4uZzdmTT_)Tqfb<ghu`T+Eezjp3xSV{8Hp^btS}Z9oCguSHrf7&VY2^o#
zu1lZ!8QZDUR9oCf+!nYIBMoH-(s!yB)x2Dmy2Qc0D}0I|h8#q$d%t?dTlFeOUDaaS
zHE)+>d!hR`IHX(C=EXU)BWhl5by@qj=B4e6+HVoLulaHd7AB0Ky7_J@TdVna`Qr4P
zGMbqzHBGjVp>$E#`=S;)uK9Gi*Eu>v8d=S2C<bba)x2Np9U#%BTQ~L-35V8vyL19<
zFR?axO@mFCyO(t@pQ2V&&3CjLYhEGWfXkmElDtVYgXZ~hX8<-tC$F7=aZjI=_xW5b
zsp$iyOf6!V?uab}4GPpHd)N7be;77&ffHdjuOXuI5;ZHCPeTqt=8l;NZ(3NWx)z5V
z-v)5k;FMkd^{HPF=6?J}xTW2qd0+DeMMf7MYA`PvO$>Cf(8&0~@47n$pa-zikJ133
zR!ehW6JC@}S|kWXK^!^lB;alwGZYgX91ea<$xIS+Qe1HI8(c9N-8f<C8wjlUN&6w9
z)t2_bC70GQ4aEIVF3ywJh0Cdzaf#0f%o+GnsL?TMZaSad^F}uh>yd*1#Mqx4h9ni`
zW-z`##xV`V5a0tg&7*NHml6$O$WvBNJasel6CXw8-Q1fiu)Q`{mSO#@zK#YCLYlCe
zXv^sp)3E-)xsk<1cWTKUp|CjxB`^wARJmq~nHAVlVgI^tRLdS4;bzxC$Oz1oOaD}q
zFaLqw3tFG@LeF2P?&saX*BCTQPaF|oq{s79^hK!Uk$Q~M3#15tbWNY1CM}jwa(};f
z0;)%u(U28pcZBVRx7SGj6=@5}d5QcE6v0=$0oms^tkxs)Us%OnKP^#&yd*6*#3KJP
zqWHd<P-I<G7nzQ41xSCz84v+<UYV~y`B1oS)B<;3oZ-x&)TDzzzUT&DU#sb9E&&EO
z6u_B6LQ8lKcc`0n*a5~qQPV;OZ@k9U0gQw?c<S1UC{AD#uRpLt1&NC#vv3N#A+>|;
zKmN3IGO$;rM2GtI@b3#8O1XjuAXM}7$!gk(kd{+HmVeUUr!Hl#TpXgb$re@1@En?d
zf*&lb*mhMS@I%fB)Mx8I7Pg6G3d^MrzZsPBA`VYSrze_m7+!~5@xpU(3TG&Ao=E<8
zsI`WVyGP=CUxHRP2)JlX-JfIdr@2pi9?3EHQHdEpVX^YZy!<$VSMZEK@5|5YF0)tp
z`+<GRLWWcKKbK#p@W`KyI(+bjJUF=S$j4vGFVtfhu{F?K3+y9+QVr~b9M}gbun%%z
z9~6LnkOBLk2<(Fl*awpXt0FZ)sKT_5*a@bogCgVs4e0=DfJb^G2j<Uzindmo6RTTW
zZ<~z$-==Ed%0SxO+zWEEB+nv(MHhR2Efg26Ps$d3#&Vfa$0)X=9~l|KoxuvOn>p_?
zdO?7isIGPY&%ZwS^RIKi&d>ik|KQj8&wri&^4IyVZZi+xc2IR0F&w9xyMKRoM{4N)
z1vPX}TQi@1bK4$%vv3Oue}xfeZvEx<XWvx5IQ?yidVt1=Y0QG%>m;~;SW+|2(@;}9
za1@(cQ8HBKC4OK|a;jg$D~>F|(e+99ueUHeH1EO=V5X84WB3bQ2woN!XW)#{9!k0<
zxvYlG%yh2ZnLFT1hg~}@&cSCeI<8#mL(iX?@MEM~1H^a)uX^O-Hme+&HQ#hEq9Rl^
z5H<Zw283;&bnH-WSc*SS2TT>3LKbNWUfF7>tm(*ljZ<{(t8VY;c(lNG#Dx_Uwl9+X
zn0z-v(pj4kdM%{2Agz%5&`mlJH+&#%4ms1j^$<Y-U-<b~Gq}c;#PX0!fd5TW8%UiG
z$gaEI?BaYGl2(;HxD;4MOp6xQFs<RUy9(LCnm~6Okb?-3F;xvD3wJ9Lp_W&cR+&7%
zx3<bmrZo`p9F73y?>2EI>H2fLlH13EfiiuL%VEf%`y<g>;OsL1o^{SEpMQ7JmCC`o
z1<BlsyeDN<(!cC3bv$u60Aa(rS2GTYccJ47wnpIK(3^(V2!zl#dBQ9-{P*hkIvEBA
zkt4Fy7$6tMl_^Qte;s>DH(wdAxChurF1B&}q6t9p5|KsJ3?ZzCuTc%lx)G4Iq+WNG
ze7UpFqa->z;?_883rAM;AySv-G#^C*`ij+L&361Q)YU^x9VyHe%bGTk?r;tzbpQ3o
z?H=8kX<bAGLj^c(g`@<ZE_v`oxG?@YerA%NGro!A(jh=-6Zvz%#zz~#uo(ux;vxID
zQerD<Bi+Q%u|C)ULxzgAdPpiULhzeO%Ils|PWr7J(JpUNvOMmkn+8Po7OS`Y?(u9i
zbCK4~vUdH`8O3d3=_zI4Vn35^g4#79GhQQ;`JFIFy`Tb_DZDhjY_(-g$InjSw&vja
zH9V7(*t}puTQzmX@e4etfcKya73QEq0dA4$9(Dg9MBMC5`-cNl=H*Olr<&WUUSu;1
z9Tpoa+eL|<)RJS^suC5Q(faGx!Z&}S6w@Ec?y8#A&UDgN1}X_WF=TJ(Iq9VF5CWP!
zJnb30M-uLfTTL)q%`tf5aI@P%P^==t<Jgt&cgA#$SlPW-W6)ZfEr(rGVf4|*H4Ua3
z%}96P2{bT6n3l*GK+cIN4z4w54VHYvC;~}{h1I1boQx1zP(b!p#qJWIw)?*NeW9-e
z#&MX(M$z_-JB@n+;p7u4DKQ^wI?zZ*YjO~l@;Jw37(*XtlyKXyBq~(>LR9WcMN+a6
z$6&3pgF%<67eX7ja`Zj+kl!<vj(vffb9qTrKKC&j5!Kbwoxq=ye+-AA61A~97||#P
zJ@MGWd(zQk_MOjd^3Hn06%qF-HfU~C!=ncdbUlLo5ONR4s^B-Ndbp9Gd)ez>442~$
zpG?@}#-JQ`MTK`kvKlPKt~AUic7(#^<d)pFg#B;5(tc?_G$Fd(*VjR+(zGxQw)PX#
z>e6qbqjgIx!bcZJ(Q!O=YkP3ba16sCqOt`e#DK%WB(Z{mF$8Vp3~`J8#nCa#mx~tJ
zxPo-{>Ec|Lb_GxDuyJd~?!fWaj`W#ZGk%>4wHL<;)#e(0>Z8H<AOZv|5J=&0CC8>l
zQe2w9HrnBn-)B%Wa03sydX5d_yY}C3g1k^o#G6U-xBaMT!ZVQ7Ew;L+V4za51eF?)
zM4J3*hQ3w=vB=&=A2txu-jJz+8ssc`rB4Eo*eUKk<)*-8mWZdr`loGf*XQcvHIx|`
zB2Bp@uXA%e=>jVx%mUVP+J!$R6L6}}zbXKdHBvCR_-UV&OdE(I6>y!9R?;di4+<^a
zZImDG16<aWraOcn3T5h+Pw_+jN|NUH7YR&9Vmd?}BTit~n%Jr>9<jxKsRe1=m~yh<
zxU3gGcMn|c$dAQWP~UE7^GKB%`266r5z@e@h(>@;{Q>gxZDN?C!t>JO?PM*6u=01!
zzy6gjNu6Z+iNl|a+I}bwtSwERKETZbIz-4lu26EA@2JTGwjRUQ1|2!##xZX!>oEW^
zCrC%czOGuibLB(;^z>plddAc5a%9oSQ0UFtBu;QpbhJ8@{azh+Mivc(x5Pe?<+IEH
zPh88!@fhIQVUJlBTwS3Z<aFv2Jr{4?34?QJ^h(Ikkkkw;xozUDDv2DSx{X1vk4k^n
z8O}D$xc7?Ofl(*2SI{f-3SDHgO=`X~E?%*4Fs<O$+mP*xK#vVN!Rm<?7X%G!i+OsZ
z^+6(pnN?gg;|261pPP|@QafB01;Xrqx*fQd=r^YtmoW@$r0cHOLT&$H<4>^-mHKn#
zpxD7=0%X`Y=^SJgAP*V|%A{eueLs(<AL2<hAcgrR%)o|hKSa2Zoa@oGFVBQ?R4NWi
zgkcomF?vz=a-S~?K*c?)%b#qNdW6}y++u-qUz~Nf`%mFVhY0b>tPkF6vBIj9`hA2>
zohdo89SAOS-IvU;8g)e+BSWC#OA!3DQn<vlDvRDV)gF=CwaFvT4KW-Tx0fifcqjiy
z6AMX%vv#)SlUBx2R^TVt!y7jVfv~hiKB10wfD=j9NkXD&_9NxOCM#+;h<}m@0?d{y
z9;qAR^zpS14sdp9@we)CpydE5sEwUjUn$#CH-e9^9V*#{I)&BUgC6}WtU}O|s~f8j
zvi4lGIuTS6Y`&3G-MEq)C3#h=ofVjrNeaxhhh)w{7g7ru2t1*rDMqftsrHh}G%+wu
z5lWwvDy+<qexzbQZLUv*d^*ZO7+sm}q-ig?qCdT#wA-(5Cptsu6I$jc;mbruwD0`|
z2h;VEy-%jo=2fk&u~FD3Iiix;s+v1c1cD{1JWGa6Mm63Z8&h%aNWPMB0EbcX6l7Y(
zmBZ|!JAoXSLIc_>{)lP)Mj431&muifQ11P_gfj-GL_)@DS$?5@JG&)cn(nL)x3D2g
zAEj$7xvuhsu68~WbmaY-PE?&%vbd=e)eG~UB=s$WFhP3(^luvNd#DhDNGmrfRy7hU
z>lwfSPNMEfQkJcSy6)`<7M8Dh($3nYJ=Ss|{VMaDV~II_9pBRKvoG|60Ekjts>BNa
z6;gs9z}3A($@Ta9AcYKUpuzr$(dM9=1u`x3nIS6993Vri{01j_0U1S0lp@+%WB^zO
z84as)Ffzf~23#|tbpDV=+tLS`$g*gUi39<}s9YKBKxW?LtXKAjAh}{z=cteY#VS>R
zQby~{-_lRa_9LH4?OC&^QHbd4ChHC{+hBShI@Ovw;iIh-`a=y|>?mnuSYxMknx%;p
z7Jsf6$QsviX=P|cjuwfzM-0KP#W6`mPv!-?VOVZ`SrVPIS)=Y0aD5E~$0-M16D(zA
zILSMisS2x`@Pwgb=8WNd!trTw$Dw^C#);^3#e`l!TJ<C|Mi);hNVJTi4P_5~Sw;17
zWe?a13Jq4tEK@wym|3!s{PQ|>Yyyh%rL9t#d|}DxOa*fo3s}qui<Cl8mGB&X(qB56
zF>L8vNrOwD(Ongb^R{pdVD}iVI99l{JlY=xpwYSU&|snXqSr5bZtzsBv#*IPDXLAZ
z;jHJ0s4Z9x<3xYV68du0=|vLGG<>lX+$O5~*k%>@=y1at6=7C>8qjG=@=4x~Upg`!
z1A@$e+LWrJDDbu)fr3BM^lK<QQ{}x4XZWEhWxMMlOW<4ubDhu^uCiI3(`NXGUP2zS
z#wWsaNZ!c897h;Z|0!ntq>gB#jo{`EhB+8L(k&6wHl9RMugZb7R%9_QEp=xLUvmBr
zf|7rVpm4Wqn>2Kmca-ec@>!_A_4l`Nb4vO(iad+|Q<4EWYH`rLy;0u+Mx2$AbkmS+
zB?e6LD{!KSz^ODEi3_d8qXDRcX>fwl;mQr3mF!Cv#{6Abi8qr_Za#egl|AbW-byEi
zfi4a)(XS7({rCizLi>LLv7MHX_r7#T^nZM!BeS)Qz``sYR#IZ|fIziDNMH}jqO8Ae
z-TrVwmoGy<|5jxG0>vmTV)B0pzsxQ{A5~m75PpKbe;v}lh-><9PW&P_5x};Kj01%R
zkqr?h{}jbsx^AuuRJ8HTwYny$w4^Bhu;lui#aK}jad?m##=90q4ONK>nok<BuS89v
z7pXqu{T5M?YDHz$$3%TliLAv~-NZ2)b&RYR`8I+VTrjyhQAA6SnThJ+8lUjs$FRy&
zRdJ0(Y^?wJh12y4+##&a$JZ%parAGj#0CcTzC1M`{Nw8sf3PB3Tt?riMo3bU<2|WJ
z4)=FXtY_xxg!47(VhnGThnA&Z$`*<M(&2++(>(uDex?0?I0fdQ)0(m;@TRGWk<EY&
z^ubu3w5k9bxT)D>o#w-i?Q2_Hw)I_Q@Qv-MO24qW+5TObT~``$k{y##em<cOP;m1O
zsM#_FusKt?4CVe&0qF))F~igE)||0x-oplribX}fTfy8exMa&DUmJ<+n@gQug5pAc
z)SRZO0o0MBDItGM?hR@5!_=a>#Y%G6YEmClJG!Qc$^TvZ??z`(z6fqyo@CH43;mD#
z=EkOy*c?;s3hfapLUnc4_SlVo*1-{5uqc6i*p1KXd)CI?!6a>iOC&a)030s>89ypO
zV|%9}71rdC$~<vBCpoDVhfbd@ZZAFEd%pN=1rmT^M!4y57S$fFKj&ivHIt`!P-wQX
zx&Cr>nJwqDmV28k8*7V8d~rX2fzswU;<~boHxIa{qgWmI3Zj|_-i9DY(gc4>0TIDY
zT^n`0)pnt;HKT2O7mu!2-oboe#JQ$n=YW^GF`0KOCX{Bn0=xs7ulo3;bA;TsWHQDS
zeRCyme40|cs%B^uq=sR)sinR0j-_nF3GTJSry3MW`j&G1Sj7@IP1O~ahwYtSocgS;
z(NmLCK@sebESgU(f*Bg*AjSp~`^aL2iUWd+Y}x|TG3}xf)v?bc4JqM(d8V&K5dze5
z%1ug_FghXeOZnCK`mf^>fs2^I=|n_DbT*YRYJRcgEK5ec3_`w-YA5<x07+;Ba|-Fn
z*G4tgu!K++!_TN7hEDP~GW9bdWQ6=>OG^T?PyHMfNA@JEIFz7`9B5$ydwKiu`sTB}
zjm4#J7oV(Pdp}wE0U9d1l(>-H@2wNc$F{4<&6b2ezbMC%+4VgujgVwpID2h>CS($x
z5jo*UP-koRiw6%%L*i&loL?*c88|d!80V;=G|fvcL}I2)iJ%F^uu&RNtB@u_A8bdC
zB)y8J7sh1mB;qFIj(J;UfbxalE(r^26Q)G)uY>~S1*IT*)p#bKV7E`B6fX`2IS47t
zWs!pe)Y@~}Y`i=hhmKTUN|cc>>6LfJ&w+tt626?9L|Qrv!yf2`!AV;cq%p-GK$JC-
zQ+=*z9_A>HyF?+0u_VtlIaetmK#|5}?d+i}ABBQ+8uu!3n>N{U*6;PFd<}?Q-6*&z
zO9xsEf<!NxW4d-=3yPc2HKxBB4#@Qm@7&>}wLfcPNV8wXw~1H_X_R43t=q4XhY0K^
zvjQ1K%qdn#g0@_e3zfs6LKgk$G9SLu#SCz9dnyk}MoB>Vi%^y29we)593=r1D`;W^
z6Ht70ByPbi7s=y>4|P0W#FY3`omq&&=1_!N+A$nL20twzZ0HK3C52nsQ8@U3#g?W!
zP5I9rax72r*UzZygv;Esceo1)>nY8w<Mu<_5B7VJ^KLOLU)i!mIw8|1XG=I>BUK5l
zS0uEtU;9-@acP^eWS2Hf796yfzcB(#A7)gTz6kyJCKllorSo~J`8g#BeW$&sWTtf1
z|CK}snS~z*5e_9rM?r9i?DBjQEWPk@T+Pe!ugtp?(V}~nKg4hW%MPajgzz<SQ}VOC
zE7D#(YgoQ(;ntGeq<DslnNT6SJAk%=OVfOWYaAccgDkuwEqlSUh5X9z0!2DL3c0g#
zB9p&n^mZi=c(+3;4Y6ha6s3AVHz^L8k|kfZH5l&TtsG(Z&Wb&aKU?hG$d+?p_}^kT
z`I#dU2d}S%$s?LjvVc)N;--aD9D6W3hlIznHh8{&ecl-kAB!4BAtBF$dXT;4+IbR;
zF8u*)&x=v_`|L230+#4##nA(i0QJ*2sf<xucKkDg@;0yG^yy`+)P#izU$HWLKd{3r
zo76daZDU!rd|ocV%RK{D>+nP8o?|>LItcp`N5bM!p%3fdS8&|QqRgZnH_L3yL^kI5
z6(%7HI5#4xAWBg_!7aEM0q|>#Z*))~Aqw0uzDT0*%cOqzZPG0KwuEFYQ;<BJM4zxl
zlZ`qrJTr9Q3{y40oC3*wL&JbAUd#pFggZ(DBC}&V;jy3daz^e{78(SWFWJ9}Gr$9?
zpmuJGgRO>h|6tP_$*Z;yx{OfDfukHq-wd^V2{J^JuLs~cIL}8IJMBjHpqDk{9<v=N
zCp1&T1wxs~l;=+CnLthOP$o5zlk2aSeA&>0C7rTPq8LQs`?1<nMMsesWKkEXZMY((
zq^0W^tUo#HDT$<df+J(ZJUTpud-h_l_3-fXj8wu9D0Jsj^(0r8mWog7I;E9U+h=j|
z*7cMSA~<%)!S>pEMs;eQ$H60yh-|*dx3o{3b(gC+7Z##(Z^2Vwwagw|-LJ!IxCBvZ
zFC~;L%d_-FWn(8t)%gP=3+4UNam|^C-g5Gtya)3-q6a~i+rX0y?taOL7u2Zk;M%Z-
zRN<;G4jkX$f*jroK0<P#mi84hM|MX*X{rktsIrHR!>PJx49MPNX>+uw>?>+HmNL8D
z-%utM&e5t$st;Gk6sjxoII3@^W1)_jKv_Q~;2&LEhNPcm0N^1@3g^mGILNAWTi1gP
z_LT6_w5P=W3mNEeWWDaaL%(6(QX&pVFWz>$=W_aDAZ7n$X@bwC{xcg;zC7%o?KW`%
zuI79dt|Zegklu+rIm4H5vW*lc)%dG*0|0f=3jhT=C=M;5klSWFlT)9Oq?8HgoF{Z3
zzC;Kft*4f8yc->t!bgla{REATmUUD!SglJobB@CCYSd8GO;DbFd~7R&$nxfcDMIDH
zs1M36N?12g%iZ`MCP1!<VKpKvevY0<%Yo4(iU~w+cx*n3a?{|vK<ge#=3JJWY(;mH
z9Gz(5`dg$`pt;D(?1MVFWP1tc9gETuCl+N`k)_@xeq^g#?bj78)K2lKAw*boTUUMo
z9uhWQ@jLoRN`XEsY_41=zIkwoYbXGmzdc$C)(!wD)a9`H8pV~sezdNWZ_%&l74*|Z
z(P~vU<Q#iCR=PD{*~<-5iiX-RaH=kmdpHr6{Nmle5+F<ELHeEnJ9+2gfEsHgG2=NB
zCsPL{Yz#R>S8fK&#E#bi4az%gz5N>=D5E#n9-9D?b7sVP;71V_@}f!3>2cw-!b?Q4
z(DiuKhf{xXvEe(OXbq}A&DUbj1Df?~0%DBQc07rB_isFH57&;JwsG5x7!f8GVoXGI
zAQIV}Ds8}c-)n=Zw*jsplF2Kf>Gxmo0wCpH-h1s+O~jYpO}P&s=0X=`04JjB8+$2C
z^X9d7xmvZ-zaw9TPPr*dPZ;zOOFo4V<v&;`uU)NavRk;rfGnx%T`<t;WKTaGqD~GE
zo=Db0`hp%zHZ3G2r2+4FQcu)Va~ZHU26(G}@7hJN8-@~DmzDoQumS&0x{x@d)c3@_
zJbQZC1+FL_(h1C#7AtQ}HThGoJmFt=r<!m2`|-Ds26EkaXOa~PQtTAw1n0Eo4ukcV
z9IlF6<O=vyN+=QEcA~<ImYJ*XK!`jA*QE+%ed6gPwOiD#L191?tP0l@S{LD-EUO&p
zk3}em(@Lv{`jA+C`;&uv#Yg`N>n#}O`suSBMKNl4O#`**b7C5g)6I=XJ1FGu-cv+f
zRLM~DNhEHhk)i`nJ~k78nu-K*$~|Hof*NG4Gun2o2VY9mRC_o@Oi1-VH$(W6v0ap`
z^lk)Nljxl7PBQ~Xzdt{QJwzR6OoR3p5B7Vb@y(8pFHe!#2x17;|B2oA!4KDxG@;wq
z2q5&vDg?1_s3fu*)UFDOGqfm7$T|_$Fnu~a0wA|`&l>orjfe+sMKwGi&LN5#OD9&$
z-_Peo3fS&%v&yHojgLVXu=w_I^8-|IAt1Tg3dOOSPoKVgw)b*nlXd;|pt;FUfeSZM
zUZW~PgRR2-Y2nUmD>9GxONSvD0hP4{aZLj^uPDQHs}`1{JREi|f<6~#IN^d~ltjl>
z5BCA%R!q4mjPW1BW22P6A4pSQDCDL#!m2U1byDzE)6~iXh`#g?|1!j8RElyWyyU?S
z>Ell<rrn!Yv1E4}4@gCFWi1+_IW7(aJ?-&cLl!H;6B1Gr+*jgSR|8yz7f_GMv=aKU
z)~MAm*Oa{7;CB#5056S(N=!7$ZBDY3dciR1z%Z9k9%F4kU7-8{;%Y>90PmyFA3P2+
z?-{JfsyZilwQu)~30oX*w>hL+0;7Uj!`QaYlP||fZ{L=K6o{qct&vys<d=z2%W*XM
zN$=o7`9tmE;Na~DH^0;`Pe4C|U)PO`^P@rMP^q{JV|R3t&Wa^A#;nj?IOyKOA$oz$
zzyN3dIZT~~k_^#v!%|Y?$BJT-gATy@7`(*SQa1W{Liz(nuihL1`=qK#G20h33Wfv4
zA^uFmlnd8Onb{|Bo_YRR^OyhHS#1Bk)Bbs`{pY=z+4ieD)4#FqTLq1Y)ajvM<x%Hk
z3Abxfeo9xS6Stbb6#HUV&6@6RTyK{i)eGHZ{r8hYiKF8_`PZnsjPtX-5fIUT3##gY
z>UyXTE2b~d*gKQb&vzTQu;}S5a7_)-On%JUz^^nI?}PJm+@lyJf!R4S`}Y#Nj?`41
zX>{3Q%*5|_8g!tlF=Eif{>O?@Oj!kAPxuyYmIq!`ji6$xy4RW{35F@V_{By>kOKqB
zjU+19gexb~6Q`I5;M<bf+72$pkAz9OE&hMR9WL9(1;BJeyf4oFKla|eudU=t7yh45
zA#>;>NL$$MBxhzww|m?er^&bt19mzYa5yp&pkpBsN!U#%{eJfMdDf+Fdv8hDxRcDp
zGru(2yLQ#8RqI}>R*6l}=Pi67GvHjX7FW#B^D$IV#}^=e*c>;me}OsN7m7RYad#<L
zr_0?FCRU30a2g}wD>$%${8sMFWX>Wpbh`v2<^YJWpmPNwaVwvIl1Dzai@<H8T`3dq
z5~79)qDH-#XyW35g~AWa2y{0(?c+H80sTK(&3Ui<Cb4gQFGUBVCkmzw@|fExlmq;~
z3Zt%W`t9k#5W&4`VL16voXFIR4rfWD=Zf?ke5OtmeLl`6-(t{rblREAW4@9?T)YD6
zHGFS;t*;U1(bnd|Tb`i4)cI9hRTkYpRF7X=-Moj-KNs<+(~IGSFuP*aFsrcHDr)AG
zZWdQZZ!I%~x4AWWw=SR{Gjz%kjnmpM#=%(;O~RvN?iUz5Zd5|mOvv{Fx&f$;r_0aC
z&k3;>9<E@_N8eR5QcHQ5?Rq!Ezyj?&US7}}x*_Mu`nnu;z;7$yN-JC6EjrQxZObqY
z$U09@-giA>RO%LRS7Q8|$03|5euoGYSmP_FHDQxYSW|isF+9&{5bvF9zB+1FPjRJR
zYvG$Dh8GyOVhH$G3`W89W1%70+@4IjNMTKmg#;#(SiJ!x0@y!G_LDIT8Hm}?ZK@r{
zTo4G<!4mPTjo^1K=)-*GX*<oX#vmbl(a?W*Pv(r<4nCSph49H?mijsIb=td6XtRyA
za1BAD7((89IBG|D!FaEgER6`tPYmY^)JatU#V1LmGTkQ-caUN&woEW@igAO^p1}7J
zf$8kj>%23HAxRXT0{&?%`#?Q=5MWd^6vYU%R<u?cbC&pNFd{NL2HLy~Gn>c{F+Nh>
zvYXe*NKor!Vqa<PCDgqF2}ES;ik5Jw>@7Oj4iYM0DD4TnL-Tkc7$(=3fI<cGr-8zR
zeGL?9PE^+h87H3x6t_DvSj;yQbsZ@#NXXh1UW24wmRc|vCL~fcWHYDqg4IkzKPEFs
zNuU@%i9-69ye-`<SZPyLh?F8ED_J}`6<$J+XdMX3J_^6_tw5*G<<F+8I+dKB42P+K
zSO6B_YHRBb9BxbODJdV~8mv>dvBEyN9Csyx_Epcw+2kS8`0e-7CxJ70=AAu9tSS>*
za5NyZR_HY2sOZ6~=Q}66FOYih;>q)qC(obneSh+5|LK#LJ6P{<>Q`F*ZtokkJbZ~p
zcoz=-hIa=$-|hXLpZ(I0{>m#qJGi~~;@R`v?U!iqe`*^X?EL4e-GiM&H29a=27BKe
z?mgdmDJ}jrYhhsRJPjum)3*PXwRLz7kx=2spdW<;&l$+O-|Yk9-Iph6Z;x`UJl}b8
z@IuZtKgv<!<2-r3`-F3P1hfRSf3y4Y+3wEsrzbD}xKEUY@xu$hmbMjU8eTo!J$&-b
za~%_?lz&FGo_r@tC~mJ}Gv{p3$2H5Tu2MhiX?NDOKB^Kf)myeH=(RZ^F0_O22YbQ~
za0Mphf`6rz-|CUb3hxQ66R<N6kzMk22CBH#7)6Y@ph(#a4@Ep$A3kfmT7TI3@=MFN
zso)W*F39fHF>0FWN=TzGe*Np323*5>tqIxjEq8rI1>)`+PUd<<(Yf9pa6cH<*Jbb5
z>(RKPIi_=R^kB1?zO(F;nk}}*u8z2+alswT8nZ!#R=OE%o<+cy?5>4hFsHa;O?@3y
zF01s5<pPe8d0mekL*GX0G2A~%Hx3J+ujUU*5=iG!p^xUzW&g~t&XGQZbac|dkk3l?
zMHRFl%uJ#ro_YPi%k}88z(w4=ZpGNuN;!I-5X<ZTO5Fq#v^$&@IS2&F_EHN<VmV~r
zW|w$Vps@s?m{rN}0?u%OW?Af?85Z^3{sy5bmb@SF51dOyG$TQakAWD`YrUvRV_oU<
zy5VDJ%QHVK#xahtNG;bjCS50HW*|!D11k=~qyOFd{~8EY*cfb%V^ySN-0~{j-j?A7
zA_8N~S_8p+0IGaK$ZE;FH<&VJuJWar=&k1NDg_w?nE#`iIfoXtfasoM102gQ!Tdpx
zs;cwV`$8SRZx9{bz@k-$X~1(BrA{V=m$mc6J)5$-nn7^y=NkJ7Zo`={*(nu>k3@Vm
zaAt#U*az-H!hTGG3MD+u0g(&g!1h_@4+IBuBJh|FcnnVDvPMD_wO0>7R)o7cDG8*r
zP4xCx@#iId$DdVhK1ktqqXDuhg;zCki0dj?iL|Gzk$7Sv@oNp>5cUfo!A>f_OpM~t
zes$M2<8c`A5+{j}dx6~tP5t7`HdY%tQVTlvBjHAFZ!O=AYZ|w<W9S-mqB(8JEvV)s
zIO_zFthZ3mPQf+yU)<uuM?9zQlFz-zM>X*qgoELL$gZ8Gql5?{ND5i|2HF(W36Vho
zKU?-dloGw;FE%VbTv--JgDYI1!6(Pr<g|)7o?brkT<_&T3mNBRdNDXlZluH1tf2un
z4>4JfL<$T{%RLGVm9kF&Pcx};P_hfDhUlh$W<?Z$hPnl`&z1PsX3C&8&uBeY7b2e!
zJ{QgHV+sJ$85y)LfZswgkYaP*vV}=(puI%&V-B3gkQo4oD^oxmtsyfkq>i9|f~@^^
zNNG#^vKMJ8TlqTVOldC65;<Io3q{38CUP|bV|q;&5$Gb?G1l#}xST~8Uo<IN95)U)
zBv~7)qp>LxdoZP&5V9=e!%Eg2$p+^xy&n<Cy5YHv)9y($Bg<{Bpp!{b7KCOI#3VHf
zoW6Xj8B?An;yD^XLijs;C^>=xttC%NQ`r+c8=P<AWW<b=?5Up_@ym)Y0C5cPCxc!O
z7dc~BeU(Bk>6SBEO=svBOE@zG|CgU}NGM$h?%j+RMLY)-ONl5AxqXt_7YRH|=|kxk
zd;?1Jq`tJ>-(50a#)V89qiC#4=!5#gw4m1b$7C{|=td&Em<5=L*>%7Se$2s4P^>B6
z<Uke4VE_hQ0;)UB@Ve-}(;Pa?Sko<=GQ9kNkR~ZnhnHE?>FU?(tL+d3MjE`tLOdb@
z^;jn?%WR@0byQ3)Y5`$P?)~WOqB`hCV?8O%Y&aK}veTcZ+z8>-*oki0gpgz~7ZWqJ
zv{L89N0<5B#|$CU@KJlaJ8F$bxR%~baQx`w)(Gs*Es6Eq6btr2IRgiw8j=mL%$dIh
zX+zUCXzFx$%|u~9x;f`dA6KHd;rrx+a9rHhy3N|X?$<zq+a_facnw=N^DHKdwg!#+
z;rra8>jqOhg%VVBcCsNF31&wo=g18Nu`CzucaLDj{d3PeXvY}V)ou3bo_QJ2OYctL
zlH=?PY7t<0+(k=cAhJ_#At0PQ;FkD0$n9f&avd`D$iowNqD136aBN9353*Q+W}A=L
z8AHg^_SxY)ANFzlo@OLZLrh!8yL69G?(oU^IWmY*0O)=zt90wTTMV4}mJpB_c3I>(
zTz5PRyE*bK?&h#ct^D(y74q~qH@nzFNq8A)eY3%-MhfFr{tDxVUp;QMU%mK`7kl5o
zXy0GaU_%ZM8kl&=8wg-{iT~y=0(5w3f1?37Gl58C54L~e7V#Q$<_-XT0lT*R=7u{I
z?4meRmhSx|vDMg_BD6wn2(>T->ebjZh6++vK~F3)?+-1QonnzCrKhBYLRB;<A;gw<
z68h%GiG_hp<RLt~u6UO>KCDQ-q?+2^^=$9#8)j`!cYa`DUk2FPuSRc2<M$(t-LPX<
zX&gi0<^Bsm6g-OkPE3k_ksYe8OzViZ<X!mLGMs@#?U(NG>Gjoca0*%$m#yieG-Pm9
z^C_bT)#vo%i4+6XwhihUvZZP=<zKU3i^R`l5mk>Q;4WTVYoP%R5(FI3*a3OW6GJ1h
zNh*`I<$t97;Y+Lq1f#Bq_YuEgHmFRy26`p1Y$YNH)9h`aLct!_KDt{tOx1ZTG@X5t
zPEC?@OVU*@JJ}^L1VrG4#LCY&67f@CS<l}}Hg4_r41;p*()e+LdWa3IcR46Na(SD6
z$4U$A46-@Aw!Y#p$|-Ytd^W~0BCeoZrR%neh57Wk2BFv_u7ti3a8WvPD3c;9>KexJ
zL0QB~{F1^Fc2~cPQav`XgsO#zB;3a4k%mL>_DF^_Q?C{SGTg#xe~$xr<7>ngKW^$i
zx0-p+ou!p+md|Zs3i5}~>?Ie-Hh^rcQrU|nmJ#K#Nn$9-G*YbLR@5NGiNIuOYv`t7
z=K^c*83!nxGof!KM?hCod6_VHSaV52(lz7i+qW=j@Yl&zw^ye??zjFG5e*nUaB)_*
zTG<%7B9y}XOGr2NsZnemYvs!%<Z~z>>uaVq>c^W%;78}EC>y)LBYPOkEqz(bx`jDu
zs~f2$!=@mR-F$@AR-1yfuloogN3bEGN$S~H+oH)qC}nzFEn5wa#=p#1{B3`v1|n?e
z#1Og~L&M%=wh==T*#lX??1`p%CEcXEO<BuOK~6T#vRANG>|sT|riNB=rwVmBLB@4K
zfKvmxi7lotDffJ4wzQm;KIZHo)AN6EF2NT)g-D$B4|I#-0Mxl&t{Mi>bqu?N=tiz(
zVA)EI*oQwXaVPL(*xhiG54B{qH<rQ6g>vXtnACrfmDH>XF(ujf^nRk@5;9rbTk*Qs
z1};Ywq!hv~ln!f{k4nfClZ|+^4ZyuofA36?S-A57PxynGOmy`sD8hJAQyI4OV3Yk2
z>EfT+_E1FX@>2u^MWp}(t0bMq`RYdHBC?4@VskDRK-GRu8U?DB0~CwY2rEKAvhCOa
zE%KzU`;)r+xttPHh~h?7qJfM|wI?G|qrDm`bvf?Vgw}Xl@1QalAtANiZ?W{UGMn&A
zJqQJXR%)R^2FavF#oia=4M{uw_k5UuFGiR@j{plXn0%Sw3mCq1z`P+q5*tdI)gV8}
z#%Gi9<-U?>vy2grct4)J{m<+EwEzce^Z4*;D8FIVg?F9D{ShMavPG=DQkq~iZn@t}
z+aY&D%4Jj{UrAp*M)%Pd`XCyzq`5Ey5#-6K5kLEey`E6S)4h>-9|7l@+EnFpu?<Wb
zyC==izOX!YSEd7`{rG!IdbMrp0_^-ux_c=*evF|**4VRZY@OIk#7;SsST?F91eOGM
zVI8<dGMK2B4BfMaH0iPI0J^@I4{%TnFypz$ONgmRLIQjOU$tv-fD>*Lx7V4cqA{{1
zW77~HP0wtmmdS<`Wl3R_x2#!PW%ZQ&0(u-6VOLsP^tnQ=ZqUfaty6eN3)uxrIylGm
z0sUwl{aGBU2ml%UZ_h(+1}(`w_02z!Z-f%6XySU}V3pS)h#0i`{5Ny%!3liiyd|<F
zBPy6?pM+b<r^fY0DH02VH@>x%h8d|URK8p5s-@>96q=U6s+QQ=%F^OJsm26}WG3IO
z$Vhr~g9gF?!b%f{z-nakia-nv<_nn0s@X(DN}>-qR9#dDL8+@+Sr|tl5T^!|NJEvv
zi`GM1={k64s+diz03PqQb^$!OD_H=MBG^D4OwIO`gFLwcOZ}FT0rQzANQdiX#UgW@
zS0(S=XrUbE)loiDzTY!0@Q!zR4WMXK$XSu(K?4B0K~&4?s<8*!LijDBnY%DtJ!6tN
zTdi>HlZL*t7&dZY)!ZHZIOLc_aLt5Yg4Ve+z5Iz+EUd@fXyoNkyu8S?P^IG=wo1(!
z%kDAh$t+YZf&`JNlkSKlQ<Fr^CfFm9ghjaN^MYcX5X0_%J)Kr}-j^oWI&3xxK1XCq
zxGCDiUaGg4ZpqVhx~=EkSWMZTG!Yx->=33OnZeoo6$&;7Nka*O14qT!Fhl&2tW~2~
zDR&W+K*-m@MTD{}KR3!l<-dlZ?PH(Sk{#xtZpdb&hZ*CHmvRb!hxL;Ha@79HtcVt9
z+=F2Qgn>eCH+~~=?`Nsm5uA?*oULCqB%@FV&o6a&eT78|zky`N+03O}06|WJksm?&
zD150b5}fK$+t=di+^xkd-5cg8jV#X+=1$@|2h7ejDXLvTX2?VIW;PtUx0Gh@lLQnI
z=1(=$=scpErTMPmJmU#n=Bhh&+I-0YTosi<2lSEaL}Vq3=v^T|nJAg$2!bL2o#x;a
zo&i3fbcM~5lA&Z=we1xA7GkZiCIfozW+4{iy%rBibkQNIAV7~!e9Ziv4U04g-%EVP
zjY?1#gTpB;@Mp#KhJ=r2=o6*{8relyAFcuSCO!5C?=8Gb2SJl;(CGUGDN~%)Nn^(f
zK-QmF(UIJZTta7!BCD?9??i4#kTeK|&A#vbN>^;^$wqnTUw{SiC3Jch@SEdxJMK$h
zAXD*UnNiKM5@J_wIwM&?72zLzpF4up+2)n+rq^aX<r`WyJd6P^q%FIr05f@H>HB+U
zXIPVLQ_9|O>$d5+e>Hd4%62ydnq+%Qev@<!Orrt;5s&4o6W#RHPk*zbhs71Q(k>yA
zN&o#|goAf#OcZSDXB0Sm%@TP+Nit)U<k4|5{L~+(29Rm$L;j(nk0xmW6O%b?M_kB1
z8)#l~N&y)MNH6CT1V}WiW&*9p*t^CZqT@5?I>BvYm-$?Dxv@A%m0RhJqhd=Cr)?VG
zmqn`?Oh3wM#3fx~K;A?j3?k>ryh%zw%|ie?EeINh|FiJ@BFJd$+*=xR%JY>acN(Q{
zQMWjX*!&fjH{hLUsJ=G*{BuR-&~mL<kV>m5>t-a;+}jF>VY#LSAv&EaeSC~gkd5Z3
z7{2H2VkmZt;+S&kqK7nC?1w!aL+IQ*$9_(~#~P`*ei)0)H+Gf|OBpro_usCsX}eT(
z2U!c7jqxt&14anX`D<%E><{xt>0E`K3TGJ!|5Fa8#-)nH0GfR-A~w*zzI~-BLk3-V
znZbNVh)#8|p!&Cy?$t$(&G%sm@_aI%F8^Cj8CDmLY>NRd-M$$2a9zX*|4VF(5a$nZ
zqYq$y@cL!k!xRsC(_c}ACIlW#Xp1}P#{$_StVQzk3lEZqqom^CWmKKa=*SZgG8tXl
z0JPu+tISELQlPW7to^F<7#U!CpjM`@3zy<9{ky~??3t@BGrgY=WIpR_z5eL{IW_hh
zb06Y<36@Fa3!dzd4KTVfa;{3gq8u$ggr2(%Ud*=Oex%O$R9G_jN3~PO7;DlFxuSuw
zOh@?(lT;FDCVl{_zcMXI%V%xOg1LfqHqhmUi?SCUhb$IaPBOFDDiYY!P)I@4Uc?#l
zYLU@CJ;SXMaVXs}r?fO|lm4sSdMXO1+vQcSBXspH{YFhby&X)Gk-fw0CfH;ozYEs-
z#_~q6vb7Fu!hIoK?0bxVCVgS583OCx-%C;9n=l!u6mgE*x@juI!qn#7=@le!4OL*G
zl4xHRGOXDBc+$4+OH9`oCySPd3<)&fc^rX`n34r_vp5iql7o_Q(t$Eoy8am1nXCAi
z$cV(PIzS$kd=hv&IwoPKJHZ66RWOpkptEuOTc&OzGmp$<EEhg$7GtpqsJKax={y%h
z7S}TaBbf)TEpfdU);?<+;i(sqA8SMuHpB}E7wR>2R@MX*7@gT~R5D9FNb(S#$iGrB
zr`d}dQ$puP;{1-1!oGLL6zRcdp^yQ~m8H6awJPi&x-l~jqCajuWa_Lxz=bf_sDPi6
ze#lxCH6K38K73>!f^z*|IV-`+7ki^`2j{3TKU;t5I;BuA0UAzI`MNSOO$Vz$^=0dq
zg>KiicuqBj?&yI9Z6C+cjc*ZfqAP?kIeNjt)PT}e5HCS!=_ij{$#~M0!mm?jkA`Vr
zSe`jxDhtTvy6H?DmiV)%S!#JF=n)B_sIU33PM5i@lz_;(Hvz_oNnZw?21fG38ZQ3Q
zE_4z`1SKd1^NTzRMqAp8078y0u_z^1X5$2il2#$&6o$LXx!&D7-rZ};v`*5f<DW}3
zuI9w3G*HQMI3A+r0ai%D?yYYQp1j!p_T>34YDzwNeyEh3zuNiz4ogTv>SSWZlbBP{
z!(bwsMXeFzp!s969mPn}l8+Ek4awm7TMfi!A80L5RM!V^YG0FQk{*Sq#52gB2j0sc
zry*CoNqa)4YR~SDdi@{r&v2{E=oDh72?~76Tl5;=5BuH8sNdVyB=FjaSd;HBZn8mA
z2nV3d2F^%j{=Our2O`3rK_z1y!TONDpEL~Ex!+1i7Cz=UaYP4W0EkCH%xm|2=;UHN
zMM}}fq!@b`#*kjLEK?uZ--$3vDFm(DAV?o$uhu4oKTvFpEX#kT#4xFcPY)4+ONF0F
zf0Y_Ss@_Q@OP9A41^@2=Bph^)I?hgOxEOfClfhZn{u1i4BmAYY*KM3`th9A7)Whd5
zp+$805Wao<4Ub=%N+;<XE%hfaiQZ&#dL~X!@`aB4uDyHmMd9YkjdX0v=g~W%M9r*S
zS4y=uguf&-iJ1U8R!2x73^?jM=AXPJtc#sz|3Yt-wb@kKhGzF$^tTByRx$%o4D!lQ
z!L2gDs+odnCmf(h5RfJDg%iK-%1EGq5kq9BF*>U#>v<i(0+f(O7T2ek5zPb+8RkWa
zCo{mD^imy!!g8Q2VMNniy;%vE2`LOrj@nm|<soU~BFUwoD3erj1%nd;49v19-uOs-
zYv|Q!htB<5eQQ0$(;w`wKoIvhb)dO`dZXLIDZ8lAh*P)-C6ebMEPSWu8TJ-A`?&7d
zqSf$CS0X`t$>o-XiH~w#Aq(AG%ijh;qU4g%Y0~t$dd$L9X$Q_kz<7LYjY(3KhrA-%
z2q@zzDf}r<DxXvtp$6OWtLD|u+ThUKTq^L_2|LI=qc6vM5;N8K61hX<EJx+KxrAhx
zUaVhGk8QmLcXryJU0?0Gu7Ne7a$CU(3aPL%N+JpEgi!W%tJY5f*tUFPZieRQHc##3
zGFf{AnTo3lH4Ki^w;;rW@fd}1C4qn6><ol|D?8j&(Iw2Zs3egBe-O60c{*0~Kh`G{
z;<Gh_k}c~5q0id0Kb6~TB2*!Jq1f#v7iD44X6&huRj7h>SV6ECsQt-zu(Atlq_+H4
z1zYK$W%`NH?&1+`$IQmFE^%Lx2#V^eG1&K^x$e~k5ah%?JC5nOj1=%6xPn5}A10LD
zcC#Z?;oOLx9+8uFQGD1CrFsZ!IM2Psd6sv$`k9A?P(vRXw(<w;ff|%Ir-ImL!r{Lw
zk|=6~NKkvCK*-lhm;4FM5!@W<j~aui46bppddys@3ZYt{BbII}76On#u$p3h>Jb!&
z$ZP-uDSB4co)h?<8!ql{MZS*HR&hfFB(gAe1^uYoFHRWJAp>9<(Pe@^EqQ}sp5+at
z=Exg^uqRNMJFN1y-ahD)8hGNo*T3q+n;V@D;JdNF$o2U}>+s1-)O4qW$lRQ*t*1NR
zy!tH>{ie^UVcM62Q^@53PCT)BTmxL)d)S>74+l9IQ}_|1+1YygsRMC?QgwjrS}sRN
z30Wz^*h7YD>$A;Yo}m;wjDg;i1?cl(;W!vb;GY_}akud|dxtv!57te&C{03b7nd_3
z;Ph^c!@h3E=rRU6Vepeb<knsj-MhV104GhgAr47~hGLJ)d1F&^A_8>`XLSx83%_XG
z!j)Vs{bK9vAJuY=63!Y9ij1aRzki-o_A)++gsw-py<dpN+CIa>Fi+w*gso!7!r*&X
zJmJo5F3X6)a_Bz*UYaC$q|pYL`3x^od;9)VWf^2A%B`S)z$99sYvNEwLF7+(G883s
zm~Iqs-~fmL#d+1ZCMn|vh|StuUkL%uqo*!P$VA(_ylLCRO|S3cAhqlRI=cp-BsJOB
zAOk0)vILk&D2{k{JB+MqH#RCxyxet<*_YUM{TQI437;NIHH*6zsl)Dc=I9^_6ADlI
zOSs57EA(_gWC-Y1NWi<37FE|Ud)@ky-nLK`uole$L$$YMog6fp%${QwvXxm!T9|i3
zJVPY4(d<1d-Ksr!xJ*xG=Fg6@-eOAxM_wOkeG-pXGAQ|)*;8>cx?OoSzH%}S9T+h<
z%O(G<kOH`1-^uMcBXdR-V0T*|{ACWjXl>1Hwb%j}RpAjLt>TpEr8<jU$x@(8(TboF
z!<ncQERZ~p<Q4eLY1QaTgEZuWxS~RD(mf;Ik5VR4tanG<;mwbIq-{VW%EbgSS}{MB
zoh!H)8H9kz0VPJ&Xb2f9M5lieANcG2#SISEy7gkCL^S$L1wQY9Wcg5{O5sViwxu5U
z=K4IUfR)vklN+jZT)wGP%fIa6=9Z#hb#omWcO!zlxvICq7q8{#=IZ7u1yFkG2XJwI
zbxNrK26uvTj;;4_WGr&4Ln$f<s4i<4mrj!tjfQPZ2wQy$6gn~}fpyVxQI$N$h?O(`
z1Q@nv9-R_#uARlW^3n8aXf-EKSnezo3fPcRURY8N#c9zV!|DPui8CcOnqvq?k~f6y
zMWZm;nk#t%(Ac9?pO_&S`bl6;Bvjz0oG1l*ks`T{Q8CL)c@p&R;2jcGT1h})h#PC4
zjC#8o6bO_n@+I!v{U=KgV3>KrAm*}U{xU5LmPF$~DC%>D3dGm|@Th^|gBCZs*Of{w
zHI8XI)nOpM)sfIxfOn>cV%Kz}kbQ&YQ(0N9m~TZ}Ob(lvnx<x*3stFLqPrC?&gilT
zesh%T4CV;d0=x=Ng$%S$Su{w&(qt;73@Ja6$@W?bYEU#0Uq+_Sg#%@KYi)FWiMQOO
z9zKeHBY};#>F8Vmh6=%eO|GoIGkZ4LKD48oHKY^1ap6AJJ3%T2swjhuYo&drFq4jC
zL&0xomdPd&4W$-eTstTn_Nc8MijL)k2GO!OBu<HLks9LcD2pj-hDRFG1X)%Z0bf*o
z6gILwRDr?)l4S#HYABS*;^^FErb$34t!qd{9L2ax*4PNk9Gl=Pa;$F1nU=GNgAHWk
z066$Z#G}9=!-Zd%!<wX<NyrgkzzCgi=T2wlmb^VTP`!=a=OnYaCj`kv=T1jw!^SHu
z(eb>}4|1)r7euK#jzE4?tNdtz{yboXJ)A*zo#{Qa2E8b3(4CGp9FO0Mx;3X+5Fz(Y
zcVQ-IHW%VnQ&h^N1Iv=<&VZ^J*Zk)JFc{UIkGk9Ibb#m}<%rM<Aw-`gekROm0N@*c
zG1mpL6<t8L!K?zxtu$T@^*R~)x@m=dEFP##*SR1qB*h@dG2;`1=U!a!bh)DkUr?!V
zvEAU44Z?ASDtM8PQA2eG`6oPvC&UXvrXVDTROnc#+i`rWH6p9<1OK>of%LjLJ>fx{
z5+BT!b`f#eWVpw%1P43?rH{t5*4Y3z0D+9z2W{d&&>&=|+POq*RuG_d6UnURs4%kQ
zaJFX%rANCNlQ5UnRcmE3Xc<He{<iVL_l%I?U&XO8;G?xq-dl&S!+o}}ujby2yU}bU
z`}?{86)Q1oXDbl=a(*mUiSB^X4dJ7zMG6visr$Vww6VL<IG!C5JD>&MY+^r&No5U1
zB_=&_b1QNT%m&U?Kb1{}_&~T*WlTMyd|==?=|}h^P52g0OOY~n{<byC!B^n9%&sRR
z5$Zlf=P`3M4`C%DawJ~3jh*alrh_RsORm4{VaC!WL<=UDTC=k7Hloq<zM{-MY}%+8
zG{Wn}R5BVMRojX#L~b!1|IVN&!~dwYSuxI3{tI!Won_gJktj@GX=vJh=z7RfR@BZy
zD`>7&1LllF7k!HacZn$U=(==~q_pHMQo9;&o^e~JqS!O=lAPaG)?QK=*R3@=AQ3Fd
zKH3@#pDYxf{hh22R!D6qO;!+LL6nut76>xaN&ZyLZ&^{T=OXKA3tpx{$Zf4h?FkrA
z4!{Eh!=>zTYv7yQ@2Wv|v;7CW6&3jviEpgzEX%2=l3|nzJ51ETf_UUMU_(nr;N0)Y
z)*EHAytQr;0CdF4ZUQHeZ}BK9Whh$QY#{PbMiZ3iqhtu%$E!>xLj@;La1Y~^a4`i+
zdxynH!r+)~%szG}77v44=UwXMm``<;o~!yL?j&+=UC-Zp()Fm-!gD&uq)H+aj0(nf
zX;tG2g=dxA6SE&Q|AZe1Ua@jqUPV%(Mk@+jzF1PRqEpJ4HM#%B5JV5xe3l5kv4LPs
zm>_vT@PxyXXL)9fOrRV?eW%n6mrn9#aFvY65T^H=KO5*QDd^MOiBB?1-_kWCr5Ul-
znOn9CnCVlXO5(baeCtS5)VKamkjWC~Nr^SQIlR8Ii#954&__9ZLhKT#L6;CFYtx(Q
ztbgf|5ExJlN`5Y{Lf~nCOqI$s_ZpYe{ce6J_)ax1J}WuWVroDkSCMfOy=yrAED%bA
zURlp<jM8-B*O{7&<`vnqCDfx6MaE=;00CcA6WA?+z6>ie4J@tX5PG4p>8vRO^M~vu
zWkl-Azp|1zNi^zXQyC|{22v@`VkD8{w8NlMIbdSU$e0y@S!>EvhBD?3oTrFW-7SYa
z$aIyy8@9Cq*~%2P2(rPc+@zu*pDqS4C7U5gf|5KgNQg2m9B6ifch_dGLXM7xG_Sy>
z{ayv%5wF>ZBFYN1HfJVfUcH@fO2#WD&q-u`1xV|{XfU`_z;m=FC-W(w@|UFO<l_&+
z`r0^`6%ZlHN`ccgq+1h2&!H)}7C(heAY_OS4WDS&WXs6HN`xY&xSy%uut^|O6U`}S
zbLxp`gAiDA4pm2sB3*ipkCStrsWgEINkOOzkVK$_Va05c2~M}?Wq8vJ@I&0d2xHGo
z@gDIoH_@YChXdWC<PLWPfH29}NQIenKr2z+5FX`MX_qE{4`>JkA!0pchPlj|-&Iw1
zik~XcMa`pUeD!}LqlHoHM3=n`(M&>7vo}v>MLvq}+hhzV{O%wkb_%%x67kbUGOdab
zyLEr{-9%Yfg(zo-LMFI*kxIBsAZ>soaB#fPbU!mMl8++Afx0q*D{sJ8LdKehJD+uY
zjf7ijgu9#WD!`K@3ye%df)vcm3)|L79(^MRbv!3nc6rcHpKAV#GD^aK*;K;Fa##bx
zgXst4%t5EaMX()mRgJP2mTZIj$GGi>Hgp5wkF5d(HWvy#<G+~`8vf=P%Ubp8R3cc2
zyCmGS5W`O7LXJD{)bN)hYi*D{4%qVI5;@M3-4Ef2WFQV~(Pw41h!?g>MlBOQ?0_2T
z0L6w1hg?k)iC#wI>H}nxbNGd-4yH)x!DQ}{BMW<6{1$s5Q;oLOXtn0#5Yn73eO0Mp
zIIuuU_a2FlDlM~CA*(Y@_A!Tw(_?>Wra~-Ca)9GT$YTr%74T2Wvr_39vbD?pbc&ni
z86LdqPO!_fwUxeK@dmsR=jFv<ijnN1Hpz*&93u;NIC$H~J*1~y+(Fah-OF9v;d;*P
zBHlnfMj4Jx+?=zyvbqoNb;^r;uf{k)#l0kCM_zUjrROc%U}>a1;$i;>B=)uBDqnrk
zKkcKM0WM>(rW1ar9sfV9FK#|<UCd@z)2#;&&T+ld^_xw+f8cH6A27<0HVl@ZU35op
z+4nsFhUKJm{@``VZr*t{8w`(kM}u$sL!=3UQ1CBZwS5h`#{2H(FJ0Ij_pmGuTEfxv
znq$E_#^D2uHE*RwX@`I}M4CNFyO1EU@|k(`5_<K&0~4m-&;&_;6O0-{GWONd+uteC
z@p^$T1pCRHLheS<R2*$Y&@rA~&cv(ycKjA(M=Rb|A{fPlXq4#Yn$1fDs9)c`+VT@q
zEu<NtAzbH;O~<~(Z;<6No827R#ZlUOih&KgJxXN6y?QXCq_3L=PCY7_$apHm5VBG=
zS@!{^t!^X<S;j>bPA~{!`#>44%!2Y8>ZaHmq{r~aMcko<Q%gm&#v78`I(ER)@n|1J
z2vO4=D&J4Hjz?guCoL5FV!oCzvZV;OMkqHbuU~?@{UKxtH|e_Xx`UzY&T&@`0P*im
ze<Uf~ko7zW)ass|jwd}!3zNIxF9)N+<@IHU?FZ1WH%P0VArO9!aa#Dh-}*9$g#OG$
z!RB~!{-8HLeem#Cohf$7FT0zUz5i9|>#=NrUotnm*Pj8tDFoQ@2n&H-_<7ypT<G8w
zR{et+QqVDv=>^X804R*2eV$!UAmK!QU1DC&S(~e3B8i5rTi=XCGgb;5@GVl}G>+xb
zWV^>21X0Qb1c{wt(xbB0u>&O_Rf4z5mc_7S&a3Y1g5n|wzpX8d^JF-bI4-AV7ML`I
zIR(A>h$fGvl*(gn@N%-MOIvNZ!x_3CRIH5OYP}q{G-{`klk@>;21*&6-B7TLVu23A
z;%Pg{yceR`1kqqu80$i6#lH@(kkj6A@f`2D<p~AI7kdfyuh1r~<n14!uTF6RcXv2t
zhTy36f3GgCTAdCC_`jh3DSVd}!n~j?FbEPDNsx%O;p#&fUZ8Vrx&Q6H=*s82txF6~
zTJZTNOE@V~+v><fPv5|V7qbU@vr>pD+Nbe(vYmlu#CT1sq!C+i2`#Ll9UZOGBUr_j
zJXTR1e3vJA0~2F%v@NEyK`dceJ1Sai!J`xd+1fJ`=y#{1Ej#a63U%9$R8uD>31}8Y
zb?fKevq!}(cBv-FbagOIv!}^9h6M^dKM4=YOh=-f>bd!;7MH{G;GEkX+1C@2*#JTJ
zhhf-j_z_+w^Ff0UDX14h=}8KT5S!6=eVpxhb=5;Er$M4hmiZj|SoUvL)<Rq@9V+ya
z7n4}K9(niqi-huZm@p}!e>D~SnFwMJVI&Bs&3cJ@xyF$x%4)v|c`_8Ir*C1H^Aj^*
zus<K-H_0vW?p%_BCAtuNM22)JX9L?-ZI%2$rCJ!7*RD>VIz!rDCI}p**@~4>PKt+n
zVM6lB=oDpATPXtyW~bxp$*IhFnT|2zNCNpx0zs^_4Uug*24z3)ue5b2eX4)sp^7bi
zsDJAv-zzP1be91O7OkB|)biO!YuZ1;Y(3_PnT+#zRYI?~okIh-7Hf(U{qKja0~Oqc
ztTP&v!v!cOmj&2uS?aEb1M#hR4uzf(^;I!ikY!QxHWd+fxE3C}NMs|+d_%UuKzEAj
zUO|QjfyTcRdBuo>X+x6+Nnd8Cxl<S%sHDklhodn5v)%tEGW_Hchv`Gls8Hu;;K;TP
zS$G(Zi?<OU`w0iblmIEedB}Y~WXIP}UhlG4dxu(*Gua{wYQ*;|{TPzKsB34d>5P9O
zXNUAX_|e`X?G=QVEIjZ^>T&|HA;a45_S76z99&)VFopDq^|4;KG(w;bfXu00zRl@I
zOr&sg2%Tb=50Q-n+_0;yPyiI6?PPu08wxqg1d#G#HQ5x^;|e5*x_R=*b-bDo_JFT%
zKz4(CByx?}Be{f+e1(j}tOd?2*0)V46J*)c#^ma2o*OCze^TUuLzDLs00aLv!KV!X
zUTw=G*GfE0+0ep>pgpHa4fwU43G@;%h)bJF9igo7Koqnxa(K2LrBsU9M0g<=N}LhP
zhN2x9?CBs+mKO|=ONIc6OvMF6oF-xn3_pRK(cqN{o1B2{q0G=aq-;P;JqR(Guf3zx
z2rrVE7UDw&tt@?JPen{y*Qsyx_sPv7-Hd%TwZlLshKxOI=GID8DtW_95!VpXjL(~>
z;FVbYz{5OuAgY6i%};0IBqX4NFj01aA-=3a$VkHzS<KirLnkkWZUkRv=m@<kp9|qT
zuzEl(`cC^n`+AG!vuVE2pmX0-V&A#|&WT*@Tt7GzR15mI0B4?xCI0FQek=qLm(rW5
z_z{A5t~g;Yf?EC7o+?e+S2B=`g&MCU`%(`?+9YepwYHxo^0mg{b}^U;D*8yv%Mye$
ziZ0D%ck)(jf?cm^RMWe_Q?&|Z|Cz@Y_S{s%wgjT3!(n*P)LD&$R<y0JvDPLgHXCQN
zTWa)IvXaT*HGMVVHc>dXTG{v?iQQLjX`>#U6Lsu=dHjZcl1Wy;ZJOdzYfY)Bq>&i+
zH7osHg>XI!q%g5s1l<~_q`O66Ex9c>em@$HyFJ-(%eLe8RoETde?tvS*mc;5JiNI)
z8;@>|x4D}*!gj{-Gq@=IiTy_#$xG~EqPlBd;8pha@IS}|!D?W~Quq0uOinMbTlqA=
z6n6NYA@Up_18w+MhUR<rS@Q+$piIW3@t^Va?CP3vaYht0>W5hRc+#ERJRKzYj?gk^
zgL5%7*4yt;C}%wB^ahjm#%4S2jkY(Z*PO`uW$)MP*Q3Fo2fg*S)z8_$M%}~vt$&4-
zhfKr9mz>x6<!{^_^M{%{vBnTLZ<qsQz7-gzmGD<K3n~z|;q&@B?G`p0h+l61bQQUP
zOQOO>VV|Tx4o_(YW|5e?<i+yN`swwg5B%UH&UkXO^8;!j$Ua%b9AL>lN0gf5L;Bgo
zfP3WF227*<PuLB>;rQ12aoBlptx5X!O27d&hPspI$Sv6B0nP7O$S8;#4v7<(@9|M8
zZ&JXoN4!Ng0wEZq0Rg(rmK)yTv+)pQysa;}377$EaY9Y>wN07Wkwo(c(-Tn0y5`|C
zIoC$9J;?Zgma$5t%CPI8Cal>8t0A!-P5j8KKPATw;!9EcoVciR{(&(<U$O=uLme~4
z50V#nWxX^o9%*g1+7H_J-#17*h;ME9+CEr%IuJ(%bE0UfL$hw`Wasi~hPxZ|EvB;s
z%h{$X(keQS^Qn|zNRuTyB>{arpmzlunPV<{EVB$073gT0ruDS$5{jYdwLW*~W~lVc
z81(TgoXfj{FzEMw-yL4Z%o#QniGI>o8PWc)!IcSjY@SO;B2%}yXQ=y`$Fa%Q*3QAn
zi@pEa-4{AmvM5T2wcD6fvQ&HxTIwTv#!e~Jf5zR(-I3(8wtgI3$(~|sy}h}nM{+`X
z#bz6Vn*y5^RW>*Alai`Ggz5{ju>`5E?`kJdpTmCtYW>#_`48Qae+MI;_J?zqiL#LV
zq%&2ewE!m)l%RI3#tJBAObjp7k<^K1QBNR?E-7wj7tnfbY{>uGepQaqn8~&Q^@#uD
z7lYB+cpa*2+CRCT41>^?)yJ`t*(*E*rm>NG^6KFE$^Mg<-|{E^v-{%N-pS4%UhceL
ztt>o5ms|v@{<_&=!ly`%VTYy35+aTeZh)fnDgrPEtU-cOc%(@Xsqe_*6O10yqjcMq
z#Fvz%Z2#p?lHTw&GO5<T-g^K3{qg3;*ZR@1Bvq{+{prE$&o>@~|APMhih9@^Z{S%x
zSl|47W9_E_)H6ZE|6WoniYP&5DIP(Tw>^QPdo|dM@H}8v*i@voeszS)B1C>~S~o{u
z*aJMaCcl&?GBm}!8y9uBaLN~Uvn8%;N1S~WM$ANcu=)A%=DK4C$UNRW|8e8%|GwD4
za$<^-ey60O3mOQ;3CB^wRRdapr@v>cn7mwuYI06{Gnl=(K7HGtG3V$O%oe^omSm+8
zeGq@bWs$Z>SLDFs%{M=Ou{`&Maby_b9%0r-w&Tc)n2Zd&ne_E5?*CBQnG|++AQ5xU
zvnJJCHj01@o>h$=$c%D)Y$k$yN6}iO%T%mW9H*Oz&&0?WGP1sEsqeS7r5R2LK|&cz
z_2u7hAFhkL5g$Drix!&3=O{rsyI6lHO;Bi#y~Cxkc}w^T?#9V%d}6^f9aeO~r*m=@
z8^{dnPtLm7SjT;Fqw_xRgqX~NFRYFzo0iTfkoJ6h+8yd=)43|i@c-%NlKJ(eFP^=W
z99$X+dU6NLApg`ocL)4ClrK1!%Tt_I<gSbG13CbDl-eDnVJS$(%R|prsJzyw0}@3|
z6~~|{7CiJ4P5$Jed&<as#K1AsZIdXMC{5Nb?(!g02n0Et_d)26nwEG=iYAOEMXzEw
zr+NcO-d)yxjdhM5)BeBSk}@iBbW2y%nUxcR%c%HM=wWmvgKMxuoTBtG2TXrQ1{BXB
zku8m*alw5o{#_Z#D8k0!uMZf7=GoM<^%P=xl#W(MLpG|kZ63D5$H3*D#ESMO<8yf9
zh<B(N?!P&148W_?krE;U`SG_H@9`(A-!1B2P;bJpT#GtzpN8leJZjvzalaLx1i_WT
zBxH-0paCf2*?2g9-)9YOJOa;|=NN&W;`Ug)l8T7&=`C;y*%(2xapuE78^RL}DkwH^
z#8i$2z7Vn(FYn3Z{Q44co)BDG4^^jMjoyxctX7-^Q?UD6t<R>LttTP5n$M>4kj6QG
z-4FS$c!K{IQDc{aSHMHuM^u$7v_^a;JO;il*(^iav+*SCYnZdstYL&R#BCoArTB6X
zYHmP?F>f6vEkgEj`x2?#X80q+i-#*yH{c~uqE5GxsZ4}R*6xAF?Sm=+1uswnDe|(Z
zR$WFL4x@-aE(1+PpWNwCilt&OIRDF#*h<@01vZ_d3dS8$nQ;LYQ}^GnUXP&!c{(uC
zu+=oYX-m0P%ZjODOt<SIxi+pFQJcW|e0)AQtunw$JF;PG^|R?JHcWAj;4{>L$BL$G
z!NWt$*;de7Fd2GG&2~ueN&Ef^mhD6g1(<L_2ElJnel-G&R@LdQ6pRyl%~_?wE8eh~
zC&<X%2?=d5!QZLo104osUh~|x@e_^8YZ|kLTTaO7QwUNeQc{#?Th&s6l9blakZe=e
zj4D$J64KVX%S;{dwc;?sVfs^djy*|+_>;CiQnjV9Zs24@!5N{ozLplX^=xo?4a@MN
z`wsOGTiRm_N4+^<Yl?8<fYlu=)I%VV|C(YOK?V;M>vLj-2Xazrqk;tbns;6pFYf9d
z?!I&;Q66Va3(ec?usABvrO5CUadDrvpu>bzv_P9TMx4SR{}UU*hl!`D^a<C1^R<kv
zei}xu3pIZE9>w_(oqYTPn;#me%$Sm!#4YhCwSRh;Uq0}U6v2)GF+>bOsRppAV*<9V
z<E$laU!A<JQ%yb!x>2f&;(^iB*e5PGvsXdvr<^K!)Ptg7+o(LZX}d=<5>n8it!$&l
zmyrjTqlC=2jS57sv>DV5n0|%>O@-)dZ-%@oVP%jPj<|6TZ+&AEpVzIz7SzH@HR7Bs
zj_LuA<<^_|mRJy|+dWlC(g9DGA&jabCb9Zd=w&#i<>gqv-KyF+wyL;<I$PGn7O0W!
zlz{eqqf!9Za1m3!t||$UT-KAz%(P=l2bY;9^SzlOLxxXV#W4SzZV$!aSln^mj=^iK
zx#MGV&*Ghio0php#4rneai@z}(6-u9qhUn~D-#k^gi*Hg!DJ6&-aAta(azJ)_*)R|
zV8d;xlNL40QzJQWX^JGYRAFxfZy9FyQW&}wBt*}xZ{R4=SIX^%or2lhU&_Jh$7Fd;
zxk2<o7`}dk=*tJ%o}!e5ngY<BzTM_V<^+*JR;?5{^1U1-#d$h8_f26<sXFN!+PF*4
zL)=%&wqhUh+~m=hID|<b=J8E``X%!Y<j3o;>3c{5Ln+Nfa3uVA2h;*bN9AX;ku79Y
zFRP4Ls{+a9nl1c?OC;;f)+0fH5N{p#)Q!)1=r$84Qxia4WLGp!vo2U*5Ja;?iw7bo
z2!W2OH%P2iJBKEV%p<&NVbcl8C$q^3HVlS{Qm(g;kAFd*bd&#!&(2T&e2vo$)TYuM
zcjmZ{q68}qBy=4bU?B`3k&CLwgnvp=9D$<oMJvq>+Ek`>uV-U-8H3>^{aHvurJj(E
zKQTJ9Ww58!?m_-ekx-~KALYpFrwgMhX|D`IO!#R~ctOR2cI97<`(7h3L8$hxZ11!>
z*M+gbaZ?{3Wgz5zm*6eI&+|K2F7CBdVTrH^KxP)qHa4uuU{j1G3;FJJI%hF!qcN{W
zJk-+f{frTQ(J1FdZk?n=70;eQ@gYO2cse)o-;Asrr^D;>!RUBj{<DLOxEac7)V7%P
z8LoD$O4FjHJcH#0N<miF<hb}<w61JTWTVPk9#5Oeu=(}{mRol)Qix#d^3-~tx^!9>
zBLyo**>5`>eo{<|E)K91NsGx@USTA9h*`5!rm$sC@_;fdV?E{=Hal33CnmmS4u5FV
z=Tbym2Dw5?FPV4l#)s0F7Q`N$-)|L*0_Q3p3E(nZa=#@I;yX?+JS<lPiQQ+MfPTW|
zupN|iw<8I_s2^Y1O+m^Ga>US2SO>G{s=$@-2Gdvp5qPk@VjSaJm?hjEaTqXrTSt#Z
zc3fyel>DmQwlPQ@Pl_<Cm-wqXUwy|kspL2G@#*gL1q%A0?9k*U%W{$r*g&g+hKyBx
zE)aEQvmtYu40_S->+Q6`W%q|EZh*&zdo+Mb(m}+mPK|+8TzLmQTt|6JYc;^BTmg)F
zimsa2!OD=}oh1-O`<H3^&7uD`tEimAB|!ZUH?pU=OL97;?ad?`<|~=ao`j$b7_q*b
zB(n=OQTwKYyfQf?z^=YzKaW?-#2PiH@O1Ud2hyUo9&Gzc<7Z&2eV}!A^Lwr30yj~<
zbO&H#)pB`d!Eo$tG9+(BODUpAPW@ZOT1zI<7?<6B#81^rlH^~vxpc0*u}kqZDntE_
zWx|&nJ~(J0%$$Z{BF>h{(Ies;JTwi!Y_*Qso0y!3oBOb8Uo+IMNqm}BxY=fyrHrx&
ztW4u0HCbFJOuk{zIq*Pk&_SddrdO|lzIa(d3(0IrTH!p1kktlTU6Ull3>G>sK}+)j
z#XAvkp*+$|mXakY&7+<QM+cK-HRT_y*g!*A;AEZYnCVhRA&2NEW5bA}YgV{;63jS7
zfo!0L6i=Ge&#-3C>|set@&%xzc<c9GNxqwXWl!-e{YVLK?E+=gMj>N9q!}h1#q7dw
z_YPm$IASF@S&NPXloTG|ID(kHT2IZLt*!loos(y~&v#Cq?tTB_`QDSKz>>}nI)xZZ
z2FGx+Qu@A7Pszd+Zx9GIy9iURfhcj#WWXxS7#KxwP02NOQPFcmkA-@n$4}o*<u0&_
zNIAUdKH9eT+$K~l_8<ME+&b2Jo3;QRrps0!$KqigIuU&&{^VL1o_a9+;N?ArY}H+l
zrvL?N&xV2!gEyLW*Li$8zPfoij*Wmar_JEn`X{+DmaI@2ED9zg<bBS-tOUC7sD~jL
z3#9O^2TCp|C;=9KDjnrO@&ot1f%TWXbeCyLdslYVE*7YanukWEytYCp4|Sg_6{syN
z#J$1Dv#`=vwgt|d=KHN*Km3CK^Gp8Eueh}$3t|`{LQ`306DpBAPvjyx=iF)Wxje~Q
zV`7cxTgvq7C`}c&LP^QG7uI(;qnxmyq(V3a2Hj=!N#c8kt(bc96H{wZUiotz(4ti#
zdkNBQPH>_zNYS2epObg86Wo-skJIRrA$F*tc4D!+DYocjpBHM>`DGI$^V)|Nj{-4T
z*<QhCNZC{A&mCmmGcHx>ygH?9^@+1R>f@-_cr-%e!8=r*OHHi6;vh3r=(;%Jkn3%-
zH^vMAu$VbSfVYnHB$i*}{KWF}kF@030OyWaArB`R7<J`vA>P~bphC|!a5fb|8wzo=
zxdYx;GGce?+j#J`sv=#G-QskUTV}o<VlmYCc#GpJQJLq`gxWfP0+ryya&9st#CUo6
zWZfaHM#F-U&_po`QB%NVwP}`NJrUrJNGbtGx4Udhb%I#K2q?SdHy$P~y`+y1ym4{N
zs<ADYMBG5ah$O&5r8`}PuRu987Sl_1e&qW~6SwCI`77>>Y=IQcjKI|Fz6e;EkHYwM
z?&1KbtVQa@I5?Ba-j-USZYA4#JC`QATX{S0ub7^h#dRL4joe7LqHF7qkS=AB3ub9`
zy(=)AYH%9O?(E%#n<%!RD2b(oQz2K&J@)nXcF+`XM-fJwv5s*MclqKHVM%)8X(8AJ
zZ(xXEs&;l~?zT$m9V}V^$`2PxTyr^@#6O6yk2rAj&6C5O7f-(15tNZY<X?@nGdq?}
z$vjx;rI@I@2fq+UXrXdW<is@&2_x%kG@T+YL(!L)fy9T{$I$~ndIlm<<JIM0RZYpO
zu>zmeXx4iy_L--LP(iic4Z2zY+ElUQ6vLaGubu|jcC&uo#4|e|81VZJf~5}8q^5h;
zGSM|vj*vwT9mpB8%<@J2Gwk(KHs5;tKo&cOidBC{3t1MhH*Lj}%9vOw{L|>FmjK6H
zc_19IH@P(&Msn<e{$l^#V0`UyQPPW__DDNheHKg#EUA?e-)gZwxQ}B)P)J#yLdPF~
zi7dEo>s+&uk~NbU{JuKBJ3sVKvGtdg@qWhIS3|em5r21l{yyD(A+1zn)*lqC2*#L3
zY&(xZY8L|>-vQTtbF(c$WU+jKFtyqSV<1c-X;`viDqCCq>1h|Kt`_rxlXo_+RZ72_
z26Ets8W6mNdgduFt|>Z^C=2vrd-F_?gq_R)Q4wg|9*&vyg&Nx#?bnXLNFl9ouo`v>
zMzgc^w)T(Xxv1K}R8OXE+L6P?`8uG4nrg(HLq!>I%9st}K(gdOkBy3M^$<{RPcrDm
zQkyHpVtSSJrRi%v3z!Y8B`tx{Vr;06oylZ8(UAnbh<^Bo)#+0->eN5+r0JS=hg3`q
zg5W6j-xC(pPBjt$HGkCd{_SU~#ziTQdt)(Vamla7NRl9DNS&&_3h7c@G-VLV<-e^&
z@$4&5do8=W3@ao?Fyvu+LuxGOML67}VQ|F1L+(H733qw)4OJzvt?ZN>LOJ4kS8Wn*
zwrrkFho^7hPpW^7@PO~Oh_8@XimdbP=bJG>@UC7{!MgKg<iV_igM!moU!Im;zEv}a
z;jehP9*is#f7e(4&8Z$3yxsXj5O?Y!ELwQ86S<EocA{ykiZTfLMvhO4uF>E^Ra`{V
zPYVucyNwoiQGFO%2Gz&|Cc7;Zf`bk?phe0Gs+OvKoc<CijM>g17G48`iZT6G2jQwg
ztMff<^UetW^QI4yquQj-OuD7fa4W^wgoA}{*Bz@!)TSGAADy=aXkh^UcyQc1?rnwt
zIi7xgd?bnF$&_iV9KU{m?~>H^LMx*_!~oX;jc{Nu?4UC1ow(&pi82_f=O~ej43KpN
z0GO5c6t^C=RxokTfV>idb4La6%kO&1m`BFw-#S6b9>&vDu5n@yg4oi#KB^5)p6wkT
z?mj=Uc;M|_jm&bsje-5!0dD@7w(uX?+jU7Y`#9!B(OCPgG|w|Q=4J@ZPry-Y8q|P)
z>Old*MYevHfzz0S6_Q-AnujxlhC|X7h_yUGj5crsPrDObfoG{%HLXx=L9sX2eXE=b
z)=7E}1vVCQY|`x#V;CdWWVkGH@Q`B!;#Xe<e^_7m>ugGQL7B@S=Lp!XG%E18rF5X}
zLN^&tqBH$XdRO-rT{js{zI-*NlQ}m8Gc%A3Amq^?s6E;lA-5f%vkf4&u}4(fL!_GQ
z<FOeNXe2RUK8h`M5)!)fToq7BsKAC2Gn5`7`SmYHf9iIAeAxN7&DT;%rQI>#hUMBQ
z;baKg7Q-ka^<ejMK%0<^RWYvc>DjPLLFv4;wKbT2kMalO_pCgNRMZDg;jAFzfC3R#
zB9_UrHsWk->12qI(t)lor|fyPm3Vfx*0A#~5I_4CG`~)F6S$<zOnZk-g9}_Fszj3s
zcekXk5kmT6^Ou_sJCoCgz(xD<R~k#Qm8ir;YxmBc?tJs=w~+#ikAi2(w0WNbEKLpR
z<4zgTmIbFoUplNhE9EiOr`Wjj!EL|P5TopiQ%g6a)9(iK{M?U^{VA@j99yI=-7qL9
zAxxqJ;+js!rvu!eflFzCbFB!3SuopvnQ7c^MX`G(maqT>;z8Uh<kFNTwWicCvXvMw
zEKFqXe5W%ScVv@`2Ddo(jK!vN-E4zkPJ4Vzir#F0_Ql$ihB13#@Eg<a_Hf$M%<tA`
zkK+3?nB^DY6&XsGqo4>9G{_|Yxs7r)vLOg*eLcd~A2!`Ih{BVk_!C%;f`qm{VivRy
zOg;PJ_@w771H-LzUQ-UudruYv<UbMkg3>0}BifjNECe`wUVvSqwI&KBU(Fa5&+Wiy
zvO%xn1u>{BhfkgHy~zOiwR~s2)d(j25?~hl#@N8xM@n+~@wH}WS#}{b#(_#ZotzH&
za?FVhSi1q}!DM_n<QA2>eIr?{`1Sx%e^d#43YNi)+_1|4XJi&bhw56DGvpPK%Wh&9
zm|EV&n$Crbl{YIpN*gxmxjPIH{8lAXuK#Z1Sj`QPiAPON%{JUWrAJ**CG?aTT%6O2
zT}`djWcEr71d3!bnbLVjRqSq$P};@{#7<)V<1ouus2~2u`q%qkf_5JtAAY{^(`ELl
zw)sw~qg8uwxi!BJZ)Lp%l1=(PYGZt46$+(h?fng)r$mw{g>>XEG)F1!x|aN2#i=HI
z%XTukPfUjL`-FX(1TRbo=43HCOa;JLj}&@gyePLLsLcVp1&7mx0;hE_^C0D^MuK{H
z*O5|wP>i^d0=XQW4r~uz9{1VwPk&*5gV|4#y+s<$wS{-0PdlAi_k0=#*`U?r0J&Ht
zt<|av16(E5C(#pFgOAk%z_6HVx01O@Xl@_MFUkDHWRsT-5;tj)OG_=Wnf6eMNwtc+
zgZbh%X;pTUN?V<OLF^*dQ|(h#c-}p!R!-Tz5=|+lk+_#-g!e;YT^}u{)ixnefp9^v
z!b!*ipJ6wz(?Js6X@A(o?VOzsE*C~6<hT7^M_jM3Mq|)CDP-x*lw5cck#lv$T}O({
z7;BEr)hx@c&qyIAQ@b*vFhnehK&$s4Z=+`CE#JlXJq>>>+69l_c6$F{D@7H;SVOi{
zN^JLXeY<|t?VQ11Mp@F<w>wXsuKjc=6-e0uB#wmTL^7<vH<(>dMoRBQ%YmvXX@bIs
zmF`biqf*YNu32+<I_&oWe@?C5X?3{1o<qn6gg>npq>}<<#Qqdx;>J^`vFrRC9-3&h
z7+yGnVDK(b148eoZ{c2$N4%*1b4>Ry`o51`2DFKtN)2t)y6Fvgi6YyPta7vktm%Z*
z#KDwMlhX}?zp5#*@xWpfb__RmjD(m*Y$vk~A7G|Bj$RHQFd?`PaZ*~iE&iYIf+f?X
z->ICh;;e1mN|@J>jX)?gbdnG%w;%J$aLFzaTP^sE*k$fsU7<`nst1h`D95#bdOi}&
z6CIUkbGTC{60<irBfBxXjMoxVdZGr`<|b^9j>K@%DS$8)KvS{*C<3&GQZo+CH#%$$
zD<9G7)T$=9pz<&7rq<iu^DHo>&&;wG?^rn8=%*Z<1E!>ADxqwqUDWQ>NK|my6SG)o
zwH~+?@ZoA9iU=}Iu!u|D<jy8WHAIky!rW-T#xN_3jmt)|SD2vHYkg-yOGJ|xi$+w2
z<wK{~PM~)nAp&A^t_D}@C-L=!ISCy1D3~N2n5<>}J9D01Pf*Ajb;fWO;PiqAp?LE<
zl9-Jt#cvkLb`Po40$PTgZ11qd3;|EnL8^8c2`VrnRBC9jm5)Qs$CWA-B1_${yq5@u
zV?KW$BFkFqstv(1y=XV}q_Z%33lq$4pconW>{y~7Mk<j`h4P&%l+t9><U<6Y(pH}6
zl`+Qr{*wjMGknDblZeJ#Cb*HxBMA5LF-Eqs5VAD~{d_$A#qs28q}5D+@e_rC5rgC>
z=>&@4o)$#X!Y2$z7%GGsaR;{)@6wyPx6=A3#v4Uvb_NmmY%y&}=}_)VdGye8jqD(t
z!-b1<@eb=#-=RV$IkRmPU5C*f`rTmJr-#Bhay`A6;k-6nK)w$S8jlf;Tz$7n1@>-$
z-4&|2GbRoU5x%k$o0cbOXDNk|0nPONfT7V;K%fGQR(;X4pifm01V%l&xj(x9?tbrn
zdlN!N>fvx3VR~P85wK(N_3N+wlqgoF{k2B!PZ4QrwQt&6j)4ZE7dhSf0)DlAF~t`p
zjh~M((~JJ_N+{AWT1&8HZzJ?JS9ZP98A<xDR&z5r-bKg@VsHWgf0&Z7PkQkaF3VFG
z3xT)umS3Y~uYcCP9?rH_II)1GVk{xNb*=7o1Hv#QhSg!x9LyP{G?NDMCDO%`Wx{1Y
zdgd1=tF<$@dYbdR^oKBuIT{kDTdWL(kjynq=Np9dv4MtO5Y2&{ze0gUA?|k=K#0>R
zSDDRU!(GwWLsp+Q^45bc)7CoeESxr#!iuWhWSJK56O1qvhK2CV(8CQGkPp(^98Z_6
z?ip&<$jl8%BJ1X`>euL6G82Ui$y>IdjD=aawwmu@W9Cv7Hi9Vbs9c)8M1!cZpPV3~
z@(%r4=yT3|l-CdX9YG2Mae_D;qcQ=OND(f;eB0Hd3BWw=<gMCj1a}@6uErREE>Z!^
zmZVd9lrY$kh^JJODH-Qbg#gT$Fss>N*EN$cO|1c_bWykEpnC&QL29$$4{^Y_tnlOX
zYV&3QqWM)_IEdBGS-LwLa?yF?4U#38fh)KUduQAp<Dl3dqYL7}OSz7TaRHcgx-$m%
z<ToqPUlTtu?nlER{8O+0uJc#9#NvmIuaUy`Zf!$1r@5ZNk?}tIo+O~Q&k_N4?V{D0
z)YFFzA^lqXU8KG4QJWVJzg*E@apUR2ySw|BUp@S3O{|2_USjxRh-x?<VR-k53r{$9
z!HQH8;@6oV0I(u_*x)PHtWoD1+w5#wDeiHKBiXorYYWl5ci2wm!2pZm%dM38>*SGb
zEO<4D7cFUV-YFI8qYZWFC!)+*%Rg#Hyv?+VbrN}VC-3q=Mq@9sbo!QD^)3l0sFks3
zS>5Mfv{cX)YY@o8H?YOG!&XKXR+vvJ;Fm>XG`Gme!H@xnRHn<Gk;<wdRBN*1b6Qw|
z2wOl##dIs=G?#C3*|mrtJ{5&ijZ`T>uGym?ZbmqBUJi6y<|B|y>WmC$tRre&{BpmC
zrdYdDAp-FUe$eGSiy#!--Hac>=0hg*IgenhBcRQs>ehR>^OHUgPvex>*BePOSTUmJ
zD*6x{nRZzX^v6E8OxjWVfAnz*1p&ld9{N>%UfsmO79yiIhL{Q$WNfOzJyb{6okxvK
zL88xNQ1;JiMmY~bB<Rwiuf|k}npre7#{^Wk=3)`bT(A22D`;{9_kFcD!vcU*@b|aZ
zZ=Q}XvGFrK=%4q0aJNvS9ZW{c?<V?sL+k_ghFh+v$1y$pZ;Zb)8E`f}9czIfonsa|
z!M7LJZ#tnFP3biTU_l9H-TXU@8V}c_I8ONW#cMrRr|*7=U%ZRT_B>+2hxQcF695&4
z&~z;gC?_L%4!;O~A5W5M7mvuqrpxTO5+Lyqv(pFm)i5uV3}>dy0x~hUl5hsVFOMr-
zqJB?)REg7Mst`4nTC|+C+;kZCq_Nc0JIHXpP1d=D??N21-bJaAkYb)PSDSHhQIViY
zY!H9NMBnagYxT3wF5pXj_E~p?t13{-pG!3-5gt=BWtCJ^slkmZR`iuK8s+Fb&*Ls-
zkC+V}R8uh2f`yk+vn;joklQ$iOZa4|E+wLGhn&@Tc3L|!-;mzeR!G_ii&1JHmBn+N
zG0!K_?eZ$=*2H-37(*H_=yV37)8TcmAGg0h3SHXk$D!-6%phNctRm)g$B1G`M)qDm
z3NaE9VpktRLq%-Ns9DLZhQaD&7LY=#tUfy$egzDxVIOz9iBv@lS3|hGTX>(_8IpfJ
zyMW4~Ex@WX>HS?6YsCdXa<lT9f2KFd`Y$oyQ2!WPEPq+iz%pujTnx?-0=8c7C$|~<
zXQCSX2Zlvv9k(IZ<0v9*IJ@vj_7Hq1mrR_RH1#Ny-kTbafBGP-;-_~jv;HV=z{M{J
z5d#ud4Sp;3D}C3CJcK+NPw6d7uAXe`Nal247l<j)ND2^5d19G)#YPT(ijV|0m?Gu<
z4I=DX>^`@sj*x?ZOoZ1>F0SFhu33+EwC@#UDk7NXeEP9R-?!zEcyW?un*|VxXHr8B
zaWUWxFBcWF-PaAI<xbRp@NphFv2!3v-Os{#j{@%U$d88Z>;w-k9Q89G$#H$V$vM8k
zaz{Dg=T=}~lo>&3HG_p`YYK$kvheKwqLVo#TEKsyGHum!5ea2yd#tK!F#_^TSUbS?
z^U|pn=X#1#f8cyc(5D2DgFY}gBQ`?@Bx)rfe+;sW`&MX6{!}<6QIzY;M=nYsW1j9F
z>}<c>JNV<t;m-b(gD3bCPbq~&z&SH%e<(@O`6Xqchm&eWxdeBj``$hWfwVe2PJo^Z
zgv28r96^L01pnNIn)&BbAuzXc^{dqeRBi+TUM;}#<8bgs_tnD5ZbS4dmW18;AtJmA
zT++J}y~NG226t%x{${SBoeivV`t~!K=i+cRq#luptmOE7z6(xbO!bvGl_beKt;pkN
z0ZYDEv|Ln!cD*!bge5xvkwqY!pV|}90mBD@o6zJ;Tf}Fa4%G8|Y3t>5VcT${hdn~s
z<9J9i&i&S}9{$_^5g+-Cp&i&2u{UtBev7D=$@r};x0Lh^Ir68K^B+;JgBMXHJ>$ih
zJ&ip|>yK4=x6)0yn5Fpe<DcK2M+Naez0!*YYzjL&JMd?`RDb-Cgc+r7h@D-W->!^{
zs0{R_LUl>MzbaIJPQSll+W+m#H~h0`$tz|iP_81!;Abyd0?zu`PFU45w$v*@oDaS;
zfC==jabb8kAH~Mi+8aN2JkKnS>o-XK$|9;+;^8tT2ZdRP3{b++X}vgCqt5J+P(Gw-
zJE9F7<N6>dWRK)Ux80?t8#i-uc54my4+|65Qw0OSoe5&VTM`1E&Te4JpeeM6K>sNM
zz9lApZc=f)+tP?J4<n<-MOqLBF=9vq91@8UYPpM8fg4UwuP<?%@9~q#q<f=(7e2Q)
zEa|(Unbit7w*UV2nkDj4SuZQy;e8g3_mPlaX)AF7H#v)b^j&Gzaj3(a%d_$5=6IWD
z9!JP7KBgtDMI>T5!{gn(P%u%up8H+wLT$%jKJJ5+YYa%XPaY>X_@Id3HZG}tlfDnm
z3e10Om*Jk+!<TS`cYX_>_>bQ`-+l2PFxwwJ{7w1tyWN8Wc*|e4zQ8;FD<c^_s}H?Q
zWegQ5=uDDD>4-pz<0oABc!nC0BUUApY8m7>-L{Lohfhr8>u$BA7<fY#SU&?|Liuld
zp{3WI;Qiqu26p>$rELl5lz`!WXkG!W=RkP6#-^aVaN!huoR)Zw$map<5c!nV@M~~l
zc0JBa<ZFGnf*}#ZMQu1kk2Ojxd7e=kCuB?la8YAn#Hm>RrdSzr12zh@qDJQcS3*tY
z@L>xTEO<9IvL>?y)0tH?X$09PO?SMg$X_mgD8?Tu_P0LQJBU&L{N@tnklrG1{jnTu
z0wxinS+fNHLv%gspYz&<X@7XOwN=tLJO|{+jNkWr)^yNkhMDHS)_-!p>`J+T1Fv05
ztqt;~l!HT0P;Y%yd1-I&eYd}NxN~r#txxu!y!`g$U<c3WS99gdk2RR22Njy<q1^@Y
zArEN^e~HwifU4Ka^=uO0129Btmf8!PR>F@T4U$iG`85k`1NM^8;Ni^_S&dj2mIsQ`
zRRoK=p_CLVwY^k&wKBafm`X0eW~~Ps8LRgCqt9?zsifj3w;b^qZ5ybM$IL{o<Ei{%
z%{53X*dr0Nb0IMA!StCWf??kBLMg7EzCk_}{t=ldS08anT{UJ!e%K_0)>0h)KRcw(
zZ6*=ikO3mpKv{Fq)^7dIKE;NY6oJvga+LMS1VfdN?i*ZZem(1x{V-7(L?`MSEh$)o
z>64^AVn@j@P1;DGC?}u3?@8V9`p~frh$M|Es|oa{Gl!%~SZAn>BwIt)Djz>eNvxoB
zi}7{8On%*Otv*=gD^5r$nzepJ65=}cWc1~64Ksv-bS|o6?VZGmNiMPqx1SW0rCbPD
zDza8FT}uX1g}$Cb(nB@|H|QF;K&A@lTXG($vS^m6NQS3;FFS9uP(cJ)lW~u2hC9tI
zqgLAiLS8seT^Y2g!hBD6H#6Q4&eh}^IjdTW&?ZjJ$U^6Td**`;pp~O?ib$}5ym6GY
z((^c2+k!12cbb>A-oMoy-Vd@F#R_edO$-wlG`R?H0b@JwlwqQZ7fJT~ngW+_a-jNW
zy4p?%gs|4hHbrF3<m#6MvuU-yWFqF{H1&p=2CXnR+IcWTRT`nxG_t9XquGU#i0xb?
zMr&a|9IKJ~DuWpgXGxEgQ}fs)xhY|Ss-kdEc;$CKc4|sX-dyS$sHLe4<9zAcj7Cua
zIz5nR%vI;Oz+8i)ghX21sb5Hovc-te;6y@1)21*7(;dsTb9G1ax8|efiI4Z~f;GY9
zjunXaCfGd{S<aCF|5{I+XzMj|rGt5(d-G^wjd5ujJ_@HZ(TmW!;l|)*a|0Q{cEWBi
z!gt!OD;Ba$+}08-+XqL?0Y~~@8&wD>$v$@9MLh3k!8lXLQ~2HrYVPB7y?HR$nB;Eo
zcmj9Qx5SVfHA6IQZ8C;^_dW7K!Q|GQB(#sH;ot{MUv2;P)df!!wWp_(!4-o??f38B
z(^L=riIi1C5&PN`y$3=bvQZJb%l5k|gGe~SzHmSJTQ}N79VXn;d|%W4&Xa@SUa|vn
zV{|^jtvsW{fz$>pO%S7;XK0~wC@f&o0Y8_o70dYFa$2?-%**3+m(6ZEGc4!Wq>Co0
zJdRX-@MbWQi&x*VHiDynLm11tM?t(`Ny+{rVwf224KzK{aQ!|$+x)YM9q;215=yRh
zoq-G&zDLM1TtXv+Js$^rh(oQGzKj0zWl}_fv;j9(t^Bn8!1!tNhf7>sQZ8s&vT(h<
zJ9>v=pS@PPpqs7TnO;|dsLYH7a*|HvK%?CkqZZ&cu26d-bpwe&lksn^2g6?SrV6TZ
z0vMT?r%dsw;X*c&g_2~_1)7%A4a?rp8L{5zOy!-CP(x|Q5F_J|c9(}ZAs3VJdy$jJ
zjlAJ*SC(WTYn>sPQ4ZLMNEmJ%pg!N2iNRX9T59MJa!lCCI)Wj4kl#TtcF$SQPjal{
zTVXdfZ<K3XGK3v@x$rLHmxW|Bl#-4nLJALLX&@?~0j(5^U{tEG@Gs&d+;-*bXQUZ=
zWy7O3J*f6;S!ccXN~ReIIZRZBJ`(0XYV$|?^_;I*l2`kmjjeZ_>=rt-$NWhhP>>g>
z3jlG>>;g%rZ;1&bMX`JvF*oPas#L`wAIgL&)8W`O3w3Lm{@DfOtI1m)B%Z-WMx>09
zF_TMyKTEj_xijM3fFY)^^8-SgQye|Eyvo^F(u^o{c-)B<W_U_p5TxD`2xx!-`RWRn
z4Fog_1M%)&(gXzuD_p?g$LgQJt~s$sRso)bU?(vdZAe5fGA9DH$*@9g{7a|}zB^7a
zk3`bonG+Ve!{Sks#y2DhwKA;K4wv`H+r5_Qhz%0H#DuANPpA}VwM(=@0$LUr6>@2B
zzUj_%3oDxgFP{|Pc;lp`P{y;S$NWZ4kH}R+1tgR`j=B^pS&u@U5qzpf{$GM?069ac
zUBfi!IqeG=XIoo5S(dBobXkNG>{~9Ts-B1f`?!#U6_rvW|41o8-qGERL?Z*G0xQ@u
zL1I-^iHjDol}y<?_ksn335sQy${qxiq1@_q0t~?yU#}s`2I<vKxRCl(=_x^qy<dct
z_P+W6xP%Ryj8kB!(?lqok3|aywzW<Vk*j*yO-y5Ze$;-aMWvm-+mALMvUW<?waC5{
zzB#>sLPpgYvWIoU0Hf4~?K~zZETe(jZmpddWY*Bwin{~A-bT{F2n+|-vvv{Enx<ha
zgB_pgIzkpDK*?B0!S1A<mUS#LV~X+-Y_AI@#BnHl0nY*dqlc_0qBAA^)7Zv;!$2^k
zL4RssIn*4HWGEa+d4QtabFj%;LTK^mL-6e~@Drlhup8F)lc4#F4@a}!oQXtiobyVS
zUCbJ;guY3sk5f6v+V&@ZpdB6>t+z*FFpa`jM{TtMjMQErdp{rJ2MFj3ezyfFZ>==3
zP^=Lt!*HNH!H^Y6jZ~yhq&pWgPA^b)iNjP8hV55Sw^zLG87gxs?Cb^Md9$tYO(_&y
z4!CElkcddGSS1<E|9cH?CqxTrz7FclluH|kj!hnj$e8p|M|RTh?FkZ-w>U4Lg;hY$
zJ&|RS3X=|5)1>owylk`1V=gKx?nF5<e|Yw7A8D0;(d)?s7=ja}+aALl#Ay^F$fw8I
zk~IgTYxOwoYY~oI_9!6PG}<Q_P<1j?@TzOVFGac-?o@v0yFmR`b6jTsRWf^RcGTtz
zv1V}`x$b>eY+VjxCXdm>k=$3SCXdsIJv(aCP9iII<?`S}4r+GcM*(UV_ZDCz+_TbB
zGbX7l(_jKLN_q7i4l-E}2FP3r{7#gaxcACcn7D>qpaQUi6xY_(h%BM};0kKcN`5Ab
zM%CMf4K1ge*6@eIN~AW}6_K!|FJD;z!OvexoPnR)Tb>Iry5TUdS&h&ps9$jQNqcpi
zq!s_vHqW5sg_Y;4&Qv)CG&**}-0&s<s(-Udfo6=GWBKH=KY`zBPxU(iD^tKa>k_}@
zFGVo_;%Jy~1&^Gr;fH@__J2Ys>S3}IV+!{9f<h<h(`*S}96_fp$V2#~r~xUy=fbs?
zR_vRr%W>lJ=G0M{%$#Kh?i^%vOOB6+!Q~C-=E(wT2sK3=)vwC|OPEE{u*4dRO{j}(
z5sxZRtB|5dv0$e7{tUMXV2UqTVvsxGp`rv2W&OfqokUywKxu5eN^+6^3(?<sI<hbl
z3+pAJ3LteVwj}=0$M}N~;XesRK2m&$q!4Splp~ZdONkGc3yNyeRba+M3d|uN{etLC
z+6(Sgw(w7F^9%&rEHwfy|MpU&P4ZM_Rf|`?wh32AwlecXg(mb-D>WAKe`U0|YLTP3
z3=PsCh@m;_i9q1OVKgunv|XX<k_D8Dtx-9H1%o$XDQ}MYvj@Fd4`FZg(V?@s5u!4*
zkAiJu0S8?LwpQy2@gK@Rvu<?K=;^3VM#O2v?-5p+4BjBQ0@pdg7L&-Gi;u<a5z|s}
z@jpgNnw9~IZzm{2ICj`S6VGRGeler8Gf%{lK=duoK?mJ1)WmC4{kXt~enk3XH{JG^
zJBWJ_E<Zh%?Nh%BwKjRSKLO#NLb&mto}}HkXG!WRHV676TqchG;#0<!VrKIi(uk9;
z%FU?{x)U2Zb*cN#aAQ+nvQ$-MfWMye$un?<ilexP%q~F%9sn<nIkm&3V4~r*tqCH=
zsT?(7|5Y?$xO5jPLLqQmAQ1?s&S%&zxTKZGkt;2(eubJWgHf{O7r&Ncc>BNF9u5!{
zKD_Fm*0zGPHbT7@ZB^TXSC4H&GJONXd3{aLRR@Zc@w~puheJ5B;$|<zFlb{9bO$kx
zC>e8TIeT!zzk7nE#th|L7hY3h8dLdEUuiFWeBOE_eCsX335k-%vK2CB30c*_WXYEd
zfQlSMam5+p_xbXvDVRT_D-S2XdMBWPN7*XDe7k^{n#_zPG2tV3%1S{+;7Y5}+29-v
zb!LPI=yJBzwBONb9}$v_pT#`NzBll(Ivhp2O}i2_7IVa~PSv;w3Y5xopRre93%NuQ
zXNC;TFu`N}Eyu|l2qgg}1D_!WpJO6alcxb+HWQ@+$f<JS@AL$kO(_vr(fDP+Ofk|3
z3t5BZY`x&?B03c7*|C^ihTZmcY4woMfSi`z7^WxF3J^IveJcr0-0;HvM^+WxuKY!+
z&}1^`Avlfbt|an9;3<0vbrOTsIy(S-`}QTGpd+oS6bZd#mF4>cKApd@v*LXUxaDfI
z0GjC<$YJ{{L8Cts&I?LOcU5ZUz1&G0%QXaZfB7|RKx#|>NBdRcv@PHd6piCFQAuK2
z^3PN&^C0Xv#)2Xh_^Jk2+ojiR0OLwd9nN=6%7r!qYW-!FXraWMr&kbGx{|k^WOlws
z`L(b&KE4o%BpvOJ2H&Ds_8pxQ3M(dkAKb3|2<LpzojjSq!Fp#V6cMOvjW<g!bzEn!
zkN~PFsXk3w`e!{7&@L<2fCE0N9tkkd0()KwL-jA1Nlb#q%z6<}tb>j)=k;jPowR<K
zUn_(zo)#6MS-GDdSc|ns$dFch<N<0gKOi_Qgz8T(P(QuZIcQyf{(k;I(gTQx@Gq#(
zN+F9m0E5+2^NFr-igXlZc|847IH3YjKnapjrBFAsQ!`sZl)3b>cRj;aJFOqcWWlsB
zj2{SVN0!p80?0O)T$bbmoCgW8wFUhWl7BGmHRwGY^uqnbF?F`r$4s%iHAR9E%Hx}q
zyx&4$863ZpXZzpox44Sp<?wc5)N2fftz>~WaUsbHW|Y7o6V)W#goq6H_nK+vjzA<t
z#Bqd3&^d4SO|heBU&DM&ZUJ2p<6B#?&<IsBmnYfYk6+RFOa^Pr|4K57*>mdvnUb~!
z*qY)DggFHGH8?pFJ@b{or-Y+#NHir(_E7;)b?-5f1jrl1AfGw`2J|3f(?0=O#%Pek
zge@k4?}h^vXknY1pj+u2f{||lkZE;+;YJ|<3>glze2M_~vD$gbFoI~r!*?Tww(uYR
zEHpYT`{{rMfy+PxwV<g%d|IOh%RoY+0!5&2W1|&Vg_}hmQjIV6S!%_^N?&AXo81m$
z$!3$^dkeo~nOK9mc7SP)Q%C9$H|9338=avVhlhPhjPm4d|0amY>AO)!@YI3F=7?e{
z5qnYVEA?G#P`dtRJRTx<C4PW2iM732#<&MAiik<|8rRg>Eu$~Y7&5gGYK)6}xr(uo
zLbr-78U!0gv$J)a&3S<1Il5{xvol0>)ml{s6GpA238}}?uQ@LOX-S$w4I>SuCDDPf
z!9nI-l2-H^S+&YcT@%=J=Xn_mCB;i9u1#VH+ZqV;umU#{9zW$(%Diu#oVU8FyJ_={
z2~rO3EU;u7(xIK%aak$pe^Or=cQmXY_i$(O*6}9(yMZf{V|#okjigwn^@dDIl#G}l
zDB<;)5k`8n`7emN<B!K*S>~Yvya|g8fT%^&!F6Id(Kor_t$q~hwiF8}m^vv7)zwg{
zUZErkSO#81w(|-0LqZ0W3x3*Y5p7vyO|9FA2Qz-Dx5SXc`E4vCPiDx4u!ck$MaYI7
zN^+4dE0)O(P3t;a-#@i}arCD@v|hJ9PjAECK>mu6MQh6<U4SM17nMNBc689BoarkG
zL`B0@15-E$3PMc;gZ+|-z%{fMf8#9Bq3STLjH+FE*aZ{PKTjnRM<K-zC3?Hf`k9G3
z@)%($Q}GdR847Vs>mf{_25iG)7@dTflpdwB4>|ayJ!1zw6BDd#fg7Vt$}w$u*ptW+
z?z-%G`~kK<j-?=*C$*wiI4YS1|7zAFyZWsvL{0X`iD}7wE`<`TN2)@ssHA;YtJZ~R
zvBVuo$zIVVBAf{|?`%VjO=vJkqH?ybGShYJHy76&Qc4*iTYGEbF+3-Cpv4Iu5@|Fu
z3nFQjo@~PuZ2F?9uu1KA7pg-4=T&&9F(@8kdiUHv##MNCd*9v9dP7}y?_9k~)J^df
zIobW+>c+X^3OBu6F3GiPlzR|U+aW5g$2<=FTq@=FIQ=Ec(L`RR88?6N?=*gJnKEr2
zZrq2@#N?UNI2l1?BD1aJ`*#D0SX7Q!WMj409#u!ADGs=A(?JRa4BQa)!8V_n?P5b{
z=`XMRg|hlM`U`Q{9Y!E@R70KqfaVb`G5<~j9}}PE*2p3&dgTs7SEePyDHVl;_Hs>>
z!R7v&{_K4pA&`9G<wC`jfLcB1E}VsP^M2wh#PTG(KLY_OSmjX{rNzc-RCcW{GwPg=
zYFSGR-=EAFeca^g-45p&I<Rl!J5{%-{NKov^UPk%&os4Hlp<5I+#YOPkFeZ&lS8Z`
zjqbl(I=WwPNZoEDN6*e$q-@#EoDbT^0kk$juLBiNe@c85Ss=%ysR?XH!dtsQVPPD|
z;W=!-hXxMS6m{69*d$3Fs;j}g1aN*C*#WhDs~xR-kI5dNjRB7Hm%<TcClK4-2|vz_
zyZ|RCto#t89~j6C<tXsU`2L6h_}@(sgpR5p%>5KF9EvkW455-t_&b3kJ|raZR}(Br
zhgE3u&p%z5QP^k&Q66m0;@}Ij8s=8R)ArRx!s<#!!P501h?QG2{o<WX^Mj86L24=H
z+fPMIx4dE@o4tgDpD_bg3U`5P@eh@m^gOr9jV!X;N@^a)C~%0n-EXPu#6a?7mONfa
z5s4HTj43^$s8Jt2e}>oG=gS{boQ8>%0^nBLt@}Aj5gyS8;7N^3M4UnUGp@$R2^$aO
z5IAR&IjrPqpN_qVJ%>T&-tl0c#ksOifM6d*z&x&D14tNv%UdD_Ot-7Yox~E+xaWEz
zOxjb)8JD12T?Zb2s0q*J;RE)m{=Pl05409`pOtWUGC9Ad%^9@UdWeP~=#h4~53Mk}
z(S%&R>o8hgCRq^252lxG5Y;<QQi{mXgY=u$&s2XX7GSF^H==vYY<uc~W80hpCor#z
zq?jQ?mlM%3q!#~X#cvHjOElD|h22ubk9CVHRHBQt#9DNc1ho&N3pd8|{zJrAc$M+m
zsNe5Nn!O~b^Fl%pJqlTY8No(!Zy?5!J;yZk>}iDP3s$IhBAv2#eBfz=@*a)l1}AS;
zH-abh*lvH!N6bcQe`fJz9c9vek3YDjOpS~`l}!u=B+m_LX8IeU-ePXVgEe<lAl{6g
zsGD$#MB|Xjww5msX3f6wA05`c)({~Kly0Fx&0<)?nQF4D<_{iYtNRu9Ku(@bkhJuE
zJb8N(K8t*`27Bx$1>Pg13`{=lzqvlQr1kh*QdBL9qnCJYwT^HrBWoNCyHjL-_Hl{i
z>zYs_aTiV*BSgu{3YeRcRXu-6L_$YlX4?&UjJx_Hn5L~)2hWu`MQD^AYNS-BMogPq
zQIrPKfL)DfT6X6s3axpYLlM`EOdl^Lvy%QkP-HPtjSQdkq+(AQjkyHe9TPFDL;PM3
z)}ziMPtzyg6|XrTVm)o5qU7bx6>rgE31GUDc47^Y3j&f3sLew>M3Q6S1?R)oG>WAn
z7!OG#!Q9TJP|Ux80^xZl-y}1v3>7{BmkK6jvkA&19jL&IQ#hI4z{{G~)C+ZxQdHJ9
z0|D1Fi6|l20xeXKLfu-eRyx(4EaDGoEwSh+51xI=jSDE!$I=Ga%35q-gH}W*Q0d@g
zdjvk+AigMfiFa~(?drZV<%4V<FE)gBx_NcS3W-3V?^13jVpz}!QEzLEStyBw>6kRG
z19q;2_NkF`TvGe950TjZ?cGXl?=?tpe~@n)eKJWdV%n%2t8{fWyh%1&<kDNQD%4Dj
z^b$)<7JF)~_-0=Ch^w-OtX`It0)e#mVXl*zQ2fiHd0RAF5a(1HmU;<Ld1;I|6$>@O
zB<1hA*HrdoDw&!dC*p8w^CY0*44;6;tL2vmZoZ&%a<4bIjrO_d3SJ4Fp~Y{EL_#2>
zMD3mzAeYl}7MKzB&MKxbiTDa5AW`oJy?9c)5u8iq87#an08F+MB(e+<OjvJ!i7;jR
znE$J7qhKmF{$(`CZ$l=9Qmt9z-3T!%rHSw&DFu>?8a+NEg6ho-D}`x+gW&iETij6$
z52zldyyod$xEY1!D#d)nPA=Yamn|v-#gpPQI-gBNyd(C+9xCIDsMsEMu%V8whN$C?
z%A`G1B4Pe1{W%7AL1b?bM*yQ*$q+?;#f3#+ZBKGg;c%QI1P2>c!`l8k@n!}6Lh1Fs
z@a+OnT+cfzP->K64!e-dKW;%%Mhf&F=2PBYo{`)lrU4>s+a1_e)k@<*uG}@0K~iUD
zbP69!{n}XEuu6_9u#<Ew5m}@{@J+h~)zU(>blEUusT<Q7IaQ`XtB~(ftK={bqhw$x
zgxhNHl(>z{U`gCpNBe?~%@C?~8!V&hs|HiaN1Qj4LYHP+rC8u(AJRo7#Jv5R*Y7-~
zPcHui(P2p-Dx$2#eX+;I@1nI~cDW09^H5~blLT;YdCJAGs?ps){20Ty;i5!KB#&=E
zu#q}FPKgC7V~Mv4vKBak?tKTh<Y~CY*qj8!l$o(*2SdWc{TwgV5N(MS?CaD;YB4WH
zlF_fv%R{Nf2zJNr-{6v6${jnM0uF3-e)!#)m{bLT@!SdZvz|NYY|n3*8m047LX&O!
zzuozg^n^l1i2M3U>`IMUkUG}4)IN|!j800Zq$=dDG?UoH5J=y58PzCx(B9-9cX_ew
ze*-bK`}cK<54S=T*M%9g;I1&{#`37pGu#s5rw;#XaAiVI)s6~BJdHB&*3vOj9`|8c
z939H)P&V3Ljc|AM|4jMG)NdwQrmEuhca)Q{i%9R%+fd-_klrJ~E3i1d8!;*w*VOAk
ziq(<`0&1IrWT=A><*iau?T2XXXEXG3R@)o()^XAhL}0s5;#~>Rg(D%6BoSkUYkIG?
zcY?L_dAhZ5R*eS`FnV%AjtWK~LUOUH8!S}o4KASY#3kg!za3m%(HpBdt2SRW%fKp=
zO`19#gk>vi_OMF&Rb-}MS~q3fWO~5t1@IbCT#G_M?)rS>*B}2Ans9U>s-*bq{Q1vt
zVNKpBk&GvWa>l-72N%~jw=%^nbWuMzBU(PTM6N+2w?0@3H$;=%InY0V0TA=&!#0=g
zj+Lu^o8-VEWnQ8K3pjF7fFlD(NEoPI1Ixujnv@}t$-_xVNp)u(73L|;P&k>UY5WkS
zK*P=Pl?xCWPi<OdBMslc{t+ALJB^3>IN&%j95)%5LkPwUY@Wd7?I;S2awCxw;LLzC
zrsTa3Lc)viq=2nL9zoi?hnw9UwNy%3PmI!XNASOP_obrreO!;o!%}_RF#=J1I-X3f
zujrHGi0}23t<l54u6@u=eu}GNK01vI^5j6Kc}NYm<2>9!mFQ8TuzEzmDXR1&OTD#)
zn+kXSuzUFOaLbKLPq$sAW!h@vu7p;1h~Po*1~MAefv1~1IVbCb3qx!A+kLUU_oAjh
zP?(?3)oy`1B=914k2Qx!v&H@X{>k$yn0~PaDtSs;Mr-}EDff=%f%x6dcY6natoHRC
zFD}DP%(!C=wfmRp>BZj8i<i}oUW}18aC#w{ji|zbWFT?@R4?{GTGhT}3G5bvuJmZ9
zoEIT%Z-3`SwY%-^s6A`RiD~e%P%>}CSOD9>&XcFrzPO|$4P%rgFn_qS^B?v7O#A(}
z_5Glddlv<wtDohV(fd_6C&SPE);TKLC=WBOBAsR~Mi@+ACO%UH3e-aPv{rcuw#!^S
z+0WH>(QrVT?Kv~X@|JSk&!5^gL-^*Zd#>qFv+>oH+y`M9Eb=|OD9q2yp5?}DxU7f6
zLWb)Z3`kP>1B_xPGVM7-)E3Q@ZCwa`{`WWj(0Z{+SWE9iJM*Y^6*x$I@pj8NxYm@O
z``r^`aO4jGLU)B#ktjkvlv!egZZ_vw^PwTV>C3?s)>E+pHN8}q$!Ai6GQ}z;T}odu
z+Fp!=wW4lgUmr@JS(4?`J`^_2bf4ZJad<pL_Wkkx7+25Vgl(cCs8{>JxIC<1)9L^v
zzi~je=+4sDFl_isn5JyDhugE8K-#uKNX0ibFzlg{k8sX$_^YW?W{Fz40qq~h$KtbM
zs6)9btaunE<e}8siD`Bt=KsxL)SUqB@W+Qkcl~{8#}Z2gl}G^@yCAL@Pl3?ru9Q&_
zpprqG;|}{5S5=46g^B)GgAUSLjDmvcK<5_1qQNf(sj{?trkN&fhGVAdXOHk!(`D0U
zqW>+g-vnQu5!k~SB48>J30KKWsi1S*F{@PpyWdKP2nj*BerP{OX^gX0vSz&#TD<;u
z5P+4GTALIb`=85X&c-=Lq3m)))}fqm&mz)Dd?R`&Nbk2~l`Le9emwn`hH;1^XnS!I
z<x93>+<?6#ovAIFOIB?B<fCdMrIWf<aUhjKlis0nQF`b*C}IpRIt)*`J-yeUJA`=8
zlvABrY6Ok5Nm(KvR&|80BSzH+Bnbzp0tz08Zm>Js<9z3?j0JH^#HQmE5*H*RA{f+(
zvKjm}MjGn25OF4jnA4`>qDRqmzhyF({*VaUpwD63so=xy6jjlOn3KP+>d|DiiPGlj
zEQeB`n$4=5L4-WQ#2YmtfF-fbBe)<<|6Z^kxq2a@=fVEzP+ib=6v-j7mC%D=)#ME4
zBVP64j4{YIZ@EX-l0+)(K8%yYorB-+9LU957ifmt=n)nwULtXRc7dy8L^9D!_x|`;
zMPX|VcZ4Cw9K#sfIv-04(YBOTV6kc#dFySo?VyqgW0?fTPv~hEI!x#ej*I0~+iU*@
zNgV5I_3+=j!WB6uq1?!y{))gE=8neuPY#~E+&jQVJa%Q+=+D>Cl_&`z&?C!jeXXvs
z9M`RmB?zfJCJwZ}a#T#dKYFK;4&V^eM;)~E$!7a2B(qd{6#im7h<8%;N6^6$&*+mk
zJ;_s(r)ch8z}5BiB4sprW%Bu}*2DC2x!5-=U7@Ndss!tmhDI=cTUhc=XywG)qA<IE
zl_@c&2o^1MSa*IvF=K>b3^@M~Vx>=Y2RIo(x|W;mG&P@*b~ZbKu{=2;?Z@%LXQ(mq
z2-QU}v6xxzdSt5Nx<7H9=YuN6xwgqluzHxK{TR--Ur8wt3Ms3TgwaD+T5mP8YPPn(
zVc+Z>?r>65u>koUme9{E_}BNP(6P8|T;MoP;#8rJ;KjuT<~S*DRJ(nMo*!Bg_P;x4
zwk#Jr-J__nIwcVs6;CMy2vxQzfhh@ftXTuD`@E7QNtBh9YmvOg+I!#@*x5Yu+}BDp
zOzDc4!XU`V<tvWwpfJxyHmjqAaK>rF(R|l_eS~XCkd9MD$G;iadioh$^TPm{n_}@8
zv6z7vOXJA!8b5V<La~<Q>_ecGFC~=mSc#0+{GsYcPck$amc(yaH3Vcx-G~&tM_@7u
ze{kv!&u06}2ihBx86%Y>l%W(z6_Maw$5!_Nhs8tad%Z2O$d)F*A=<TBP8ZTdnrMr4
z#&I1OS%r{nudA*+lZl|jVog?X85WZ}&k>M){KxTi>k>Xms|zgo2`mst7Zq$q9SLtj
zMBx4~UIq<}>fPr2GDVZA%_v4}u5S6_KyBbfz3f}jrruiqHbJ-x5ycH<l}Li!z4QUk
z%7_gu4;l6VA`?TkMmT}<bo?mQ6lvn+H3|%H_S6cjNBU@QZQ;}sB*kSP7i;$#KW_N4
z^?B<H5|pfQw$HT|fSI$P?XZyBJpV%~3!mxvI+n2RUeCt#JgH>vL$Hsuur&m{a6mw;
z5;(ro)k@YHAR;UsZYgsu<!C7JSS#NntltT$1XxpI5mxW=PT=$#vTM|vQK$*Q9pWG*
z3u2S#04>FZC5MW}OjREwE5h3ANM5kpn$bv9a?BfFRcdP%P&k}@3AxOz5)o0{Xz}h=
zL0b*@G<mh@^3b!Z0cS*G!cV%su)Kn(!@6c?wL*~%@)0zEr<J!`O!%Elx;HTOuCz!Z
z6NRk=cU^b}AGG@@{&a<#vm9~4zKMbr#&KoiAVZC;n_lKYrZ$j`|6u`BP&eTM&?LYF
zD`_Dw?!DIH%dxb`z92jz&;l|RcSm6uzJo=GK+tpmZj;Q{Z>Vyu)T(ynW`W?tTuQUp
zy)jJznPqKPB(PXEwYG`el0(Hm+~I`q0lUkr_2W0WW$r5g3el=L@pGYQzLyjW0Wt>b
zwzhh(wn@T$eX}hMars3U9z<;GED(eP1b_KcBM9OM*pM_p^rV9K=m`JKiXZ)vn6OOJ
ztq_hVC&_;8_81JBzh_fyRSc5;>1KOXlTRFJnI%R<<Wu}Lhy^|i7x<rcmT`YDD;L^X
zia3{(^&1?d|Af$7#9WQ9M0H4pfN2@Cg49`;@JxsihbRIm6Yl7;YDe0+@?oSw)DSIn
zlPAsYo^D}9s{nix6<QF?pdXbsQPh2yOx@bTu5KpoRq5bSFidg~NV>O>v_8O`-a_HP
zIONt=GVAENhfS{P3dyw}4IZ@T+Ly~h{u_)=4CS&$nd_=AR@zMCZ{J^OLk%!_n14bl
zcDOCx@ti9>#Og>69BI#KOF9+O(%c{r+u2aOdr$-Yfn%KN<pcO4`&}M9hhX-j!jOf6
z?Ix3@I>_1#M~ab8E7x9!p=45GaE8nnlt=ZJ5|p(g@DUhkUOy8EISo(<L0dNo;y{~r
zOMuaRGD7?o@sapLjz9<V#GDQMCc0(S#2?35y7O}cIK+-xZzIt~q8@U(FZgSSfanE8
zRBC+X*o2nEsVLAPszSHAGg>D(hMvkx_)5<ZK;eRzif0;0Ftxj<G#imyBc`|^0zKo(
zIeAhw&V>YU<^_2x|37=*-WS(#r2Y5uDMkck23vsabuLS6DHxDVA_*3dll(wo2WCJM
z0YhfMlA>ro`};g~>DxJHW<YY1-FMlIHRqhZR99Emt*fh93G6K#{ukwxy0m2>!2Jhu
zk#MWq`^(Mvy7dm(LRy@tmdq*o9*%kd@cbQ2n?BxNUtf)pCIW4`in2T$ovrkSCo5eX
z_QKhq@roJ;<K>Iq&y>VvDfYSSH;-Px$$7M~xwE^uzP7))0dfnxN~m+voOG;gGW8iJ
z@w%z?2S_KGg9;!$p^4V13FWw0?|mFWrJDi&M^IoA{{jS+&dDbysFXl;*Q%gJ?pPJC
zd?Q2lsETj4VTHG8%VCEK;ThQi2p<q6K}Iej`y*-i|G4455I2;`_(`cYVS^cInNwTI
zsz%$W+)7v@7Eg}33nog?*p(o><1Uvah~X1O^OsZvtl;qYFHl$DK{}=nnIH5glz~LF
z=~t(D>|(*62EpakbSEu+`9)|84Fxv_oOKvBz+g+dezLZte^-M}jz!V3#~-^;6y7p-
z8aoYly}1fm4aUGmzUz<q=DY9y01}{eI2v5P1B?%43sLoNFEKt)0Yz=$B{mMN^h>$D
zHtbLdl(Gq53&h^dF-7(<^gyAn_#A?Nkj*d@A{f!N5Z{Yb4s}(F0)rAagNRuIDYs|i
z_(1nUu9>tc_PMeTGRHd5CLEax9eXFPv{}($`{%fnp4xUK1pORn=&tRv0mLznWy{mr
z1T8V4et6=pR{#S98_}gj)b(OJg?AjSlRySfcxezN&kmE&bs)-`4Qwo`r(qLQ2qHBs
zRk;fVL?icv8>&W!Z_$iWIb=Air{F}Vlf9w@D?SJ(_M#|2P->7h`+78%6*#>Y6tCUH
z4c$RC4Zm%jHFSM+(I1^*=Xy$U&}e}^!eH!!j;y$yYU~l4XW~+2^JGkuI*{Qvta2P)
zlbL+lJh7qaY)WGVWi%<knj|dLTWmd&y<P($iE0a84UkqzBNYO>(j%M^3O6(a_rVv6
z$x&R&_PbS*BH9BukGU-DGFdBKGm$AzPbf)<*8C?LmY33`#OW<P!i8l0Ga_}w{;;eT
z&ba&E#kDXEBUrP&mhZeAE0R@JO${B1U`V<SP0L*v8@8*Z^)x4(?Q<C@hLt2hA;V=1
z*L>E!L~b1nK@ccD>G;v!A_#tXIRKT~6bORB^5l|*XWc6fHv#bYW^mrKqWJ30_zVEk
zd?n&EfFO)K-rOL+=+VcGGi4uOrsKZj6L~`-FtHKVG+rQ6R)C3Q=-5?xDxY|%%~e^R
zNmAkj`(kz4tD&Yu*w!#tI1g*~79mlM3&ik>n*Bg}rlBio8J;q#DZ|~yX&|_rtn`)a
z%|$qHH?$~X6tLIu50N|R?jXY#1dR4?#aS+?JZg)X)7e>4)rA%VVk%sqPfvv^3?_XN
z)~oBKaHO#?rd{V+s!J@aEbd8B5Wed(zdjh9+`w5QE*{1|D96x{0qknX(FyKjFjR(T
zKLU)lyKPgbESqC<qL0vIwc<3@5K43$$4Rl4hha6vhGow08u9KNoBzU<XNjOUIR!zV
z8wFhu@KpaS?iSIOJS?3td{-QUM#ypB`RIuFk}PnQ`&Kvfo5ihF96Uq3etlN-8GH48
zRfrG#X@Ad$dt+>q`3CMs_ky;wV+I<Nclf*U@KS$jPMacvou8!)9fZZViK(KE4<Fk3
zMZ}k9z58cNFnMVIeuwi2u;PlsQ82)hjEzOPa?rT<EJdH-dLZ1H)?D-tuy@F6V(WqQ
zib}GpOk8;4=xkKF=X~D1=0U=`nmc)HdtH$-4@q5_Jpcf2c8RB*^P985<)Qu~YYFF<
zG63bJhl=BbgFI`Qz5aPmX8I}CBT|`P;m8T()bIQ60ufrlO0ERZ4-TsI51+{O6j;oM
z8b1`C?_6>3`kj+q^{9k*tDPO%qdv_t?udt?w9EHH{}B|YqF@6R?@&o7CnsixINYyX
ztzZbsrhrS&TMFDyd4mzOvBk*BYk1-J!;3xwbM*+*D?C6Oflkb*)AsuIvmL}G?jC)=
z{cO{taTi#Kx}V7wc^0`rjj{1;IX~X>hSzp>HrDpnG<EVMNhtdpR+=YrGl*~<>W_ob
z@RHd_TX1y@xI|kQ+%&*7W-x|ZSY}K=0G0q@3Gq3);U$tI-jR52_o(_X^x!ZQ3#b+q
zFqts$5XMe>g>fPLFVlFfxYD9Q&I?#cBaLy^yZ~tCg}!oq+9R30BpX?F02LdK)%qd<
z&)NeD{jp~$myxyMZSPE3bux#4-1zPYCwZT2eP^@CbVLDE1K{#>aE3uRE|P{=AOI-6
z&j28)aU#)41}0LS5)<eY+8lN-6?=~UFxl&C>)&r4ZEWoZ8uZdB9r`<gQX1a8{4v2q
zXs$xdD3>-w0%uhjJHgWVol4$yV|co~zV<XD_gIQq7BU(NTgpBZS_ZNv3s7thQko6T
zR%ulc;soXSNUGGSUD4NB(ub~~ws|ajK5+?#QiHV?WMZ#Mv_ff;jw^B<p;(!5ov0du
zwMoh7Ra$8#*E%##rBhz3lar~_Mg;e^WEGZhBdfI8RbC{>Oy~4=vq_U|!fe*h2#1|4
z98L(?bt40jkg1jNE9@#4Ttr%$JSj3BnF&1sUCJmmhn25&io<{alBeTf!o*8H#U-te
z@)qn4q>dNjeZe_EGCNj8;DNQuhy`r}HVf0=2(cs!<$D*Bu-vtQrAO*tQ|An#B7}v~
zm8@^(+H3;`>mzNAkrSj)IoHCsK}4+cbtEzr>;e7Nayc#XW;z8{E=|Caq)SEn4j^LO
z9fgB?at^q>jlfhsh3?q_BxoN1!WAT4@aPAAkXwpCXZVL_7~%)v2+OZ<;*;k?Ria5j
zmwc?u+ODDEhD@08o%Y3Re4#rL{~+Y!SHJpIvRhFWmR@l*23P;WG`zf)=$&fsST!5{
z%XckFtd3sP3qe0~OVUah5m1k!#tx5y%%WKJjyK#0rf--3f?`4b5CR7O!_5Y#lcNL2
zQ^_LnPk9|enOT_L9c%?b84+q^a~dnC4P%OUoB}O#bp*J~rK@EyFCgi>0Cb*pX;E#p
z4r!dj%U)_q4JDr2;Guhi9J>E9Sg0;M84G=5f|m+npsT;i@X(ALk6{vKK=Q>5s0d^u
z4O+$83!I)4Qdl6Sg!6sVBEF9NK9Hi7K;g2MutLbo5JqgZrL9CcZcN4#^>o5euNCt;
z@a{QQ^F<%m`i@^ox@2g{T`11ygI?=*d)v=jx~-|pME$?FCTId5sG9c$1o5c&siEhZ
znT#k9>EJ4-t`>RK?A%$msd~aCO<`cQB?&neRzJ0DIAIVF?d724Duqw#Lj}@W&yPHd
z4GJz0y9y#;aQ5mP|GO4BKzb|eUa8t?BNz*I#a#PMcXSD^Fs1Y{(wM)dG4igDJU4Gx
zQ;3~1G;F5UOHu7!ZM`u~X9^9KL`|cD#3R*ouVJ<u!z>0N2h_p2wL$U~!?;OUn*P=?
z)?a+`b8lPwe>!@;z58tKDUR8wG!H$7fi=vxLcO7>WrBAGaqEqmg0cXYh8&S}8S1_>
zzYz{(^Rf_rO=dP0GZUL(<dWA$zA&e2khJ(Qx(EZ%*@+Vr{E@ETD1#~CMkAtK8Q2ak
z#N?RPuDWtKi+&V8Dnh7w3g{_n@Li%Z#n6)L&}&qqBo(&U07<_Ta0MJnV5_TGQTd9K
zr9-Bf>~p70Nn6edxkv$A+BWE?UMeMx1VR}MNpGTI82hit<fNM3GP7K)_1542o&lhx
z972<LL_j<o6HbwU)%nfoJogB8n$?rMOuQ5Vah8}EpeC4jGc?BYfXebLzY|(XMngo)
zS^UwBL~4=>+WSLvC%5zr{zmQ&Gn}C+4_e5H&n#a`ZS`<Mx$EKd3&SpY1e*OfMOU`w
zzr(F)RAKfIS<KxB@DTw67C<7j8gDLgY%b$~=h5c&Q(lt;i^qnd%f_ZuBothYdr*`w
zx*aMsq?^^%r)z&a+JkH0*&50(P}HhF42W>y?hZ1GJdS0S9$nE3*P$|X1R@T)q@4iK
z1qbDr-DxYUs~em9o9p{Wzu)|m_*9;5t?%y<RP4{lfvQ@sD0XFp=CMzMjAWdp1x*mR
zp)Y8ch{~9p7LDi5;9n*6SmlHE>;Aj;OFeql#ZAEGg*6&Wr~^e1-^)ufvg&-n`uRCI
zZW1ST2_o_0%9Jc=D9uKU+~nh)3Uqhbs#H`t?|akZc(kZnNonKws@y@MReZT{LRER1
zSyfJ}sw!0+tKQiB_QiK8Ff*ygPC10AgYj8*fwVVJg$XL~A2GvpfNM~GSS_JHX$7Xq
zCAI>aT2p5tbsl9lY3sHRFO%cV=Y~gxbQ*1oq7e|i8X2=8ugTuchwdaptOOze-U++X
z$wbp612L-Ma!twtHvf7<^RR}tF^W!Gzj)3F+IK!(HWllp6v|rjZxXWL#H$1gs8#WG
z$;p+A=17$15TsiAMiaBR!H7y+K$;Fd)uzg=8ebw}qOmKq?8l9Rc2#V_MB4-MT*O@%
zo4{Ww9vieCM31RaTrMoa0<EHH9Y21=`gp*Y&`RE8Y{~;9RL|jPbgP>Yay;`EOWMds
z7icp~qr&IMBpl<kpzm2+mXL$u`~Z_Y;!XCRXr%rd=DYpH^26na>azt%!scCdwQWR~
z3JYhtpJ-H)Z!-GyVmU%{FLvfQAbL);b>fKR5hl?$ez+E~-L*R$Vy#Ksh`vFNRH7mv
zU7Nuh1BeY?#i%1~TS+jJPp|@<v^`~6+8G@x&^szqXt0Fv6rG8Mfn2($bEvGnHXNoT
zh`K^(i;ar$>@#a?saNU6x4r^B-I5|J^?TAeCO<^~*bIXW$*J<NVMo8#7&&X8TEOpA
zx4>$ugjO*`btk1bGi$`K0-^=2=wz#k@7+|9vv77kJno+Rj1p!j@G1gQNkpGf=A~NH
z%WF8*R8%X?*t*m?{VwYy<4FJw=tv9;9pJ)MWy;j>C?_loFo&S7*_@{o&0Wd-p<5++
zM`h2<9ddEOdq!Sl9L6`C!$++G+~T!~=6NcC^P=Tr;`QQ%01e9^i5Ic~Z9RXojjX!+
zo6q;Qwx2J~Jr3tr)7Isv?$TE{SH&LMbJ$x;Khqz&EMKq~C@Ejpun4}?X~=XBDX#ED
zZ3Czj&OwSg1+0bYn)YoXSm#7!#Z&2D7znylBxWcS>Y@E0QpMJc{Km#EV+Z8-cy)E_
z`PM$_hp`GZmZAuud%n0ZJa07#n&L|E*Jup_M1=u{D5U^zCPY+0)U|eXgZNY9V_0EL
zHQ0q3;UJjkM^dCH711)E=w8qTQ)8LuYX#^MBV0BhaBn?BN(_anFUtmm!(auL9yKq@
zCM}zCC+|9V61gfkoN8cOjZ7{!Q<#gc2%a%wixQg-1fm)mT{$_V-Pl0L4ca3GYcNdw
z8XVL7*ohi}c-wVuRtBCYuc{QQ*Sl$?wFs?@Rkeyl1eX|1ohNZ#eSrnuY=&j>KGgu%
zUGGeltc-Q*9;y}dp;o3B3$4QRK?Omwf|A_^ttRy-Eh@bqneh;bt3Ww+a9>FBpCXU_
zL_1;P^78dnIK4`_J#~@_4rB3-8$p8GtdNo~khd+8o@lc6Ro18Me}zdDCYYgE_s7U#
z$4l4O*g9e)xv<Aj6ZF2$y}L|;+T5nqxv_;M+wLy??cvgY94#YyL>SM4C_-E|jt6Lp
z{xgxRC>?UZ<H$7<9`W!lO!N%k@MXysj_yTBYWFs1`?|e4haP3=?no;@+qGofM+cTB
zKueEt;`XFpDQ||GI1-B*U8q$kA5whJwL;zjGxq&#RRB|Yc?J+e))+GFV>C2r-T?_-
zZ&Jh|iS#ZmATv%n!q+7l6ZZ{J`dVL(VaVL}8PBwfdLQ2&emA&|Pub*R#d(Wde7I{B
z9oIC=d#o$X?)`ZAk<FemLCbly<tOs~BO8BpR&DtEvyW^PDN5nW2Q5{wv14BvI%vWy
z^1@f!LjL-OW;kIGUdbhJcq3~wpGum!4Rbl_a$CAsrMunHPuC;-sns6$@7Cjlft}Uj
z@;ki!>4x_9Z`F~dHwIQ#+ec2`6+%*~J?`M`UB);8nG_V&j>w8WaYwXRsvXgf_YoZh
z#khx{WK$s<M7D)%yLc+Ks0c(U$)K{P`)QQ8v}(i*!zZ;ohLp(Q;fNf6j*rZe01i9N
z`KB#rHm<|raHi@Pn#33vx&jV5^9XbgMGtkvi-2-VkDxqzDN$Yh;Sj9He-qM)Wl45P
z;ehmRg7jL|T@~8gM6OYlRBw*<r{|xB#aZ<Tm_?Z@y}*9Zbdx<??b*V{19r?uqrHNi
zPB;0LvTAG-9iUnH`O&MmsV;P3HbJ<0ATOK{Z`A<OPO8oscP`YnN$WAiqpSJZPJ_6O
zXc-NZR6>hFE8*$z>EIaBB&Hw>+|F^xNf9|TrJ<dE7iV8!9x}CArIq#-J};eXXV3z<
zf+ho|6Wr9|EG-~AV%PkYSu?C_45en?a~pobAq%MQk}N&6?dD}V|6=K58lZ7l$Qmn&
zLLU4Cdi2{MCC?!rx+nO-F*tR&dM7<zWW^Ni?uk{ERT~|_!1&Sq52+uOTNTbir}!0W
zJPNAPn_YsQo7d;UZ9NRNrf8*$IA2qMFD<O<U~<_I$27d+`6%v#YMjg>4Knq0v=$fw
zBi4;DB9N|OU|%*)?PX%F(dm~PLRMkv`W*3xB3{H|33IJIn7pwRy`8w<%H*<$W93l+
z#^P8GfACOAa=EwnbY=hP9=Fwb=nmt+9pB<G?JbvEs&Om?pd5<6i-}(E$pJ2Ia2$!Z
z9~U|zf&S%q4BL~y#y$0L+~zPp0;b8JtWt6lBSOD;x4?i_Bi2T%W(-W+<abm3h)rsG
zfZ@5E!~eEUnSlw;1x>)(zm!833=OFC^dOWpoFIY#txBdUWa6|`PQy2st<hlo`a$dX
z9S><p5CPjeq?yXrIdz9^v@iIDn}WEqyV0>EbIX3aq$b-jwYG$A5Cg4p&4|DVrb;^}
zT~XwGq1X0w(B_=HJU~3J6n+Up&MDAaA?uq+sz@|Sb#B(~q;Oq#+6udyhMpsEFNBZv
z@$bnEw$pxu_o4x#U=6lXg(rm6Z8=i7tf50yxIwCLp{3?N;yke%cVbmfK*5r8el2sU
zXM8QSam7*PlcT5&4lt~9%|H9X$avU-PadHqAr5kVt%Y-Q*Mn0z+$Jve<(4Fz=CLA>
zYbbeWeFRDc1JK6l-w~x%0y;vUYZhu*(*+mblr@^QPO4(u3=+q4bK+=gqxE&`7Y~2=
zP+e5j@ovfb#-t;i950XR$x^+!IvX*MHkKIEtp8&9m!G%Zyc!_fo6B9#6>I)Ab>OxD
zI%!^#Gb1FBNAJ(yWsor>$rGQ4krRK?qOR{x$`(?jf5;L82L-z%K37RRKak0u;GFbR
zRjDo^qUs32H*io{8hO<v@XG1ZPn_eGYzuv58C-dO0-`91MD^Ri97L0AK~vJ}lS;S{
z>hQwPlTqEYQwn%3F|FybrhD=+mm7zb8_wl)9jpE0j@DI<v%oQG4$;}pLT_!umxmb&
zf|=(!u^(ETysXRatVMRlqr<1eGyV_xC5m&r$;YSg9AWQuuFGr#*v)C_(aLWTMYpvr
z*_rx0csNcksCv*UTE#R)y;mZ;QH87}*MPY974)q@hnQG~%_KF_AY+fU$V&v;F0K<q
zww1gsRc`gDH1iZBL9SU+9JqE<2oAu?eUpgQUIvBHw$PfPuj`SF>~`HH>!=<B6$k|-
zX3koyInTl|B1<E!VReBDuaG_c4N^;nca0;naHcpf<;WoLya)XBhD`&O?FZe%oQB(=
zBDcLc0WN7ghdOi>Ko_#Av4xp?ckQybS5nxmM<V)-j%GTEaqeUhP-W6e6&f!gz)++Q
z;_4vc0cwz+Y~w_VDyKyp-EAx4&yjG1Aguo+=5o%{Zp_Vx8w&oQ5&5g40@$gE?vFzp
zn1J;+VrAM0ezPw2KM-DtC}6<qs_j^G?)&4!JeWu3#C#RaS#CrF>W%~!X1!dBG$@oy
zdgNMen+ZuC53Y}I5Q=cUE#Q|*ZTP%2<afw3oSQHM4(bkeQeNF0*H_hVjrCw{G}gnn
zN%i)JultuIrjvAg6#cQ@H9wYcGM*VM90@_nTKuwf%uhuftfA;q2xw4{<viT*qKjy0
zp7eu{Nh5HSxUo6dw9I-^3EY8IU`xOt#=Xa*B?<4XzD8^{I2c0Tgmxo1DcTEIWcVa{
zE}#6yt3Mws>t^K3&oyVhmSy2+W*#=5H-bJlYH8Okp$2G>oCsQ%j8&GgApjYpLdY{r
zeZN2Z1TIjBQp7~9wwO0ws>Q_dvtvyvf@!~1>hO81VkZz+6LxIL%`#QqsSNo`+r`c1
z0=?}ECO%<rrlmSL%Mj;?YIAZ$iY_|6Gcg8X^D+>kGf}6&xiTJdbrno&e{j(s-duN(
zQ2ctaSd>tMcqDY`TFZ!Evf?KT$^%p@g!kZzTO`0jEo1@<_ac(0;42BTQQSu!RihMi
zK0NC{5fcp#m!m)1-2HCz`TCzQn$Q?I2Ac^yladmDCF|_M!;0jYp`LI&8QSfDSbMs;
zyFUW}W&b#hA!gPm0AYQ1Ykv#b2&O|(0wHGYCxBpccXxYtdiT{?u!C`*0E9oR?LOan
z{@qj{N)+HqqQMEAB_B-)&$lrNn^XI*;$U!jI{XCvGpuoH@8#%w{o@<tt^Mf1%ZZGs
zy;l|?*Ym<T?Xk4(q@^f!MYz?#s;ah9%NXKITd46xTsHBY5OQ|gTcSt0j|uY45SCnC
z<S|qlB&Nq6$%Euvm?*jetZ28fKciaFpOwXB{n_$(l=m3k-mOn1)TS;;jYWU@D(@`b
zY&d0vB+HNQ>reEE*;R`1s9gnPJ})_}eq2a*aAwH{7km*@b5HGELtJ0zDOk(pbo(q>
zg|DKy)2&M+M>s^EFizz6v@XBzpCeS~Q2&8ZsQw6Zc-)*K<P0tk&oH;Hhj_QP;oPW|
zndyz*2=E0g5Y}x(Eg;Zybj<W@jEhZCQg;-%IAt6LkHD++JTkXT=?wFJ=#5TBj@?lL
zfF-6sB>qj>f*wluw{8qzzzfh^6jW;>OjzfXTOFM?a!cZ#i3@B%jNV~1?GevjCxgQ8
z2pV!PhX>6Dfy7hrG$UsN(jH5%X?XxRheIInDXgwm@rRNMokk*sdFq$}X4|YK+lFzi
zSQNj}DQ?T)MIe$IS(}s1QJn~;?Pr4M1P@~&+9JZdv3gz?L40Rnvw?16xrJo@edqDl
z-~9RShnIgpeEaa%i=E}qA1`+J^Yg_I$X^-sKgb$6lOY+o84(DA0XT!JqwDS&T(I`_
z06C=rGWjOHBIlh%531YMUwF|kHMzQJifun1WY3yPDQ$Oo{_t`hoH^j0J-ME#=%X%v
zORwd$MJvlS(E7A>tFxPDw!3X*ZDZzIRaOCWq18OB|HKI^dOOuk%R~Zr@R!rZ=j9Il
z2c|&(ls5%d6qu8`XUUlozpJnSR83y_1g7s^2^UoMmlVadir3H72+7+>K5UMV&W|0a
z=N`Eba44GQkkqi3f%_r(FiY!SxoFeXpHl+@9LURX9lUU=FlCWB1jc}*Lq3?q-ze3w
zlC&ML#B?$9y<^2bgf{psJcN4bcu+!l(VS{F|M_BTbDutu0@jq%OPL2&K~<qCH7rep
zRWVsVrH+;Ie()Hygy}UyQLipa!jyWd<;J8O<qYc{`UZ7s9I1h78>l6jtMHDv4-abn
zgU|~1{$MQBU^0!(IH%GnmYJs*%8E6&hHIc)T2<OoA}M5Ab5lQv%+FG>WOOGoNDDzR
zfOLHH%_O3+>73Z#yHDR|AcW?QWkPFqXdGXV#<LGoXsD?#QT~HD#=C=|vdB??&GUg5
zKPH-1W@}JaQkv)u{{ZJgv}u!eAc?!Uts-p)(ofwrCFv!Wi0jo!OQ2pxZ=@~Z)m<Pe
zn+X7cjXK{A`oqMrVL9PYE{}T#EXl$fpN9G2qEjq|B4FAFvBFE+mPqTs0h}__lCWvs
z=&Dpy;;@?@cbmo{#Te9`{j|H9WwQNZf9J)%Wz&&fs2VY>P!EVIc!8UU!KM1_3;NhF
zYJC!&?}o3i9#?QT{Fh(ZKV<$8YkN?LsnuT$PO!U#7kZGSwd5!WeV~F_!6D*HuHy|;
z73g?Tu7~3z6>I?rQxtlO=xg~#P5a>f@)xJizFoP${N?HRCGIMwk<n0JzKAQ4l=Gma
zx-;CA4xSPe5k^we{UxOo*<DJf77lF;Q11&g9i*7V#FkCgKh%WbNK5hxgT{3XSd*Cs
zU`ANE#sU2MB|(7!JZ@2K<;}xDMcoin5-iB|W5Br+EcFb<ch+h-6Z$bmFlK@qJs}Nr
zg=LCYjof7$r6vv%P2Ps}1xRIH10mP4zn=`;Nqn)6G0ujB1|Eqc^}by0G~+np%LV57
z9f%YO(L%e1AT!%iW6u^LENWeY#VWv&>&kmGi2aL;1GGw`6!&X%=4o~E$^?NS&U%e(
zpsrrF6wolerX=V@*c8q>*eI81%hFm+N9lMa=)@5-@SOXp!%F|jNaUV#DGd|e<#zor
z%26b(1bndDX$@iF3W%n5B+Sk+P4?ppwJL3cXfiTO8Ym_gD09T1QplcFnQ-DZDcGQ^
zbjNUj{M9T{%)K30L}dxRr`c$B4%^dCg54wXywmpT66NO)YyplZH&g36!L$aYT-(IC
z62hb1m}zHE&=oUT0a(2&6;o7t3L5~8K^C13%*L*R0bp`csO@~*K+}%?*@H$L%MZ+9
z3FlfiW=k73q@u_4g%`ZCHVBM=@YILt69=ThJH+9Q041fo&#@GG)_uEXU?VB0{rX2x
z5fJ+Qpx0l!9KYf55(g?VPE(!VKO65!06g65<I0Hj^T7@hfnHt5U#qLH2EAVYQs4U-
z<a%uY!W@D25eiJ^$*Upu;wWWp67vam76=u8IC`dz8_atfNQYWCe}6pBa9ayW5QXvP
z{Cc4Mpqy0aScSvc;>_*lL=|md5<OhMJUnYd$8-3#p$Y4H%E0hB9S|NbBG}3k+~hZ`
z2$QWdj!Gc%v8uW{n}+K>>fo9bT>+%7C_miNf9j`LLewV)h`&L&A}Bc+hjE=?uj0C8
z&ZyQGzp)qYY-)Y!Z-OtW^;PmJSpA@jU;BG=0kyvPB^3kym7KWZW@##>_9l$1=9W+-
zx-*)Kne$;%f#A{cLRG*Um1B&*)HCr4w0*j|8a6R?lnt%UpdvTw@Dbn&s)GvtNwn`-
zI+|f1GwS+DZ<TRP(Nxl^b#+tuobk>m_CyiL+L8GyX&Hcg5&;`~s_&sz;>1ezE6AB^
zltKRx;oxAEBLQPw%fqGP%C0Vp>G87OPql8Pr<$<Yrsa6Y=u)hAQIZ{~a`01wOv&wM
z-c=Xib(7~g+*ir=bKK#nqNQNG5CA52M5qUWSSZeeap1$ZsS(-o%(W^m;wf;|y5T3*
z`4|q4Vbtqjsk-Mz78H<MIi2fUz8(AiH7<Ib=;@i#$?UBAjy+A`+L2CNb8RBY+u3ri
zTa}7mLFcz;<w9z}d6XJ%C*P{eTHYR!!8V%^@lWFL5jY<^0i20{fj@FbJSeJ=2$FJg
z>Sb{oa7^K79)ezX`^b_HZRl`AbIb1xao1jXJz-7pZ3OYi{`%pIE1JXMS@+okNE55t
z@%KB`eSss>aEzay->G(}6n3_!!>y>fRfQplEt{blbnN25G~yO!D0t)d4^G1Fo=oTT
z*0dbVVw{4$7>9%2WknpJ){L-al9i%hkuFVdbKD5o(u&eH@gO_<C8T8eKBkrJkEFBm
zKFQeT^WpF+JT<nqhb?j7z3<mFk9Tp}4C&>|%Mxe1TJ(I-^`s6(kCy5IR^$Z|<6nCs
zhlWzP@$z-kCr$9X#yioU%I}7P8#fmWv6F)OyC}HVzu=OG<_@ncPEEjH5`Gp%9eT~@
z+3+!BvjFyxE26xd)!Rr?6F03Q&Ve2j6nyb_KfWuvj~|?j`pBIF?8FbvHT+7nS<1#T
zGr%1S*VA+Se)xx(?|Rj{st@%aCx6qrRdo#n4|#&WB&cXGR6RO*1z&}nV=YQvX;zX-
z@t$#kFrcE|D9l<@2^@V#8rVw117+d)>Mt5cEOwezt<naLh~;^fQ6d?trufr@g0L%0
zqpcJGtKuee^)nL_Tww3WWve9Btrn9>z`i|=vLUCH5uddUcCSuurZm>myX|YM_eSq<
zxVrX}IyLHhy%8y0TdfCnkpV<={WmENk)T8=k528%*TLB(P6);I^jDUuV#p>!E;pV~
z;OX!UxN>cQ@$12raqZggVmpvg;cEp-&88UjlsFX9TShX9t3SK~?>SepP_;238l*1J
zk3pd0BA^IV(crjQ)D$I={R@Y1hL_!QU(dbBFO>9IL;p&%I1A_74y=kfqnKEVOWsdP
zA#G{Q6Q7hc4)7r43AF_cy*%du^%;CiKEOljl84&zQ6raCb;UDAP&B!e4LoQS8bwWd
zMJL*jVI7+<B{N8N5^%v7cG}1d&6BQNkT0Dd+6Th|zXX<Op9(avNbxE(U*U=@TzMIa
zD~n_qXO`Aw4R%<n=5=VR;F&C5yb47t7-QkeOShRocqP?=IY=ZL_n!HGXD7Jd2HC<+
zaL>)V6C|~zV+t=V`--}$bZg!QXuOG+G1wc$0OS6Iet7vHBAW8zzQ$pQC)+2=9+T3M
zDJn~_MlzOW%*sC1af-Y|%4qU06<(Tuu9-8*__w7C09#<(+RgM=AkW~sz)}J_k~HPK
ze9XsFE<N3+NiY*fOzdEnUnqi!6*%L%tzt9rxRojztE)S^n@2BrjrQiz`m+s0{8T@0
zJ>T13d-{~$1Hc8z#T4yY9XWF;+6e*deg(iJp91gia?T?gmOGNT*U|)Axm40etQD&>
zsZEjZ=+nA<gPKGRC{Pg{kyBZ^vP+-O2!cv2q#{MAkRy^IUGta*gJMS5^Q5>i$d&!r
zroWdfj)up81TRq25TJ&dB@MYe4y6bgQs#JViWNXWdh?=urOLM-_Q1S&`OpqgYjoTf
z;gpO6Cy%cCPR#Prt`K3<Nn?B@q8G54jwB}uD&TFVWLnergxPQ;B^}ZoRaBY5*|2II
zSbJ=nr0@KsQ2-~5qOpu&$uy6uz34O~H|BsFY5FQ^4)m1k-tmtPF%FFQtoMe7sT@Zc
zZVQ~@F^*8yHad`aONrTZPZ?f{56M$pRr0haY@x#K5|kJysa#%q#N(-_o>?;RpD_4W
zqv0Fy7WV(s^<sB7WIThBlVQH2rkD(;afX*3>3%X#aET_Ogbga9mp3^8AGm@pF{i?Y
zRFAGkNTVLXnK=P?AK*RLF(~fVj=Q*c3W+^b;(;+B>xZUrl0+U%sK!)4prZ8BL+eSP
z!b<Q`Ms#S4cZ#%Mz*pGDB>=HQuC|)uyeL3+<#E3)?-^-7A0lU1XtwnR$NvK|ak|OZ
zXh_GU-tW3BCk7JYXBipM_2Fthzbxy?&6xR42G`3dTVymzhYLw?ef^F?1jouU9C@Y_
z%wR%^QQ0uY5`?)>N;1QvK0RNeuOw_ntQ)I}V~Wf%;arKHJfuF9R87tlQc&$XZ*c}&
zbI!HV{}CS&7I@26wLUx6hUXS^h`bkzmXN{fq5)fS{IG$GzdLhsjs}|9>1o1)go@U?
zm%~dYU9$;8+h9Hr&;CqgCy#jY0dX|Rqm2)5jCJ9cF5DD1&u~A)#m&VY%)`OC9L}gM
z7QR0@=ixj8VQr~Gpw=aC*{pV9I)fCc)Ks%GG&fx$^URHqRf6q=*Az>k{k0rfro)F+
z&B>7Tc3<P}5uR=qR#RxQnOV-JGdiyNC8qXnb`k;b+LAy>+$RFgIEAN(fUF-;(xw{?
z)s_N-@FKxt6Lky}@@&{cL|w+uxkmIweXVtbZ;nfki?>nhj+xft8ay-JK<=fPR&zHv
zNW6ZcZBH5p?FFx*G>26N>QLPizcVrp&ozP>hlEe;Y8M8YX+59jH8YE|o0zmn0w_%^
zENRY#$QQ}1qNVILyh1mU0dPLtwY{<V!_l+tjm>Yo>O7D|4LT_Ikq0%I7u$RL6YUFr
z|0xWNL@d;*L3a`ThWVNq!d;rkyfo^6JGktQV7h@Hjd>4Ap3->^yUxkolZyU~9qD(!
zfs$ju^~bpRCAr+l0^R3Yzr7yboV-fEh+@bIV#RydZDks<PW$`se|Y9OJ~c}$Om;mz
zsD^`wH#}6O2rb|*dc?vRn~&3K>AW@vNaFYc1*r!vx=+;Irf*S&|0e@7^c$r?5OnS_
z8W+iZ6qEWP`1Tf)^ENIa6Q=`Ns%4{<uQ^QCq}PKo^_2EkQpkI?<%5+G(#}>=WMais
zn0zWBH$+cxIZwb6Rl6Q(GXO}CsO4JG;MaU1{^HcT*|B?j#nSTfDS%PE31T$;oWW1&
z>21^t7VdCSQaMwj0;<+2r(QB-TkFtNHT?nnriuI_AzjICc4R=vQw0N6H*&?cyc!PW
ztb{Ey`#e1Ni3=Qylm^F#`sR`u#yHT+@Lb$DM2#8Hxg`OHT-a?XHHzTQh8MK9@Ye5&
zcNIUNg2TY@lQi@K0g=QcU9kb_r54m!CWBMxBCM?H$SE#r+bfq!(JCiqq?_L5%#eiO
z{der!1fzb`E!A<5aBiI<T{>u&x$=82XvXo*db7~bkEpsivH+*b*%_?4IhbiO%^`RH
z+Q~`(N|njXprR?F+Hy+$L2-rhKgUG!N&vNDB6P9k)@Y)g6B#-xBy!{wGz|H`i^<^B
zK4L;|^Mcw0@D62SEXbgwv?c&q4?Dx1_BYfl(oy}O_vF&R%=t+Dnii(QMcwoG6Od_p
z*{9pMWDBM$WZLD)`HgS@f(q+q37;3+a@-JCV5SS*scGqv(!$#LIVT$Y3Y+OPtklfW
zTcr*0=_p8bYCd6jIxp$xjB&BuxZJb@t~36!=@1fzF|U({slz8hD=t=oj1WUAyi3q)
z2gzKN86-NSL||dckx!lJTXrIKTVXW6o`0*lG`RebA-gB1SORv6l_7}B#ucj^R7aE&
z7QA<k(1wyypY#$DS&oQZ+Z+g1)85m$De1cK*TibV48tR{Dl9<vzL134lfeLJhG;?}
zj3>TImW!U^4PAI)cS#M#Lj8TRC0zHMG)ljfB_?iM%dvdYqh^`m=9PtUfYj;FO=wD&
zY*9O|S_m&VqLKIyo7}ixB8VMVX`(8`McGltj_Va<8Dfco%VMVoM&vcTK(Y+6j(Tf!
zPx**4atPRYiJY!2a1&eGXko0bsZm3D8MgX*lHoDG5!34=IB*9)oI-n&l?vA`CqtxX
zOMGAjPPFNO8G6YaHVg-~7Ud8g-$6hntGj5&98t2RXtN@$F<eqesfz!b$+$ND|7kd8
z>l8<$YCR8`t2|Eq6WAxOz}hKfOj>!mef96fn@mFP;(c7M!4@=kKRkT!7PoG$nD(Dh
zf~vP6%XbTIdv5b1OY0?~u%N;PMNcj!wTlqC?^?(G7Py*tP2i+uBfU&~EVp2MqYG&s
z-T1Fp{foqF2vVcq%kENIU6sQ+GCc_^QX=0Cs+>|@qpO!e!n{g_iznqwft1uuWXr*{
zz*^xfDIeIBt?Q)W*{U71rQ8w1F)r}yZ~a~yxg)UWGo3?NU%*9DxMnG(BP^B->sFUX
zTA^ueB*<4fa~j0NPx7srzM$UsPbPpxgt!`@nPi~>m<Ygr2Y?IS8LEU-iNwsnkDy6I
zufbzrkfbI&m!t)s4P?(rA2(=0Q6Yn&a8?{NZk`%buqen*{)qdsxmmlhxwH9v1Bqz2
zHusM9wx9l>c}TToLt6^BjiZFurjs+PQNku*G<Fm^qmR3xzH<#UU{>_d!-mD*t&M`r
z7_-XFGGVR&%qms_i_lOopcROnGJw_0Xm6tASlD|Vjvh%?X?6#3r8g^_3~$bR!mYq?
z!1xGJk8BIKlNeg$HX^vsTf@_qku^6aw8hB`rXHGw;w~A^zir{EwQ|vY+xY_M&~oea
zkuJgwZ>9{9Aq0~nRdEN;409m=<dm!IR+6DaFgTEpCSyANJ}n&sleI7w#>)H}!mr+e
zd$#76K^B)VM@ip{%kyE{Nf<^XnJ_|<hZ-?s2R?z7dOAouw1p-@XBPqAND;ROPTpbf
zBRr4q+BG<fI>-_fG)wm9Ly!>+HH_1!pZB)tWb0-Hk~d~+S;<arsyn|yuCej{@W}|<
zmz|^>G0;>i#Pc^)1pL|lhr%Mt2^gBQ=cvNkE?Tw#V&5QzjE^@(5u`pv<D1fDY$^7L
zY!--rjva0X{7?lAAZApU!XfK9DuziFvRyy8A_hq8UZAll5}s-cl3?BV5s3yXFSa7q
zQFvsVc{fS~x`n-^T7MW_2EhR>!8pgM?!+ua6zIu*?Q(k+d4xm1HKm_m*oB@!5zt4*
zux1Ef<R?-DKtoF%d}@{(1<eR$U7fqZ>T{IiGfrI{#Lci$Bt&m+!Pvrn{Jic)S~I*|
zPtWq;s)YMJU-0*#!Qe>|XAfeZcF*toJ|V6bF%Ib@h#lE4#1byw)7}%a5wgN7*`$<p
z{B>;=$Je-0D}G5T%55h?DgnYqt~YO_%^M?@J>3-qJO1BQIrKW2!&}r#7{dD|p8rCK
zw5}_!J7Z7Dup2wurIQb7_G$(eDQ)`DKv~gO5~-%jJ(N?o4du8=h!^OD3bM|BH$2{d
zcSVIJ&PRwTadu9<nMZQOyKVtkFp$QOEEm2Aigx{V@TSS+t8`M#N{3f!&;`!QQf>V{
z1Guqv(R;xcQeJfnY&13>2FG<S1M{MEDcf(x5kU$WW+g6y+l(R|OgYrS%RKD%k#bwU
z(*?*j-E$K@)LN_QyeZyn^&|~CL|8u&jQZA;am_pn;zG{}sz80f!YoB0mx8@sN32_y
zh}#IUFeW2H4fcRg!ztjP4~<t12tvX`MylMUUd@84g|X73Y(C5cEQ4Wh4{BL!RzQn-
zwzr|=T8OcB3%MWn5aD-44OMB6Hdz}!yXm;h^$#kzbVrRD1|4X0Ne%Uuc}41NJIT$`
zsME)!WJ#8o)m5*=<S-$!j{;lFeQ=|8INNE<3Fq`|)o(Id+6?@iwe{bxeYbhkUWStq
zd;C`r;fbbQa9CZ}zSpmPr90n642WXA(E*Wy(-GHFmukM0XONgy?`a?jo~L;UqS+s2
zkIs|uAbN2yUktdnxKuVdbv))izQo$g@&!_`6WfRoi1alE7~6HY%~U3y;tY_edd6C0
zx4g7OP^v)OgTVXgfC$#L3D`V9*FsrxPgOby!=+6Wh#VIs=(Tp~n84m*+6kg}5H2|h
zjpcOao{1pofM*k_c7X}DsmJOn&70Vj_bz}SETou>PK^$l!j#)%Z2aTSSo(@s#3UGh
zs<=BrG&vl?=l#{y-4{<ckAAoPtw#>6HfUmCC#}}*=Gxx&^Hp3fEIy$gFN`K%;C6&t
z{n%<)PWx3(vS*tN;v$&94)DqL07@8rE$?o^AHBP|Hxr<r1{B?lxE=SQU&cZH|A+mE
zvem$QM4=4s1kbfAYzwS?TFv0`HJ6>S*B7I69&acd%%K-isbWA32`_xhAjmz6%xVJ8
zY<-LcMh0@(UnRM9jTc3{j|qDQua27~$x*@p&V!l*CXfo5%f%YCeGX5F=X<ADiN3l~
zhCyW>QyUP35Td@Gx~Cp;_2e8;^@-Cb4VnqONJD1AW5#g8+nndIj4bHD^j7WpqC0vm
zQpN>&ue~b7fKjEuEj>bZCYz8_V)<*o|0%DB!=hfmhL;2BAJr5x?qkpevkyQi`~lt<
zZc$?Yh`cE8RJW+T=_sWIE&iH47X692d)xL(ofge<Io$$-g;iiyOFpD&cr4qXt>UAz
zbTzB#ZI-x9Yj6@WYINE!E?-~b3|s3O_O1w#e43&qtVWP{I>8mf0-}+MZT&z;Yk7_X
zKdc4hpbjosA6W${r<o!oF}pexSd3xEd9e+7lShFRGY<o1y?QD@jQ^|GXC)I%kZMpZ
zJ(BYj);UfCO6)VMDuLcyVsiztK<VUwUFnd>w**9h&#*IQslhN!hnYRDkE1uO>0?i?
z33I+h+#lu@>Vs@?m5@{Im?iGb>lcwu(J7-6>MG%vrqLiGO=jwoZIf~|CJ$+sz<#@u
zYn0%t>Dg~5Sk2dRo1IEgqA}$&nhu;)-Agpe*S*18ZBT-f-FZ*)qY=@(*L6!w0B0D7
ztdT^_11ln6^X1I}GtO#w#kDJH%J3?eZemQeqpXlco;2*v>74l+8#BS2>KpvRzlxqQ
zVeiCLgQVARKWpHlQ*{tY1AtKXF49G1yCLqP!mblt3iX&u`p(#$tuw0&$JXmJ!9*FA
zCb$h;X9N1z>oc-|z~Pr30k$B_i*{f|eP))bk3jHFnGj_4w)^mtfIsaB)p3-I(Ub1k
zhob}TU^-6**Wu(*de$N0H6akk6ajCb6P5AszI%Z}ax#G0ss>FV^Ep-!nlLg$sz~tE
zdoqW7^g6FjjY4Hkeg4r&&otdSBCOfGG~$a05km5*92LYA`EaZt7##H8O2}ch)TVB#
z88#AxCs@li!P2H))a-s$TPkJ-*zt(847WnLR$EA0efG*1JT!mlN_KqP;?0~|I3N!P
z+~2!QWcaPGy(3&qPSYou`zXY~+9H-IO{FH;7#2`pv5eP&qv3ZguuI*mfdG8P+XLm|
zzD|ra@-Tix(>z_<f3m&%ESS8zo8N8z5p0&7uou%>9FzlvLvK$9v(rs?+YIwUs)Ytb
zfQlmNT6>Xz&|>ecw}%!}d*{evn$xC41o(y?(4`!=mg|z>0|yEF?)5aULg14TeCT%D
z8JEbG(95g{sYOR<r5+a*>7(l7mWQ*_ob7z9O-n|h=aLeA6W*j9C2J77*=(80mT9N4
zv_#_V8nvhTzE6BFN#`Sl!TW2<%lYVGSZVo@o*8V6+MysMDLwp)NH?DDNs0oU7_LSp
zU&zwxN9<ZQVbl;_qu7|321y7d4jipll<-+v$)rTqTv#&=gt7ytOe1|{=f5~?L@>BN
z!}LV;iOMi8iet~HV1-ilS3Mj}g%@W}<fyb9g|4p{PMz3sY1&TCgiu3Py7Q2X4u#hT
z0*YS!6snZO#Da4Pvd}y#Xgr>v<W_@9>di8yT+x-v$(gNGTXnn01}2jtdm54#k@Hs<
zP@ke9Yy*<$W7WF_&r+JzZjf?*W8<Vy@bv|{r`JRHe_Yo9r9nz!_C*tkZE?t1WK~T*
zH+cU)R!*YBu;C>uJwlLbFHgfcntW&nq+1CtiuGh3*Te8R=|PJrQBzmnuxJ966D=#d
zLcWk9`C_B^&%tBYqFFYw3_2_%PZqBj#+EzD_?+jgHL}azhU|CxmWXmEXu>s$7K~#g
zYh)+sf@mpQZ>pi)3nE2$v427$DpUrzER=dUvbVFOVj%A7dQxbTo&t89N6IkutxP~N
zm03PxXp;#>Fn#)H9r)bU*rlo&kP*^7lL*<w(M?j=QK>%N)h+e|NSW;!ck$LSPbmXw
z;yp=svH)i1KV{7wpzE4@d^~i3{WH*sL6)3vDJ#_n@zV%<8&e~ZTcr;Eg5%y&)<<pN
zN!iH^g1*vs1=LIoX5#>*G<0qY9x<t)B^c8^gbQQPCAJWr&To)1DM!xwP>+Jr&p%Q8
zxC}@|TcV{c>aCvXq|RY&`@uEGL9Ogsl#&5YwPh;2eCd(hrjEmMfyIX-ud;sVHF=vG
ziuKef!qRwylXzV8>Jg|gZrhHd6GbQIQ=k1<xJH)z8ebL{ex@{{nA)(A={PK+f?(X{
zo_u7^(O@JIX>I_DG^9qBd%xC3XE!|RknBr5L!=kIK(Q(`LM^3kA(7LBGVR{Zs-P42
za@@`Y%(9A1bZEs_gRwiRlQato{wF-p3c(wd9SFhXOn2jcMO<!Xp+|Q^5<CaO3pxwd
zN9-jTxOz|x4@5o3K3gcEASE;<E1=x)Le7xC-XL%@P&rPIGpR-5lU3QAHxss$YPv9L
zW)NGoZ4SUk%^Ry^<y2;qr81SVm{a!ZYD<3Z)wqi|KvIO_6lHdznOKb{D;=%Ls}SBN
zBaB4_i^>H|Y0;@p-vhd=rE2tF$^@f)w8@oq+pr=+MQmkJu2Y{+Px}fO;e&|Wv<j5j
zQz(kp)jIJ@rBaYF<)FfcN)<46=7_6y6F#s?;$keBfb$@qnaUU|froYTR4v4~skjvD
zLMD~|%b^zGgLR#iMJint1x(h2g{3|^veXy3hc&ECXbX`LML_^-vW;R{NWQXE_TXAQ
z*)gNpW_yWmUUhLm!c>L~4=&hlVUA`hhj*2GO66@_&G%x6mwYj5b#JI-E*Qbz;u%J4
zLzxYj+tRv8yr9LHg_p#}ZKYP`Lf_Bcu%psDCQ4&3Tav|(2Aa5Iy|7S;L9BB?BaMbT
zvll}<hM9)r>^njHYc!4tI?T4!Rd+6C#!ANdL?$~(hV%WQeZkFfI^_os>CrWpSFL~*
z;9{FZMI?xfyTn4XbC;sdOx&fY6X<jTcPXmm+@)do&RPlt@Ub)3>F_Ej!=?<k%0tNB
zrxvPET%{djG(}4VlPG``h{__Bu8IO+G;>Nqx8pStGLlld&OqrZvkS^{)ar_dChVdA
zq}V$DpMR%XCr!?HHbBr4o3GXQ&7ercb9D+t&q&nFxbNA>g3YUbWg4EUFu^U9>Ks+K
zV6q7tH7=P`=SO6pTF+>Zs1#ghZ*%RW`1F&j8`b`1(m1NsW?~D~YO`>LYOS0(ByrXg
zh?tm{IA?g>r!$3WpB0Lmg(yV<&#9nx-IEVhyDS!ewZo#ooz#%5TdZBtJ`znNGl(PC
zQN5&gtB}E8xsIL$8er;%#U6)1fLNfIL53?uvjDgyt{lfpqKBtt`aB+{wqv6%ggKg!
z*BYrma63ys*t8*s-t1k}G*4HqH;`=%$ut|CiA#qnMUB@GL)nE>P@+HS`@vSA^5euJ
z7J+W@uJ!CUROg6!;*}~urAC6jv;tg-F%5`)KsUiiPfTWSlDtiR3uu+v2m0}#ZLGEZ
z@@07Sqxul>D)cd@$w+XR4yl%;SxZmSe)+(@rHM)^#nO+cAw(Qzr2<@}+-hvmm8L?l
zdAns?wN0H{vnoM}ASDp#SG<0^{Zmoor;5?&MZnbE)>q;BBPoX3C_=se&t%Z_7%rk`
zwM872|HDiMIGo4B8~6c1$U+4b1qvUiuTck9`DP`fz9@VweDlJfU_A-SW@`(F5<wq~
zCW*PtsQI~)2o}gyQLFvw@iSv32|P&`>c62!eL^c>i2ya!BIpxq=z`Tns*5@Dh18Hw
zkq1k!79AQ3rC$RC)xs3$GAEc6l>;{Hbl;Q5{*v1P3%g=MNnzFjk1!156#^yHreV<w
zY=cHn@MFZ}rK8L5AW(Fr&2u!*2F#)Z6*;X4p_hKEScOV&O;z)+;(b!BrQ(;b%$kiQ
z*^x7NF%@H9D&aRLEj*;N*3)T{4Z7iZPXR)K`b8FM)+_~JLTk&ctA-Z<smzxa6<RWI
z`I`$kMSE;8j;X#(gv?5Y7bq`25cPs!3)BPO+VB7T>iYU>{CIU`rSo_d|NrLAo5SVB
z$H#-~;~PY5TpunEM`wTUAZU)S&j#18ZjKL;9p_?k<->sa5gdD!;gKeYhfo3oOPAl4
zMqLFUqv2JZB^^LTl?r<aTs1DKoBhtScb}4h^bPbi&|`HNlgM)9tkzY-d(jv#m5VAB
z1Y9w8%6N833q*TEIpIN{cJ>2|#srkzHPV7~K2jfar~SV3`0H<077w~hrw^C@<K_FW
z9)9RN{$}Ox3yVrxVT>S%hK@+ss*Bf5kSL!X+i9~ht_k);>)Dw!*R?rEQE5xHWrId^
zS5tw<9CR%0Z)KC$9-^w7B1pZ!2>@A4c(y969ZY%b$UxDEetVGSb@ZwvfR%w`x7L!&
zYEzjfMeC9X0`cj$Hhq`+#?Yzpi7^gBdgzGtMn0B}Dm@q^ty&wlz`{gU7FxEVtx}cg
z#}$=wnZP25lVo$WJgSxU(X#7khL`7(h7!s;bDh3<)xT_Y@sD2*i7b~OXr%JDVESZM
zYC#ZMkm^sdNa?t^hL;!#Bw;B1I2`nhD&*yE|E&L(rxJ4a9Ch9Op?>TvRNDeB$t9CT
zj+K)%Sga&do!PP3y_`B*RQ5&=qNo6=H+qferh<mmw18^?ySR2hW~D#^kIf$i-f(#M
z858g<^S?5hzupY5kt(oe$QZLjKmY#E^UUDd+|6=pe&NGlK8&YAY!ffxku7jh6e4NM
z4L-rVjtf)D%9QgLQC!T7b0#`bG<p1lsB^X<+Oo#)BWOpdC9JT&Lfx(ulDN6>b_X+h
z{SLl8V49adQ-jGILg}CahGf!fxU0+rFFSi7F4^qHfOx<qggK~kt{bD-%Osj~h9z9J
zzU?GH5!9M@C#FH#5lBYjL9k8h{zKAlQy`_7BFT{hMoXcpOo9E*n$aM3`;twYpml*4
z+8QIfOP6VNiEkt@SvQ<`U&`Raa2S{fV4x#8v$znzkt6XXX(PCz7RG(ByVP-H_LYo0
z8?BeJ^BXDZ*-~5*&0I0zAtgIJdEGth|L0BrMq(xz*s^zZF3*G#3FQ++PI8TKeq~^n
z`0-44<IFwoqrpJpMrZ;OxMtlfhhw$2y?Ae-xoIJluw;z$&&1>tH~@Y`W`bsiUrCtm
zyoQf8Tu3=8*Kr0Eha9Os?Z7rYCm~TOFLOH!^Rv@1$(-R<#BW*;Ref(po_ueZ8Rfh<
z=?3DuJfqB@V-&~c1mj9^Wgelg-*iVf>wfmgNT)TQ-i}8sGV%9(i&%l+2hKg)kR`Sb
z&f(?O7IW32tu*H}<_St^oT_BNA`s%U0BJKUG|jt$5aV_OQJ0jG_CP&|8%9`gUVbp%
zlzVPWAOwZ4DAohZVzYpaa>!gYT;#c%hFj*EhdSL@RCZ8P5=%yBJ2S5S;5t%XgP29@
zp^iY}c3W5+8_69J#7_=~EkxCt6Bn2*W%14@s1L<uRaDuwQGhmA<5#u%8FM8UiIZz3
zLPJSM_Rd$G;k=d$1~zVWq(j4(%xs2*bOsYqGPMz#q{j6SQ2)-BL_W9#C|w}D$?<%Q
zT*;a*$4b-B^q_TobIqF`nFk)4ECMVB*HEX!EcbhjYbP6OLB?Tq<dDgVqb?Jyh`Q~W
z`iqpK2n=_gk1kZnQ~GKW(HDXb>|Lod-!1?u_myPDUujF}avvXbpof8TEC%bAD<j{=
z=E9GpX052Lhr^O0DwC#37LKpfb0}7s5te`;<`+7vKp-QM;@hWW72X<@OY}$h8tq6F
zk!D{OHpUMgW?sBkZ8Utf8C)0vt8iwPSe;|P=4-)46GbA7046p~g;%=LdUFiIgJpq4
zStqS<Rhb6&caj8cI7Ny)?Vf@S287+HM=~%whyZ9qI6a5b4B6+QHoJCt9T9v3qyj77
z8``TMHuBr8s|6}_SuFxA;Z#-iy&)=14^}NCY31N9f`NGQs;(8kJTP3^hz-q%-tszl
zdZ2qfq)Qy)P<EK9IEYUCnHcXe%D$$GEJ9_M6U%`tlW_f-hQP3G7KH;QBvLbqjnO-7
zFGK#yzi68RE&(OH?;Jy5z@Ehd)_0BFl)YfjL_107%XqW5<X$aYyeKprh!=Rq(t-C+
zJ;PY=$x{{D;%fqP`lS)_=I-fe^v?OascZXI>6q_t)$KO!s<}W)bSFzqd|18%7~-)=
zCaiiHCJH6+RKKf^<eEwKJDDiY_}ZLdO6o5qRY-)~8tAVMzx8dwsZn$9Y0Yq($|WPN
zEuw?AA=Q@9VaGw&%!sU&NdvASXI{UWPHKJyYt0BN1(KoV-)==<>d5HWHOlC>SopdD
zmD7&CyHz*vL+Guyy>R7f(NX-649s3OcLg6!l~pjkQ;J*WQi{*jF;>2(5M*_42?$u~
zp4>MDk9JHuS_L5cmUJH9H+Nh!eiO_359K<gbiy0-1;DURw`v^93o~f)mQ_uqSs|JD
zAsLFjZ0xEW#$fCnRmDYUYm&SW8?_((jpD+Q>rW8NRk>>g$S65zZJ}gD)$fgitEEz1
zdBJMbL<d1F9w+mq>n)T7tl|JT!s&@Ct%jT)SDiLg>uX$Z{iZ+SCZ#LR++MzgCJ?#_
zb`h@kasH6LtCDfBbBbaK(sU^4_gevHN`PNPsY-yKEiM>}iF*H<0Fi#VL-%@+*3|%g
zsv<)>312JS>zC&4YiWj8s`se?V^hF?R=Z9=CH;oix9oJYOh1hD$K|Uk5OqZoU@D)R
zfht)8G!JI4vJCkhaV*<I6%4wK<HrPm<U@1UG^h*t(i@EBvQ@qJMGv`~$4xi)Qv*cO
z!)3*r`YOBOAaBlaTb%5fi9dKuV}ch1bQ-@P2BfJ{!cQ8@0Da93)Ik1W?(4@_uddAS
zlz-^Zgmrb&4~8YqWWt2RZsG_n+#s;iEzZNOZTx?*i++=TG#k?Mfj1R<aXmOcd;%xo
z_#IGnQ8Xf7T?W*C({M>8@gX400F>%`XdHQ&3BCP^p4gI`&EJK=!DvS8<u%UIz<auI
zJh<E#pi2IVBk#fhNZt;H*&|ZJg^^fqC1o5PL4JeP-<(`mt$FJh>>m(lcMn<-JuMsu
z_YN@Ge)86|TLqd#;NZs(+LnFuaoFbpGv2^G0N07GgshXaOT=H%r4Rwq?YPP!8z{+q
zYnR7&j2HqD1nGu>T3LAVU*U;f?(NHMs0hOD09X(U7?fN^Phb^J1aO2dTwX$2(xr5d
zIFRBr0OAf@bWHho@pa)roJ{zafCxTx418Q}uONjrzFFJE&r$y^j6q0Nt1`)V&!pYK
ze^1_b1sbz)3%lWs0BDCfyYbI*Tio}i0Pv;#OnjNRjYzMpq_6T67k{xqPyYS&if<P1
zXf((zJs50xo(yb^6xqqoLDsM0P_R^-H1sYBSg_y9iV*)AH)rD{pdl<3LD4Z1nv;P@
zOEwW1u2@qLrOx0ZK4TBh-zsRnU{_Ze#L#J<yt)|ns1dQ;gVw`e{pwe(pSS+|<^pFe
zBoTz&E>e(=%%^eo;$l6{$ETYHju;n$+F@9|H53=|34Op-D$}G#)U%lo$EU&v6#Dwa
z*3x|U=6X0!@L9yU-|FgM{0F3s8@|DIK{_@iQ{Lx<DH4hg6B%$us=|Yj5oRXF=o^N#
zmA!WtL~>I@te73Nl_LTL^rwJCpbsP@;^Vde&fS3wU{93XRx?0{0W~tBM7$!)74J{U
zXnu|u{K>fwaX<6I8jREB!*quAORd7wk>=o%J=wzJoRZKgwU^t=$G9swB{r5(L7g|^
zMTI^f_vGX%_Un#%1Kjixh#Y(Zts(Bw-=*!rzc?7_wqQ|%)Lv4<=mJfXDP)aX|3(C*
z3%%Ee@`ZF?<YnE_yLC}&lvhFe69+nCiRB*O979OU*9WboujKZd_A=C7y>Y4hOc>2w
z)+XD+J<p)V1>6|ilf+4<IJgA=ET{;(y(6*%vNMJUoI;90jA%)`KgDhe?ep-0fWaC`
z;$MGukCCr8<IxJ*Tj^i^*jjvy7YBb{dHMOu;%A-Z&leXyTo}t_0I?k8m5DSMv59kG
z=X(#n(0q*nd}xFUeN=OTBR7?}6W5Zv4pYe`Xs0zQd?eALLh~jQXW<nSdshOq+2oPd
zWV%$*z>HRzDDphMeYZ0jzI_)!<&0?!6;r>N`>n_$T|9NtNXoUAaXXGIA5|GDi!1B>
zj^c^!BhU=s&ZaMPTS<*K*z@h4L-zOHxhjGb`pV%|gQ|yF-d^2YQ;WUErYT65%p|PD
zv;HOXi}X!a2{z?2*b<Z0)dVt_CXH;-y30FdCZj%r%mzR9t%*`1X4qHJH`Cn9PHoMy
zFuI+g4Zu7r24ic@_^ZcWDZkKZjTGo#j&DYNFkRR@!_m7<(`O5Ynlac+TPzAv_yz+i
ztmuj40HPu7i0rj>7R6sWojmKj_lm}|g#18NHJU>$(9r@pu3Qm-YzU{Ts}mS~w1Ve<
zySe`Tc5C~|lf%*B<^1y8JZ$!JZj84!w)WP(eY&}^`E>11`0pQg)}HTeZ9n&Ndz<^M
zZ?~Qw?XT^Ax4Hk#{r~P=JzSYz7M==v=T*jTkgl%w$0uFf9O%@K?3C^K!*&~2<a%@O
z4g3$S!}<HM-TR00t^1!zA(15$kXu73p9;02_<%+*FHJ*(AP!1DS*0AN!Nfd)vVdxt
zQ6S%GUSKqD5!TsbiLbx@dS`e0kAIqjTuX)8_|@DT8vkaY)BEOd;r+w=2P=p8d->Ui
z-?UC};ql=@Z(e>8xr7m##nqeM;@sTNT6_IIm5?`m_*!u@(#<7fr)VOrzdL(_e{f|I
zHjqzVi-L|=sgrkSS2X+p*V(9hK@NB_ytsh)ZY}-TYTM^_>(L7KG*%>&)t8Tc{>5*2
z@zUkozx~YHWg}zmxsR^Z5?Ddi-UzH{g>fXf4(iPSDGDXd3_nGmZ|!`)^MGw2Dg^*v
z_F*0Js<%=1att=sm&=`i2$*+(5MR6=j9Ezb<<|G`8N*^`6R3}fHQWx4o561+VHW!5
zfYFBm&>JFVlxCxaqoDZ#a?(rZY=Vo7xZ{;l-Xm+vpvIL^f!Sjx9;IRP`1H;F8|oh%
zv=%x#W~dYrH5ZZbS?lFXN8Z2bB>NOBUC7i9M)O$O|8_b6!*n|6pS&8jFzvv_-&<p#
zsXbn4tsJ&i&f1G}pojPP@?nMKK;oN6-1Kkf=jMdC;=+%#r|RNI`M^vKsd00u`)2?l
zIXb03?ORbgcO>}!q#s-%;LY+^8ZOG6r4m#ouh#5I5pnuaL-`c$JV%Fw<Jd#tl#3x1
zhw*1XM|0sVmD73=ipDK9)e8F)?{@oo)yCnrXMc;o?o`{L2+bAnPV8BCjP6#|uQ|F?
z?PrzSoj$G^$Qz%GriRq@dJX;ZVx5ZO+v_J?Q8Ta(?N8mXhS1dOjwRto9NasIcq|Bt
zaAF~RKIwvs*M0$m8=MZrcV80`SK}vu^iOvKCfE0!%W~Bs<j3iB6z`A;loq_5{SlQO
z-3CN9s#k+kUM3{&6i4ld_ezEbYB~D7GUO&$)b()-aKN)WvWqIGKs$_{bJQM<``ytA
z^sWF<Q24_ez$?C5QLc4pW5!TI`KcKh8h_CGV&brTU{$Y9%_dHC?3SI0gLfcthXGdF
zejM(-I3Trx^ajTr*&Pa~W9iIyX9>j0o^U$f?>=xMuMiq6=+G_@!_%{(p5q6Q9k+n(
z8<q><%))Sm0uH3EMm@W+oe+q97R_-QYHTf!94Qee2y4V`t<bw+<q4=P(%c&c5bLUN
zP0z12C-?MA$HM%?>&K*98HS^@Tyq1zkk0Yi3o9=<Dw3?h5%%Gb{8-sS^8U&KUNIlU
zxw}hK!bs&>iVHTG-;L*uQ7Lu$uL6oQ8id8W%0pR2q!;MU$;Ud8YCMPy&oHytL_`$r
zYm17=5{<W3IDjTikrU^f1fWDX6;4cP$c)oO7)6F0r2@^07b!)4>9TE5u{co?MxIun
z;agFyd=dmnf%jC<q+93pk)b-9K_(`&HFW($R{KT&`qfZ!Fo*XthanYe;PwkzR4<+Y
zp@j02rqYlz2}UE40BYJY`H@i;Cp2^^AnxcQk&U1@3AdT|bSfOuDxDe~@uJ5k!_jk*
z<|m>hgoQZ?1sbR_2?K>TmW)%T$aVp$2q?4Oa;lJtxX)nH0#}uLi%5pzoZ(_+d3#J2
zaA?kpXyaK4WlvdCOhu|CSudw}c2u`2lLmmwbXp91J}Xv`B40+z(%N9{^n^m?E>q4F
z((bvPP||{ZkDcY!lhK0bsSFFtB!WOUd_BX9J??@D7ZwFi&yh0+RrjC0QX;fID=uLY
z;aD=Oa82r#>+6$N!~`=Tp=?<lv#<wKu!zTc(I`}^pP@H9W62Mb$Qo~)!Z5xtP5s_X
zBTUMF)Wq^cE{CVOwhzI%rvTqTeUX@V;6RyR*OMC?d5ep)b}8F@BwPZ02h|}|m+wXk
z(rpfDNj_o&d5Fxoj8-8bl&WQo)ud<3u7*gnk)HjL+!KO5U3t&Bi~bE9PNQ-E=!Q91
zh*?yEruO~L(Tm-uN8jV-NeoMLGad)SRm2DjhW`x*WlS=#PHa!z{Mb&{fDTjv)~8ef
zYh#7IYq~t?yZq%>M~EMSr?Mj-4%!GY5{GB|B@dLX$}vn<VY0dL_|0(Cvz!xn_dWJk
zWY*g37KbkXCV;NCzU_{2x(A^h)>)$4zy`n|lJU3i+w|Vo&De-^dtXhY0R#ekv=p-w
z#6}8z6TT%Ju_yj<-bVQz&`$aRb-))IE$_KDyo|hmLlfntM5m4X5YA17qG@WpbpB;7
zlJe*e((^iY)@GI;cBE!<XRfWPko4W&0WGwVo9~q%Sl+33pe^o(%j>0U$idZC_lhQ{
zAbsWS5;6rX0ecrqz>iKUw!F}a?PW<b?DpvRW^mp+!s*k|JE|MR0vz|)QEmTRDln1b
z@a9@7{5ul<NzT8-P~)uX+$#-oWCuE;m-ttKeyK<9V_6{r26fY4g82gxR{STctDs>S
z<6F24Fb+`)uX&B?Z)Il*pOd{gDIi8Y%2~@O3XaN$!KG%-z)TOv*BjmIE_yXQBGTlN
z*BRsP1|<8?mvrsVKz-iWK|400gVko{+T1XZcAWtUIb@U=D5zR4CmW3A(^3vKsGM^C
zG6v0;gu<7BHTaNOC?$?x{u7BKFi=5H1A&A>wM1gVhE8isy%Ah<VT-R$2i%#}xR!ZO
zwR^ChEWW8KQHt}~iDe*lk?1pJOv=}I3e;bd`uTq!<t%5s7SB;m_y6w5_20o1oL(U)
zBzJOEYqE?i47SE8y~~sJJj<39EKXM*%5za>;x55%5Wk0tAcVWeckcSl9KU_Na2w%r
zh$iTbThb>q3;abdZ$NhTnh~qni+J29t*!TR7e$#&q<H08k~ocuZ`>*qd}aLw%XqBC
z1seHJsY#_lzYEx_ySye{)}7vc&`M&eVIa~wWaEM+l+zMOT3+x2vZY8S5pOfYe00mn
z;2yMW{weKcU?ICr98!=3xWxqyd2FLzUazj!4+Ol$Ekf*?%EI7+^WVW-tV9~y$XGZ`
zp79(VnygzHWN3s_r>`X~j2fcZ<ne3vBGk3vvOt(p>n|I5G~_Y0YJQYcDq>j2D*Fb7
zTQrbtW9_Z&ntG(^`Wlk6R1V>BKJwlI?&_-oRHdAn4i(K!ok~qyAX@EG7f@q=TW)Z2
z`PWo0rl+wTBh;D}!u&PX4UAN?TIVQ>)2;I0w$ebBc^xMkEAvub03{2X={vidM;n_v
zo6k2kpRaFi?j7xIKmB3r`F9}0_7$=q^^)u%OWd$TR02Zq;-57GowvCiC~!>;eWa{2
zVP3Pxok@?#=);kbI{&-MlO~sM81*1*oI5Fvf(#89F4tt-tuPfpoMEBO>PAxPEBAit
zfOdCKvtAQADPb~7|HwUw&^(Bu?7_TscY{Cp_Nqe*PBbILO25Hncek!>jwn3zuj~Na
z$6B7^BjFOh?E=ndw?UnoRoR&|{z88;G*4ShGn|X^km!Iu<wj2j$0I~HJ4mC}fXSNW
z<ZMVr5|`4032`Yg;w+Km$xEm)z2Ru1dQPf*<yN5Os>x1F6$hwjwP)XB7-07L*pEz<
zZ_}X%%z(T~+Da9fipJ&S5LFmxVFZmF`&<zWX8?nYG#tL>V3uPxP6XqUQt;4Nae~Vg
zCh8~tJLoP27Sl3fMQxPHSwhW<+GRtLApuFI0Yy}^?c}NU*q2obM1VRY4J&uZ1*&z+
zF(VHhIFLv)ONi5gJ>9~&Dy)8MsfwKn!VD1Ex{DMZ;tQZF&j14YN&6SSNKH~41DK^b
zAfALE5lD<i#0|i&*FWk-QKGpS_oIZ;<nh|HObrfm#LF_{`=-7=ofJ#p8$ofa_uF$z
z8)rKhJI?qTM}NVs-Q3pn=w$Ha5**m3jjxcq30{C~vtptQ2X}sO&LrEEzHlf+NZE?e
zCUD6#WRDrw#U^7+8`1j!nh_)$8`?`j)V@<76f6hxqYwlgg@MSCDGmFM1?+33NyB5M
z%|NbALk02DXUTltO>=$AJ(_Zge`3`HbKf)o-#z32k{rrSdS|C?g4y%G8B+VbReI^Y
zHCcgSmORcrBR8MK$}|;6(ULZ;K}D9%dQi(MTdtLME7PxfTjT7TT1UpK^ImLT#j!Tw
z)>~j)g*H*$zYEHvoSCRu{c%I{qt|~EoW68E4Gnl}DGey<W?iTy1;iHXXCi(_(pZ!3
z6(+W$u2yU)^+ih#d{-B@?<QGWBGr7ahvHWjxP1(C3@!-?O(#84iX|Nz!|;QBWi{oC
z%J`y)l>6nv(Y{I@0^q8wANdWZ`zoc0{3ePj*xb$}AM~^}%$+|&waYhEU7*{x9=i+)
zuXy~q1y0dkWGZiiT5z52X;9UX017q&L6nx;_YV|V!Mzi}Smv2YF9HDMM0$K7UAuZv
zm9z~JPqScXI=it^H1Zq!4(&-(^h7X4DX2R$<X64~qMi(?alp#bFve0Yiv*DqCxL7w
zmL;8<3kg1(v$(PPe&w_Fa<<jF!i~a}&+%U=X=&iYY%MN=TpllcU@831U6lkAa}PDZ
zJyprL<bAZg!sd$=5e_sNN$EPCMGLYO;?>M{Lq-(i2O$g#Bk*@|2sNN4;;jm4kayKj
ziZ^;7U1V+~EU9@YQI7c}&Q#61vp55rw6vyNk!p>nkdQ!_!GL7r=S^#h;mjxw+gx)4
zI#5^AJQ(sfF_<fDpI140yH&8VW^20jc56P(i40J-oeDH7N<~3Gro?7Ajm%0Y)C#7|
zANvL(M`}7I%TS({B7hAw5ufRkP6(la@FPW-E~XkvOa@scw<{GJLT|kUu8h#ETHsI4
z$|J74YFIbg>8)3dp)EIRqRo?HTDd%AOWlFaAgUsW;T47kW6qL*=v0Sqg2a#XuI<Kc
zL1W3YoZ>MY*Lcb(HxR(U+>)(l{p+qxnYTT~<pt#7ue4NdZLc#N#e3HC@-?mggDrDe
zgLPnhGE2Kj-S<_~1i}AKLvV`yG>dUmM~nZ%O`}?@Oe2unDDS2pwPo4;OY$b?PS1zK
ztDlrD@x)crn)~0zd}cB88<$*4fAt>Q|1Vio6HZMF7R|7U=UW+)Ujs%z(FWlJH=j={
z0sVM4pSQ*#O>@cA5BWB1%$46+odE`2#jOCBL$D;oJJIAE_()c{9%5Ruz0--Cq&2Z;
zAuQvL@ok9b#omt{t1os}Yr11-)bvsoXHpd0jf$%dK8vd2P67W^5#3R{LE_g!G+W^P
zbzyBG1jQTD=J4NAIZZb!7yf62k3DFGc@C3RoFbVVQwEdHHxltPkAn+wZx-_AE)+E@
zP~3up@GBsh5z=l~hL<$z!h_4vD*mym!)R(kk+K$UzsO|Tt3;Fp62&m0owfDfuYI?9
zwDo*%f9>g0z%LO*&Kd@)=qJHLN$5~rT}BN}t&H#`ulr(WV{L!4>{dnR<X)4SX{Sw9
z(>|X!j5g@WQ+m$oXPvu%5>4`>B1k53LE{ZIrfkMZX9)V($@2?2=P>HQA${2ejC<pO
z6BX|vQbTyH5e<Wx!@FSrdk<5%QYD0fd-rffU<HPz@Eb45NV~Xe=}B$y1VULch`=_3
zb5s*1)N%}A4WYNBha`j{Iz7T*x%9(BaB&4-`8hlYXQ*)`fiHY(c_F8q-~w2ZK9U|%
zHk~=t>^dGBHH!o{5%&|?lI9O2s__umHbx6}AVb}h^a7;`X+u}Cy%SDsFn4c~Y<UYZ
zw2RbVlAJ13c~Dnmb@lg~e>z&*J9@J9bo2SzvrXjRscC6-b^XQe?&kA-FS_;o$@Zco
zU?7Oxcak$eKX%V?Jq=VgRX^p?wuvnG-|;0!WZHOQ23~JNHO|CQ?xmaNDLT=<wKTS*
ze+y}@e#9XcA~er(r!$70K-KWyyKGuwx<pX6e)e={IA4fNg0HTyTs6%Vh80r>PHu_`
zZc+GlAV{vk3K)OT1XFhb#2PeUfa_ZFpOEVhz<iSAZX}Tjc{#;G3Hjeb?#)*J9I48&
zW$-VTfiotnt0v(LP{#q0yo9WYhfo(g1Z{rd=*=54|69q9H9S4_#Bx_R<5#`bvOgka
znxv}pXB_E~qm_~#>zn)Q8}|<nQ>rEfcv+@fTP|cz$Zgm2tzs+=*#SzD+p&|F)6ROC
zN7G5BXsHwP;av@{dfxASp8XU6?@!E*7kR<0(eO_inoZ^M{KRnufgCpcoWs;EZE!Wp
zi2dOY5&#ti4Vj52?yV+MO!-mZAwl@XD`S<?<_Z+I($tCA>_JFa59v4G;lFRMkirh<
zMP)jzoRGuD)~@DF%)b192>I>ZBl$?8!4J^nyd*{UF&UkBV<@usjgmNDJ=xxEt!%XJ
z-~XYt^}N-&A6rC2pDngFwp*RhCDVY13Oici{txx#SeCRb{oUh#@BQZf{r^7Q$y!?s
zPw^L)DbohGmYAXa<ou=w!2${YjW$xGcpUu5e1Db6t`R5ypsB8;WYkodd%Np0P5*^_
zi~pssY{I_zFJlL-g@^VfP6E<6S~29y)XbRUSN-#IYyyfB;*(4UA}Kiq-G}q|KY4^E
z(^9%*D#ycx*z(F}ziIV`T8eW8Zh1hIqo%2GNT+ET2V*IgWSeFIX`|Xf*&vJZFKyVW
zHMOi2T1qnn<UNP;VHH2D<_8~`7v;aFgB-x#@UkD@_}BOr1Xdv~ESkdw_6FD%ev8_w
zGW~QG4z>Ea^N_ZI+8-TKuQxkr^DS~)G8<g8yTvVlx|q<SmNhz)apl?cizX_Dmw95M
zS>^DIf<aZr*<(T4-z4u%#ZGd>W!?@CHNOJRmOS2EcBCUC2Xn99#i15LV*Y6%;vSlB
z8hnmaZg&iqo*~3qm(bm#tm7z69MN~<`g-OJ6XVad78XW0HU7yOfJ973j<)K3w?F*`
zfLRH_ZxQm<m<<BZa?3CjNPO*enz$LV)5CxqiD`u-cmW|ce#->t%U4Tw_XI!$$2kI0
zqk4e{SViK8j`bO&<}*VpNh}Hm$z^YO0Km+^P*rsL*h$;d%1Wwxz4lxokePHVo@!(S
zsc^(6V_f(VIYv;>a}f+R!$C4~KyE2^C8!T(%*|USn=p_pA%TLx*H~Rg_@?g-vxLT~
zKp@tA0@8RtLgkZN{OqxZlnl*<_?4w*&&ok$J1W+Y&=Glwr=v{ncA?HPxI7&O&xCA4
zY~T(YIdu=e&D#F=H206bUE32E0g0KY3EUn%kz<h=m4YC&YCu)$B5RupjeS7iwsv+p
z%!h?IL!DVMkmzo}b_gQv#&VORU|Sq2d?yIhBJz_bJV+gaCheuXcbpP>oj2$*bMy^O
ze{R8Mr?yvDf7smpc6)D=qtI3&KZ+M;m<|v}%x$!flF%+o3h6e{HJNO2xQzQKnVU<_
zJdfri*%>^KwtggKOE@t0=|yWQRh=nLk}Sv(V~2#UWKhY$r5cF2s6qv19Wvaeqhg?M
zy&$@MEK=ypfa7&Az97O0qc{YE7#B@8fnX&*^3*Dg6j`^#hta$64dDqy0KFRhi>qs-
z*^z<=ZPB@qL5V^j+0lJ*32RbrXZwN9!_BuRecY`)yzI0$M<b~4NHmUg^9DB8W^rv+
z8i&1%Ra;=pO#tL>e<k_TUv<G6ySUn(L1w)^jApsBi&@bj>2y#`kc)Kel4_l1@3P*X
z5i}J?qHHCDwHO94Y)hP}%prh6LebQ-M))11^F*vMm~7u!X5f^;ERT6K15!fbOxGrG
zl1uJW6I|h}i;%EmONbJ|*+5_qC+LRf5O`<EeRcAh<MG8&Ra%nayc!MPpi@#)JxTuq
zbgMvUsw@2i=`88+pFw7ZTc<M9C8-Qca4qK0$l#WQrqv^zeXMxe(H-(>hMI@)!e;cJ
zK@McwWp@)@Q*6D_@GsypTe^>m)NK<_PlLaea*KLjvX54T0ViLgN<iDDxTQ3gL<)>(
zl&u0Kh`qlM!_c&qSQIQ*Ej7fN^CBx_M%qgI@R0fs&j09tsp17PDPSh=?%`(6iYkK-
zxdrON1k$=3&@_?ukB(!LqbLq<39R#Nr-C|#XnIBn3cm#DL}O)=YPRo^DsrBx?s}v2
zuC%m#Au}5fJI=$EPcr8zFAWwCd!eKcKLaU(d70O6=oN815sMlbEs31IcwN5Mo3Rq!
zVPh|7tNu7iI;5sK`soEqqI@(X66wisC_{3j_j`&d>Z&D0Md7(+oaN3m^)X=8Ko)gV
z&iv$EY(JM&L&hI%yG%4vW|SO>K!`s>^by#s{HaRUcn33<E&k$eXZ1#9tmLr!u4UD2
zFIVVRraoomh4(LI+~E%1di{x-p^DGCWJZ-6p?%fJfH|B(z9i>$7skVz(FvVqF7Kg8
zbdlRIfv(cNY|Fa?MH^5*hBbyHh@yMwOKW(ptqto-{Vj<zeG5}~L;`vZ)HZEU{Krp5
z!wbX7WU{ppk^x^S4l5y)HEAuSjCBW`OAvz&L_!!0N8Do8<&dCs&T)!wf`$i-CM8bi
zgqJprw=zvgF^V9B`i#cLBsJuXq8RtC%9K?mD@vY>iXhQL8%|p(Y11^5EEIO&O>z=g
zcgO&7uVr>yrN2DrBzD3zshAIxELgQk5wBk(L`N5WBw6p&zzpQ3)5;(dt`WgB7E_3R
zsaeR(O>E$Uq6h#74Zoa{cJl4O3Hl`#N(D$6F(U1fgsR3PbHi&4ZSnH@6uE5g!_d^L
z-p6i+(dfeK&i(NsoO)T`w!&T(uWgN$uWgmRoU_y$Dck1in$64fJCG|_CN*yU+`~mX
z7u`b`-vhYHeCuYff8IOf=cgE#{McavXWQ7xzw!&l7>v}yOC@^wZvO&zus;dA)kT?!
zLc%p;*bV!<I*s?1B%K*6RPIlZ*$bGx21q_#`{U8x`uCg9*02P=kV~_XxNUELcWZq=
ze)tjy(el6B+kSrZ;`!#@`r6Ls#?ju>wY~2*_fYqjUwLIIg4A#O`+q_R!`AbCw*2L<
zv(K@G7tgoWw>KDB`Q<NuiEe^jag`esqf=xg3?DrOrVOrRm#O5z65;*oC4QmBp3)c2
zb)NY!LFAN3fvG4tvk|<N#sFjB5OQLbE%64v?5bL3QyxEFU0L}sfWaXZOV3WDawYka
zTgYHM4p-H6t0R!bi2Fax6T=+bevzqhW*s6&8jP5j5ov<Bg*rrrnl7KEM>-u^t14IR
zQuw4ARAyRw?zCogl>thR<(y&d_M}(!5D~+16v83@0@CFi=!{@c!cy0nG+J(BjWb1&
zF6oMN?!)POk1|Rc2R@Q}YQVIZZSwf$EK9{(y25t>t9aV5W0CYE!&@*gIrbxnog^i7
zgh^SnUjtKJwC2W2G~V4C@YsvQ`?t`S;XlKjn_Y+pn6o%tz0haN-tJ*w0yDIS^x?cc
zk!n4sf?mW3FQXE;6JK_!6KZq<F^PQitE=#vjr-&?9l=F58Ct<OQb_{hsOyLXY9V}W
zjAhFN&Qy6Y?_YiOYmxFWxbTGk>gnnTB--Z023rk9WX18N%!bh@Z&56lVxyB4`9~P_
z2)`$Mkw`a>YcU6UiDWn0kjrfqy3+k`%_c;>%oZ$6)Tup&+bX4N=~*wh<P^d4dV*=X
z?)SWJEKI0ch^&(!*?<4doce{ik18*hmIQ#&JIH#$MqQAnt{0z3&;FlJ#Uc;g__X{S
zu|d3ew#xD!<1Wj~c})$1nUQ-xC()*?NrRP;-i*)t{j1I|9`X+jWJ-oE40pRaPpWby
z#Me@2_f>{>7RDmWK|Ts~Cgp^<*bBG`N=#e73*0wDIWG63UjIbyNwDeww-6Ch7>uS(
zxzd(v=CDm7eQ~e~PJs5^<#qQhVji_X%>uH4JOPgZ2uF-aIy#c=)8*wA{eNUGkK4^U
z?q$MCZ<ju9R0mKqXf4a0s8q?Y0vdV0xJ-XkA~>3){iR2+4v;Nsc-cLN>jif~T`PgA
z+bb){qbu;r@OG$RWs<55E!gT%w_DJ}xQ!COE<KWaZIQPJ;{wj&APKi`wTH&Y>7^y(
z^nGro#NBb!3Hr0Lq($zKnhq$eSGE+~K$7Y*cx_1RL~QM_RosuELw|>xGmt{3@)G<7
z(E++vz+<%zlzDRY_DgE6Q4)}`8U}#Xz~MXX`J7|7=MfeLUO;YvyK-^=F0Zxqt)vzg
zY)er@^+bWjpn-5>=c!|o_%m*;%ppeVIg>D~J7Wm^n(Laetk@amRnkXW8?8s*w0`mM
zE39hL3Te`pQGHv0N?3T8fz6_Du})kzf}2gbsbrI=>+tgtwy&;_WLOe`P97l%N(xxM
zCMaM2n%jo8V1oZpL8z(qi!Z)hlxns|)@ch9T=$V8L>Uz=LIGL;4u6dL9Yjmmt4^D6
ziXZxm<$quPCEe0rBP~^m6;UZi{i}1_quXh(93C8wKR<jayZQ$&Id6gqtOEWEESBZp
zv-}4s|DIWQynLKxc_Y*AZ3y<#nfO%r$MwU!Z)Jy;AYknr<9y=cEV2XuEZkgb)Fc8z
z$!e$8uh<u9BYPl@S0X1T0}&fBGY5^Rj**Q8(|oT2VD{2F<kZ^gO7IR&eREEX{0@`&
zL~gdmuIK71_zxCtP@)P9U^4xUw+`UEb!}7Tm1{RzRps)nEFMUmp4`kielsA~pcujn
zKz@SIb>SOFo4dP=R^Hs%|9*9D;bMFS;M>o-ZwD7Q7rZU|<TW;quV69tv0061=pKZQ
zDF9_7ILx=ddGu^+@7db^deYj9%O0+eKF8)N_&zz;1)vzWE(T*^X3Y(*@8iOx_3zhq
zBV<w4!shqoHTHRN1v;*_IRTmx1$%Qj0lpXePktR+de%MXtUyNL{nD>vGq@xNQJ~|&
z@%cM=HBN>j?k?cBuA`)VFpDl`J2vSzfC{U^u&^{Evp=k;{la9lSy~{vcMajO%xM65
zlF!*+1H-PW=al-GK{t8HZ23FqwMtp+pjPuC*9e;)c?+)l2zyl0dA|L8Q<NZKHW{l1
zq19_c+Jgi*>NuU&*?u-~QahX}nBE?#5k^TAS0=*yd!}P0ttKddxlU&e-ym-5cNp7_
zybz5etXC_w^ax=CwoU<AK6ow6<-_;6s8C#x!c<ly2;xh@e=UL%0*SN@r67wAp+3Nw
zU4|xLCW{a)BS75JlVQ>o$*wJi!_XSUf9#AnNc~u<b9f&?+u$1OB1lqDtu1R*J4_LB
zy3%e<UvD6_(BK?Ol}z!i{B=(DiJ!}PsqSoefhJAn*8CRD8tmkMRII@R@be*lVek2A
zpCZ{4o^c{ED7a92FfjJymbErx>R1m{20tPfA&n-yfd6~@^gnMv#ca@&<>Q;vQ@vA7
z>Oi|Y=}Km9{Z54y>>3w70|L0QT3<s67BKkq`R63UIYeiH^%1U{V|)zKWuSLbo00@&
zQdhqehA3OOwBFcA0o}dySr#HmJ$oy)4E$xdaM&z|9f!bu`J~Iokf`D7sv&@cWfyme
zCx_j%&6v{0eVtWABcTrH!?<G-!^1;OT1f548VO?v&A8S9O!yHMKXux^IHrpSN*&Zq
zDxp{*np_mP4qc{-*%8mUF~|rSUFQ&gpD*5Y4jwN3M|bJ;<@>KbEUwUF=>1?uMdAqx
z?QT2CpeJt+zTmyv!dn*3TlZT~NrQ@E!-SWhqfrkiv4oq)VJFtW+(En(LHhnPVnQ4L
z@+EKe60X5$+ZXG5OJ9DwY0i$IYiQYrv{MQk`s(16D=L<K>H}uDZjy=BvG8(1(ez0u
z9Fo&5T55B}=mo1JfKK~EQkDvN`(^S%)oHBDr6mv?rs2o+dip17Yz+uqy*I)NRWSGM
zgW{!v2H;F4aG<u9#eT%({l%8!U$`6r2iD?WbS<Wp0#_jIh1+9e!yeLNb&NX(_7m3>
zFr{`*`$2*QK?_C{a4Ih_n(#u%t8NrRTkYv}F|^M=mzASz=q}@?_To}qVG(r96dOz-
zjg`<8C*=|SaeEwv@fW29Lahc@H|N~q0b}=#2GbuFHa4GZVbcSX+3q~<=+ojH`8WPL
zkH5aWxj62R9^ql>F`mC^9elC$A20v@u=DudVQ;au*gEWejuUvx_-7FX4*HufTL+&n
zy?o3Mc$2n8!w7NKY>b>8C(gx#jekWgON(NSgWs@ZYaXrsd7k{_<!3ykj=$D{JRhH4
zj;;?^Ui#acR>Eysi=Uxt+NL;Pq{-Id0SC9r|NLA8SnF{7nY?+Bzd|D|j6yyizRa4t
z>J9`i08uOdT!98k(AH|vPJ+_mF(Y*Zki+*84z{<bj{<}}XApeIARy#y#$m@`bh==_
zvwOq0JvnspNpN8w6dul{Vrvn6DA0*J>(V_cZ~|zCa;t~ONr>=gwIVDb+p!`$<d#P2
zkyJHca1p$M4Am#23LVQ@W*%mUI&XpG3d5hPh-E#ZiqhyAeL|h~&;ES)zO($<;^Bvd
zpN+}+y=v(pZ{Dohhjm62i_5fiaf9tcxTm|9Ek#(X^Zvu4>Xx>nV0Zy=O4r<G2iCzs
z;%FW~6B<DmT7<()e}Pm?IZ~Iu1Bzw>(24_$n2Dmnv%>a_cZr-5__!KtKqzQ)y*@35
zEtw!JDQf!QM&&$VLucJ)r)G;RfK>Kd+ZH^dQ2JhDn@f*w*<QZR0q_AG8reb+Q85Y;
zFpe_a2oa|<-KFGx^w4H|FFBT|C4}1@c&8<<ougsjJCN=*>fCv+Rt0>|0K@P52vJuM
z)h!FN=j1g!lM)SrAseW+=MD*V)_)72fd6H+<eYgfxBuLEyz@0!>gf6r7ULn@Q=fso
zU<uly!}F@VQ*MKw5AeUm&mJ$fm*>EEcfNK{>ZA6u=v10wIwHUwR|J^prme+A)`dZG
z(|_cwOkUx%_#;sNhw)o1lKLg01Hwe)2*nFgeEMTqDTcZjT`4&rxN5!FL0kH#KUaDh
zV7s<54dDt;oe(GhhU{I#Di9TQSJ7O_=pXY}f)B67ISGKscXK8tI%v%_(R57vg$XGW
z*N44G`N+xRJOPD1a~}pcD3&R$i!I9HBx;|l)M+JZsnw~oo+Y9D#X)=Qz3R2ouuQPJ
zVpH7pm?wwBgn-+=CSME+lrTu0x18r0@g44czWmwa4u*30f#WeADDC(J%k*L*<R#`a
zMI+)_fW8#*stiLbUv4wcIo2GO(aBgHjvXpxC2xo?R0*tkZpP7ao4%jw_lO@wBbhyH
z1-O_G4;SEbHmZ<qOWzew##qlgomKvHaOzYAW?ktJ%$uwfjF~el%8_WWDUUQwR`!zg
z@~>i(&SsajxjeMbv<2>&#_H)3vcm2uD2f5Bf=Oh{U-+I7abyY3&h)jC7zo5T%Rw9b
zNf<`ylo8!EY9GN^%!VQCy7hBk%49~$!R2C88bhZ8a_Ue!8#wVuEDLoDK%J@*y}G*d
zbZsBGB%XytX}g=>ZT=C9CqYiiA@k*<@XWPW{)~+gFfsmm_3Fn9`-zh+_6w)h?APH2
ze}II&?!%>phphR9{rjc;`|tMeSNboBK7lC@NNoRN2|O?P3^D_szIjcA1|9}4Z-Ogs
zG9CvmC&(nuY^uQNF0tw^u25UPbr!ljW$6#cA}`6Ts#*qEMzm8p3k-BXO|0qD!5Kw=
zMNaSC96yIJ?;y?;TGRk68oR92=JKkwooRaZvcGCKT5e)bjiXr?(D$(YL)FTm5_wqd
z2%`_Z=$Mm{Xq^cIsp+cPMKPAVm&Q@-*e-5$09+^TP*ap680VLIp+Q@YXW@igqz&T|
z=|H6r#+kSBOHxcS<jTYc1<fWpDO;+pI_7)~#LV4bFEH2yzTz?mTdQV}H=^7L0UU8~
zNfM7V9jM1OS#Yu_2*tC_M~@=I9IH`<WsMNi<L8o+O)P5F{9^EH>`Pe;j*oKXVRp*(
zrE^PnA3cMzL7f^iIw@_;JtzMh5k!G>3<8h{_vFuI@$z)0t&J}gBMln61E@+OCoqhN
z-OEM}I*eo%knJ`7sJL`TMP5f5n!y{RJ7H6<BHX4$uP<xWC_o+6;CRqAE-_};YNyx=
z>Q;E48t`T?)cFvQ#}AGL7!9x)MA|gj2#$=>PB2oTrC>IH)mE^Bti*lQs*YpBgw~r4
zsYnb)LyX)j_x&t@Po`P}|JF$|%3%;{`yG-Gtp9$07ymhWy7g>pU-xJeo)bFJ{Tz**
zf~NqdhkLlG1azoc2hJhuN)f$7LMHD0_AgjSG04fP)DD3}*X3rmj9ZLO?PJnZN2?H)
zM>1tXozSG`F`;H}?0lWf4#8RsbMxTw;s-b54C`#`nK(KUATTFBYWy@+O9m6(1AYOc
zVl8FOB)WjNVm6O)V_W3tMVOTkL-(4K(!6F*Zbm<J>F3XIZtW+NlWn##=FwKL*0<q!
zTl!6S%7PX+@TLBeb0y$G7nXndLJOr!*AWGSjP|=KteoqmN3IUrAZ3Rk7B>}^_L-xd
zfD|kx&J<iyveM~BjkB$m8eOM&N#6)`<+CwN#XgPE#TMOHv_L~2_$f$3UvmqHg~U!8
zOJi`{agcRu$2g=Ijya@_+A;jRqBdA5DOjp$Fsvh)jOH14q&Ko)=>Qzl(T6dl)CnW>
zXHZ__7?<0?1~lTI#PTF*V@XQLfwH#R>5tIRK$jc8-178cWX9*GMNhr`>b&?C_Uj!-
z%NeV+2Ge<qKA2Qqq)f$)3&&W(&KKwzi3`z7(0PC_$(qZ>4_8Aei|ZA|Ce4c~<Lxyr
zRb*_z1<#`k;LQ8@zeh)iCmH@Xs|co}*6Ro+r4qsoddUACE&M?Al$RO}>T;D90mO3d
z;|&CFRiYHuxX*HFDsC&vO#A%=JxR|SIYSWHqs<8K5n*?P0EAAR&OM=7NLh8&tI`Ec
zt47$GGkoE=KfZoOz01TcL`1(U!GzL(e|+%prJR72M^}iT(@SJSsJ)0ibAPENSu|^1
zWwP?yARr}8RVgLg@41(d>zA=7$V^6E$hDM?y9cc$gap9vg0nYZoD@-Cc<4S~5f(4o
z&}DOXfaSqs{HNNjG&MutMl4}LQRGCmsst=R>e_?~hEJ|{8T3K<BGz(6E7yw{F9BVS
zikZ>5y?{Duf;xQBXDW2LR4b->x?m+~f>upi*Jd%()^ozGbQqx5MA({g;c!FZ28Du~
zi#6_B>b^O|l*6{T-52ry*n8XVxNhWJ^t*mVOXjR@*&=0oa*{QSEr+HkTN%A@iL#Rn
zY4cEIi|S*(rJIzkvHjog^VADaz{cKeN}k-a?z)`B>b*BmC={v+g+ifFtlZnH03Xw2
z`(K@jesPnY*2)h-r+|doZ16OpN3|!Vs1D3vxzvgXaN?@2>dgnzpRh@Nb#vxLlXq8G
z%9TXYuS?L&l}%pK-hgsN>&l&Rv7<!839o`+Eq+Oor#fP3{;0>Gw$m`EPknzD3Asd5
z>4~EEG12H=wxdrrx^CCo2W46K$p4YsyXePtJG&?n95%zdx%K>ux|pUhGg}w4_S1fK
z<P<NAiNZu$`a?L6dZFDhS{VO;$wXa#wTgCmG?JFv!`Ynw_q3ny+u7{2ExLT18hq3d
z+dli@qun2|_x=MnJ|*&lb9t$tyyNOfD8E(WZtG8>6ZROf=l<Ty&z_EtWVy+iSXmTZ
zSB$a_E|-`kzFcTFNK^H=(W^#BtWwhWgBbsz2-o6YxzWiX?<4o>3TX;Tm<Qvc>>sNE
zq+16k@2Lgcs-l;u6Wd{9_s0vp;r!u(pY@2X;AuT&yIsSz&_{_@ppVA|!4*o6xZe47
z@9g(G&6&NUqG~?#X}}Kn>A@a^nET#Qc0IC$Ab{9S32ossM30?7Da71O=d;5JI8_g6
z>Y6974dkE*Gn1f%1Q)?GCPp?t?DDNRD^<81EtCNj6su1V-5fiKc-=Lov6I~u822nl
zoYoK`Eae!v4!Li+i(orVPo}3~(>$CA)gv%1RHcp_{=~wCkva^&E|f;V7SY;kVb7{i
z_Nw!&<}#g?zN|^wKJ722r9w-hVZlx!LW(cIVhL=;uM0gJDcn_mLSQL^*aka<lYCm8
zAUoTC>`3PL*B>XlqrJ-?KDm46_tP$)qc34WsGTvLjMaiqML?r+Gp90s^g{e}4NMCq
zf!08udkCd4V*os3E_KnlvO@IvLy9nx!fSaY7X@8VN~of&-N@p8*iK}<J`W{l=46xS
z*>m03Ssy+l_DC|CGA{`3-#EO&#Tufs*GB~qP_=jlez_%NZP|M9Y~#rv&|V43TP@!f
zZ8x{FKB&u0h>l6~V}a0%E!^%Ur<cRt+6N)jgy;x0KNbjmz436LN5uYs(J4VW3gz2&
z+s_|8+S#&=#t$M!YAizRU=lw%LjSyOryZ*E*$+nlB55NaS^4-{<1pp0p!_?wQE@(X
z^RHplf!05xb#7s&qnz*8M0xO2VCIFl>Ik-Tqcd=fxJ)Sl=wgTm9KP4=1u<&?^aIv>
zM~hqu3OQ>-OA7!%wNX&St6Qc5&M#FGg?kyWL_WQponLWA^`oYy_fxghAi_5g0l&W9
zZiks}peY5K4n~b1;<UE}mqwJfwzfQNw>8`uR45LY>`YH_=xc8?1+|8sRrW4}u=o8G
z2PknVl>N)|l*|~GhiT(vcCfJ0X>7j6xodwHlJagMqreG**6q_{ERE)`vBO!8AW&qJ
z%$()3F-eDW_=ZIB<!Ji0c%Gl1d+fQOhM_ZeV*rPB>nm)>UcO^b!h5jW!Q%OAa!+pC
zLhNu%BEChtFEr$``T)E3HLg9rtX>1)+ZoP-mr!iCo^S21W{b#9HQh!lhA^1RsHSX|
zX~oID$(dGDPPRUS8%xfw`eNUJfp!iz`Q?g{V1eRADrVqWAQ9C>ClMbXZA6u=>@2{J
z56;9Jhrn1}P@N#vdJ2Ge#~!}H(91bc)Z9V*lLW>xwva`9(D;~!iVr?O{O!Yn2sm2I
z5}8lexkkI@KCM3Zbqu+s2leXk^=!^)DJlHB9J^LD<xR%gVDBIcjrBfVgpCNuiMD}Z
z6gVpKWR03H(CmhCYeQcol2hq7>sr%|x;FMYVKnRJ0U~$VYOEfUf%@A~m>U>W9M$dv
zlb*e8RK^ewH0`6`sOWq$P#cO<MY~Zw@zOy3ZPNjb>dt)wwO?pb73RyVsqXwbP#e@5
z=zI_ihasHPVLPmnUsM_vVk%d}cCo4oQU;`GYNBT9*`foy(b)=Yf)@32TEAT*ZlWaY
zqy&VH0{)WcaL^g?0i^Vz{>8%(mr)*JE7{K2{?xNLSH(hI!(UupsUH+aNlsp`nMm5%
zSC$a9t#SmW#7Yld^-D_&1rOsBo+^<gJsw7UQoSP8b|i-|)9*=R^6sUizU4bs2q4E_
zLLZuJ8ya4bbV}a}nKjE$RT)NIh-^t=^(E>#nk?)~-t1UmCy~1j1mXG@A^(W(<U5jv
z1M@zF5tzFQkZzg(H)#{>W<e`4aV-+i^CWab=uOG%_(>4~(gGl(`w-dSyV6uZ$>sV$
zZSiKOYrt5UCJ5+{$C_1LZZa`m)X5GqXPlOh8Y<p|4{|l&4|ls&p%1tsyu^sag2TY%
zPOgsfOe8>$6qb~`7F26`Xu?9lVPG$&NZkPk-*vpF*oD$fuo%g@ARJEC!oc_p*Bfr?
z=}Z??E2sKnnPxh5noXoZ@!|uZyJh=8@`ko0SSuGB0(3oFr3>M%0dpNo8Jq!zVu`>Y
zgUnu)ZQ=%|D2px2O1-t3&0yqMVdu@kXPa-Phd(W@POnjY-QrEa!2hMEy?(CgVkc&a
zST@1|eri?r?54dU-Dz&10d%`tyssbZu?xX+a@MTT1aI_r@WUuau)Sp1k&6YdW9U?>
z5347$!ztoeu3?5$st?S@)#drgxh0V<*HKYr=VE<;VBF_I!_pN?_JMn>cVBNK|MhjG
zDvS03#L^|5@w<j-Wzl|sNH?;WK18D2YvF~Gec-<4t+QNx53^V<)(7aX(|13g&yTJH
zl#BHNdWcL}^O@e7H;hzKvJc$1=chQ6&ZOIe+>wj*0ot*P5Uv4~i}e8lzrI2C+H2sX
zQhi`{^gf?!z@$=rV4j^{&R)+B#U{Oe;_Q^}18et%!oSVC<ge|lUh<ktH^I&p8?TVl
z8yAgh^7$UUQ50(ef_8<lF;6ScoSSd)8bY=zG=cB$U!Lzsrk@Bgdm{5!WD=f;wi}(l
z&MDlZ{!6|poJy3f+~qbHtVeq`2=rRh8;-0aa63Q7|L3bQwvk{9x&%2HExVG|;XRQ|
zQnJuG$-P~IfJAxfC1zePR`@ON+rc5P!xXYew!`ci#E3jTJIl!NN=XFmabJPPI@i5Y
zLsI^8H7B74>Sbihn!y>~030;BD<#n(<Bt?`>n~D#Oub-4klq$=`Q&=h2?IATNQ71U
zLYxPZ=O-`LGX}C^e+`n!-3j6{JxA_1#z~@^z$|3~aY%5T-noAVK3Dn~QcXB~*`Va&
zPROjBI4qgpm)&UU;R9|)oMSU%uXD+uamjS?4tIH;mYlNl?)3Hf**k2Up!OLCX73Rl
z6m$DWcTsu^?OdFIcczbVH#0`Y|1_ajhvf)1u^iw>wgu1O4^Bf1R{-T$JOQo8H0~le
z<;|0YEihA0oEuj3iV--0h^SJ4yoaL?I`ZUJ-eYmbOqD0Izhg58cW?MtDTz^pbY(fX
zMMulGyiA5#P&<-q+>_7W%y|Dx|92m)b|J7%>V)1>WsU?Ti}`wVq_j7RBR(WY;bbq2
zj<TXhTU*zaLD@C_LM*nrqJ(QHl27PolUPkKx}=6eD|+;mpR$iFQA<MeC8>BvO0l9r
zVmI9RP5J^)IM|Mp@Zk8|k7b;GLJ<e+5G=&AbdV?xXJ@DqQn(Z3L`DzEHDut7hNKD)
zhR5h3rI&i-^tb4qT!T~qX>hbs{$(^t#y^jbNUp1(+dgWTLcn6wZ>Lk31xKE#(ttuc
z-~#rem**OyYHxvNYE(Z<4q$De82?)s9h|C9x+^1TU15P^aP;*1%^k_nayPwscz*ea
zk}%Q07oeFZG5Gv3(pQHmTuRwZJaVo!EH7-wksf9N9LZ26&mO5Ry=Ep5Qr-4y#dKhd
zl(hQ_j*-et2jG-L=}e*6kikrwzt{M<zP|I<$J_f4pTFE8ix~9V3F_(tc=TtmL}&$k
z(7fyto-UGwQuS-1+IPyjVz+<Oc5=^FcGz0zE83h25@L)TlY6sBLlA?1QFl)bO)&_P
zX0C40Bm-r&eV`^^P?n@J{z92^_}%5685ueZqmHeV8pz34(GGEh=RcouwP{$2wzrwP
z{6aRY3X6?Lw5kSSE4X<z@gIq}*!s2Zg|_^a8Y`)UH_+-893IaL8sIXQ;yHjXKd>j|
zo6N%^X+zCI292{wc$zTRTwTZ_*vZD7J05FSDD5#q*bi3!j;y(>`#*dnN4HjK^jv7L
zhG-4ji6;|K8%lkR&O><NF0$>SHWKcB`a3gql~w8g2tGEo4Ld0LE=>C1`&k7V6|6=}
z>ZU9+Z6ui9|K%FnhEhqf9kMMIDN>;#Eh=#ejCaI3iT$VXnK07uR?Q#1fIBbcM-r-c
zE6deX7BwYM(4`cX%Td!Ut=Bf!rFM#JOBb9XU1QpTvI{CocEUeIn-~{6W2zITAVe0W
zJmkNQ_^EvL^Cd=3199;KN3X%su<H2|9Q{m<at~}GzZsHRMK~w~iuI?{d?Z=RQy$I5
z#q<!XX}{TxnNzM7TuZrz05REWv@cW<CFLON2!Be3mcLkikgupt-HS}41L{0VPUs*I
zG>>2OB-EDlA;eA-4VU$ZNuWrb@?_$7MvkAIgF|o}3u;QQYhYxnA<c^;6>EyH(^|$_
z#zl>8(x{lOeFd0t^Fb6#%(`!bn$3u9)h<bexzv!keB0IEd1VKz{HNTaSgk&Q-}7od
z1s5Up==uEJmRvH1s2680z_kbT07qy~xDt%td3|}x)yNjtkZlNg|AmK8nT-!OhNbQq
z>N!x6CUUPISuZvIh<T0xGuf1iszv=@Ud_*3G^D!(<n++nKCuG^&qT|55W=ffX;ymb
zFf%9x3PS{na_FGkw1@TOg<R7On?2oqg}f-jI^#85Mg8AuTrEMC&;pvK)HP@K_BGxa
zJFQDH6jVF3<uZ*R>uhK+V6Rtp>Gbn|>{OUBvxnX@XRa>39U8fYh3SMrVfp<CgSPqr
z6cy`I2H5nIm)spYrG7vH30*#V8kD3CU_ERoJ-2L~t2YR+sU84i1pZHGh&FZQ2E1-K
zY#+ecsa4@_S5{8=QY*nrV)H;45i0xyB<UIaNQA*;P!ooY6`TgwlSrC1!1f7PggrCL
zZ3nuPz(fJp6Jbf4_!^Dv06Z>k=@y3Ar+yIyn_vwo9l9^RLT<)9ZQi=YUT89%WX+pm
zjP60Hfj}CKN*x(>=E<N{uZzL$n~QA>=p#@%1>)NtpRsDT6J?Mc-x_dZu2X5{{3cX-
z&+gLC)#s|B0=aeTfERAy(xH(qa~RfZLImLinL2K$tVR?-SjJeVF|??7W)ZJ-UmJcZ
z;RSl8RGZ}%s3-Ch7*`8a?VyxUU00}1WBgk;4Ne1My?zT|fGNsSLbO|Aw9<ukkIIF!
zgC@|kmtcgm=`psf@B?(6T|CiCS>-Y;rqzqJK|h&_8U;ZMK%`R$2hZ`n##v63g}Ght
z0gnC{N=d!I@Wz20p{oKnVxlQrZX&p=fF-H30s_#9n+H8P*Wfc*OyLBBAH`$<QBPCK
z2>->R365B<;!Kze>sqs>?w=l9Ol~EB63>P^zR1J++knv6BTS$S5+nIMAE+l}0?|*n
zapO}l??629)3FRJZ@A7lXoiHEs5wxJccAQ}z4<<|h>cn|D2tk=c)DP=*MO{XK!EP#
z78js5CDJ@V#^Wh=2f#qvCPWzdD_Bdi{<S88Qi=AploP9(zr=9y3bHmz24m5H!dI+Z
z&w8~u<c$%pP&G)DW91`^k2x6=CFE42jTA-(Yi_GW6Ha={94yS9Jie`$1|(4G5|J`d
zEYMZGAjR1I8mRXbs`YThn_GUoj~%B0xFTj`tr9eDW+@nqL<fVRfk=93p}zr~8ClJ;
z%tnM_Hvz0ROz4nl40VO?eRFdDYJL6k`~@>47_sk-K1aK9&J*eDzg+q4zT6rSC6flS
zXmM=b$SL-u*W5|^0lq+Cf1Pi9uHK3}18Ga7XEy36_MSPV-u`&uk_26aeVytcPo8KX
z4ZSl>TZ$0WZ)#H8P!;-~leTFGB~ZO<axHC^B@Q*p^>h%ON((KmtumRgB3L`dHM-(f
zH@0t2o0#|X0~+*6LhgHx)?fwBQ)x?02$oa@H6BZIPoU1nvZtY!hzz3$R-VsZ;mGOY
z8Z>oQ!ijLiQS|Tt)c?|biS73FXCvX}eG+9%vDUxZis=%<X23kTwwW|dl-I*9k;?F)
z71a4DZp+054p`y%KkjptQ~z-umea@7A}G&<u%pjK#mh7Daj1?*l3`>rUv&=36LVh&
zwECyKUJRY6X7LZ&n;5o|eHWnG;v5u4**Y41`_2C2Z=OAWv9&KL=rMJDCRdB$KAQaL
zn<vk|#+T34qK-QQriGv4@VtHegXKFhvuJy^!`W5t&(8P>>Naf;NZOlYA4!=4KTsEB
zXid@i_PDDFu3_w0CG#|A-ROn?=*NXz!-<R*a_!48X3*c{sQLx2;-EuMZOL?ko}DES
zf05XmHPzm5jWI-IG6C2+S@p}*v4QEzf+Exa(H&g2%PzR}S5$g5KR=}jk#&dT*IV2y
z4O0f{WHC*acEyt2Z2Vm{LmSfV4_IhOBs*x-NK<Omd?q{Nn7?y%m~ClSeZ2eQ{U3Oj
z;CM}L6tq%MC*FNsNqTiS`?*b?5_AnQn2aqu5($8wz)P2rVPAyV=IQ?I2{IV+x*Y!9
z#qBlx_p7@f$wtB6XM3OHe|w4Ttb5C1``#*1223pw1_-&IM0Q9dm~7GXo#a&lPHysw
zTXXv6CCjbx-)9#%=Ja5+N5Jdre|x-*KnD0x;>(vCFZAWw@!#9;U;X{_&Ib--7XvyX
zk^5!wNZij66Es@E|K@MjYz6-7zkRMjC!k9`ADiRMO_>{dg~;k^QW3M{T=sYDlAaSQ
z^sc$*D7IH;V1M4%<sLimsOEST&l=rg)ad3$&{DGZ1-a0rvA^?;3b~InwW~>(bU4@7
z#jP8txoefu%}@#qE#}s-f>^_AP0=DjU--<O58(j%Q5<`mO~bZ~|2MU%8wZQr7UY}c
zAjJqrxif_FF;FUwhuP-5g-I(uKf^WHJ(bl*B=A{YgVbJKAtOD7I|Y4YK$sD11Pl08
zWE<fWB}GNwjJlj<;E4q!L=7ICaq+~lz{-@ucHQVKwXL;@EAr0H;SMy|+|}ye&AK(}
zu%GBV#c+A_gxidTJG>*=D}D1QPj%THJVI12^SJKy-8+rgF|kN2Cj?sXZM-|1>DuW-
z2p^Ktm^hEde_D)3uQB5CYaK+mL(^ZR3_ond@k_(Mo?;v{`kEp;MCz50nL_ra_0k0J
zhPrxnj;XcGu!}gG{uM(tMCSt*?sUcdZFXU^K^274iLBuqqEkq&GyfxJkk9^sI=LrE
z(#Ktxv>SMr^}wwa%yXt<a`)+?M%Ww`7r(DQIKQ9^%r$T>xW2ymVr%2&7RrlhDRr=l
zq8Ch9`F-^t7{WZ<3V|MN{vuFSjxbil@q9U_-<jWpOSkcQa(4}pyn2JXynKYOvVDXT
z!+M=F9bo8VisPlM1q00Khs8dvFxegl47TZA)fRF1KZQRFrDDu-29HYFV)k;xN--x~
z8the6v5-_sAD$pUSfi|yK`SDz9vqid_5u;$XP2)he}Kl2ma#Xt1GRm%*pXmv1dF3D
z%jgYunWEJ(ME)ohAq#a~1^6t0WGx$Sk-v*`tZlP+KzuUg+hwz65>?=Cr19U98qr%=
zZ)B-L!%`J-{*(CbbfW0pUjCW(^eR1-OYC*8j^jkczs9zh?IuPAX0K99=&b2&osnS}
z=;4*AT&l|5yl8(~J_@m8b=`sVCFxL*ye@K^Xyk159=i}JQNbW2_#-s61qCmH${z>|
z1~jwS41%&Ul5izLO798CRyc!ztV{<`dar;ay~`>xSX={4)4;%qem)<e^z`KXf(Kvj
z);$inFnH}GB#O7a!zF{nT3<hY^7!R{F~!!`PqrZUpSrsye|xcs_y4Q={_Cxm8~E~B
zqsqpU?Qi8<quRr*?@{Y>)ar1L<;lUhHIz^mxy7X45JMq_I%+aziZzJ;Vzc@E#U{SJ
znqD3>zY>3ZM?lY;b^+-HO87|B&7&u2O7_02%#PTxwDgx)yYdZ;ScC^Y!%C%m4=&I%
zg-qfUAYjA#Y(dYV9jr~X;Xuk?Z9Nqm1e~x!JZWwna11;1iuvLRQz1@q{>(*%RK^6S
z`f<jKj#uH)Yinfac=jruj|z|nYHcTqH0oxA2OzLrJ&h(P2mUb(5^h|J)2z^u%l3ix
z5XH7rP8-=><Z@F?5Rw9mGCqqj@KPqu$z3Mi(A_#vr2+3n%5w7_@NAY=M?I3SV&_Hh
z=${pF;_7GuE(}91!zPT`J8r1M3dtyVGlR89R&Lio2okzw+&alW)6~GLNBuhiiwUA6
z&L3pv4UfzwZ^M<}lJ>z)cyfL`JH+-G4xC|@o{Tp}7XE`(5pFx@LYEhxAK_MQuH|r|
z30?-rq__b+zosDMr@lUeHC`l04Qed~p?eIyL-ktR9iOBwlL^u~H!qxm7;!y=ADEy9
z%T8OV{uX8nw;388t_TuCrMy-qSbrM;^BtlBucke^1K;(_8Ul$uoi1K}uUYlnKC9yR
zks}SR%>FoHp$ru|=|IM`l?Q*uo@|6{fT9Ls9nk|);`wXEL-^b|a>l$c8f@7C<7Lpg
z{p<ou;+P6kK!`v;HQUEjvl5J?7AW=H0m`a6PKJ7|7E%cxmVC%HeJTeBlT@RbM{s^=
zQ(k%ndh`t54UKid($6%9$H7{Pv=PExZs)5)ljh!$ka-P~_EP}kH^(%GuyKpMP!3?J
z;`$Jlpy`Yb!Zon0llYCQwIV>DksSXNPrqflmmOvL=P~|=oX_s56pagXv=afWcrX7&
z75;U|C$PEU4!pim$MQ^5NN&$(xV!lBT@-mWeGufOWniK0h9<lwCc`Fl*dr&<!~)Q&
zWo-!o)Mz355{!(Su9CzUQXdQ#OCd97+Pd7<=<&F$sfwmK49pj-u!J!B{;%-^Qe!f8
z2D+gZD@H0*!<;So6N;Aynkt}p7nVZB`!|;8pm?dPrflVuZQ3fV2crW&Y>)32utd7m
z0*cASvxeB5VvUA5C8M!oKV31^(8$=mnM$g<O<dAU!lV{?w#HMyCyo(%R}>q-tbk2$
zrI<yi4yB2sr$;e*8Dj+eGONIh6s~c24SG4WE<#n^CN3!<hauy97_jkIq?G?2z$ZcM
zB2?FIV&s`G3?Y+&fRDW)I!Qp<<|nwxp>>g|>NatSDMoN_fifWM$$(9GON_$f0F5Ht
zE@llFP3%(Ak`)*JfS11_QvQ1YqXesqPF=f+%Z{ZZaUGzK0iUdF-1{?tjYI1qV_^E~
z7Rf)dgic<apCcR2*<D{ZB<0ozh=(U!0Zvv@@zA04P8iY4I=n8Cz>H8J16fCK?FNho
zWUzqeTW0yx!rjqV9mGfLHXYt@<HCIbvlQ*$H?P%mVVv+}25l&#q`bYEK01V~NM5u7
zF6p*`GFjgY6e<L6pj;LMFVyTn(kfo6fqUZNzA^=`_7&>k8nW3<;wT?ponA<mjKV=|
zM$TTyT3&gDy@dUnSfHm^A1n0~ZyhZ&La24NF`pm2Q`~7~Rcp5+bn2Rw8)lqr{lLd(
z>8`}j$0ufsxP<?i|02Ks{D41XJ$Aw*4(DfT7Hj$Vesv*>2o}dgBFW778!kZbhCb4R
zls<}^76q`09P=$H;P`U&YJQ?fNKz2)FJZ@(pk|t!Y?Zq=C;?jzm;36mG10nQS;~vu
zF?~O{^&6`HBAl_-wS}(6Qc)b#WYYP*(!S)%u&u&=%7b#$ts82oODCeC4zawC#XjJo
zZgG>X;)r7xT&N-XHLjjzIjp0kD>_+yD@+?BoRO0;?hqtaO9aW4I1)W|Bvn?U%G?us
z_EiiEHuEH$YKqQ!uoSnH0kGLCtEz3;0|^9cs6!$iC+`sG2rtLI<jGZ1PrX8uG%Q~d
z#)}V-g_R1T7z=w<99l>A$u2orugp(5Oiz|?QPbPG2`z1>-=g#3PY$DFffA&u#n*Ik
z@6;ErPCntFjuE_K7EjzO<L)#$nrea0xtx)QUS}%2EccqHe=TL{XYpO%==$L<^7Rw>
z)VBDoHQP<{Xp1mOgSfD?>DSJ0RmW_G)ZFabPYY~I=tW}i2LN*oRj&B7M4U^(lyi?b
z8W2+@618MvuEh#EW9?l*7e(;`w`!&IiDA?(XDx+Ult%6x36+~XL;%A6i$|M({_~&z
z0=0-{Tn(Gu#(4=BxLk1sYS?18ol_D1A<9(UQ9z?3v2O|$WUJ}S7Q~z`Y^}$iwy@#P
zu#(`TTk`x3P~vdl@3?a1v#}h064UqXo!?U^>OD25cY2l=(aBYuij79h1$N1ZT6lXT
z91{R#GrbT)H^>XFFdpby;iIqC)8<UYH*s)U28#^L>FjKFdUcw7V}%HUlb=@KIaf<Z
z>WYCU?bE6|({NcS!PYg%frfHAU2ymlg07^q-Hn91qX)PU5b|V-HaYxP&CTDXSQm(}
zy}Lii((<Ac{17BzTMdz9K_(TIjw7jn*aA|-1U6q*Ov3xre7dzRx)1Nz6H+o<T4*Jd
z9lW#%ph_#o0|#C~OS!@%t41w~k<rT<iZNW#N*p(7j;=_(UtBJB;1<<w;-Hq8T@1mu
zdfw7TM#`#xS*f2cN|Y})lr>#%Wa<OI!p0xE`X~(sD#@Q(-A^!yfb{qa4oWqPv|xCx
zW?SdA#0|*Z44qFwCw0DEpzD|K^6a4n6t~*?FE+0gJE&L1=U%wUA$3yjq{RxGH%4G=
zHKsN?Y{(IESQv2ym*)(^s`0UE!3&@b<2?YU-Ee(LM$(16B2^XAm$^_T844nxm>JfH
zYJw67#}eT&6=~qx>4V4s$F^Ecfx#bU4_9>R`?Qw74#3!vby1_b5@n#~OWP>VZkvL9
z12r?lG}?`ngr4MIqw8I}1!z#Oh@f-B#*nE=O#Kt~o%>>KbA}vB7h{KE*J2BLdNw+p
zEr8?e+40pJo6*Q{i%&<0HfH9-LK;TLNUL+UM#dyxF?Rww<k~`6t*+BYwZZ*A9V+Cf
zP5Y>U+2U*EVJ5OoB{R7_Y3(q{chu$8ogNekdmkNv#Rr+>bnpxk&VnqDcLVJv#P`U)
z%W=+(2~Bp5THa#n*<67^Aga5o4}b)c9UtR3zTA#C8NYlpl|91!{bw6bw|2HSHn;Zo
zxy83vRkdsu*ty0XyvSpO9n_<Hl1YG8r^F0)r;8c}y&y>6KsMak|88gN#s0%B+)cU(
ztN0;jNBrqDQdji*3!A-)8*L@!lKzH3-e@M=E8$!}0-5-!&6^OrY%YO-{RnG=xBZQ7
zM)(FgVvAlw{M=k0LOu|U2L9nRetLCsIm-=aF0Ve^gz#0>_u}#z#?o9~S9Ibkm)Bpz
zosd1F^A8_^4NGm42jM2P-q93yhq?Bs3JR{6{PGR%(jn(#Lc==K_dl2jZZx|b(jpby
zM$1m?ZkPV=QYSYdu(?hHeN~%QWW8oG3DuY8<ZjDO-w2K25{w>RjQ%hNMrzjHy@{Tk
zgBT~(Q-tXCeBv^G2?(v#eh);WU42_@Ym}_$wxs^a`mcBmhRO4&Y=S%yA{A^cZ_+da
zy%CmWZ`G@>jT9jvOoaPfqyErJZ3)Lkm1?j8FD}}du?XuATRsJFc&S-XfC?>VCP$O>
zM)Wo0dN`R_z@&w0wsnCu*Em`$EpN)kaDF9p1(tB#8GfIbKjBa;Y*@}JO<3tAlgCU<
z$!s4&fsjYT`eBYN1>u`3;$2wA@E`l-(H9m{M8hA-kb5++!;m=}RzK`#=Jwmitz*bF
zz%+qe5Wtk>|B*sM&<YVFmnHsN3K#1u4dk*rFq#c7gX*YR`*z&6jX|*&q;407qDxuy
zWp`|^io)ocpPctQZ#nL|maz)k!bKsQyOSq@1O<ePoWgp9Rfjqk)=e?6fWaj*2$@Sb
zU^;W3F>V&VYPfaqE5o7x8ZBZ^KCby4lz&|x!6*A_elUN>0hIBp;e;J8C0fz4eB~}#
z!ctdTx<w`(Zb604#x7><Wp(WQtN8z7b6aw~VQR|Dk*uv&xOJ+oT4PLie0f8;z=!hX
z-WcoT+1UubzR}{1oJ(co_L+v!A!c)pt~riV!d@E(OWT53h?PBw1&$JIPuSc`K}DfV
zQDWDfbSGGMA+H(=p;bB12hA0U__g1rRAJIb#ny(hQndj4nDN2Iesu0?8$@4&x^yl_
zuHbDi;5(*wHHNknfLutn0k0_&_WqXEnSQG9)}3urnmBg67-_>5C#0my9LZnz?vNGs
z?yQg2*4FNg_U-`1@RX3Tcfa4;JA*Gq`GhB#edJA9+-hj$-LVq@hZ76MgY?WEnESF|
zNUxB{iYYBJ&W<?ExObWm$W@fc%^G>*v{5hh732M{Eg{ZLq|MK+9#YhN=d{sGu1l_K
zs|(#^72vIj%=chTcixh??ULubh!W$bPPdAW2yBRp?yTLqb5l+tujyX(2eFn>s>N6j
zfpy7NP86eC#Ud<Uu#^{XG89W^cB-Z-uNhg2F=;L>jqetp2F@!uoy`>6_B}26w!ehF
z)y??F+y=R-2U6zG&ZI|VI-Ko&$@Pq{(*$BVDTYnmaG$<J)po=My|i70d!8?m2mA0f
zuD62mQEl5ts{U&GUB9Q?Y+^QZ{j<QUq-QWqYU~fxTzskC)9?M548Ce(!RI^vA*)8k
z_LgOtYs>q>cjV<#pu8>S1DU_G(}}Q&pVD`koz?D<2^y#LVJ5Q@N4#+twocJjOeorE
z6A=hvTXjNvDnjV6XHrqOdq6@LQ<(@_S~i7;HD-2?70YTrl<u3vy@-UAHGn?MF=$=5
zR;dqm5o6YXUEF$^h)~r^z;93K)6Pg}l!eYQ{0&ndZs=oVxJ9Cp$@u<{dq<yqvU2|+
zcQ}05tPZrz_}{{+#R4|g>w@BHdB+tec97YCWYl?x-!IW1B5Uzqz3Y9?U;d3pPR8H8
z+{DSBC|4No7yu#wIm#)oN;+O>iB!_;SNFRU-`Ak*vobI+6RlCXm%y#yqwbVmhaV+c
zH>;<{ToPhHgsZ}#sFj9p_P~|CWlH|uqaXB)2qG0=^R3`WJ9l(pzccGJGX+aPRO(iO
zwyQ-c`05uxJb0?Yeh#6!i+lpQNn-YMV;AOsdy)S=(nfWa58oX!l}zVvvC<d6CwBzS
z4o<lBa&H8O8bXpr2XD`Dq)8kChyjOhv}c2*T665!ahYy!vqq}<vC!<UqOq8j*mD=W
zvXH3KG;2UK9_HQP%&TRc#5r~QDGn)(f8BT}(P=rwXs0fR(!FhBiRTeU49n=wNufZe
z$-nyTwyDWby0=M>$!bWu2ex<7$d#gjmEOYSu@Dm<r}WA8gUQ1+!7*j<&RH;SyYKmX
zr|TPbU2W=lE}J@DnlmYUtkG`|icioy)GP#dRnXEY2opP3L$cV)!CwhtEKdy0LRmTB
z3EgK5;p63v%urx%CfG$^BXz>z8*vFcGpEIhw-?o3%QGA)XY)?&XO7{36*^D@oC$4*
zWHSDBvbXk^U;pLzyPxpBtMS?|oD-`$GT_O5X2NMrOJ7Vo|09}BWK~yPaK$QzK|?qt
zvFP>86gkAKJ1f2PLgt+wt}eXip%2z|d8eN9jxRWv#FlIc^wb5gS-yA<>I+XNG@vEa
zM^jU&+(D~mdsT|mHYx-Y>WrZ~WC;;Rh<VU({*|AX@{ZAzN9s8T3r?g1SFzV<hk!X=
z;IIG=z>ATKr{FZB#}Z2Z-sl*)>`KR(BPB<JCp>ce_5cwEl8RV{!U|Z@PZqnIUDFLq
zoC2AE!XA?mIND4vvjA3pUMZN%A&9mZCo#dMNIyy=?*o7a^n`J63Usjxgi{ok5>WDX
z(1_$9I4a-{{fZwo3LQSG8*;8idx2x?&5KGa726oTGcn<Llhv`rA^HJvv=+VXSh<am
zPXC`o@sBlG5c6jY6kq)(T#MRp!I)2YcMPvKwB@R3BPeAnTYG?ei<9k@2Gad3O^$j`
zO_4KY1=ZJL^65C}3qsbkppBl!g2?laPNFoLyquj~A;S))38bV&80L%%hohGwJ>HH_
z6&0tZBQRQS;zTW*;%x?KO<iye#F}mPVMuinnbc`!m~n=e;SGEoN!IE#K{{h&3H*BY
z%dluZyFw=T*`dxbQr>IZDBQgT1IyF*y)VQ_L+~RZqTe$_DV2~RZrVX@NW1?jyc6~4
zy4W?z&GF~|ZgQsXiEO}P!HGZJRkVu&ZQ9MtLooF8>}WCCORr?NcJ`J_)D?;1l3W4y
z;^%|;8MJhK1)k&`?X{AK+qI@46j}<Zc39~JA<3ks6So%UfIAvy6?sK|^50k;Yb^rd
z2Sp1S5DT#K-iSL!i<EX}hO56}d+|)&`e^U$&YIsUc?S=MonM>jeMk)>WLhM&?mv{|
zL!i=#`c%vy)Qu`z^Er&+MkOafyV`%)urw-W0=+jA6`=zE4kuT2;jVHRKPcZt4x;rX
z5fpW2BY$YM6feN^kE|7VL0@oO9Uzq={%WxNtar)zOeFW#MGKjTGT|h}&`fgz=5|WH
z+D`X=sehQ=(vu)cA+--xyP=|kTVI#ymXBHo9I#I%&WIF42^xXTBs+odg8?SrO0-(g
z6O1~-R%x^PK#8Cj3U|K0_P`cviV=JwqmqUR!)$-eK`@~=0~kHOL3jft>b>+`!K$Y0
zqBFcFMn)$L-OsCvY>LrRy>!T}#a2}m2kR}#Sjcqw2=}+)qRy(QLp7`4BX+*QX)SnY
zmMgq15eZe<CJycUyCo;^_Wb-1ZEpQCJ;aG&wCsO6grV2U66rYp1`kE+JLL{Pc7vN1
z&T#I>K&7|Jq6$B~G<zX+yCdUKb>yyDe{cv#o05YTus7Wed#{Y`@dzGBu*M4+)vehS
zis8>%j7p_7r>Lnthp_oPWLQwMq`un}rM>1C@4tO0d(HyY8yp3ry<ej0t5NkGS7Rr4
zH#cHdbkfps9TH!~a8P;!JJ1tw_%j_Cq^WD*+m+iyH?X4`^~fbu9PjVX5d*isFLCWm
zqjGND6~~CB`=R2j!5Xd17o)~Dbnx`-?PR?9{OR^{286$O{`}?s_ggP^9zTC(R|{D^
zw0%u!*cfo4p3eM;4zyHqaTpLz`5d=`;>JRV-lrs6Z}O-K#W1pH9kj!tSG1FH@}Gnp
zqJbhVWf4VjtZA%-PX1kHmJE6P+1jUTpYkfTn#d^^q%mk{mz^^Zx@s_zgjp*GA)v6#
z0u#$*z*P}4NZ%TBn>Rh=om*KoDgJW4Gc~UppC1ZN9Bx=kADoz`WoIrdiUa#SW-G?o
z(-|LQLQ)o$CBCI6H(t9DIlm!ujVIevs0ICnlA%6r5{XStj8l9B*Z2pf^u{IEKEm*G
z;hr4M2ROR37|CdxU!=90&^XhnB3d*Dpik=qqt?DoJ?raf>6wn4I<Mn!x7LjZN6mg#
z4`N80OEF02OCOya9OrBq{FJz+c%4@4%6FcOnSM&k-7J2FyRu-Xi;PhVq_|pQ(KRSe
zHhwl=F-t$jN&!rB-6?oIiY|45Y2G)-ScQFa2N;gLqcq>PyA+vk^?`w%<gN(gNcC%v
z!fz&5*W_;FP!Y|qFdnAO#P@Ni(aAhNE=@r|;FrKAksNyOj8|DTF@y|fhEP@m$b#iS
zwGPXPk}O!DZ0Ut3y9jvZ;@}_A1!2&QZaUrhbfhC6{GhgS+?XH3(QtN|*39_`9Vf%I
zu-wCy6nkT)b;V98_Sq!cHMY+r0Vh1A_$AkN@OYD6(GB0__P5*nThE`2)^3d-V%P26
z`PJy)@DK^U&lqX<PPgswc0S@fnc6yd&4a^;J>-gBcJNHIeI=8!x!!#0B>^ZyNfqjr
zg^nbWO64J~%hY37j-5(o2{n#IF$N8xLo#!YoahH+%r?=haiBkCIwvMbZZ8S6uv_Xw
zS3{7lhLQGi`fpzon?kUxxKe4S9ISe1f^2@8Nx#01ZPEjrB$6+3L=*h1=77RTk7gEv
zWJzmbL<pBV$NVJ{t#z%iTv3x7)ii>_&Yi(br6k>8#6Tpz_)Kw1r!|dlq@!+>r-Z$g
z60k*{u{M%O*b8SMn9H$qmhN5q_umZpktrY5Z4Pl&j6qDX7r*{E+5Pd}-r}PlKDmpV
zw?5kaafDxAjXnXGPxKR>SGXAkA~+;GNbH8rVTO-ftLr9-^-~Oy=(hw6V#Y-FHkC>d
z8$^4hlU)({EzRL7xdIJBk;WB<miZhpOZqlOoC_%m!?Cv!0Em6njQFC`+?8k@9vs9<
zl>5fAlyK6uG62L+<E6lD-5E0HT;+UNRi_meYVURmz?8i)6Hy&tP`TAr%%=3DpU-!A
z4<_0bHnU4Hf5r8Id%x}d_7(mmFTtaWi0aHxl#z^v!yqf4&!H`>F%elDly0^#`Ra??
zKmK;G_-*&g2Vact!6E)JU#;HVy*K({@)dsY#qE16vT|1#x_})KNNo@UJX9!z?(1FB
zMG>+E9cU&apoFcfat8wd(MS2YEu*_X{@V{9fAw!TE#ltO@o%7_Fx5Un5-<^mLB7kw
zBP+kpY;H2*l`_k@3YfJ-6I|rLcq>?X0Q2oGVU4<^GS_dMgc3}#Pj)3wy~ZV^7T(2A
z*v-rzNEQ$&?b+lnMU{Xp{6DIr&;?&|VN_(Hp}MOA8Tc;`hfWvynRwh9VQ4e!0={rM
z<`f=s#1_7()PIBG<mj$_BI<Lz@>w|`V0mIwYGYBxs99@V-Nz#hdDjKEdh(FVT@gT|
z8XQ@j?4;bw0cFcX@}4}9X-`d#QD!eZ3iks?8QYJzIJDMb1t+{0D^Maa`mVrBEC=;?
zFtGY`c(A+PXG#tOb3<OrOgP9(Uwt&$Jy?DH>FQsGrgFw_*i|O~1KgEqW|{z9S9!%G
zZyMA`E0@&|f0*;S^5kx-w_?Dh-OZA)nI@+mU?!CAZ&s_LDx1v3cfEHosdM;ySRUBn
z!Y1KHb_Ne04i()s+WdBF^RN3GPo99xfA*_-^yPQYzJC7Qvxobe&!6qQe6jKP*~=Xi
zO4l-LA$)nf`LZg*3!2@AO<PClMPbYFDC!NiJ%Kcw%b^nHso9}~v4@wh>9-&bRYsfW
z!mhSee2WW{VFrora)!9hhrJ^0p)hute|1kZ{aJD@HW8p@%A%4vv6*lzc$t+QB!j0K
zrv+94N6cngz>!dFRCQJT7ayA`E2Sj5dxH(4+&kSE1@;0hpy?Z*jZfMld+ejAlWyHV
z>T#!)vh`H23f3&9a4PL?tp4p__3zYM`RD%H>JJ~^&(HNIF)nd61SpgI$7c6iD?1Vk
zHSuOR^enBmfk#kk741Rr(8_|x?vpjB@x*npWa$Gz%>0VZ<m9kWawCY+U2IEVCLsWH
zb|x_l2+%@~!7}wu8TteB564i~>bp3hb#M^@;`0VDBOThKB8QENQCa=r6Gnnn#httw
z$RVdzyM;u3O)T|p_cM&TODn=No!tWV(IdR?{wr-V8<1+4Lx@}h5D!Y_{D|F#0wyW+
zmqggR`1RTE_|KEyM{D~=2;p<vjwwkTVs*jw>Mu!=fkQ>BKAC_xxMDOuSsTIMs6*1n
z`k^=1!)UNV8e68#1hLqYw#3-+4GOu53+AoXxr{fwNy_C+YYm>$xRLA--V%bN&E8CE
ztb<e>zN{$u&sfyMrp177)5?peLoTlpOU6j0WW@+(6qkm+C7ER`a$LE62wFBO6$7>g
zB>_kX7G00wqOCL51`kE$&@6Vgn##lgBMq*co(7L#Y}2wW_3eP>OAtJ8^7IrpYC?Ov
zULMbpx__)5c!kPcI`658nSR){3n5rQmz3DPO!Cz#vm{wapWL>PsZhfAMyFReQp$7b
z7Gk?3T}q6ViBv6!f;N=1VzOSDIzZG6+|Y8yvr#hi&8KOdNwn-D4f>j&pJ<9m7cvAS
zbogxa=JN95KL4{AefQ$YvOSNTV#gcZR#vb+YiIWnl?>Op97;fjW#JY0=^j0P@)8@*
z-)}s5Ea`c-o^CvTqFzaro@*!|fCjs7O_hD;6pZZyt&Pj%5YoF?y8i|dqM@C)!*V^+
zZgRvyT~gqvkVw>FxI(01h=Spua`RIh1Io>dRLad10K^#E_LO?42BiDOkDq5h(R1a8
z?eR+qh99%}A-^1*A4-)uh_3npI;>I*B{dTgMoq4Dqly?V9Rw=sPy?F8sb-@fE7vH8
z8oe;7G|h=z-X%c?Ac{&~m%PS}Wyia#On<#K+C}pHWvB~Fcr)sb-u{wlde!G}E%UHW
zgz?O2rZq~n;-o=SH4`+<^w1f>ts)EFt>3@@g&Z54Ke+$J;T!C5PEQ^T>DU^iQm3A6
zml8G+t`Cr^?iAHWV>J+haJ7jVOq62o9i*(lE(yEre3fbG7R;bRO~v~aYA#}ey>fN-
z>ijBhD`2@yufS&VCQZg)Le;tX-?zq;b0DV&IQ%--L(J+R&R}}rT67ZFBzwI@BkN|u
zbs-qa9^BQB5gv}Lt-sX%O@SVwp<3E3m8fP217}NvTuRJe+g`Morno;OW^m<GPkw~`
zwfk5hsc+pt^coAPMJRaipcOA{KukM0zh^N{^rkP8JxD(eLiu@m>k|89_^p@EX2*r#
z0;54E>o^;s4ywyRvLGE{>RAFc@ey=qpoq^w3TDa;_82j#L74T{Upcrb9u`u(ItL~<
z^=7E-g{=a~G@Ge^^b)ChD7=NM4`lBmm&5*6fCie~z=4u*?;l5NJRY@UVq?vu{}i?z
zqH8DIRJ$2@fO1;-VAqIry5-Inb>rdMP~6tKzTWGJl>;yyU4>n4=A(AG<JlN_$aGqL
zAX{4uAnK?X0d+e_#x|TRm?BQI_j!GN`^m=3N6%k8-A6Lvzixbk`!2S=+4@f?B9|69
zUt?od=<9@5m&}GrgWVRQ#PiUC#Bm+Ulkw9d%1B}~&~b=_T|f186!f!OC1c9rxzFwj
z!J&zT)U|G1ePDdEiTnTvZ`2LBlW86f(5(hL{^Cm{7z|>)ED@)_Aguj;Nl4y9Le`|K
zUZ0=f0(T_Vf-wgkJ#@HGQ=*YL<MDrDV7Y`AdL4(ok%}6(2o0M4T4YeMCqSCh!L=!~
zIt0%h2_~97tn1_FK+vEk9d@7o?o*Nq8J@@E(a&$*6_i6t2kW^oW^)1Mq^Xo;FGby}
zX8WEYesh4h-NH&{qY~W+)ij>tMEq`E2%b2O8?Y3Z4SghOr|c3IZ>P@qhr%}FE$rv7
z;jvE94OrzB;SIfAVc<w}5E^ZYRXoVl@Gs~cZUcl*Wt>EaV$D%7<BP?7^;51=`QxAX
zLv~(H4=(td=bdEjI{ihid0TxwKR;bfC4gZv+Ku);M!8urfas0am?r#y#=$sd89DPa
zxhzF1k|;()Cl06sZjU2sE;CftM9)}w6Z8ra7ft#=PL)U1vPv&@B9H)uwR6&GF|nW_
zwCA`oE?<C^KhgFWk7YM|5AmULD%LL1PK9vF_Ot8gu_g#LsgmFbY0{Hfe>n0v`gwNw
z20p>`37QsD!~xFx+zS&NY>ZT5ZM=uSuJ-M987rgWx-li3EpeMppRg6<Oe=2p$OgF#
zx=fplk+sNIE8b0<{5rzbC<qotR&VVbrxPDnBNcQGx2EqQ829x2?Q|4CUY?JxE{^Be
zD+QR5VvNNHc5Rx%>wtNins<+uGHb~HS94O>*C=vQL7u9Tx+-a8ra)0ZCQr5&DYjDu
zbY7K*;Iq)BJAlgduEcb*AB~C)^(w{JVp$AFax$A^p32f}4``kTBh1VU5>JLwDMQ5$
zOAk6CUh%-V>XvXAo*li}8GGwpVc;koi~ezPjxEayi`=E`Gg4aUYCqsZQ&H$baYMMb
z96jjKKRun^E!w40DmE+7-Tw>o0{jnARkk!_6~>4&oH#pu1<5~|{WKkYzV_)qgxHil
z&nU2sn4kae<RkdUj~Uc8RW+9MvN3_-VyXsJ43c_GG`i^w=!vG*U0O_9%LAQDs3r1|
z4m;7YtsasT>);DGt&+T4t|;(AjUEUjRK71gv#&s7Zu}&6RJW{pa<Gu&GVwj#U^p`g
z0r!|PTW3fMb^H*fx&K)1%zr2J2vxV1$ueNUP$}|Xh7B9!y1|dTyUuZK^Ci4X2o5}+
z{(^t8P{R&7&NZViRl0rcse`c9M1+y*ip^sru%KkKhofTv5v<V3((u}N<@YNXuExP;
zj3(G-O~CO*7YRqmF)#dNvd-%IYg6H{8W8^lV^^ekjWEc4Cwmzx!FIuEEotBV1i6s2
zltyXiVom<f=x`?JP|03+EO82G>x|*h0MvumvNZx8gZosWTn}A}!Dyv}3aOx?xWLJ0
z4EO`QUV|I{p=66_VKv&zLQFNT=K=a~&;<e}e02g`%MXsdFrgby@O&+B%u6inmioJ<
zCrb~%`Mn*7A<<oO^u;spO0!1<m2^=k904Ih82nanM<3z;X9oYa-ICdd#UC>6a={*m
zM{oS-;Bv&|o0ss(PvKgWi*5eljFA^9iPc=c+x#@dE4NL4T+I)#SzF6v9$nB+p$aUl
zvE%w6VW!?Vk1`Sk?!yqbD(r)8r<>U^v1?GT)k-w|J^aYU)B|ZoyeI1H5<ZiB&x5wM
zG*hM!B=5RSd8L&vZU~ueMB4@jd21e26hhC+DJBbD>|1x=Qov-#N<Hfh>!-eciO1IT
zlyq6Gn?74Z;)fjJo3P%{1lQ(0nFfr3L-5j&z+34)@;3c4J@7J}^#YNl;+C=`yq`4?
zx+#O5PH!#_-$364k0H^aZ%`tsu#JoGYvD<$7N)*>2Ev6Ta{NwKY5U0f;Bu!{lX-$z
z2uUNPQN}@FvMs*6IfoNp?Ko?oG7M+Qn^=MZ$5>Y^H6(?i1ENCuMwy3TP!uVLn64>g
z$11+rT`AM~Uc5EssVBxM%H|EPkTs}Xw4Jf151Q!Y3?0iXWUJ)vkX*58%b>l7xF#NW
zN@4xoQ&7(?XhHp56%_2g!C_0BGQ&T7tV%q7zW4ZfN-Cuvccv#tLdGZ1uY23bVF9OI
zTAYFC`lkm5o}3^4G@XMb)A@}Pym<sSb~uPJPidGE6;tteQBC(N?w)!R0C9eXjrMK%
zR@Zx>7_S#+x}($xv>0<Lg*U{K;h>ub;QZ7l^sHX&#ls?jRHK#I`Ms2WjK@Z_*Z0Dy
zbn6b+Xr6}3UfmltxDX(*ukRQ7q>@<dDua-yn@RvpKdb7@TjuJ@e<!zQr)8GvyiCO#
z2(Uuqb*jP<LYdQ}|HVD{2BxWG7(7(ajcu%ltxFu!>mcYrqcUS*L_U=5?WDh*)UcUB
znIINcVG5^Ij4iCpcA0o<f)Q2^gE72u28M_`qG?&TR2O(;#`=%em}+kSXzEwGb%|yg
z6MTumI+?!3ChT8viWgN6Bry7x1wrRS@;K9u(|%X8`IISs?5DWl@Q9XQBc30b<M~OB
zq4FoX{}vnQ{Mp?NlkYJdgx1w`YB=i4TIvvzY(W`O8_$F*r^jSX3i|udD~MQ{9^$A;
z>|U{>=j4LHz(dwB1I4KF^45x)I%vy_>kAq-oFi=kWt3;8er|-xu7L$N<P;Q2W{2A>
z8rx|W5QFNc#rZt+%pe9UfwCej;J5l!o#LJ)_cHibtE1WC2{kS<Rhb#!cE5}M>VxSo
zxa|OGESvMKq~>HlRC$1T36GkPa^dr{(G3nIq`fj%7_-C=;q?gX=Qo0usG;0O#tf-M
zfH;~Fd(X@aewLA;W?!W`u`#)YeWJ^&`I+Ntqew*|R*y7O*HjE;pT>8_w1EJ)1j`Q*
zo)$DX9eYJbr86u>$LZ0Fr0ORt+~mWhD=()1P03A8Il-G0&m^jb!CjiHmQ-DUOcDx|
z>5&-?ESccOAu|tIK!t1(WP{WpGP#my4fEXbYte`bFucczc&bo<!Pg56<n+^a+5Nls
z;tIhfr_*Fx=#}LJGJ<)GTQJXI2h(rOq360h=rt_<pjEIiLgHBLC|)6<kItTX`*C&k
z?3|q$;oc&Y;);>wt}ejWvd`!>a0vnoG3RqE2RQoX_zYZ{2@z#04VJJ>McG_s_fzcg
zaW~hvJE_EqilXh<H&J(vlcb)pv35_7&W7&hmMSWs&fi{H5E~@>e=(zx&)#b?06#<P
z9ztM|^BU(@i0$a}Pe1#7_0zwse)hkhsky7BgJa>VOmuek?i2;rMi|)99fxNGzr&r~
zW8T6->SOO4b2!{9M(uvN%3)|Sy~!QJA0k`ELwZ-SPpB=WWKz+YC^8yZfPE}vtp&T8
zH*6u9q9i=r=Txcb{$vi-JUn=5S~dtxZi|<Ev_4>6t0sWmv!zSHAl~An2uxe{BFZ)-
z1FDH8gVVw$g+I14)?~Xw`myX3O^E`nLg&P9&^T2mN&$#vMF{HI6ngh*l}sYfV5D3g
z;YvkUGi4sCt|f&<bupNY5+J}>!w1F7*qUB;JW&VSgvXksIx)!2v1DU2q~UqDDuJfZ
zNhOY*?|F#&2+1wkspkcg6Yx%ye)JgNVcM!I3<57FBP$Vv8!o0Q1A4lF2d&*0WX0{t
z`Ok<*Ty}t$zzs2y3~{W9XSn&->u@une~#VPIn4179$VP=61TgG7#oGy#jsK2`Vro+
z40THmZr{LqG)hgj^!|EGo|w^+tY8klgmNVn7MAzH5N>@=^Y*YQ9VXBx=f|_dV6`Cc
zCwX0Lme)MEu{QedOyVLi<<aPs&zK;6Mpg^Mbw~@TEHi}6rGH3|Zj)vBr{;NtS|u8Z
z`b-hpNE1h;X|_a88s}VtDS<`&@uAwU$vbI-gc7F4l3q$hlI{>qMOZ;Iq(>FTRLR>*
zQjnMol+YOu6Dny53rb>|^n~zB>`th$NIESf7WZg+fuJ8Ou9yxDp-_^q5sTgbgB2%@
zkb}`rXXihk;m{o{5l+r8I7GBuMh7@|bBWZD+yN4QD~?)0^Sddtc0ND8j}uV$KmBa=
z)xlzRxO&Bn?X}aRkG#k#@mNKClKvo1u!)Q(@clP2gK%s_tSbhy(T+W2hDkVjnYGT=
zI=URvSS#O#aR(SmC3a#7ia>ZLGi!M7(v@k)cDi$sG}am0TQWVx8AHldLrs@LRLfGP
z82wbjlw&Ai3~`I5ST>FgZLx~2yHdx)+bG+)w7-?&_8`Rtj9Anho?XrV2Pei4vh(l!
zC8ac1p!~FQyydhfV_+;6(+RSMwXg8(ArFwgwl0YaMV1^gr{}d8ZZb@3bz<&fo38F+
zC=1sdSpf#&4jpz@F@viIXsj!0(a}kUDbaCRrK*5;Pbumza|8ayREVua{cRR>j9?8T
z%v`a-R41wIVCY7oNqBbio1)kFo_g5-&FYe>BzaKVs-L6d<?4(lJGy0{Otud$;U_69
z)x4!*$DrKJlSkj*(0`JSH`UI0&l2N>Q9Xes%C}e27wq#0)#I#L*to(xrHgs>!O!#A
zr7iq{N1y<WW$~q>Nwvg$c1r;<9bbyO=aVb7;P7hNO7j7xQJ)cjLJ)pu@u(}(SlLH)
z614M%Q|G2TpCs^_)VGQuQS&n_;>C%a6ilns_j)EXySo%3$bQ>%xD&a4t0LWXKi;+z
zgm?b5xHE!#3N8^3@1;+JP6dr#$N|};Bo3f(w$;jZaX3;pUQ7?!FCAHRlCM7Cze%Tn
zthT|-!YR8^q2EY)rqp2ql`r85khBz9+%ligUXL%58C<|lX;MU@60v6sXbq&-<`K5^
z672an?2zohX@Pu-b_h#Dge3+j<7K@Rp*#EuGxba0oBPRRThayUlTwQkFOh(v3#1!9
zi25nI5qC&d4HXF(78#o!X@D7FX<c;@>nPpJdkM&3fgyV(4bz}Fn9s0P00J*A=Z732
zxarh~#A5!NV?pdcRawpz3CEp5`K@4()?ERzPgu-v3*hMy@PB6;4ocBhFk*vt1i2!(
zF>u;vOk+x+7QN&u!0-Cu-o)WPf*jBEo(OANH4D~wU<w?d_>#u;34gkxjpJZDXd=1B
zH==r-p{Rl_!uuu8kx(cit0{Z&iwOG>Y&DKOikW^hjs+gz(!2|<qtAL~+)l>Fa%u1Q
z%`rNoOeL)QV*0w>C_YSab(SzUSu~=>yXnLCf4#MU^XKplrlZ_2+c@Oi!P4^jE!+ED
zUq3`t)7cj8<j5=|M@Cf{kmlGW<qiVxo<XWoAdaQW@o4qzv(f4*EV;|ecVGSqR`hsn
zyWA~XU&p<_2cVUq0Thu8W(C&t#T9~dWV}S20VE`sX4>DQPw-zobFIadMv>rf6{c&s
zC%o@l;{mPsa6YJ8ufQZffqj)C7=(3`mcg#SIo6)9Uc!!maWEa-`IRXB_U4%X{C;P&
zI{zcdFaaIzTK3~iHVOP$f*vU~=5n+sB7qX2p|Jh;>H`?nXnZpMuhrAlqtUnPvvnp@
z$!?YpiqtTL;7Kq2Uz#(WyOtz~VbgY&pUxOP&AEqv8iHWJ(mhDwEFjbinEDdyK3ENM
zmW&|)`GuSpc{r0$f%G=&5LE(2r7UKR*NB;FyiM}j_@+IQ(1e5mZ(N}QDT?!D4W94X
z!1-n8&B14zZ>CrlU!8V9wc0g}UzAm$H*2Q&!d*)!nTL>u;=Sw&L$T?q3YsQR9A{=R
zr-kUI@-x;SMe(%0EZ$;AX!aV*Wp3&SMd0(xgKAFSyjf0N7rw*Hlg;qn0w~LE@!s#Y
zgWmfcoxsY7e)XbAK*0_G_kH%$Z9Oyu{0%#Y+Ec>LxN)}YZDW4;X7-j_ZzorEUsbp|
ze$*HVBgQP$x^WB@yi5+pC!E(|ykJAOy!}zCaXid0v+A4$?Ppq46!2J}c#%OQ7C{+n
zUDb$P3Ap2)ZbZMSNsjj2`o-QLO*(O{1VB6?3M#8F9@MD4Yi6CIlWr23HBI)_o%^~9
z>|-${L<z}SMyp(!<-vdZ>cX0mx~x_<D79fBj8A4b1~-4#i!nfCLUiGzzW1R(^j%rP
z*7BdRBmPb(>JNSJ*d;GhI&s6KoSR8@tiUrGoQM$swuslGk8nxFkH78mtn7W}SKnLP
zTl<PH{S|Wi^RfB<?#I9LKK1c^E}U$9982Esbl?284VD0bwthiUmWeG}MF|G-xKdn^
z9|L4jDBU3%tcTXWHqGh?xTXy__$!W({~o|3LF^)PU2WiF7PlyAokhVnVDf-ZccK;E
z8J2_w+%A4y$bJ0;riz_NSdC$x>zBYR!0BR?Azh0JrZl8ynad4+uA5}NfDc3ipZZ`>
z%)O|HRjDo%34rOVpQw?H3PxzzVkC^50!oF5RVvm8JK>h9^<tPx^ubDania!40pwU+
z2S+PF46R5X%!HrA>>EV;ek~l(xpkQ^+L15y&?c3m#I;tcw}~><dvY%nlvW+oQV-3x
z>_{UmMuJ&_(-Vv?i5yBFbue<Q4GPpe&YF0AJ%@Y!cH;nXP1}ivZQdpxsmQXNO`KAt
z245FIufZ>x?O0C3h)0OB81aF-2{0BUqOoXS8{j_T8mso88;kTov(a>kaS^NVw9iEG
zRK>Ipo?++k+9PaVuYn;l)&)>WAIXT4Nt*a<@QZh8Y>T%hb*$7cb7;}*(gvZE)!Vij
zQFj|E{PJBtgvu;dI9-z!BlRM$*CxrELdBhyabeuTTa|%SXn^jqg)Zp>NQcpbr!Jn{
zlJNyyfn8c(d9V;x(7iPD<yd0?9JtqRVzlW?Mn@^FAp=h79qE-G9B#cHz{qB(U0QYO
zH}TrGm8zrVE7yRPzamopdjO{dv5Qh&yNMBkRC&qK!ZGr-a=^%6DO+jr0gMvFE=CT?
zYd0}^m{;SDR9K(~yuw?e79JgbV+<qoE`kjhO&oVLS6ZT)`v(Gc{)(vi?*Y6L#4b{G
z?IuR<U~rPTOCn(8uZWTV9>6F;>|#{cZej#VClI#A0(t#az$?79ar8n1*hT1F1OrHi
z(ZuobnRrvg@8Xn;zL<bzd`&dtv$xUUx1xnEqAe&*OrMo5w8={4Qwx|@Ux;k=Y5?E5
zoi55fu$owJluj!r@zk7H0qKLXB$4(~1K8H^yVzE69Tq)i^=O%P@|oMcPcR^gDLWu-
z#sHB`c7d%DiYhEny!apVhG7;kce-!c@#=Ijwzx1w@65!*ule*f%_Uiy@FUG7c2{=%
zWQCjjI@wu_%a2hsyDg=h#f}(gZg1t7AA2dg{`#BWWpmEj)HAYbrQwrd_?@gZ%8dp1
z40T%zAAL=^D;K77PgFfZ+<#3EHcmxn;a|-8NWI^RgRReGTbp#zx7;3-Z9zqU5HW7$
zU5ne5TC^nhO>yBK<&re|_V}il=*Z`-($rtj4>(&2LNXRvT)!f$2YiG;R)jPBfS7`@
zDF1!zzQ38hzU0Lu_U7UWCx_tLWp!CwGK?{`Ktjm>k_VLy>dPD47Y9iz&g{ZE(Gzj+
z!cpP0&yCXZC+FQ>GcTd-RizS;2KlR%2V}~_OnM7!%gUfEhQd9d;EHf`>JnHRJU+s4
z$D`g7KnEpyKbt?i4?b#B^!w|xfcseoOPO=DuOW7%8z7=iy}ml^TDMMbS5bUNJZ#tY
zNQ+TKt_Oxv1LKq0Osmi(Kt9!#_F&0}Sh!x>4;H$v7glDt4=l7?4_Gp~AFl1hGM)8_
zcxW_Kq&g+rQuPP{w`GMD#JZhfUy?pFJB-{by~SF2fH}jqjhAYPcwXWTl?qGFp<;Oz
zI8-XH%!W$AvR)3iyktq>NW5BEK^WYg&uk%7Zj>y{`UJL7u(YeaWL;34D6a4)S@$P(
z9?#Yf@ma2#`%pKAj^-VhK8#T4yR+GUUroztpaH96iT8or-@iP!3@$@mP|jbXV8^KL
zYZ!fbiR7t^3ml(*E+;lcpWHOi1iYYeAurC)FE@B(r4A4&UNWV$9e`tn#spmxGwy@8
zs$e<HR-e=O&Gz9%^yASAZtjm?Uy~*a|7xr#7vX$Px$$FQB@`}|8i4e)bttYMWViSL
z?4;20LAHL#b>RWHX=(x2bT-N(0b;=ccwlgtokZXSrpqA2vgiOz>fhP%HPF0J9~{v_
zO&2M#v{j1xHEBrex@E4eIb?TvZ4=zXa88DMAp#S-NM4sdZHJc#7+d!`DA#666ps2o
z`)zDW2nfT+VlTP&TN3A7325gBM5?thkH`cM-w^(J$~5c`-5DH^Pw$pTJyikb={NDV
zu1A7zdXsX^iAFCC;hg4gLm7)&iD(KK`>W*<_8P#~fA*le?QZGhaUMc8nGuGp@o7wP
z<FA%S)@ukitB<({+0;CkSqol+Y|4N_kuhTc5p99Cap(ZPufl-PBn!uAa$xaZJ&mN|
z$q_r$2M9r`VDKS2{;H>mSc#I8iVyUnQr_0q!UM8I|MxUkRmDZuc3xR44K;ebFFlS=
zvO1GjeTa0r{<MxrL{1O>T2O<16kt7H!-_aT`U9B@f+?WyYApc9m(s2mtp_B-dzeSj
zK&ovc`om79ms5Yggn5)ag7CJS%}%dQ)8GN^e5@Tx!vk?)3#F+QW$cJjC{PYJiU`SR
zD~CeVnzQ4ckJ17|=(8uWtSuxgzlvSeF|06$b?lo?PXik?h|nqC$G<S*I&BejX&}`t
zIJB-&2;9?PsJV5W5OyeEiJrGrvokQSTI%9m!gfeatP6{Z&hEr7fzt6c*q|up&>MIb
zu)A1Bs1BuxBWGn8b4F3+#(-U7PuUEf9XqTBMh2;ii^A}lO)OrJG{WePNoM_kO@2vA
z<tK;Tz$!rQV%CMw#BRfuBt|3?x`0;}LtQmZhxYLuS{EGy!`{4?Ld+i8=86ug9w*8d
zlB+Rj$F%N;h{|7siV5d|Wp7vJy>OSP4M7_$97oX$Hqb1EVxXY&YG0AKKCn^u^#l}f
zn@ra&6L`qEw<Zx{a=Wy9PN9((L$#8PYB*wEQ4@KfrDQ6mekFsk7rR0GueX+4S1!^A
zZvPEn7cDeHu(Z>9TdNeeRW7yl5G|7p<u%O&vpO}mVYC}|6h-$28!>)l2m~oMWRyLt
zET?G@-tB#VxY+yl_#tv0cMcjM7wR*(bRR#jYVS->QP>;Q!v)-J)c3ZQ*4I_d>+`Ql
z3GgD%-EGOKJDzSTdL%`5-XbYEntXf4N9O`_bQucY;HGRGpHf}qM>5mo!_iMK!R>5O
z;iR|d<pM`)yen;fiQX?IwqM6_Z&Y<eB97S3y%w#Kb4=R{a{>ie47-H9t%(h7o$IX!
z_Nl;~1&S9w%aI0Gub>?Ak=tWZ%_!*NcA;mcwR`a5@q{Vy<EFB2d-t7G^@edQDrS#j
zE7-gBhWtop*@SL$^LVMA<h4E!6gOqv3JkBMV*^cod;CA$;-mA$Vs`QmY1R9${wx_d
z2*h<w$yrC|aUZ>r)D8n;s;h70x$O?YD^JjZJx@<@Le@_N`#9tmuqAWwW_oS+WPPex
zC$%@m+SS=xyW2_<86M(7w>eJXBB$ZOE1YS4J6e6qb1LN#TiLwVYUSc)Wwv;PY<^es
z>DO{Y8L+W+7tqPk6csj$3nU3cK0DqRYo~1`&_Y{aV-jd#c;Xy1dK`}T@}SMfqr0Pz
zCwmhlEzn|ncYoWP98E8odHHC>3l2wnJD<qgZ?`rcj?yRO;CZz5a`W4LzUKEG9O(IY
zvU{-l`qS0F@O00bof}rb-0wfYS!g*gD|pG{?q^7%!XsT7g}Y<sIrpnwy_8<+`l%%p
z$S4jJ{ueJ?W3EvYCaIvc+{8T|tr?2DpZ|b0e$5`vK={-hEz6_fX*9hoV7FA@>Vq>L
z;60iB-NqZ?ETj#HfoZwZa*f_+M6Zj3vMf>tz^Rqa)2ZuqmHssnef$^wlWZx<;zkDH
z%}kOK2{ql7ofOm0ks$0>rcE{>vks7QBGVP;RW}rkKHmKiC&2E@!2C`@F`jj%ZF1^Y
zWYwFnnL(W{><S^Hv7!j4xFA^(k{BRI1<r?Foltw&_=!p~8SPbSM!wXYDJEArrVoX%
zdV+oP(VuWH<I#)<x&O4b_9yN1pV)ty#W1X41Qdmk>Y>3JhNr~gK$-1Z^>G}ABBe>E
zV^-r4BprJ6CDRu=GD053qUMf24mahX#7s<G3V}TK+y<W4;Le}E#i4#=Iv~p+l8Cqs
z>uBLkfCF(GS~xLok*t6@M?8%M_`~KjnSLfI<wi_#h4EELnlVa$<E1K(G#s8uE6+EO
zM>Nox`Fcl#hNC1-9`3Xd&Tc@SVm^|Kr&mQ}@pOsNZO=-U*(h+I&NcQzt-zqRCI#8J
zmRO2b4YJgLQu?I5l*7*wCy8FMW_*yk6cj<tmen<;!%{Rwwh5KgHtkHyfNhJul)^kr
z&aY+GpSTu8)k-S?_;u?2X8JwPDcUt*M+z0QP!+ZaMboJ6+DQsSNgl~c4Y)B~urJ)y
zPM=6%rt&=vbWMGsOSlP>k0?QOHIW|8(BdShXk{~LA4%q+nzTvTD_5d1B`5js57pV8
zUt~VIgAi!sZ^A+e*Y_<Zd*h(N_h3g*UlpGf??2%i^i|AHa&~?+{e5rm>`qkou18IY
zFrnNS+AR@Wzw3n6!|T<g)Wh&5$SI|xlNY349G-5T9w8$dBWHazRnGrFD~~-$tzzRD
z{0qB$xA(SPNq%g}sFtg;>y0Ae`7PRnYTFd)`3p!q2Y!c_SFBlo?~IX(nTCq?7yS^+
zVy?T16I>DnH_R$6kTHkFYLWj=rf0{J#c2E6?fu8kUhY5L_)iQB!t2ip@emH>n2GRM
zTDt-%WP&<BTYate{i+k*GL5(4rK{9`Q)ZF$Sh^hTg(gCo*9zW_tO%0TL0ahGt)zEq
z1{#67s|2x&8@*l|jqhq=pae;5d&}s<m;;%xdKzM`waUsn#ief3`EQKV`{}o%>FWbn
zVEFg#^yK{FH)bMQTl=JL5`!kQn=}n=5|YFno*r>=gfR!-AUmn}16?}>G3MYWQw$uW
z(|tx8O(=1wwz9B*Nz@48CFQ2hIt#@WLjpCsMo57!VF6FDU0%$6fh>T+hQh*$BdcOy
zR^wZW>2ecI9x~L{S^0Z)FabuM8+=vr2P_`q>^*EyHH?H3MNqzFJE@6YcsP2LVN!CL
zZfa1J?SK^e+Bo?d)9bY~?N?sY+^6t%hdt|}KhR$2A4BC`s4jBYhpqpXvX+btur+bx
zbRP{q*xu}nR7K0>%*bIz43<4-hxaz>g)xIO>>92#KXU2LT8;%5@I>>j30NM#oSyWs
zpXk?Z7-it^_H-;IRw*T#oLuzue_tU)Af;#Y$tV^Di9Ui!ghL<l7IR{<;s%@ZscN|v
ztJ1kq|G;*+(iv=%1(s_sl+11MCG9kxj6pk?wwMTV{LvS9ez5lO{fW$)nbMdnf75r4
zrC`U`TGH|5ta^X4_VL}jU)(3GkJt(&SH?i}v7Dc=Mx%m}kBD{RWWgFXw)PrMElKg`
zi;St!hm%(>LeE<uUincV!_LlHbIIAp;s1FV_ZGZ0e}?<w1GSY~(Ekt#?@>P9P4(g}
zO(n<_)>kQw02eb$EvDa3J#ilXLpJknwDVSOhN(Lt<pgZ()xq82z7cG7@=7;uH^K11
zMMH9Lrj9FVg}-{+=z^&Vp`pCIL}vPf?-k)DTx@nu^s2iEeK$VnRZG_ug6yMB$*4Bm
zwpcDUy4L}fveLi)PV!`d8E)GLGb~Wk+=2($wY%VJ;3psU<yunl@bNnor7kf2yBv7H
z8UhJRXL$E1)Qep2jNf+6vw*w%uQ(Guf^{k{`S1z+{l7Bt>uip)qO08C#u{s~`{S4E
zBkT$6@*iq9SvZVWJ{zrkj#*0QZUMc!G6Ts~{ACM7Rwo!cacZ*&lrVIpq;=KA#$Ipq
z$q16i5ABKzz?L$AfMj7A;WcP(WD6DSQz6g>y^=AYD(Ygu`ZREGJX;T)n!sbz0i1nt
zg600?&gc$Es91lV$sO|AYsO&llPH*`n-Nvh6<zM~rLVxW8*6&X`q`X?X&}4PCi)Qm
zfKB`qnEvj?ba8%yUhi$6pJUs**cji|UD~~n@3W_bl}zC}J;n4g$$|S+y+yr9a^Qwm
zQ(Rd;?B$=TPato4O2tHL2U%DS;r|I*Y2`xtoU{}MS<|C-lu_K1C!xR(_;1)z=Of0_
zAadQD{%XbcdUkg7HC$1fzKf9TF)9MEnr7uQQW*At&1Fpf`sSkZ1rm!$H5}tj;U=i%
z;h}<SR=>ox*$z_F>gp3=P}@R|$=%flU4_`HoU_;Fg?Bh>KpG*z{hI@%+|gJdO5`PW
zU4Nd=3nFllo^j20byA2O?L+|80dFoc1CUs7X66@@@t0rHkZ$xWZQ?ZmQ`plYJL_NE
zI09P{^N|lAc8JopH~R6@@!b*S2tOlMf+r%igd!*;{!c%VAxhhX-6l2Yi2ugL(SSh%
zo;F>0A(cZ1F-B&&)U861h%yml>~=t!_*jd!Pc7~q2!VL2Ktw1VOs=sC2!hzy?fL8+
z*A%=%k)RsZ*Ha`bqV;!a5LN65i!l4PGWELNt=wd2D!-J@CnonIB;m^86$Ic643;qT
z{`6PEvdp|(v2UCieSi}<%uAgnS_Mg4FmQHc-LQ#H!)&DNnzDTwtBmZwF46J>h=03x
zH3{CgQHm<26^1+Q!^njK$?HxT3R7RQkC0HQzl^^~bG$QU#=Dv#jc@t3{%fh=7q}hx
z6!-ES@k&uSnB^Fm=1r3sCx2;Obf|TT3e;YgNUPa%JAbsZ`$61H55W*VTR<RAy<uOJ
zK~qK0+mO<<vJiU3uSpL4#uHqKaBwuG372j*)p-{hPV(>u%}$Zg8!BRPoqa`~<Ol})
z>Q@$`^4GG_C1JPN%aNPF>-#MLS9G;h7fy9R9kWt=!v3GTlf5;IZn(?v1~)O+KBf8=
z823i2pUL!qt4+NbF=2~lFtH?W@z_BE#yCJiZC8N3BVtW@)AXt`45P5k;%HQ3CCqq%
z9X1|WPjkfJz{kNxKe2Msdj!dR)FLa3=_zijo*f$3V3L-_l$)!#^a|CCOW8%$ZD#tn
zDl@`n^iQ<8WpHF3;jYT3dk?>oxQ<pOtnba0AOqbj@~MRdeF{_m*L34a1APJmsXX$=
zFaBti{G54Wnyo2}=toaGFhY`6rP!*?h2=tCwTbS5q@h6UV79B!GpF{5ixAwMkO$4G
z8|d+m;wQ4PY4b)9P7JsjJR-GXf<dY=R<s0#+f|<Zi^zNC3Ph?>gu@~t7Uzd0T?i*v
z$FsA&&4UZ$#AIJ%89$v%)|LDbNPj3;YVQCVS{l=#nT+$~a44D-g<orVkaANi-{hFP
z^fq-dB0>V3Oeuz)zgE*uFBn5Q4oj`uy!m1ww)GY^&BNK^g27K%tI4yjv3A~_zCJ&D
zhvg40xj#cOSqPn<O!pq2;bIW|Z<@-ktMgoL6WP;$LM`pEuBVKqf<qb&c~a82x`PET
zBga*Z(ldRT5w+@_?y+RTqsf}8aTVqZ_Rt6vg8;_AzF}jY6r{dM<~t`)d;OXhy<&3H
z6*#u1dvXToqJD)?09$H7ux!o<2ymVZkEw`Y#gEo+tEouxz4zTMm%8KN<i{jg<6OIe
z9H{wR#<Wp&iB>oB5H<i5eZ84F@o}~bMuh8yhZaG;U_m3??(wt7FF~X+jy_(}k?w!`
zb@TbtryI{6vVft}EwlY%YyZ*XCtLdupZ~|RC(jW9i_)0uR(W^i>Je_Lq;-ROH_=?c
zq>xdN3hu4enTN;EJCc;u+qQH)74Y6`=eD1G_s!#H`y1Pj_rKqIvGe%(Gh+VP8t%Hz
zy{wt#;P6uHjfrC<OewhC8wFgzj^{EzyuV)Z7Akj-3}mgYcCKD64(GF1(|PJ_RSm~O
z0(#ZQZSzIhzdc`V_A8H)RAs|imHe|0CvR4)-Ql+F>vfP2&Vaj;3f9@t^Dmpp*<CoZ
zilR$8?u4nrlpTI135ud;H+fM3=0n7gC*$bwa&}%E5k5XSe|2z@`5=@uGY9smNJaQ^
zG+S^cRo15h_Bi=a=0ds<v~$rcV&LJm#j(tz48*iJ`secu3^v_$I5}7>HgRn*YM`}G
z6A6Z=n81vbm3wkHks=-i%DX;VDM^8L%wgB=t*j;0Ae)_|Fd9hng%h5y{dfA<X{Ks8
z6)fmabl?)}Scf(hb#PoM--sF6n31VBExB773YjQ^;aWnX-KyrWz=TihLb|V{@ZQRj
zI!mXqCq0pv3ZI<gBH#Q`)~QMrz9i8USVgJq*(`YWW=*C0xuk<cJ{aPpM!-w@CK^@>
zXA}pjghHgxz_f0R-<Xz7bbQRZOb!};Q_!FRq$FO3rL%86=eRtY4xH4wFgY}6KOM|Y
zo}WF%&fV<nc$1<j;di0~C>6`Hj+DF5X+Z{26Ya?giMD{~u2m(5EI{c$l}n>G?IVQ(
zOR4CUjPT4Y1nXTqlnAc;2_rWW1nUM?4=!fdz2ROv%PBsRaS?7AE)g4T=aO0TlbKWS
z9Y&v>yxplxh~Zw-7tn0DI5EXqPwpsljP7thOk7uj5H0b^-L*T27#W=^{c%;dVWyXL
z!y$Ib!%BE>`h)j-<Fak+9TpU2hyAwowL8M)+50NGd*fne>L!RJHWjLmBEIh1^lzCU
zrnvHp(wAQ-xft7?+@%iEaFIA;VV!zJghcAU*}-P!8L~(zFTvV^y+!|sE0$I-V%Z?z
zvva%CcZ=7jOvYdQd@w%)@CPqJ(>vl1=nfH1F^%txly!Z47R?b5+^|6c7s2S`|9Jf5
z$;XIY!)=U+Ad*pgeZ>Q565))~7>hUOS0_iKSMWv6e!<bOjh&6{(Zm_xM*x$t$AJUp
zD7m}##eK(CoL=IeX>(-Qe&`=dl?0#kK9nv$<GAkPr`g2?I`V3IctAS|ee_`k2b&bY
zd``?fQIdNh*95Jx=NQ}t5phWF=o~7FLFD2a)9dcNs$O_}f(tC+>K+;Vng%Wi0&{Kj
zU*}f<^$t@qaKAhsAy_tz?&#`*M~g<dbo7l3;T<`VsLX)tiO(f*Uo|vgN`=x`IX_9W
z^4y5<CAXu3E@&!CyVYqr=WvVe$*F{6WmP25<vNzG{gVkG10831_x=_rY{_+AXo)Ok
zLDy1ncPzH4MOOz!_uU!1oO0LX_Xj6eLE()>x<)_0!Gg{Q1VW|(0a*d6bekH4o3zqt
zuFt?%DkyIb-a>Ye83>aUPEBIx7UQLog%}C#6pl1-2q9BfV%4sxe|SYI$;>ue?5l06
zUgX?6Uf^7I80JO0D29U<9jo1w^=aw=?V@aY{9kA<A+AKX$Z3Xd<LAVE>GEOq!NE~s
zJEhJVf0&iaK8kkL&E%?5Wj)<W*}Ars^Ye0|U_@mIJLjT1J+PuZB}Axe{Emadj3r2G
zF?#r<7m$d$7G9a;P^4Clr)Sgo0m2Y^#?bRpeB<S0dUT)}e*nu{ss1@(3e8v}B1Cgn
zhr}&rJ2@q2=6hU$ZMiTE+&yTHOjwx)^W*9RT$gSIJ2i0^bD*Oh;57E`w6PL#Lp7?g
z(}N3=DbdX05H<`y5Z%GRFDHDZ!wlBCIz4^I?H-vQWzR={VmlEiiEKXcZL-+FAtuku
zsea;t-ixaE>H}AH9GEZ$F770~{3gMv*yO=;8R#d33mlv*raB$UcXU(Zr{oEE@8Rqb
z5QW2+5xx;&TpWx5l1ET<*qNtd5)2WjgmUy}?L!k_#}#W`1ldvN<KPw`EvDEeAmeSH
zlZ&uFrqNhcgYrmuNUf@RF6s*}#gcx`+kh2?8Lb5baO=RKA}O0pSUK?;8*VdAaE2XS
zIyl)|d~)|I9%x@6a#ld$&1AeL6)bW&wR#tN_V?2n$9Uy4d5bVK*=gnqegE?GV!Y<T
zd13hvaU>e|1SiL+*eu*vSK&S&zde{^;Fv8}>1cr{1|!hD0`Tzhi~Xwp(QJN3>s!ZT
zSwNdlHg<MuP|iaLOrC0>%n@%P@XnOvz5gSq`NiH>6NHi6{p!BZ3q1=&m?E5zv6Tiy
zraJCpT~ABgjE+-U2>TH7n|~#Q+cnG(;hbb;fX67cTHkl4>1auMR+k*?omv{jd>6CI
zhZzrZg1aOTq>~yW4<-XvFfN5srO|Be0<qjYK&v^0ag=pM{^V*Q*+6<+-E@Cgenm8z
zYhlu8DVxVEI|+KsHjZ02Fm7n;(`tonLg)ERk`>@AwTcVByoOWk2(kWaNtMv&kh~7b
z7nruLp<J}tyt<3d1O1j<pzG^h54Naw@glgJE&<Qvt>9n|RdT~UV4P5|tEjSR@o7Qy
zNYo&?IgOP{UEu>?Uw^pu^>^RoVbtoDYd{~seEI`6io4b;c1#>A9(7UsR!5Cd96iM<
zTHN9dH@FBzgTE^`I=x?@PzI1XMw0rNnSZ7ZIjCGy*;GtprJ2>~fXRjhv^18l9CNdu
z%oSo>4aC(O;wDb)ZS<J4hn$jTXIF9uGv~$HgJRy3C@{K43FL7hGjym;-CVmh%F>-d
zC+Yt{=us+2;STGzOhQ_@;9UGS?KoPoDTL7a5a{onA(Q(86)I9Jrv20D{CGN{MUX}P
zh556vrCEM>m{UDNI>Ft=n5q*@d|K8F_aubO_IzZ>(Fjn+JUKuA>FUDus(;4-Nbek7
zfnx?k4GQiP4{S_S*!7AIFz87QFTunlw)XUBuq+X+8id%Uy15_rs%g5X;t=Nw7)U50
z<}o6vXx3(JmDWSI$QmS?dJSbyzy_E`nyrkAh}rBk)C*lC9xS;WgPTkoU54-Ec6p|m
z=ZmoR^rm(ZsFQPMs0L{<ITV3$fGiY@NxB>i`soS*eJI?zx9vT7%}Lo%eAb9~RKmnw
zHSg%l*i_)Od3VtWSG^BVB}LwqN;4Hu$?+$T{elBQc$0IEOv2kTXGQYJ3sC3KV?*A#
z!~3=?^Qc0l3Hadxxzu0H#ZiKR)hFnzheq!DOetMXC|U@LOia!Su|7`AB#iX*1$-7T
zYUu>cp8%y>(3UM;y;COOeqB+KB~A^uROP|5l1(&k;9VpiTDyA-ip9N0sTUnI7(Wg5
zvrMO0@;X#CA;Hb8H1G^EQb&ViZH$OJiG-(D33mAGXq39}&doF~8x}D1k72Kk*W#Ka
zm94wW<&CQ>7w3A*2Iwe*ZYI)Bmf_gM-+5BQs&b=<uyL-=yz@?~Wj(x<0GSOHM_l1h
zyhJL<Ew_cJ4Nj3JrC1oK?q12}!5K~A8DdQsS4rP%CJ8zT$Ga?wI-QZkYHAu{r=>Bx
zce3Mn`V01}^FE1D4AK3f5^(PtwY~i2yp=}e`K$kfFS<xdFQs3QO9R5`U1K~M-S0tq
ztf^fR+cWn_B%!}x_#*8Hsmm2fhlIDrQWOWUkbNzqpfBO_9-UqxjwdBLAH6z<SHqR_
zo-$pCPzXO|<6_)}O02j^vNoo519j-LNp2kkuu0n_;Ccao7zK5gS?wB5&s_!p441mN
zfLAu8b&^XW=Ua+8;#84j<P3)=E6ngpb@B-MRMD0TzIDzZ<P4QsLKrO*$tcvJ9*XLx
z_m2&#RxdGTF%7+BUVSk=#i4XMv=;9W({W06=HH=69bESO^w5H|1nOYa;rw--brm|6
zIZ~%bxMa<NcQ?fw!G$$uFbiKZ0x$xks)J_@&bIaS4hZmo4I?UYOa(c66`}B3d={P-
zDm1vZMk_ABb94P5aHyvPqJAN$ian~V8}aw0W&^eDK<Wlsbr+Y5x66xFj9w89WftgO
zTx(XbRbHevR(pz(HCy?|dy3n>-eA1hP;^%^4lDRWb)1D&J~rm_103Q+u)_>e|1OCJ
z4`$`pVA2@bbz#eSfT{;E@~|i}Fs7Bxs*Q}1m0S8w-?u%8=26z6S4Sbl^}2g3>h|f}
zPQ}ll=Bdw8V4QFy-vSrdr|Ik<UvVB3dmxlAroaRxQC+YsLI?ntv#pJ*#;$3ELaI$!
zE$Xz6M+THd5Hq#%RA6N7_*8T<&|Wg2#7G*}#)3DF5a#8c;hkT2m_)2>y^@1MI^m>F
zh5h>iZd`jfB04kVNjI%Yps)sp6LV{4Jc%|BGmt-V>viwIdeg=9Ds*CX&E=d)&OIvF
zLaB5AKNk#fflyE)mjcCj+%!K9az`4Q!sA^tNh;;@Qq3ugDX;#Xtl+A}ITv}Sx?9Bb
z1($M>bK(>@XO5@8fOW>dVz+3#mX7%_(VA?3NK$3&En)vi4loJKyRv08W8kzWO96~Q
zW35+0omIV8K{nO>Rpo7w^6Jq7RSQ3;AgpJuv}5S7s52(Z!xp5F0ct?^T6VTJULard
zqwk(PdAaqUFJ(8@2Djw$_B6)wjRq41N1$N#Bxv`gZ-1S>`}ur+1j7|Mp_yrcrheqA
zhf_v(OKb|t==zvCCd;RcBHAyiSs03@U*mP!uvjC7(opRx3Dbz(3P~3t9<Ud2Sn?S(
z3B96-F-T~6as~09E=8>DA`==ET{MT;JzalfWhnfVQgqmvY+a)?)MMB@-&Sg;xhc^}
z<~-}-V<y%u2v}WCdi^{@pT4z4V#_=bpmV|rMWo}g4c4*c0yiL4%>9kUarYUB!4eqk
zZQ>izN;+X0iVK<g^i|%|;tY)qW-GJ$KrCE0iF9Kzo8k_sQdb2fB6~#RQd)<f2F8_J
zDfNIz@#n9JQrVBPnh-V{Xci^fwV29U(oZp+U`pF%U&$${u?y0^k#{f-S0F>#Kp#M2
zq#9Yx2A5V)0m`Y{N1EVJmcB`4(gBj<b<tX@Ri6^N=EEFo!?)9Gd-Dg22t&HQd{D^2
z`-+OH1I|YXcJPItO)H=VTGCJiBfE-|%1=E((i4y-Bn(u^T}w-e4{(J{{}|}W;lLXD
zqN*>P$rWpZF{7XY@v3H0>Bij4xE!+JU^Xw7goB5O{(W(Nw#Wa_rBE%ojJA6k`?Exj
z8Y`&(B()8E@aJo@%NdTxhQPeEJlm+2ZOJ*tE3kMu%hjAmWO6@K12mO0jqO}MS}opj
zv#vTYb8)VKS+u&LrLpGY6rkmc{c<@o1`Jvr4^w!z5z#n3le`bybe<kf#($q(Xc~th
zKkS$R(KyR46hjoRtM44pizL2#<HZQjl#FacgDst<q`_9^1K?&{ryFuq+PRCnlII%P
zNvurgKu#`NRurMBsR=oKk*L>jf@EM?y{pJ8d2wRL*BC`Zc7fy>4`*}9ZS>AnjP_Dx
zva?rR)sP4=iluhj=2o0WLTU(yi*9p5*hA8cE9CTeJ54P<#+fRHm2qu<`R3kGk@fXQ
z&!0Sb{vZ20|Mm3A<7a=};i$~4ey0To?EfOwgk!^cz6LO~upr0<9VTMK?*F)kl*1G7
z%x@QO4(4}P5CVnj*-TP6^jt7_j%r>M>uq^leSkdbc&)Q@X_sE8E+#`L$uREj*nEui
zowyr>5#bYstHwae3L<ty6~c0dEsef7#F1Ng6CeC)7yvANIb0G<?*9J8eg7_d3uj{l
z^-rg#j6i;asBZ)-8(zp+M*FOg-t~2yBK`XL&K4+8b_Q-&#N%P-1B<Fhj0tiuqc50>
z^ueDNU)+};qe()L?;M%CunHvlur80)7Wk?N^<RJ?&n{nsRjpClAWBRz9L#nvU#eN3
z92ls=;ga&1I@lx|?6ZqZutU9NP{ch_)ZZPMVK}lnsV94Xn_X;RsR}6=t$?u%7hSVa
zjF3|1m*=l{-XLxeX}+HF%s%L9wy`CH?tWkW3=#*n!J1Ff_IY|fw19IhSEuBTMM%@f
z!;NG<`Ob&OY!{W$6`BTcj~$*~OkO$fFe$|)5|4wA5Lh3PmXj;Eln1yc5#t~LI-Xf!
zQdZUm>cU_h#AH%{kfp?|$d#GsIKu}jJxDQ|ut-;N;0S@RJ9P_~H%lVlUoNN(m0a9O
zu3dE5ux<|zmOb9S=j&!}M1@!lmNrfIh{V{YL4c(429|0Zayc$?r4b=@8xKsdkeFt(
zY)3yRuNDC*Eoe5qRLN(0)YRV2#R*xAs>^gC=LB#U+7%`Hx5W115EZ^VzhJ8KB#I0R
z5q^>26OKC5Z5xj|53~2i9wrm8cUL45Q4s(ev#+XEgW##KrQohLULinb&USS0WW88I
zYPW{5QyYsanynhYt2BW`e71qC;Bj7r+5?A9=d1}-?|>HT_wNUdoeA;#M`VC*V|Osj
z>;S5hd{xwTYydq>OIwt7^L2e4cEMuG$B96KNLSTdx$TZ|2KQ1nnbqBOYfNX1I7Gxa
zgQEm_ishMJbxh{ONwjP3iOMSH)G<9P3-~UsPPNXPgU|L^Bh-5_Jz!GhH<&ubM{>)u
zg?FY+wb@o5@ZVyC0HF}noRhu2Q$Kd4OfV|@N>0X)aOf0gci4zTiQ43iSqrhRM15^6
zVX)fir`Rq$K)4;$B>j)vTF4$Ar>aWfhpkq&6_35vow(KdWN1q$Ee>u_rz+~0Fd7XE
zCw~X8zztZxSO7%1rIh#c7bDucy8o0g;W8!uHLPS0!s9Uoo`gFo_j<tVbRL5Kh>e-f
zIoXMn;{~i;n4uIso$q}}7HYfXO$CEn1W0L95I@bST5c(5AVqeG8@noC(c8rvVBk#%
zi|_)Q^K)OPTMweSnM_5acZvj1Mqmq6XWpUcY-#a`U{to~Rm*e4-5Us@<oa=tC_~w?
zw(g29c#+~w$sx=|=EQh>_HsVuc%(Nn5>kx)!G%OJhJr7n6&IsVa$y6mcm1j_&mvkJ
z+){LC!!2TAZIH70uXSkt^B2*}-;_}n5-6*`)9|pcVC_MeVnZ1#cqxRNk&zJBk#?da
zkLP5PgE%+SCx&;MySS{uNiES&_T!XE^U=(N&CdDN9DXKRcjd@PHD~zU+`L!Fd5FCt
zKoG1SRVo(@UTC4?uAbBm77($cDGY=oo<t&VxY*`Q4AXRoj!K{d0jIh|1DM=b!mf?e
zzo@eswDkLRF`y}$o|YATaGLhbv=4SJGR31*&{PnT8@~vxGZ)387;U>ryGSVzcuw_0
z*aVMcac0xX*rye7vb|_38!XWSLyaZpEs6=JrnM8gC@LmVJ5-50XmIrqe3W0)n=<N3
zdFO`?(<midv<yPQO{pbG%c7xXmY8<e!>TfOe_(vFf{Pg+gj`|QY#S?OjPGfl^?1%@
z<K?60FP`piZ*2Z`<D0Gh7hB(K{io5R&=k_V3_xeBg9P|cM+kA81j=F9N`#4MXl?&=
zOkrCN`uJm7&=7#MO(se_fck1Qes#sj)7XIYs9KXFniWY1B<2ty`$uEQrPRck{G4xs
z{(vF=;W$2l-bpi{w3I|F@oc7#YKXA~?90`Nsgz<E=>{axlh9r0O3^N$Lg^F965Kmy
zs{qRUmiad|o>62<=+qBxp;B(k!Fx|3zMXYu>}8e<k5rK+Qxf7}F%Xs1Rgq>pY7z|@
zK|Tuym@+_+kHsvj`iVtQC(VXuehS_`NWB3fQPz?AM4?FGo-QVe$M4eNK~}5gw^rHE
zOY1W_lo`f{aP((g3$#F(`Zqbwbz5bN=CUM&*FuIRq#-4&FtNb4zDlud+Z#7l@+<rL
zg#-uLNV!>Q;mWnT#c}SwUdlY0$n?3ANnZZ6kYlw5i_+&VBYUM~|Cr^~6K+Jzm)O(U
z))^zMVtN)yabvWnh|M>Hk%7JdIvl)(fh+vJO{+~_@m*m6DS*47die;;ID`jQZwEKa
zdV+I{-8PHe)_){67s0p95{)q4s$?@@_+d(VT9x*IU<~U^=F#RU?UVqiD*MMTow>5r
zJ-$tI1!ZI4kJ;l3+j)>58PM=5?jeQOni~c~5Z=MDQ#{ilEP>FWOqJA6qLjnuL*=sY
z1_vf)uiw>RG=!^u%NQ3u<sx_3$ZT6+1h{W!kk3V3^W@;~?^4zJqt|)@*A`qlZHO>t
z(%JwOEMkJ8*4pCN@4myCn|2{}OcijciADAq9v|QxoZt?Noy7^6@^}ti1CZT*PRlI5
z1%!pH+@8*uB&vi>*Ijt8-dtEP($3LSjgm`^*=Hj9@Zj=*ZP+UbHNy5P-EtBFCC%7J
zxDw-+uO)yAI=S}R+-#?q857C6VzT4kdwK*q%QReTB~3s#wdb1g==PPN2)-v85we9%
zbtCES-p^}4e6sh|SMbnEFkqG)+xIow2=kindD{oHH-2+@N#8Dv#?EKXP~Eqv1P|)$
zbkhB#zEhBuCZj-Jd<6$8*7u%3qtk{sAphBonJ#{l?ruAM@{7d@HcMc-6a9sl;{1ho
zYeQ}QGe}5h{MG#A>zS<C#mBsl=SFvo-nq2zZ}5nb9XK^eyW<5WD%p9WJ-C$1_OYFa
z1BlTvHruelh}|*Qw;@769Jd|HX;;)Z*E^<S5z~~GVyYOb^a&%Gl*jKfW){^M0g34E
zgWmg0U?W1w>snyD(p#Z!p(V=yn)W6es)UeqV00;>69UlVQdp)2TkRR@{1V$tJ;LHk
zA&{n3A-o#r5D1$kP)M0smcM{$Nm~$idw=)G`+N9<Y{^cHq;A>$>Z2cUGuT&pZv(8(
z+#ZZc?i?f(IzbjP%b-D7%5%oSUM6#aXdlJ$MbZF%SY_u{u{SwJNH0XQLT>lRgVn!(
zy84&@hrM@gkLpO$MZfba%8dM!uqg?!+h;~?LvxW0GZ^$p;P#A!H6^Je)mXP(C4p(|
z^V`q+#wBkn*Q$~L+r7^)d$(0<<;skVj9W%TMlQYjP3oNt8gSVAg#;Xx_BQ-|X6LBU
z=y`lRx+C5PMWCUS#4K-cx7rcPuLJKqe#HNkEX3!M5DOnp{*_1$eZ+3L68O)(m7bYt
zxHeSk5S3EU`^jpgkOBgt9j<`prhqMFZfF`wMj#5(P)M)T1aKd(hZb{WxfCf!%+QR=
z)Ni~?^sA)w8H!><0WAI$$Y6?<z$N~(%Ao^xEpowDB{fhq)5EqRu4HC(r>uDbHz2@c
z{LF8(;v*B1|1`%D%rfF0GsGs6XTF=4Y;2WoZcOnL;Ro9c3SH|8W)%#`->iRN1>ius
zSr2KnB=l6AEaK{=5F~y+LZORDuM6xChOm%tf~66`5+3{x2V{~$u5^KR(kWwL0A1x>
z0yF&I??bP(ZdEgO9k&topdt28bF#)g&?LH`w4-CoIGr)CgqW{N;NiNe4?yBcUw;sh
z%Q{Qu1T?Kv-UmDXlkh<B@HktyDV8<cpom%#URN>F6x4|!`SkQ)yc;|Ki~Xjd$kW+U
z&Q;2Ky3oP#{^5y~$)X7$6<3)?!9$>O_c`kyJHR*PUcZux#OFO4?jie0W>jd8EYRAf
zu`Y4?NpV47sX;eJ6Znq5W%30^7sfSW54&+OTe{B?iQHDB5EKw@&Qeb5e=6svK7|?8
z|CEU%pCSRoxGWsk;&VJL$0ke*ho{qX3ODd(vY)_ksDku(ZTsc=)9uI4H(p@Ff};X>
z42#54W2`$dr+`O@b1Q4;U1CJ($9?#)d#4vivUTZ7l)*QItZ+GQB8nhi(-4Nl?j*fr
zq#$tc>^)DW1y_8x#3D9b6^r#Nkz?-BFg|c1?`2VKMoF+PIXETFI8{uxz(tLcYecFf
zUm9Q~iwGy%lsYIkiXtU)=TY<V$b=&KyC}D}!OE@;pkO8oTKbmQcSu_d3z`#wgF{CZ
z+Z)QH^)`-j1PHvL!<`~C?HlbHhj?U1cQF~65Dd;u-%8l?;Gzzpewqo9d`x7?wh0cG
zbAasXj>He>i3~8iq*~%p;9cALY1~)SLPKXT+ZzDcvc{c<;YH~R7%Yh+7vq4sn=qU(
zn>s+yTJ{ct51B1bcwjL@OHNl)_5alruT5FwnOeUu_4*4b2DE_O<t2|h$-eK&-r~Tj
zo`yyMD?SEV^j`^nW-6y;!rVOEd3%N$LJ;^!oTvSvlFzm@=AbBMX=f+JEAh!f)<HtX
zwhB(WsP#ngP?dzaw~+d!!MS-8btnqb6t;MlVJ&ZI=|3dSPYj}_=hCV<0i(VnhF)np
zFn!(}SCcssS$U}_zRFt|Mg8q<Zc<T`)iQpl*7%$@15-i=TGJ)D4?@4qrA0E!2IkB_
zb=d*f6aHAs-v)`(F7p-;j(IwS5BYR&ygwf8Jv<!k{_J#PsUW&0rAXwudda?=q71ea
zzYAC;rnz%E(BBX!_>tNY{ZEY(Xh%3P*9V5MM%5ZPdT{C(z2K@Yb=MD2(pS?{iBs_>
z;eX)8S?X&tFbA&B)?8?z0BlL@5RZ{c=ALw_0BlJe55)jE#^4S;9NziMSK@GMqDoan
zhOqvp#BXYNDGFq`Xm>fLL_=L2iy_)uo1iTtWsfelXjNY?Vo{m5x)LBM#Ss6{)mN9^
zUBw8CAnC$e-VvuQaw8d*#^LrJ?(f|TXDY1|c+HSR9J;VN%g_^X;cH+y(DXHioc;{D
zqGN$5X`Dgi&%@}%he>9Cx7kN^qw@i*a$(BohejatL^4|qJNqLfLDRf{vY&GCqME&7
zs$i-1u9eGEf3%KL;x%-UAq)ytU~07X7p~~VjMTOF#o{-4;&-xT4pXa4hRcpQRTqoU
zS#EeRM#9F=lEN)_Pt&<KaPs8>7l46Wl&ZeGCD+{=ic7gyPSn`Z=OiibX2sF06w=~6
z4=1N6W^G5}q2T3gCPI_Bh6CGJv@U$vbYAo;gY_*(2;m$aPBw|$srMiOs2FN6lre=M
zIZUVn%wVZDtV&)>Z+mo{Dxmq~un~8$xj#*tu!Ryy9keJIKm$4I!~vh1IQ8k<xPIEH
zS$M<DfxIu%GHMiV`jik09PjmCrMDf|^@AHsU0?dDbr02PblSRm54K#5L9u@G4JJE;
zV7-E3<{MzCKNzwOf|f2Uz7ey!0aC&%a}FZa>gMhH@@tAY(3o@KTMX2K+3?Z47!Dtp
zoOF4iOpsaQVsTm3-?3k-aesGrbcS2N;o+h6<H3c3#$f~efDHkTGl6GdhuZNVh(5hO
zd@<6qeaUH)8L2nCFgI}M!YEEpV^M3$@Sz`AUhZPsvUhR`aZ)`^!fPfd+KO_hlf@c!
z626cd&hUV+L!ar!3#4mJTF1H!*;!v#WC(*0J|<waAgnrfB=fo=ql7W9G9Ur9d)Gr0
zju|A-aI!?1s0=Mw5jb6pXRyl`Z@kJM8eQVK<uP~wkNrn7yK1M7E%K8C(?&>Su=kkB
zeM&st<b(1)%Dmc!pzuf+B4A4vNlX@@*oe<cIT%L|EUWDM&AxoQeZo`gN8^7Wd;3*O
z8%PajwD)DRAQ%a=0MKN%DzDT&r12>RsQG0h9!kqbc+%{{O#0@wOTn~m7SY&&L7<mK
znG)uQ#yiDV20lI@+J0+kWqnHvcPiR6WZA)+m_+F*j4l-Malr#3MM%@fm@4`?JIxu5
zb{C)Uu(E~^By<qXw(v>e7&Ez&9-r6w78=4N!dx0cChgv^(ixVFw5IB(?)LP#C+(fo
zNv#ItEYE842>_~Kyyj#v$D4R@w2B$7V#F;`ka5D9AdTFC>2Cr<R|=-Y6fVds&*4h*
z#B@gDwZ#F|zldB-VquL6&3(CXkX4vbNCK9Ggt=^#SWE>{Q|~s!WkE}zBUY8*obepQ
zq4g|!QuP|?Z~+7*TZ!1fi$aT?dps=3bfk%6Y0FfEantNya6@r^2;o8rE&C%0J-dIp
z))eGrN5Dn;S(C{3zz%Ti)|SJUy=Th%vQg{QT}a%y?PZ!N*{Iqm7(4$ANRX*{b$xb#
z?T4hv4t98uIkz%H0&Ro`VaU^yF|PQ7`~_N2DWTSfPf4stOR`HFzTYq+-A&QtQ_<K@
z-g}HNhRsIHB*NDT)gPaoqh5`|eyoYF)_s&j#ar1&YFn;(iUWPR**_Sc53uqDplNok
zwCTS}BVznOBzmJ4F7#xyKbqjGacSaM@s{x|tQ*d*2GrrfOqdQ2W`N{<d>DwBRosLr
z$u;goPyG7w&_pj$M!j~?Y`$(q<(PPR?%(wi-KH?IWBf`tq0gm$ag%M0S%aN0phKvM
zoYGA+lXI&b26aU5hHv80mHJ~`S5;geI0T#_%;gU-rdF*VuhO|EuLIAy{w`BYMScb;
z4PY|Ay`&H`1sZGZ+0ZZ!L;I*<W76pS9eA`c09nk^QPf<IfL;j8ORR+ohwf#lO29B4
zF^-T=8=&Kp-J=VfFw9B0y4)D25tqV^h=OtEtLq|yRhLdD2g}H+Wj<?ydGQw*Rp@%e
z9L~)?#OUrwi8+P6vD#{2l*93)ruG?^bymyyS)|_6^4W5T5}9rfP=qPya%<=z`t^M^
zu7ZY3fPHxgV`hbcW|kpImaHm-kI9wheM6Iia6J<cNzWR7#4}vp#1%CoQv#eI0?*S9
zq=ankSD)Wl`b<-z7j8=dUwod?lfdHUDqgn1R@*@^j?ei-R;-!JFq6E0MF$%_F^*Ad
zK<&v2M%Z-C<blQ05dY+uju+e7n*?;xVT>q>Io`#r!#h=Zv^8Ol;uYo<9O5|lQT^Q8
z_z~J_BVfD<U=V*r#|-2pPK*4*K`Cndx6>A^YC}TOG)=G#A8@B<KhzK!oeu@GAT0~#
z7Se#GGl?h@^%kYX6pyWg80y4B8-W_bV{`yD?ZIL#-VcMjP&<@*@mnsW0b^mmme^qv
z23cO~N%p`Po+kZOqKU*z(Un1!>=%^7e6m%%D81XKbsA5Hk>+AVVQovHQQ;=?r5giI
zpR(Uq&Ev5QqpT3s(PWW~p$rOFY`K*#ra3aMiQ)JL4>n`mN<&4yDzX+1nV6e_XvhuF
zjV>v?QDBk`EPrck9tSU{q#)`Hg~-hjE;Xg&XZ|z|S~YpXN38*F=0gT$C-OG?@wE(V
zKO4hVN&hx)xfPqkV&;xv;l0~5SDK_ci|W0{5X6o<X?x}T(<FNTmh07^J(rD+K=4FO
z6QqNLyUeNtQDKAJA0;wHg)RouW4p*InH{|V>uB)WREOjnDOJh-9h-I5k4sI{wzrhX
zI?b^27ixC&#P?)jq!C#?H1_8}gYyZ9k(KtAog2;F?poVN3+$Wc-Ec0BTYde#ZoFNd
z&fQ(Qw9*}O42#Z^-6SQx0*q)A&MS4~H#$b}Nhts@s0aSH%7A~T_X~GZ+f~TKuBUUu
zf8@xtWkd9W+=BW<=>rrp5pEfsXvPWO@iGaz%Th-e_nw>C#ls`Cdc`&!W#u3&DioZk
z5LTzNWFOZfpcdM{APm)f$#0zU&(ghKj3w_n%Mui4rDptxWE)6`q<1Na$3d}fK7J%o
z+YrdvyYO1bS#=7+Fn-CrWS4pm(B#nPqFIteymZ53_G#!;14jNmAn=2-{zwb-VHCNR
z;?p}{ib-<8!gX4V&@e}U{33e##BoAb?rQtuuh-rj0>Q8-?BQTQap_4bm9hmyHjgk>
zUfBhTQpsS7qCo4bdYSyA=YN0p^!fcqLDnIm9e?{IjIoVFiW7Sxyp?x~H-o$2Ft|;5
zYYOoBEvDgB5d?y~SV*k}1_t@H<mWqdwIqFi2lB83jJ56sfW{d{GIq%C@67D?QDW5;
z(SJtsOcb#{>3GKm`*6|o>VRCbtdNg|dK86?lllpIVXKkClJgIVzawl|^Bs7QYjKl2
ze+SX`5lCT9030*AfPxhxU(V-ZdNdlHE#AJxe?)f?;*YX&9l;?j+_=G#isSRQL9Q-J
zL3U3n<0%-(F&`JgZ|_tN&B(!n9uJF8#|J0Kvzm;SdM^+I=}iws>Y_3&^(1P$_l}1r
z*oO?xaf0`FyoY*^z2W&e>@j?Ro69B_Cnv1U>6x^>;Q<cr$qkV7YM=llmyk8t-kx6Y
zZ(N(KwWZKAw9oKo6!A2QricbA{~a{&E!d(~E7hlhF9d%s{fIy%vzjOsa$)M8r_(;t
zs@Zztmj^{3BL#6Z`C)i;f#gIdZ>H|)3nZjN4Ccvy4no<l45peW|3LC$%GMx7(mbZ!
z)K=v4uxXW<T1oDhEa@yGBEe5>wW?de&?pVFS5ooFq3s!XoL!nTr&#YG<BW=%0CdoD
zeJEdX?ysff;(V!}?2My(dU{%lLsm>BO{?#^X%3X`X;TDUJ?I#aSl~q5$6?mZ&ZxH!
zyWp)v?Rd$&8GKm)?+y0)Vr1^A`yA|u--tV5v5|T!40k+y+e2rkDEtP-2MLZ!2fD$S
zP9Bt}UTpURRoTxaO3;Nyj!39jhih!{#!@guD7u^yuVu87Y!1mELbF{~RGB>M%nu-(
z3h-yjQM8-rR&uidllxy=;g7Ht7VKaV7QmcC^l3^g&VP%<O`km3AGmxTY+xrjk`f_}
z<>kK8Wo8i74{K)(BfW?r4O)rEokS9;<b@g1?^B#ab&%v9hasM(+Cca0U)=hVvIElZ
zVsc_^MAMAajhR<DiDLp1hmDg-r}*ffMl<==km9F9FFAGuDPZ|Ow{rSl^?)>(s(SEs
z^b~aQ^&t|*jj-0E3RjxNWG3d7B(<ybK&j`_fJ2OisMx4Rieu6C7D*{-CweL^2o^9<
z^a6U<jZAk}sm#H5SD~o=@Pb>2kfv)-v%dYa)Weuq4pk_FZ$}dsd}uXZJ@{oE2e&_+
z8IVRX=uv6Ff#bBB4b*F@VT3I0ZSM!oH=YT3h7RipIih03+IVW8idALd%m9c5v)ZQ;
z8?;O%H258-qBQMP>BIz5i$wonN?WcKd2ukA<X>8y5ceTP<2aNje1JlL@i9yt{&Nn5
z*<(9Hxe8w26*YX)+H2-xO}`TBN$u41aSOX8!^tx7lZ3L6*|+XZkToQ4m1qRC#u;N;
zlinUP5OHD3)noy?kgcyw5V{us%KSq~!;oTm?KdfMnvs7h^JQo)4EC0D@|}+Fy$5%b
zfr9^&Eem@fn!SF6s0#LsdPP?N*ny&;jdK~?oiUP1r<fb*!Z6OvJGQm7`C4#)*5nZV
zb;oF|T%xW~AwC<9>4<77W>R>*rE&nZ+-=+BYmi{DxL%<fIG3^2I@-p`w#hd2oTcsb
z1wsvu(8c7iXXWkdR9Y*`fGHy}v~hE?T`=%o2(dm?ey7AxgTr9Rap6$Ri3)2z2PVMv
zCJ&Y40~X(F!oC#E8E51DUdw%nJ5etymooCs$8l83O4jKxUKF};etHgn0%$Nbspe!j
z{g&5*q{L;NXGoTo@}wrbbCt2~;{I)tLLLX8ypWAte3Ux3c=>j2UBq_g<&N3}hfH1O
z&!fRYhH?`>CB<>B{J%?^4kAfW1#L9JeWar~t{5#fq3Kg>juE}nqwbdm0wOPz@05bp
zZB?y}hv&bF_0h%gnb-H$zugS7D$U=*2|Rr}#f^iJ4uT={clf=T6n{UzlNA&ZtRe2P
zy~`3csTt(^^ptivHpvTSy&TX;+k>0T9N`FHn1tr}ofCJ(w-vWWUB^FCM<s_C+#US_
zUpd5MG%sWj4|rJu2tiSL)UV|HJD^q1XWUS@Y?SU|T%q;OF4pC8ZzT`{im)EE)|MPD
zphGgUm%8-ZH~Vo=D;OJBIEiI=G1=HJy_&-;gsU?Zna+F5mY{}xdby8Py{GdAdsX+@
zMZLf)gNH(DPbJd7sF7Um3(zu3>@YNCiT!{}XDA}Yk%m<gU_!n?+Az>o{aB1_QA%Q-
zjM$&<u66wb9BNpFPE+~|WLoSH-hi`k9uUOuj;vsZE$j);rub0KEzTk-5LlFlemr8$
zNYuov!6#_plB&$q-r<12DB)C~twc)K3ZXzJZ?7s_y&U`#RM3NYH`*(N$t+=$-li7%
zh?E?37EpOr5OqfwI;VJP6^YfC$sN_9mk$df#7odoT|+dm`C-9uOufcgoMhV>#7P)F
zW}#edT{4YmbUnY;Q~9xlP~#YckI-o%Ra{x=RIKZWr7(knOsS^4U_1gfQd?SVhsM@8
z)ndZbUZ{Dts?(0CVf3%yVq-ZRONd8WA}&um9!<u(i>R#T`APbLo?AEl6nYaViiQ+S
zW~?@oB~uB;iMJe}Q~_&(s*3Ez4v`*n7ouoa0c3zD;+--q*TY0oRlrkNwfev3Q;GKz
zCCexuD>(M9n>f7=jdya6_&ct@JAsxIR&aIL+Z$oyKW6S*?b6Jp>su#VC$4d=!}Oc2
zlk1Y*zwH@S)<*OE+#+h%T33*O8K*ro%FZ5k-i1QD?t@vY<~Q}LVZOV-ZKI<}ac7CK
ztT*!iu1Rr+qOIK2;`Pj`?<S|a7-aPqBzYjK!G(izcH95D=r})|EjLp2rP#fg(4>xE
zrWy<*1}US|tq_HW=KKwqA&B86g5E{*3HElrrdd;%&7619+(LH|3#AcMk!^8^>_&B@
zrK3c>j_u;ix00S&9Y0O)n}|HR)p1&PdN$g{MJsY#_rEdm|0Y{X{-k7{%8dv#mw9+L
zVP{W+Q9+XlutZ1XmZynRO@eY_T3H$WayC5Klbd{HZ=!ScnFMRlFl|zDEvGjY@`u2(
z5%R@#xvH@QTXD(8z%=PSLe9_Nu62+5AYnWk|0pY9BNKX#n5Nu}ze0l;O1320*k-eD
zFD(JI`X;?<GU$txTEvEI=*nXA3|-T7MLYefwj#MVlKbb=SEZW4;4ddf-!n}xRCLE8
z+cpAZgDW0zS2HO{>TofU^3jv^)rT*hum9)v#_HPr_51h}IP}CU8iynZBAK5ph2n>P
zgj#or+)r)WcC+!ZY#2ezvGj~b;^i|yz0J40I+WB9@7w0cTyobd-NEsWFFS9==DhmJ
zwb)W_?HcL><0vq+mfoF>>_g<<GYrpjFtAnR?6#XAo*|(3Upu^;{d2#rY7i)nm{dZy
zN03djb2l=zxrU<@pix;rV<m12V;4Dp!8=9$VmuSJLgHc05Set&zLKufH>l9u-y6Li
z93zEcG{G4`(CY%ygwBvD#N-F{_``U^?Hn8oR8l1;JVmDT_#8LVAVA4m;sXkd*XgOC
zx*9F=OgA!=4tj%=+r7a6nWpFGZ@>BUG2P@kw9m$fS|T>Sye#VMi^h<n-p!-m3A_5r
zL=3iBT((B-h6$BV)KnpgU9cedKd&^reLOr{46RXNX(;SzLP9TKNwZ2T?1`L09?+Ff
z<!?^yX7AwWbjRrtOn<ZY-P7j}w!go>{vW+x_3!7;p8hBL^CK3QbLwrajfS))GDE86
z-Z;h0wQH42#UCCHPfkWh>ypPq>`2Bcid8&(eE-?A)u-FfHb`7{`1lNCF0n$z$>@A}
zS3e!xUb+R^gT%mB_n$u9et7@I>UYoApR8_le7twQw~Qi&=S$eJGez~_lv&}SAM4{k
z{Z7}7?I+K^eLhcr))^}G3MsODM(CB$*(I^_!e-q>$k@mF>eJQx8*|SOj-3$#zOfI#
zsCM2t84#F9q;O_5;dez>rAo2P0EcScpNdmRWgdQ!_oOdZe=<lz1M)UYtQT}86~p2J
z*6U!*94HDJ=~2xWyMW7IgEi!MfSgx6QUSm(oKd(BWrtA$02XoU+elDB-j~+h)Z-ME
zeBqzV_DzaF(n{%YfxdvnwzhhIy=T;mH0Yk6$E9HG?V-RB&Iv&=v-D<pu|O&cs!Tze
z4#)MUJlLJGjK2+Fix4N%@n2;v4eLK1qb5l}gJT@!&3-wSym@gC;qPsV&ust+kA%yf
z?uEPAyS^b`|M>mW)gxRJc5;6GX7A~f2l&1H?D^{Y`tx<jGf|0xc~ut}YhIL3v3oc<
z-`AY>D(ZD&f_)^J?@UgA9-S=p7JU~BNVrP0O?v<R>GZ#EP}m9%M27@}4log&m))Vd
zAeAN1jh6|<Cy*U<PPkV_59NfPusDWb!SLHf1-X>}y0$hRJeNM~LWkgLP8pPAk~eg0
z*Mc0GPYndO)RO>01jgCA1P7Y&nUsuc*u*+k$9MNRzWA1NW7@#!_;M{ckQ<|>5MOl-
zFQW{zrB?kHTyxCNX=ui<&<x0oCf^UwD1QsL`&`1;-Cb#85zOcid#CC7?OQqq)|m1v
zfWkKls+?~C{cH2vt<7)e&3z-N?J)(6pw7E%{q@z27q^$bkYwJhTiD|I?oP)$R5f5)
zo<MKlgzs)8?`&o_uPO5?t2EF^_;gDIIx#6{RGx}Nv8%4TC9(twaGk<x!Gt^M8xv&h
zP6)&sA}$Aajg_p8&~mfU^y1(E3o+U=n!dI$y*T4Zn33n+h18bp`iX#8t_?!hqg{Nl
zj-G56fRSt-`Z9U7$azAvH3mXg8cVG7cWEgs5|}$|J$uyqbc*9dIFCo2XCu?^_?B)-
zyrWJ!-90^;G9;%8nCgP{obiCqn&7%GQE?Gx9B{K)!%VW(;a!L(4ju(}ry454dZ_Ux
z^ijdbX5UVBVV?#{F!_`)5rsN-wQ9SPKrn@r6!5b-G`e!3Mm1ez%2rm=di=3vK}poH
z38!;_$na?L;~GpO0)?}MyTI_t6c|SL`ztHa9_PHGr=fv<Nr12|q-@IIMLIux0TRSY
zTtAtvV7JWI%lPv%{tsD}|0UwUP@8|FH~O3YJKu-?)_-*!3v!NTpjO=Q!xc@TmjOV%
zK28o`@|~t;unPhft}{n}kwpOw-YdmpyLo^O8c$jJNmW=>T8T&7C82abqF_^3kz+qV
zsZJIyr3i$Y3l{d!ESF@OH+pw_atDhMk;+LgsnFGboK2iOYsJ`jPLy(8!_^IzOxgMA
z)0R8AP0Y%F3dI+7N~ijBs>k)mxa4*Zju_mWPCW=}>{5ch;po(Px)~{Z23?SG*VJZ+
zH@3U?F=NU?537K%Md$V;w^aDr`o#vrn`IrATwec#oUEh>Xq7@p{uJfpWjsWc!&v7K
zkFkquS7-(agphTB&iLnCBBI#d;?pQ5qz`3ONR<^{M@()!fgl$lVKT3wG1&kh*lP+;
zOetvMfW&~qC_dR0p;zo^YCzJB;My0(nLY~Hrj$+a-!_`r3JKz^N=|tsKw#sBW10Q|
z+6^CXRf^5NP=`s+2DW9Db|b#)4NlPDXuON7rYqJB_CPqA4MXM8Xu3NY%O$v2vM7>d
z4D4IL$Q=##K*TPmA`go4462LJ-3+v0zYH)uFQLGrWJvrn7er*mb#s)ot}l?6kVq$u
z%<Su(lV--a76S&70u@cX)>1na!HMNgo{vjRh4-n8#c@M(&f*5)4;t>tXo6fmxceGJ
zVU0rV`bsmpoeWLJ+y0exO;;}kw=IDYP1$n?0R&Oc;{z_8bIF2V`Fw83hQj0Wl@rkm
zqhDCeAAxfUJV`hWPt4&Ymy~?+O%HdJu#LtE<98%J`M=*`6lXpYpO>cl`A}%)k+|<J
zuYJkSXXl5PVHA9TodhJW8<!NqOXt6wQ~owxfrZa8YR51zHB5wDBCMc3Rs;_rA?ubc
zgQiZtSr;p2aChfoe6;r{7&k(-Ni*pheN-0as@{rVlr=3f+qVDz%p!ZJyRs`7X02a#
z*k_=Vpn}A={Ug1sfyUrgSdNhi_bpt?7|S9dU9x(&l+_-^Xcz!ykNJB?RXUg9-a`$j
zy!E%qtZ?RDf-P^^bH);4yn2N&WfTZmZ|+^;6KLSe+HG?K)V5p^46~_sm-|cDta=uc
zzn58|IMSK<882zxc{dbkM2^@>^K8yF<kbvUbuIeYxLkFEnxZEzC^k;hk)a=lu4u3a
zNp?xA+V^lF-kJBE3@;iX`90VEqcvzpfZI&oY>>0T-7f35Y};lCm+fZS#N`KF=5g7U
zmU7e+uvoBkVIvqfVA*GJBSwW-l0npG-L_h6SJQjZ;#cc#537&>inBN=mBqd;s}Hle
zZ^mD56KgPe%3(!@xc!#lD?D`QTF^EW#}#v}#E>&naLTV}xka?6Z$6qi9H8FxzmmDn
zV+nh?O1C~hMUAesHeOc?q)2hFN8U!3c+qIlZw<ephH@gs6CVrlx0vxNtaMv~%kS?T
z_v-3@FDpqcq&PDKhvC#mj_f%!hKn^ha+w6=VE9?1^kMY;CQp+&f=l#Yr&o(UQmV%=
z1b@V?@hj<4q^a(9CDWIN2fl|?)Q7Pbiv9R2Hi$h-i}*7%Nrq#S<fm|3IT=L%%x2=c
zR$_U{WLi52ldpXrUbcO{raHQ1{5dA@0ywOUtf;-ZUsYDWBsg+LN@|QkjbSAvG^F`s
z!z1pG`L9HI9__R5K`7z7giO49E~Wn+mO6aEx8aANxiTzPdff)AAS-sRj+GK<eEnQD
z&6ro^^KA4Fz~M#{k<r!H3(V`k61=uen<zx$(^C1u1r4J!eMK#i__yMKA_5ySlgQN=
zoO0}V`j|Bjsn_P|&-lE6_Vz6qEOruQv@E7u**awL@tyKxUw^|?B~DfJ&=?-*!U?ML
z@h)nxj?Ql1IzN4mqn$%!t*3$FJ_;sC@T33DHw*!e{!jYrK+`k(cjeu|xxTy0hDY}5
zYk75KufFCh1Ru5GC*M$~CmVHJ2|SBB-H|T(96z4mY^nQn{GJYd$^=B-M<aJeNLLKj
zhQwI=)>?rTo70v6(=|Df5iV+EGs+&YpI3hgJvl>gP-Iv3-y-b`DKpYF;q+YRA;7wJ
zuV<Oh_;n50oyLrNbcVz+3_EEbGAn)#U9k_3>-m)5`!_@&K99y@1IpLx10;|rxU2Lf
zdjb1K2wTBTngYVl02%{#vHOQX&I28bMjQ=T<(_t3<;FWj)Mo)r1_P(}nNO=AMlmxL
zK0(DRdXyqefN}a}jIbeU3*xCZq*~yFaqA`mow!@*@T|Wgh)NV8<{&S00TK<x7nr-1
z#IqiLKB)H4K)2~qLwA5EJci-cPxHiV{qy1DOirM&d0s06W&|Hvbvk@eo6d~>PN(Qj
z=Q$l>t6J^5+Hry>999}4(0ZkBp^qqWQPd3{Spu6Z6IT&>k&$BA0_SFLBuUt^_>jCR
z!3SrF*gzjAJ?n!6^KQJlwy<FnY3t@^*i{#Dwjrq@XRQ}dRS<AdlV?v2g+U3YFWW_;
ziYMgb*z~`foL-!{jN}179X5Y0ypbFW5cCI`<Wfw!#vMFSbfs_DM+O#~%jl`JPQfF0
z@~e_#*+w8rx%IMZo`|+Pw@k3a)eq7E=LCf9@t-~yJmai0wNs4)Bn>6k$VM%(Uj!|T
z$yKMwRKSA@y{}-y*M@JV<>266#Q6}kXdiynn|u@iQ1$!$5CE+vSK`3UA!0}*LL_Gv
z8{xzYq$lEsk`ad?<!tu^Xt2czvc-ODWM*_U9SPkYkm7?*abn^t(|3sp$G6HMB)E(z
ze6eIjV>@C0*>o=52_)R*pf9Hj-1>Tsq_iDe=;1e|t1H7#uDKlnKS222PM%a>1vr6Z
zlbq41pb9Wq$<Cg*B6AZs?zxBrf{Saj)sFX880BBGYbQMR>$BGLbsC5h1z~wRrJM}e
zJ5jD~!k&=IcV$~aS^I9l_%_?BCGtP`9~uf^Ue2_AN|P`}Habo6M_6VjBwg-8vapUr
zMx62~^F6RwkD|fMP7>XR&ik#_Li{okEWd9ERZ{sZu^T%Bq%IeKA5Gq_$J^I&J^}yd
zS;b_TH~YcL6zO|rb8&eTSK>o2D@Cn|pmdPkuwyYH1RA1TCGo{~!d^zze>7b*>^s>h
z)yxSosvzT8Dkk4d-<fkWeLOzVjj-;YkB7hLPu(*kUxC*f>LcGO+XgLnqvIGGPV2a|
z*p|{BQ{*CZ`wLYD0M#-n-~u;t*?1<9Wp{YaM|(>1EM*g@?hspFZL`@Y^t$gYUO3?T
zQpUH|YCE!a3`CF4PWEV?AE34W`gbK~AQ3Ajb{sw*8+0Qzy%ORJzyX(+0FJHa1n%qE
zz{%JSTxKHMN>MnUA{#$ewx!tTGy+}(8nmfZm6i30BHw62Om6!djx@=-%@IopUeLfM
z;E?@v72E1Kkj}`XQNueKYDn^G-HI}!YIh+WQ_bNe3Xn@PcL<hgnz<XppmI-Oo>2Vy
z-s0BYyW2Ov`0d8x*3#DA=Qr>m|F7m6TQ2~?>miCxw#QU>_I&g9t7}V(zW?`!cR+S<
zv&5g=@8i(qjY``u^0r^R>S+6A-uBDT7CO~^hX$j_Q@3y|@3;+vW*D0=lm8HZ_Vp+7
zAa$eN241ukNwfK$#W3}~P|PJO=v9tqnF=di@}}sh*%Gh3*^tFk%Ti3*mD!5Pt0ZeN
zdFs(EI>a26MG|u6@lU`5LK3j>%abN~t4^M}iI>Un_~pZLNzBeldL}I<KcVuSi{&(|
zYyD*O(8OS14*@&&bqN&Pl+Qvx3Ak*R{0vQe4t(!_3xxuYrHB90fvS2RMl(b}C5(~b
zqcsrnC??ES`p72Q+Q&w*zLk!em~HLtsb&FJItfeI`t7<W+a}Cy>z%I22MFreNF~2w
zN53;Z!o5ccQ$cUo;R3cc)+$LNW2?}Ma7y}h!XcrNP;MO8=9`bIRGaqHL7R^<fGZ1;
z_ubLchJ#h<B*1?a2?{J`M@2fTEp3dBp)&H4M>He!`42(;zhBj)^>F@7#QWAm4QOlX
zNl`_~ICtwI3XvR*&PVpUXw}lAQc*U2xt|SJCZ0)ImdRG}yXe<Pe~K%widTlJ=yd~#
zG2tK^v!yOW)&?I-&A05~x5Hh8-@qa9pGKzF;1r8<4tb*Y^E#@Rs}1EF9Uk*bcybIr
z46me8K=@_W1<XOAhwvOOzj{@u*Wm8ciJZB_oVplWEC#9v?`dZ7r<=WEDWQDAGJ~nL
z*k`pOg1x20fc=Ow@GC2+kg;viC%O3;?)UzU#%To93p9z1*m#Vgyl?<+qiVbs0l@`$
zz<b@kL3Jshb{oI7@V*j>kD$9X3whc%aLA$BkY1(<FMav)@N{~PZ4hBg%3!@bL9Hn{
zs`>*rj;o$2zu5Of_~^ZNFvjvp-K*jdDb3mDRMbp25zFsU?MMqiG-eI|9_mB{`(P~T
zkEA<SNTEzc4^?oCDOOhWuGw-$gM%04!I(?~u8LhJ1PFL+o0`I-4F+=T*8+C!ubFWr
z*0jNy@dNs2Tq={9^vwC30%fCNf<2*tk~1WU!#Sa%AR>tR`Bf`e8EobCXHQsh5GeTv
zF9|FbpKGs1Ce|3!qdSl10&V)4X!oUo6hTWnL}K<TMYUprU;7c?wCDD!i!ES({bo4f
zy@q#7kfqjf3U_&xo0K8i<JpAr{@C)uINKKy?QYbUAtGw7QwT`YC5R3sQwELd95dPh
zcH(aS%#xX9J8wTkxafzoY9h5|f;7OgrILlnL9A&1upL+7FUD3L-^#tv$&}@^O+bd8
zQnG113dg-U!3GNAYw07Ju&JlRPWVdA9nM8$9WJ?icMaLolVg%`IK)qa%Mg$90Nw_^
zB{g@j7?JeaL?y0UOxC1ipjdgEltr>?OgP1gZc{*%XlOw2hT|qR>J{OcYFt_QVRik%
z^Nm%24e6H@dGms7hI-!&Diul4x*!PwqXkkU7KIc<JjZ+h5OPP5Jz}Z~$WPQ;%IOQL
z!_dH8$o5I9-I!h@I+#&DtYf1I-4P*vDy4Body5DpA)q=v;$_IlFtHARTmB{a3^<AZ
zQnc0ecLv9AICcK6Wri6Zep`m1Zf%<2eeuR`gLuGx>HOdy*Hv<ARv{)X_K&CwD@75!
zx$=;D4SPJX5n@xQfg=<S$eIeu=}1gjT8K^8PTY$1)yw84)LoE}7;@yHSu|8~zkJB+
zSX;1~p$c@Pm2P1DSVAwxq;8E57Bw5K$>IFi(GnsFwB}=J!(=eA7&2)skm51K#&8<P
zOCy~(lhd_DGmOp+IB(7kZb8LWVu;+h210e!m^^3Vi{u3j!1yf~WMWn0oh>Z$X69Mg
zYh6nVCTIZs+ngj0nlAD7&$Vz-0hvpL`b2oA-1nP4Quej-_;KqW=|)f;UnHbA4;#e~
zV~;X&#;Ksz)`rP|TE2BR<tbM-s?$;XmAj8Pl;*XVgC;&!2tq&F?-B2aRxa)0wg_kf
znQ)=Ade>|iI20)oyXgiiO){#_L%ZZMY4Tghe?k(QCrfLLgEK6$R)9=P$3*M!qqVuq
zCFb_{-<8G^5xFEOh6Ed<MM)wN%czE`Dd!tUs5TK3tM=wkN7Ak^Ep=w8Px%=kPYBJj
zLEP8ZK|7S=kmN2;gBZ(vQf(?Fn3P^S3N#&UOYd`VcRD&3@>8LIG~SsECvRn6e5P9f
zX_C63{oBGYPMb)P=Jb1a8mrmR(t|fby?+St^mDk*3`$nyYeJ{y6N1e8NBQ}ER6(Z=
zZ0C~t;?6=5Rtah=NPXC5<!4V$Bsk22nN$B1E`ze!x}S6iyjr%v8Y_!(vFy*X&4w2L
zg1B>1DqF0Zp(S1!F}c_x$xd{X8Qpj|PFsjk;Lfr<Yp800o9?Gt2j$pe@E3Po5+i0Z
zvs%GPTQa5%d>sv-jWud7Nj{yeVKu`%O$Pc+K90S?`-ld~ScyqMz_k5kV+SySGw<h*
zd!wC;gU2X>e$-l|p9pJr3$H2PY}tQ9CH%sb>&<@94PWm5F54bmpz!I|n*67&Lvi?S
zfIm>L+G_rr-}Y#D4nbSH3+sOFyT#IiKvA`8?dkm&-#%ae-b>e_{OotDKSBbbCf4G;
z6|5}&cktDhJLB_T2jjEXU;R4x$NBWvfgtb3y<ZOxUmyR9CRo&8AyaM)Hh)@v_4$nj
z#J0I1(|=Bl3#SlFlQDu>7$9sc=JC`TwItt#x|WqbNx4zs#0FB72n1ASLYNU68g61E
z5bCj<kQFG)O@8pV$P+8GS*_UxhW)MT?RC1RiwYvYQk8YSYVa=N)tEAkVHBp%xRaCZ
zS;U0e+F_cabp03k*|rjiCtAzH)rkbs!aw%(uF~~u4&M8d16=uXa$a%$VoNC`8?GQB
zdU<;n<>&Zd$={2AqA{G{&`|*-_@8cK2Hl7Pw?=wNB|u75BP>H;!SCr;vh2te^&zs^
zqi;#0W)(K?vY}4<CJ1^43d)@IUXO?6OdCpX^%opQb+uHWMF>%@7{kags3p44=CcE6
zkl5#bb{cao<jl&-+S8ZcJ$bf$f9=UO96cLPo<GBcY<AuF&AKKU!fsw<+5NL|3y#Aj
z1A{YZIpBei49~N>ZsMAym|fl3G<%qpAXnKS#Zm)SOu#+J+33XC<F)t2&HN6(a&Y`U
z*fwi^vHveWEpGnw|9(Dzz+L)m;a_;zZQ(ZlHvy3=ai#SWd-RYcT-6tnDZ)iP(2z1i
z1}R5{>ZmUZpaX-WASFM+T?1^~ToQcFMW6w~={#C}@bWv^4B&+*pAN`WJIx>>?`JL`
zZ}xEH>kxMx!DN7fKIl!~9`B%VX*=+$AVolN4#)~0pU^|d$^6%g4U}boU&em;yr5aT
zW`)<Y=Y&R=nbCh?quP6M$f!zhW4J#$*DXPHJ<v6l1cc#>^5ofz?Tsh@wJPS`zxph`
z`m$>hVoJ=O##dj?IScKtm4?3hQbDF`I9rGN-=gaO>Wi<i5b$!jHHEMr%B?N-u~z5X
zXXiK>1D;x0L8DLL1=i~DP?RL@CoqP(CQ4&z_B`3koHG>)e6og7tt@Lh@v&E(x85?g
z2OXj(eMn$%Bb&n-hgJQ#nG4Za&?r^%p<7Q}XQFa^;s?E`8~u23AL%3~d&9{dG57Fv
zf=!%Q1yEc`v;bFk9HDE&-_|;QB0`LlNMSs^0-WobNfvy8osfJio}nL9^NQO#5N*<l
z$v!m*pj8yqyicbPPAY$c(O?Q9i*^w>8=bsf%r*@xE340ba9xnyT?gIL4nzMJkqxAc
zY-$b$Np;R;$U7i5HDH{ii{W9P$`D=(&K+SN2pwF?s@wm+#*=Y;%cWJ%)Gh)Z^c?ZO
zhGX=^U)zRKL3uU-G)UVgmR9g&S?^;|!)tb5e*fg*^N7;MS10EJ6Cvfao;-YsO-9Dm
zkVsFr@!VI1ff%xDVfU2HAl1Tc&7>~S;;F}&tHcQb0a&nN*G_kEB?8i)1s&nr!QC;O
zx$-PrqJx&?yj{c&PxuDLf2=^f{1c$=8EUydSzP+;#=`P3sU+Nm;|anwrtk#d+c>)m
z5pCzz4BuF8RGX?H+%98y^!&w()kluC67e3@sw|ky!3>gZ%FZ2Zidj$@dqN!WKiSxL
z`s-5=+X#Q$xVJU^Y;p6};NO;Du`F&52mg5W87vpcgASgGMa^Y)DPuV%GQCYCJM+2u
z9ke0RqGiTl8=-|Fk>($t$v6v=N0apu?&`s1=CpROp9WQd-|>e5!2yYEW^+X@Sd^xN
z-t8N`&w6*h`YWT3SreQ@JDs4!ybrhOA0VfC(EGQqu=eE?miqX=tQP=vv-f;$^%*ed
z3$y)V^~V<^oiy_mK-Zd;i+F~FmKp`5D<})MwLx=|_%dtq;rglrfU$GRy0{qcReL==
zb2-tMHjOWTL722cP$bBWr!ThKC=!O0p8Ae;6MA=K0WISr@KCT$E4#r=AON(vE%IEA
zfv<r5s)3AG^bDK6F`z0UANkNb4Fe!uUO-ndkMR}3R+10K<>SXceD6!8qR)_$%@-i2
z4blI{#aToA<|Nm)i%DT560Qa$c!fTywq@cw%GiztVw1)SiB2ViJ-g+xw<wnAjTWPf
zoydY=$?&gC!Adrx<|PQuzu=RlS4)jp%PdoS#~mBf%xh(S4Y84)L0=+n=I3%TRr=tl
z-;A)ZHSvPw;Uc-@O9+C`WLiEH5~vVONaFy(y@FRza_KAFx87Ssrs>7-NSrt+WNWEj
z=FRMaJ`;{4@0l(Q!Db~1DYL+SP$N<0Kv?T;n-IuGab9>{lDUwG#K3ajz|~07EBv<J
zB<1NomGJ{E-BwVXrg$V4)hrS|6D6dAeeZKQfh7L}{cOjPZis5kBm@{jKC4?#soo7{
zQ}#U&)#tr!NL#xaiefN~#2YD95{ZmMgsertQaS)vTpr2W$bA*=kYoF9PDdXN%B~pC
zXl7*-kib`v2tf(^5Ce3Pt%UhQs&lY*S|$YQGYf(7lz~;6kDec1j)KS|rgMT>nP9ox
z9e=PQvUo_y2G5>$=JP)kA8irBDDtFdq`=RUyk1<=OqSPp2Lfew(~JR6390`~z``_V
z$*F*DsBd_cznz>OZ|FT{Po5`&wt43-5Deqi2Eh*)xTkNvMd)Sv7FViJtEf)^19Kn3
zv#nx6;uFoO){AhkSdsyz{KOs=b@<+*3g<{R5NB?1hjN9v3A2nt2ZC!IPN81s9i`W0
zP#B<xvggPEUc3L|v0MZQyAhTo{J9E(+-<7tLhN!zIlME|pouF9av}kmM9*Wt<Wu%+
z`L%dVr1*(Nwg^|kqc(*W&hnaC@^Y+!*!oMvN?ucM4U$%}ySqF-`;0Cl@6Pz@e6k1!
zm|c|{*BcpR{jDvog9d{m-b*@|0}PI+M8gh=C0$wm71k7cn<R0*HF|R06XzqWy_f4x
z;k}3tZfHjR5}bRu_l~2-tC(oSqP%5q*n?{WR}#AWJGp6kiFaN35nQff{~h2ZabN6U
zi<FLZvxh^Nw}r0c)}4g7owin1pvb2q{!G985c39$)aQ*FdcE`0-p&Zql>3VX32;-L
z^vH`neuz!}u=np9&!6>9cm4;Pew($+UO}e|IW^oUU0P%d^S(d0R7pSpcd0mold=?J
za$UKx;|!Svr2G6rccA|B(K^HuZ}bj)Z5=5BgdAoF*VxTWu_|IOJ(-|6-jt{g)$#CT
zcrXH~WgzK6n@fu)VT`16Ki#2%`q;gf3A!oO98t)u7@=H&bTS7bY>!HWP8SPQ00y0c
zqSa++cq|`{c`oaFoj;HyyB8Rqo{{m13tsxRlos%PffMcDu4wlHlao-(@DtwMKBv5a
z4MZ7oaUw2SNC`_M9w8=zeFQivoun{TV;bB*z#Nphe2f(adz!`My*HNFP$fH*03UAw
z;Zo0JZzsoeszRFiLRJonjgWSG9uEa3OMQ%+aH)IEn&5R-_BOHuP1*ph6Gg!k3&jxR
zTsZ{;e2YaJJ9v0BoK6=rE5qZY@N~MBAqOIPtRy~?9bj8`uXh=mg1=wWo70|t*qzSV
ziR}#AIx=f|-p-Ik)Y?;azOi#+yvqE&Cx%XrhCA=y&-As#quvk8`=`_C_^1t-Z{_XW
zIG!L-u{&z(Z2qM`Jb3>;;DjIV!_m;z(|vnC@8n=mw{^u#!Mqz6K43VPC7$)9@+gYW
zR4&C%5@{b&n<inSpttt&caj1H{S@OwNn6S5flFufOV~S5($k7VaYk5g2w<x+6QeD$
zINQ!2E%vX=*wjl9nAdwKBXDuFC&m@bNamLDO3%>dEh6P`D&brolFV-QPEU^B!qA#x
zT$rIrt2eOG<lv`j2iYye`jKonKZNT+c=fKs@hbCQ>R*TDD%Rh&-{v8`!beh>0~mb?
z-Yf3K3_tK)HQ@`kz-C{zZI?LldTAh!FOZ~nIDAbjmV+r`M|-ADx%-F40L~DQf>Pk3
zOJ+fG_|y;d;zW`-gn$4hmC_|72bAvlwU>R6yVyA43zYqpF8_M%%^`OXeqf3?MDak5
zZeB$JWC?DcJL2}?N?KQnBH260w#zm_Vaw$GQ1L<4MRgc)<z30HNp89fZl@Fy^dR~!
zVK6bI{BwoG=7|3&wq%<Vj0d&x^x6CQQ%qe|vO~>^zJ?_mO(3yGiGGjX2=OxQO?yzq
zsFgL`IU2#q0AmmIuz@?~A3omx_T|&33<zP~2w`Qfp56Z*SsA_C5D~e}9*j5?+IIJp
z{WlKn9y40N{;%{{XFg-DJ`gG($i)eBehDW@<5SWBZCxL`92g900K;(eMu4(xH-(bN
z-L4C`eQEL*z#Zdq8)6FIyl?98$99UflTPNFYw?}=ij_Vx(=ebS*GxzbLC7}pSz3z0
zrD<{JO!!n7PuNQ=$~>hvw93jmfIOrKMR;3St?BUlyy&1@(zI;C?`VdrKetggK&Jze
zvVBGH{1($<i{epess=xeb+;O1^~#CQU^XmiGz1jdusmEpX6}RsZ`~GfE+>5P?cnZ=
z!T7oDw7iF0gldQ+?xz~dm<&)80J8Nt^JD4mAtkF{a*_^SFqyn4vUhH(Kz+2?#H(<L
z<S6H!EL^#2h(WD88@*kOe<p%9;$3FlTsu9n6`Q4GQjQ1h&cjA=pn1j=<82#v8Xk`K
zVL>z@$dIHca0$~`*{0^D&Q;9-*2f=WCzymcr@2=pE{1+<a%EGlvuW5^t<KSYoWEL3
za9!~JdB16Gv?n1m0;aQ3yI71mWodU(<BX9NbrdJRm@vlcMH9GT6#DXLekc-=07dJR
zXs8*d3IGG4F#5o_mKG04dMK{hgB6B*2cq!}A1*7ooW40pt?Fodf4%HQD%4(tw}{(+
za(p+o!EBHT`$yl8&JR!b(4MAQk_n~HamtC^!E8<>vuvl~?j0V>II$P;yCCR_sl)CO
z<(;5jzu08iK_te4b=d)a%2z@PIBeMZMp`{X&uBz1b@W73l_by`gcxV0^Uy<ELZ-f&
zpjiD=7bg5H<av0mUYF>L`VgI`JXf+js5VlTVoBHq0LjwSq)Ne9BnD|4MN2sfOs+7B
z_{+e#t^x`=!t3^a0PWM<;#mCqWGuFA=pXlrxmBEKrDPt3&5~=c_OUUCa%^^*PowCY
zy<Ndp6!9o1O$_^u5t1iSk1m{n31zjm!Xv;FhNFZ)88K`QTHHNg(aairI7DftA{P4y
zl0ou#CtJ-+i;K1l&TM(Wrg$ezgrZS97rQ@?&N1?kK_K{AdOFSU>JciY!XxZ^7dU|Q
z31*g@b;8A;0&6_|v;aVPmMwWQ392)D46*oRIz|2A?CIq6^!y19)W8EYF8Vni9^`G2
z`Feuk^M?$b3I;(>8j=Ou@m~TFw1A{H>XhttQI)eiy_KYR{!83$<20S3a%<^ya<Gh3
z96U5<|Gme8v2Ypwx_s}39O^I;gn$*;TJti30V#MR4`{9-b0Hg|J-)sZU&E1ugA6aF
z7D+$ALvXNxccI5+!q=2yf*f3V5+vibP(ODWegnmszgB4}LK&`R_A=md4)(@4?txoS
zw37frd_pP+zaj-(^NR&#MGBOIvz<j)K><a0=yrlLB+vsHT|5#xZoZC*DloY6_o{%*
zvE4R+$Q=%m8#95bHWO&nlu6)GeXa;j+Nruj=!xs6kSi+h^0Y=azvdk3Bcmth@FFB@
zCp%!?s0Tle3G_a|Eh)25Om@&pB<6bZo_LAF*%=DY@-~X;=>`48<}`v+e8kFaYJ=K_
z{aS*8Kneij1^5`n$xv8lWxfv{7@zw`NxuM=_}6jMOS=RjYI6*hp+HC1<OztNLwSit
zNbTh$n1UR#GTOU!(UWx@;mo`D#>o(t;q0a52#9{uiENBE#h&}*e1Gwp)+B<*@6)OM
zzv*PRAMkt`mh>sZlnP{1UF@0T6Owzw^KU*y5b5&50iO^N_>U}w->H;xV--E1Q>YQM
z@QuNOb|FXAXXT%D&&pa|VLA%GqoI|L18_#mHi0%-BJ;_F$QxKAaC1i~d9z3VYVeag
z&NMh0O<EhM4w%S6YR#sMrepUZ5>KF{7Z;e10yc|-zrTo{(Z4_ji&7;}#(*2pPqK4B
zNj`7}&Yv!23UT@N(yis7j9}p|_pjkfI8$O&S-2q%IYPWAP9KKGZ{-bGOy9=&#xIBp
zXk7m1k>*fUMYMJ-x=wH?`DHp8R=V<u!D=zW_rqD<v8G#@=uBi@pLtpI{D+2#HVM8Q
zx0AimrhaBeFRBav5OsGCIH!!t-hZK6`u!spn6)Ynm?HwxAQi-Q#>$45AiDQlTHmh7
zZk(8XF$Uqr5fG5V0|Xq8&mh;F8EG^nc5Da%B*gT0WO6)5^_0VhlaZKBypfef??@LT
zoPI&*ZXFh!QK<MsxRQFjz=#8Tb};wKO46_NK|nm`xw#GTWaq$g+-v|<oZ|zxlie{T
zsX!*i#G4HG^e60e)TV<kBCuedTtRc5oblzYFVvAUGQ)0<J_8&9#hcreH*h&%tHVT|
zvRJ321uzOq@6r&s<Mv}0i_#k&AiECEBlShx5*006X|T~t1WiMH+c#pc>O_JHmdb4l
z8CnK+ISyL<V<fBPXi2_Fn73;^%#ug|8sma<?7Xq5!xqkbE#=ZDo!ghVX^a`1506ey
zM&J7JFQGhW42>)=K)gwPBca2vbBr6SgdibHfxwOFAvPv<Mh+=5yXT|57(cHM8HmvU
z12bNnIYhc+b8H=q@?lc?+_K{wVNdSlsf&|u4M?SW9w5P^=+)K9<a7dS>ua2~IGx=6
zb_nUd$8^M9BC3aV?<MM@1c=HZap<uQ=iA=o;)La9SsMeUoG(<5VJ)hoes6k;euwA1
zx2G52wv!%63z`YPCXtVNLN+WpF-1j=Ne>QNXZNp{6;`>UuvodSMXMqjl0%5#av{_$
z_zS?Mxm&}6M%hD((N15*no5J!U*L5?&O*(d@hz4_W6{OiGP*~yRX2ww$8`8OB2CK<
zB6Cr=IiEf3@LSM^AdFdiSSq_e<2>l2aqcV$))+`qx8js{_*WWYg@1xlDI#KJWjqxp
z8S`75Tey^|^%x<*SVs3~LmAKe5otFoulNLFvre}_u!xEzh0ULEP84TGU&%31J@lXh
z5HfzXWGPM1wfZ8#Qs>~Wcc<!-xcb=3@Rq=M<7SZxE8<el{_dj|1CEMZOh#BRIaCb)
zADsD6CJjirU~almhA`s42!{zBfwktcC`s;4DYzE<4<Q7on^fi?>JZR|v=+#s;Vw01
zsJufOGUo}i#z&Gy(*ToE3qYll!A4X@aCmz=Z$-Aj)k#yJ2d14MmZ64-6N^&;nB{N`
z_4X3==-*;zfB=!t6Du1OInq=~%Hmr<p36R1!kamTS;osY;PO`q3CLSLnPC{f$ecOW
zM&tuhVU&U(Socq93WpQdjp+wneoHDs%<AwA>S2x!t(!*lsMH=8XE4^~EPt|;H)|`{
zlPctFl_-+lG4iyK5Rd7xpFQSAl{0|xCd%vi#rfF<PPJqCf@Y9~3~Y{Mf7174bc}O>
zfxc@C!0k<F=?%i-DFa=eQMy;aJs!Ta$^@TIdvr4N3jMIMg4=h7%=!v=#inXTqc8#P
zzm#HBQ0m`N@&s}#acdmkJrSRP9i$^QjSo=RYP6?IG+Pdo)<;-$u-Ov&Gva459>v8z
z@{RUx_Jo7fq8Oi((fyj-JtTbZ(F_sOf+vh%YP^O>S?XQiI@!W`q_7Weow<sHPU+PR
zHRTr<Kx6l*G3CpWfa9wpa)>gv8{;MEu>8SXWaC5Mh!hdTSSpZ2onD4Y=0ubLnqtMd
z+X2HOVcinTGl8lFUs{<|*#h*de$b$4@g%x0Rm?YS4m4WN;xlr`1lNJqv+lWcfZQBP
ztd9HzSgVhsXD2^PuAy&=jd$~=9WX)fR4%&!Pjq~<A^z*sYsOu$?SIwq(wtCc-WunE
z!{R}{fJhk~ZxA0JL07^lF3Gylb6{GticXeEZ2=;9Y%0bd7J>!}bZv(hj*dl~bzsoH
zWCUz(dxHsBVFouOljxA5)Z}t>P;nAug)3zT1S|xTfe=@j$;~Fc{l!{}P_$(U7;zt#
zwdq*q78etix>T1AY>6u5IN`<-_o3?fvtN6MEVDH@>D}r5+QSLFBmR9Ge-GdM+}rBE
z6DNId;r4G^s8Wl`s)%Trwa*S#=T>k9DH8NM^pN?9{tG+L#s13|4^bJSW9}Ab(TG{e
zpp1(sTWWwL?lsyF<LV1vqUsQwBD3aICV2(An4aN2H<bIvfgKm<lG%oy14u{~8XidD
zW&BqrX>j4E26XUWdG5rZ#K+eiIEgY2Bv2DvzPS@X{0BHF;Adgy^z`TB;pAsR5uUb^
zlVeT0`bdb=&J&OUnh;RF16%%S>obOUnI5k6WgiOj`eZj3^e;ia{+Z(6n_&oRXo({E
zI(V~UEy^Pl7Ex>*uDi)g8?cWMiJ=K+kmsPdz0ruJ41?d2YbdjnusJFzF0x>ero;a_
zMN|!O06f_I>2I(8hKpP8t-u3{TByUbv#ljC^{;9&Z7ton@h?)W5~HLdhZQ+x3gtMT
zEbEuczi!#C^%F9Ld>$1^_*{2AaZ_1r(0x>XkJ+W|AnMnww5o3A**3-oWT)hx+Y$(J
zbj^(C&?hCxTJB5{znpBxP}pVOjTFZcO@E8@W+VDMV{j?J5b|K?#L$12&xpCK>!{uH
z^WFG7_@WE{Px+i_5eO%pATKYZzl5A7tg-$p{W9$~k_oWf^Ozv9`d)n7rO_2AvXtt%
z$s_{7<4NS#V6$8yQ05k><f$wnNahcsnF?c)KP<E`UxcI{uW+FC;v#BLe*)6CX@m$b
zX6BU%eE7{i-fnw<H-3}jr~2VBJpS7%y%<r(jm)vr>>aOMJ|3cY#N>k*uM{v<3Yd#R
zUaFTp@?S^~NcIRJxPTo(vVa{cic)4gH(3br#La(Rx1-;0GL=3B0uwe3>zp9|KBl?J
z=T=4vF&@bV3oO!!VYv(nO{1+1tp8)oTc%vRHfGG%>5L(iyLKRWHw8mn9Lv@OCaEuI
z!>bJG^#T@D*w<a1BDI@VNku#D_m)qPS{xRL6{r3Olv8yH#lawBf{~aQ#CRja^FRlh
z$&PUk;d?<zv}B_d%$J+}$?(ng3FAZ(u8-Vq%ftbFX)Ai^LCm94<M-;T8@p;<VYdMP
zfK~RPxzhUgM_+3nvcBF45L|KnSzC1PCNsk4T6_PpSpSb=aA%b{=g5X#`uyGv$%NG`
z*j96uYU2}0Ic5?5+Fg9_kW`46dlH)MZ~QzyJ7YXG*l#%Ld*0jA1ZMCYE9PQIxXgA&
zyF-Q_8MU>Vrcy*V3`y_kG0X;ZM}I2K+SXeWT{s6z*B`hsO#2Xabs6<;4YNk4L~bI2
zq-eaua<%C9HPZ)}l=_;8(2$^p%nCP=;2+`M$$y%WAAZx}pO&p)!dg`%un4YQ?TUng
zKw-_fmGSSm6Y674{?=xTtfViPhE=yO6Kmu{>?%HL>{$}+v1D{?;6MC1BRZYM#U+Qh
z_&|5WunO=`!`Oi&s(Jd6_9|)msw$EuQGO&3b9*@pV2f4!hXt=e@nDPGz42~yYis$H
z$eHDPQYKi=!K*eUiv;R6*wKTLJKo}@BxS#po?{iU^z#rgv(Fm+D8psx@^go#m)hgv
z5}P(ELd_`FonA!^b4=(e5o@>%(onQbf<B;`Ve8PAO1(3d@c}~CEpT2rZe6-?-u(ps
z5Zpub4SoK%&7W3YeGWen6BIs|LYi^qi`llHSY)tWQ#qxKTi5`Az8JA#=uHZ>-Rxm#
zHktYvK&uSF^m?Qs9q%BeXSmO_pX)2vA@>lE!+%~vcPY#Vkfx(wkR5n5q8n@F&Yfk}
z4P+g(|6!qt^1ZwS0+>>?cdDWOA>#T7zJc9YjZcm!PN#2B@0z{rqL<U-9^vu0{{lKq
zQu>S@ebYKmLZ4JJ5+(0_p{CK>lilyfl4TW{5gUHRejvD*jZ*6WMrPIrj;-fG^Gr%a
zfJG<T&#;FI@&->H1lyZTtCJIXW7=H=ek(3Bi9(!cwu`9AQV3d>O(srUvUSD_N-q<X
zH`i(oACs+B78=C7qI(2cA2w%DTamE=s_Loo=?w-ar-NNw{KtY%WzfV)0a$fi6$N<c
zwh~)#XQ|(E(f<tT6XLhQAl{WYed7krA4^<9e^3YzA8x%PpEPwmzk_qV_xU1gu8Z{0
zk1&Fs^O9Xq3!aiAUsjWyv091O=1%oe9b*HCf0dmcl-Ng(h+@?;4}?w)gf2sOIU}O|
z*gRVBQNy9`LIl->Am69cmF_f>=#{}9e=YZ)$7GXP5zsImi5(|9-e}x4dNn~OupP3{
zXry;l%YTlr=rtPhy`$5E-rzR)6&wTl{^rw95m1&xKdmTA5P6O6k|HD;XKN&%MLZY#
zf9o&(ej!nWhz55{?~P0CAZk2fEEwUU;%MYPa3UWLaCKG82<Vz`6gsVmBo(2S224dN
zVN{-~&qWfi=Gdo2^X|)AzuiEd&HhweV9S(KJf8Z~*1N6WZm6|PM^sH-aXy<Sqwc(t
zu%O5^yRV}nr}#zKB{sSILC}RQD#oR+2u=S1whf-N?zPY_5^bV?4tF43U-t&DJC#dZ
z4i#x(x!V>&N2_Ru%uDj_r`~6BIC%N->itK(L`8_^;Sv^0hqCUU^={njeI_~u|M9iN
zjVm}giCblC*%W|$<s50gZnx8Q<Kbb`FQ%2fITfn^Een|IpX^oCJRil!lwt!xT}U06
z>Ik;t5A5!K#eaZ>QzR0MZruCZuglALu(hQBJ-pbVQM^1Or2FTL$+DU&OvJXZET)Ns
z6Bb#F5)`|O&X6pL>l2BblS*w#qua`ZzgYZQkXQi{854LR6MaWh2tIm}z-G*gnMz(-
zt0Ew@0s3~h8*qLImyf$pGGEJws;wX<eBrYx1JT*Q?jw49v4a)Izu?uulOi}BBG3?-
zSO#+{RvbGLYif9EG`li;tU@~JD{^=8urxW-R~GBddJR$8>}mB>L(<iFx(0VGGomfE
zL-CRNPq|}MoanpK>P<bAf}?mP7w)CGw}Ss?St001vO=s@14YT-P`Clx{~If~%*rY_
z&}WTrtbhY#EpO#CaS0L*;x=sZ1o0H0svN(7khSKtZ7jVty#6x2j$zpt$aV!Q0cih$
zh;vZGx~Z`Nk{(zx=gpfp6q6<fr9@7%VUK2uH=(SzCLd-~T8@`GbUUJ<?t_90o@Kgi
zzXbFt<KlMWQ@K^vY72F}f9DetfBSb%Jl-j#5C}<yp{P0=O-{;{TVZKXQXDor>tSVI
z<~yz!e)lwqutsyDLWYbwHjTNL(3urw6Tq@6pCrP*eAS*+k}Ld@`)6)A)f5%`d2Toa
zbuwKWJR&b9pa=9_Id>rl%@mcm&Dd9tpr<^MrZb1p#UjPIa(o0<@AHIt&PL(rRA^+l
z4_(Lv@OJt27n)d6owa`6K?Ic~i7V1Ur6-P?EM*YSoIY>0Eb}1pW0?3lpy5Ig<@9In
zfpoDOxedT4ByCKzMh_p-olt^)e&0wy)<r~h%x&rrRbXReVeH_fEStyOZ4fH9TS>CW
zKUCI6MG!rlQP$>)q%o^oSx6Gf|5nR!3}E8{Z9ke+3#iCmPpg|id<f}pdwJVPLcz`m
zp5$h&oP^@i<`2TwT0Tiy!5ZlV4e$gnl^9CcZiUh^=abP0zoG>cq*)kTSS9?5CJ!D-
zH5YOFrKD1YCZz1qmtj&Pn3N)XJ3s53g!LU@oPQ1idnyn=R&f@$S%STsm7Bus!WyO`
z)bLdQNE+7d2&Ngsu<AHMlIVcB)%lEG1)c>QPy3dq)7D4u>PcZNT>WXK$;+2qb}diA
z{89_A0TsdJe02$KK36ODK`$<~z&-fMzUePP@p!5E$eSRA6o~nfTv@kV;bs7aNVUtd
z{P8BP@sM?3$|*`%g!Y?#0Rj0a{a12MwZDTBUSECLXRA-7&Lx)*l%dbJ_eQ(O4~qk)
zJ&W||j@rs)06aSgqXY|sYW0$u;h5od)=a0#@R6jU*(|WeGs}b!3t!&)QV-vzB|K!g
z4=3=EYrP3II0}Lp4F2j40zVZL!SE;CQrC24y_^lG>?-8ma>7W-+}D?!`$5R{_!wnk
zVQ{q4Cd4C}V7yCHCgh@n3nDe6H21<&`k;5x6#vUQaQYXMpz#BVYAgxYA#z^8Gh8N@
z&MjH=WA@I4D&vy$XNZPSMDr)m5M0J#avx|d*#Et=(PaA!cUygP`xbA>yoS&^_nmlk
zsf5GGu0<kg;*9n_En~EpsU(VGdyEnTeRcH$2wxqk*LOJ7q|r=#MWTF$9Jn>iMAExh
zo|u<J1f1!T8co<ooMp#KM@!_2wrA|1ZN=STmEgi3eaw{yy`qR1A}QkRY|wx?&>!Gx
zj-Po>E`ZWMI2@2zrp&)KQc-w6G_fST#&lD`5(xwV`Q3lo5P8L;{r)T5{dXIXk+(j<
z83^SZzt%xphDoU*tDL5<M#=RbhJww$UdETvAH;MZtBe&dPk}ygL4_YB9FYZc;86T9
z?nK$!s!wk;6`UH6CEod9rUF}j&P6<ez&LKT=LwbgGq?!xk{-x>F(sw~)7)>3funG1
zNvp9)K*1}?QVKzpBX`siPeAf;6%O`Jc>)IXTL)DJ3%RtHP2c@)d=9twpCoU$YB<o_
zM^+UDW>T=lUFiDLT5&tXFP&0T$9Ce5%H1fCn??<EOLZjA1{dR8sk34gRh?lE)du~&
z>&0PmSol1%3=iB>J5OSn3ZgQfnPPVOW{jeDP-UfwxeGhGhe~lP*X*w^S1juGNuy96
zMW@=$W|ET98IUlzXIFD$54M=^Rm?5v$4IMQL`?{7VY-V#5?W=ORRE$bq%WS0P%|uh
zA@ibILs@?cXGPj$eSq5)-A2~$qxdc{uUByBcRvh|F0di7Ug667aUiwJ1{(q$gb8h?
z0@TqjXGh$}^m#wFy%ZxvAbii^@1tDO8LH;A!x1pk3J2mP3WZfSac|Dvcz=J<UTH4@
z{wFL`yXP9#5yFk!E-!}iue3wON;`=qCH{`yI;#X#0WAL~%f)@CGBqg<2O{9bZF29Y
z)Ugu*hO`k4zeT@I50^U(_aL)+i(k{Ta5qn8{d!p*-mtS|vuWGe?28puN=CkOR!|go
zP&+mn2uKywdoi30{65taVeMd7^mL5<lJJ^o_GsA`JSK@f$}G-qX)46KpMD&ydxwMd
z5fw7L11r7HZY_u=*DNXPK&dDPz>lZE5EF%cxrK0;K1n_+3Qh_?t3+<e)itubzxD(L
z5x89){c?zdIY7`7F2BXrLnss8&q#!HJV>!cS9UO+kWwA#b!KhFuCj_1K0#(Qt|!Mi
z06Q-ffKBs(Go{d>5N2{~=siWnDqKQV_puHjF*NKQ4JWu=xfi>b$*}XQqLg@LT>~ve
z;m<gzQ>5!z75&)iL);~llAPOeXLBg6*&vF=PjR6+;w#u_O>QuGKA{pwH<4t+BbLy|
zos?cpb&t!;EO$w$$0EKNg1#VLQKJRrZgxgwJfy($z}PYWq%1_C-Dy}>9FGc<-0TZ|
zVsDf^gr8&t?*}e35Iq~I;F?_+y^@&=k|9U43;$x$Lt^(O_!Ax7bLGIAL=l>x!mk-v
zQk&==O08o9M~K32z?IywWvaK3EH(x@%|wRo8AM7+1k`KhYA2$;!)gyydDH7RKG8xM
z-0?yh>t#_m>u@nbiVs%K)cFe4xp#F3U<Uh4u5p4TAl4XcggA)E&i+b*h_Ec^dJ#2L
zZ`c%cyuV(vEb3-g>w>1dD_cx-1#Jh5<39(-gT3D4mGKJW9nJ0mbdsg*tQ1$N)5foE
z7jg%Q?-j^ys#evQ^XD&%i0T!+-=Gj<Oz#*NnBQrr`x(HJmS_kMfg|dc(nS|WLU2dk
zeS&lqgW{l>wxIJ?O%k$e+|<{x;9}sWhaLvxB#%$%<4UF#vu?`I)KeCb5JKF*g731+
zZ_G#%4=1PB$Wq}J4_?sG{_x`HJQk5i6k?uGfw2-`>t6#2;XaC%0rfP)6~`++_Tgxg
z+x**)b%dm$5TfhA9@?Z8IbeQNWMO?bLn)}hq|2m+jE9Dhl^~f3+OD4TQyx#Hm<%lD
z5sMjxqXV5{sv2kruoVZB8wZn;lF@F(ln{MPlO||5OgpM9QBIYWC@~0oOSF;VH^aj3
zeK<J`UOAUsnF;KZ=vyQyxjsy|FFyz^UazM(BlOsJzJHymPAL%Ft?CNy*3I5uG{d!z
za%VrUXe!<j{^agDD8-x5Gv%j$p=}X23HlA4x%`U)h2v!#gRd8vvUvBez0W42qi?RC
zz!7&{(#-h<w5~V_(|wWU67r@g%?8*1{YaI*FP$CRbv*krDZ7$oi;Yrt)V)a4^-SES
z!ySkO{TGsP@vt))S9i|X^Yii%roGw&(M-v2YCq4z<q)AEX!d_lhFM1rPc^cZ#jvuk
zH2#u3iVkr;P6ce1JqvpcVF{c*-Cke)ZuLiyV&76^*S^+S+|^WBEuOaawfcSUlD?KT
zUG|<oymL2X&f=L&8Bg>QG(6y(^>-tiHXynYqb*W-N$IKk9p&5hS5^Y}AZ?XNkWd+!
ztBjKfX7`C*bmgaDQ4Tj8(p~C3!3LD4X;EpJd(X!&UaZNv-qU?-yS_sW!{H7;<l==x
zvZjB87!qx$IlO&EuN9Ump22tDG%4Vdg&Q8pM7(uyG~l}M>zExw3|AY8^{#gOy`H30
z_F1w@R+e5H!&>keHG$P-&w=%4A<zJON1_6E64NUfYc@l((ZZOK6EJ8Vl=v6DGEeWz
zlk@x52fq=^%K`tx6O|2tV$!cE6~y8g<i6tM6EtcmhH$TME|(aC&z~epw|oWUHJXdf
zTp1ina7JPdG^3t6NqMdR&QBfj1erLP(&P`ZY+dRQ9J3Dc^UjH_>}|8WCiPOqoDz0a
zZK#S08^NwDAS7l+`dZ#OI^AKyL<03v{~q&McqIOtklg}lq&Y?md|66OoDCR^&kz~$
zp}|1&<+x;l2g!Cp)(i2LnqcNWReLffvN$Xqi`tTYVq3~=!(a-fiFzMBD?n?LuYKfs
z1@PBxbI&Yn8KKT|Mu;ftq9#AkZxmD_c9NA7%O+fLNwe@0a-b0HR2mDfDA2Yc`?t?e
zw^c|`HWiIfH6d@~P?byypk{#bX0*euESsV}ui820*2LqSV22Y_D1AL*1-N@C+J-!|
zgmgirwpbd=`F)a$D+%JDuQotnox`4u`P)Gx6pM{ENISTE$?VGM2Lq*C^l~HKvE~m#
z-mGzs`?p;VC3|>(l*zkwpUQNx?+x<Uh(YHsJAiqC%$krS<$%!KOMu|dRsi_2Iyhf+
z4D7C?xVi?knk`GbR4;^qg5d!A9m;WrW#dVax4P_Ll+_bvB;G$d3PBorY;YaLr+2;$
zoz>b!DIR4=w3>taH2)OtO1O4KX=?H&<C}ftW-)+(O6&M_>#s2Kpo|f4lg1E;$XUZ>
z44;242*x#UygUmob?Bkk#F8m?>8RWu9-&B!+}EZ}p~851TllX3DK(4xxci^py7MEL
z0<~4LZ(tBPkV_aVbigHYnzjbMYN=5lP#}`jScx~%07CT~&#x!B68H=PW&t97h^f!-
z3$)Z$7~hF@q?UG&&|8Q-C!1Osh<jT6ad9lTkKb(S(=4Y=+PU_j&u_d(F{~M)N^n76
zRolwp_2P|S*0CMOm&YLJvV2=^?txQq8pfwhf(hcznJHMk#vNBcZR%q+WVeCW>Oiv$
z>i>0;rPrw?f$<LS7>$>jv$AyFMX~u|6Lu<8_)qGcYr`Y%{dNNz;v9mcjJShoWGhRt
zrj%?P+ZEh*k8b-_m9SN?htE@w@W3_IpE>o&;Qn)_9+?zug#X_RC(Kqzu2I~BZ{X+@
zmuZwGAmun-@-SKhe4`a3HFe^9+yEw-%Y5<mvX3D?j<~=jqN*WzwR3_;DVR8(Pm{=t
zkC8~@O231sKTI0U?@m1LpGFh6%QJ)Qwx|2(dB9CaqzW(bSf!qJ%lW}R*u2CLXC7oM
zoPaT(7d}66B4X3KVv(kWK#<K2+%cDS;LqyLI!*tHncksuM7_maqfJ|60Ur(L$^teo
zqaLBwX^;Z0Ufb84t$TlXQQwbNg`_LoQC1P5J)+@>#|NS6aS@{^p3_jpzlyt~vY)2I
zt`WP>!oX&%t1${<jArkK5G}49rj%Y@&MV?`pV+*E-GiaxCdO_u><XjI$5YwV!kFg6
zM;~O1!fD#RL@70N&BJxJNAzdBk(;h}&OC`MG%o8jOh6JKmb-U~vyD7Xhx=wmLu9a>
zGL|qzTH6Hf{PFN8OJHMhm@s$gfW-w^35c5tMka<*+B)Dg`C7O8Hh3)@ALmg7;_aJe
zW<vZzW?ojGcV$)m-nXk2dV5**^kKRk-(1cEb;$Ba6$g}zd{wv8TrRy%;txyPP3PFm
zm;UPC>pF<aYQ2ERDEOjEnLw8O10?~6|1grkXqX*Bm<3-&jK7DW?D$=+c0b#6U)XSQ
zB5o&j5lC|rDnk11gNUy?IO0L+uq@Zc{fvLW`uNaSynZ1Gvg9r$TP5$lzEb*D$ReJ7
zt1Xvz^MBV>3l@8~a#fJ3bvu-*@9NfAuq}Fr2Y-eGTWt=KZ;?H`gJYO#8Z9yclY~=Y
zl_zgOO$=N;KCs6Bl$R_V9<&}4TTb@U9SVpd=MS-QB?V_KFR)w7(qR;P`W9yej*GXY
zkxXl9<L&YO>B-xzhb+o+f(Y)Gjo|;K^Elelg$dnKbB}^Mt8V9#Bzu_KQ~4`euz%Rq
z31ZfbCAguQMoq6!u}ShO0A94}vrVA99)w#Bf_FRVHuQikfRJ6D;_+_NN|evI@Bmkh
z8&k$LXM%-XK|VT}A|D1|>izYrzaZ}%p${J#&1`!+-S#`}7h5hAm}%1YKfq-M0kSJ!
zUl2;D`*J|)9geXVfLZ(CEgY!G6LQBTmNYoAw4SJ@c8sUnK!UfjC6jETTe<^Mia^gr
zk#@*=TpxyOPv63{j+IBXx%Wd;gNmj5Qtm$bI>U9*GdeP5-*6$0cvV2+K>Wbf{!s(Y
zx1#|Iu*g?A>hp{9vy1aZgf4HS(E<oeXCAE4K1xS3QRj6e&Jy27o84#%u|uFs6NgzH
z-)1}W;^x5FILd7zZ}k_Xke#1SV%73u_0ok6d?e;$YhmgVGjq+11TYx9#t~Co@<b{a
zADrMM&Yq%oHaEr-Vv3irV4!;5+Jx@k!1Uv`d!IFiZua1k*!>xsXcQ^uewrD6a1I{v
zoEh)O#!c!tF*NC|-GA{I(V^GaxG(i^;|A{Bz#m44t<<2w=`NQ*e-^|z|9P}00P{3c
z?+hjj>qPYSC%A$G=@bb##(Od-HMM8MVI^KbFNu8Ab;XkhVUy6YleoJ7N5=_>@uI*f
z<-&7-wT7}r%>h(*z5b53*Zp?Ab}6<?5eWEJdaKmnjtVGqP*HnnH=DO4!TEpP6^?<B
zWmfkG&Q23d<K5|E|Gu}%S4XBe@o2nNv0&mJp60^N%>zC9293&A$yO=bklX7H9%|3R
zE2W1s$u$T68pWa9r;xG<aui0tPWE@d_~MIy<KMqs?+qrGx?#b&D7o!TtWI9?W}b#T
zQBP5!JOk8aFaMw$=Ybg-+Z~oddPbrUM-^Ey&!q{tKYa72ttpB@Z%sen+WUNE5ta5A
zZV(2*g8z-t9VztGSz+HDKJ*L)SC5NqPECw#{5PXFod{odkX)4QaP8CT>U04476s!4
z@785+vd3z8Hy?IIv{$ztE%lf2qV#-Ab6q?$`STmHI@(LVKn`!`BKdkW6d5^l=k)Yv
zn7u!LU?QA%%DB0HYx>!%&sdkdPL$PYJJ8+p4m%1*lT*hw;u9zZ76}Ac^)_nm<i1LO
zSCH#R3BaZ_fqNlh9VVps>UkAh;>HWzZrlrT65huAs~mhlL<(Olea8Dpe;yy593Jqd
zh=0n9BCrsxw?u$&Yuhy#$n6&{z?KxR;OvVcd?x{Np_VB27h-~reu1G46U=zZzZl$g
z{paeBJZ72yZ}n27qPw_VA@tPgwq4--oeXt2945X+68vcXtD1{RTuZ3;yEU%^Yi%?^
zTw#}~mcjMp&*pPL?ZJP9(_+A_Ujv~Gh#r$?s7)BhDI+_@5i7G&q+>2yG?vt_z(dU#
zGdbd;Bcd&HYt@^4e0Clx(a@OpPqnFqPOa<kd1WYXFVe)`6JAGBx}DZBKnm8*Cq1}(
z_T^I^2v0Ud@ty1kH&E1gc+B_BkEKq<+LsTH#xM&v&PKcSZca~5;C|9p(vi`lmskl$
z{6eN~K97hcx}}>~bcjpCRAb=ty66~*PrbiJJ_W|Gl#XCrF$&b67J`hUC<fmtQD9s7
zk`Y`yupL`?Wy~Iai|48R@W<?phO@U>c1F;0I(Np$g(eF42lVZ>x(zpDe^l8o=Wr_h
zc0lkTY|F9gL9s@@^JSNb#>R=f(;-G(_TdYdk5;J8?M%6rA?sWlB^}Q&I%`4j`LI@x
z;4g#(617UNB?np@P?k4==+~RBn}Uh2P@bjzED7G7eTiVeB<;aF7Zkt5OAs`7MnR(}
zU>%DnOn~*v*B;QceS@?M*A*mQoJH7!a2fB9aTtPh06Rnq$3H=_trlAF@7aX+?!kYy
z*vFBK=%rE#9vD6u0wOXU;Yk>rW5sq61)iXmegzCEKbQK1e3ho~az=)knEV#~4Ijla
z6vKt()G`xsYiH3H``5ECZ*1-S%b2?tEL0*@Q_)9X;l^kJ|D*>M&{p@ug=vB%Jhc))
z9diw`NA7LDI5Va!n9*K#Y8<RokWG~ZctjN~z2`J(@*WFU`AEe}<tr62T<;HJP7R@R
zi-4Ufmf~So8r-}V7TK#BB9sm`UZ2npfdBAkF?^?mH>2u6V$Y%T4pWYN5N$Vcwdd9x
zYK}p?WY1JEC!a+XkUZ=NI<j2fL#_g-D7XMpIAFA#{Jh3Xq+!&_4_J0d-#(x>@!xqX
z!s#vecmCEq+uk65DTsK8Kmf(!@QA(<EAb)-v#JUgvS`G0NzHYe7R-x|%y?mud|(z2
zCgSeW3|>SZj)zC%fAEzXm-q}O4(ui>1*4VX>UG*{*#X;V&cWG3Y@*!r`C)GOrdT$h
z+CC%ib&laHP#MNw8Kx^Ec?<h^=3rdR%0S6_R|ZKw@&SbQwtpVI-NvSkG28z1^~pfi
zevqWsi@!}|WJHe$Y#XIQEQ|2fV1HWyqdtO!1v(LI<XV|_LCB(l)q&cyQ1V`!sBF6}
z+EI>$FD^j+gEDdAaDu-QLX=OUjt?ZHWteTgk+i*(j6woJzS%<|8uqU0JH;Tz;}%{7
zgOAua^&&>nVqe#Y%W=uu#}yC*kP6c`Tt$e=0K?N4AU1BOT8)jFMtCj>d$=&wdJov^
z=$&#Pv0&^=V6x+zx`At&SOsok+=3!O;iX+`D8h?y4pEKPXrFZ<<?7aX&9F?2!mvBd
z#9h!R&58o#D#l@=<W=H9WPFGtuyHy|{iXkzA_|<^VutQ|_vo>DFcpItb-TS9$0nIU
z@XwlLq)P)<yZ|+5Z}tfEwjRypB`j*Wr=;BqeQ2$;l$&bxubk={8V?hju@C0PB4QgT
zQBQ!1s`x^Mk~#{2hSY$5_mk9>TKpSgAvf+VFGxwL^h*nf<X#{>)q?W$l0QH-$mP(d
zEI0+?cpaa}g~XmivEZ%aAfY{-O^^vzslx+(?1&#NE$qYcL*p`j>`q6j8XgLa%EhH7
z!u+8(`TI4fiB$Ya78c`dY{lFX4^=?g>c+*)faW<wFP9pQuqh2YQRqB;`Ad5TtKaU+
z_`h(~yO))$Q(adfr?93;O1JL7Jk_Z!T!`vvC}FDJITl|Vm4T^uKe19=PnTk^hs@&<
z3OQ~5w1^O**Iye!|HinXQ&3)S#hGJ9=oRGe-y?~tgEmx$LI9AA%itglGht2D&HM>2
ziu(c);PKrW8JNX7#axjOJ9!MoyJYz6;3mA<qGbL^4$c9L<7}BZ$dwmM7s@N!$`UF&
zx2zyyY;wyw0z>j&{8mB7PqVcnVoVGAf=T{d_KeqZutJx%xMMeWEnzc(Nx2<@S#Rnt
zB(zg8EJX^;X1W+3?L7&un#F5X#6<_M4~U?@Lr1R`_wfFIPCW@29YpjXbyd~~$+r>7
z%=}S<pw!m=Cf0lnwwQjc8_)^1f0XE%)#$1zWZEs*w;io37PHK>TaGkHH$wG)F@bVX
z`Z<C6GTL4i$57*<bd#k!5(+A`6)i0D*r}0P7%!`{O!BpC3Y9!q43TV3C-(J6l~j(B
zt4piGn&1f)S1|tsGRx=xo^s2(k+llG>dVV6@28MoAKbZ2XEBJs<}AV+lM1QLjJyO`
zn#eWymH*xP9H|NX!M~U0$~E{zec3!R9i5pgE92?YQ&hw!-t;gF<)&E!(al~4-2r+_
zm>%VaX2t2RU8l`V;|d!6<n&~Ka_b|OIYN!H-aBIEx9i~ZEQMDBDPlpyxI7`fhV!cD
z7n2in>oq(ethFSb4|pA!e;%_2VCrTW*ie*S@n51KuuEY1@1el1j?)~efhBqn@h>_!
zOR~wmf?Q~BrdrM<$ZVPVJ)Yj*nVue9oR6xj)aMQbw7`%JrhATDeDsXzY;DmWAnwL!
z;qjdlcL{C?EBNMy8_Q#)Gh1+@#qgv1RJE;`zx-+Rv7rue_W}i9gbCP&R|w;%ini4P
zAJ_wSI=G9Yj^b8IJi6Yu?5~sO_%!V5Y@xEQTBeAvyu?e}=8Nmnoo6q5!Vi;<u60cO
z_yE?9+2~}SKUX^Xwf;JMcvc!ePoC#ZM3#w;bLzIRBajrD6~$KEu2j0rI4O|kPGKi3
zkf?BG2c(1j|CssziJQ%g;n0TM5N){B<4T>O2i?3I641wrQ_Tqb@SK^umT=IF$8d_E
z(zg(or<1qE)6P)ShX{j}wPlO(UxG`wqIQ8fCI5yykV^m<hfapmX)m0C>y44aj{?`I
zf*K!65uN?v?g-K*5v3}((g&fo2!@Fw7gXSBLizLYY5sfH(3*a0)R{LdOXTKngL@4k
zFYa{{ctkMaoA7Uce1akkrRQaP8vp@K7rW<LKtgk2bn$QYo;(kDL1_K-Cdh+S;nh?c
zCvrlX93?R}^*4Lv(y{-%EU-0o&um=B!b*r3-V|@MdE^?8(O=pJ6z&>MY%tA0^;qKr
zmDw()HOSUz6jVZ3e~go?y1^8h#SZdzp}|;EBS;=LvQh?H>tEr8kq6)Sp9lif3?WAG
z!)FydH*t@SjHf+Ss$4MFbfPsCGDVnDRwTI~E8DJ=2WRCfQE0T`gz{wQYMExYAxb)?
z2QkSy@syIpcj)xx)lgL92R>@?{A;VW8HsCTp5VZrBq$)ZUH6m{i%OPfaJLaV=!hA>
z_AzR%BUY(fV+^{KIYrdJl9m-h!=gW!BN{?u*8>}r43J@sve?OUga?h1G@t_&AN}_r
zeRZbSdL-^_KJHVGBV+2-aE}j`K#+|{f6<CJvJjI2fwO4LQ}vfTm%&T3)nm6C?uIAQ
zb<xQs;c*%c(%?qfk*ro|kDD+hRV%+&N@ansqW93orZY0fvU-Sl5PS{>$DnRRH{%ny
z0hG%VcO-F_X)3Z#U|t8uHX$?a1U5Wa(J;Y6`bc9wo#PpCB4}>=G)JwjU@6okpdG)_
zMnOe^kYrD$zV5a~Nwc|hG`9R8HfYU;EUiVNL$>0yzFG`2m|DdZ1Zj#)FmuXAa}kX#
zM`HwhobcqLMC3Bgq%6>XwQ&RAWPIkIRK6$^0XmvK2)^u&MjAD-29@y|G4}SRm5hLn
zfnPG`z~47lu1f~!S0%84Kx3z&+tSxDy*N8Nogf)!rs=Azq3$UU2a)hEcWV$p%C=3H
z*<!7S&FdC6C;b1KY8PzoVxP07r>VDo&;#fG-?iZcX}>LPpQ0iuj&$6gOongu@9u`U
z<9Li?nka2L_a}BGjX}5i3$G-`ez<!ag{5t#tFN&;1A(3mapg6@FMbzYml7pg5C})v
z$C6XSMlfO?cUQp}*qioby2<lolc-%>Ools0qu3(A4%HBKhi({?-?rO-Hq;xH!a)~g
z<uaz2B_?t=kbJsppeK*4)_^a0@<wZ(ClBbuAz37kCdno>d92=NbxZD|njaDuq54fh
zhFhHWq+%r!_0gkcFzO>dDWvoJZYzuRMqTwl3xG^_ma^@|8wJZA)5`anvnddeM`i=X
z_Tf`(f}}G9B~hjD@M1@=AAM+M;8t97;Zs!b#BJ%1WL&27;~Nwco%sza)n?xe!lAyI
zpN1+`SqC#ei>(h2<{hn3hLbZy8A7>`wGAU)YKm<=MZNorH9G+lLEJx`PRDvSqTe>o
zx6&qjbFD|0XZJwx1&8)jN=5}Ri)h_OhA8*%jqN%43!hkYJpGVwXm#U4m*!&>tnH)3
z5G+3%+4!ELvm(rR3Tgfp|DVEuYSWr2@3WOl0ecaL$R{AyUCD>AVb%lBR}ukZXVuT)
z_*OZ+Scq%H`b4^r-3%^$K$!HeYKMSFS3_V)wxfH(P00qzde!IB7h7jpd-{%r3MBx*
zWiX$R%6WjIZ)yz#qV*-<rbA^xzI2yI#18B}?Wz$wh%2TYl@~eshA;y~=2NDG7?g6d
zU;idL1V{RFWhEFlZO{Zbxq0B{*|0uI?e2G(CH7k%d!BZs#Ghlfm~Y90hr>OykfPZR
zw#oN(6&0)u6i!zJ_gcF1Nwi)1m~t+WGfKMlfbWz5$gBD)VM|ReLew@eWX$^8YQKE&
z*=ekvh(bjy4~z}a-Vru*4>ywHzFAdZ@J@6X0X#-g)Uq*3g;<Qw95Myj<WD5=Ab-fm
ziS6oSp7pOItGglLu)*EN=0x!z+w5|KrVIvSnRm<>!fZbt&%XsvWLHN)7$wFc`t);a
z9Kvsp5jB%O2qo3AWq0?-RxrZN?ICkC0!SF8r6k$L&tnvMX7whZAGgaRWYz&aLxatr
z+7j+B&;n;#&p8tWNpHGK^>QXs-8AvGae0rY2fzg)eZxHh<1Qv~h4pZM<=1o?-enlt
zmpa8_OqZufq%-a7W%;1D_%*s>6iB{h1r&YjPj5hH&xbdHNQw%1zpg6=cTqfTA`Vos
zt{2(zrllSV@w0FR29~oDyz}h{K9EJ^%pZ;5*c=eE`2TEz{>|~oRT=F8Z=^b7ZQZnu
z@DMma<Dfa_SaZK+!dD&N3zrTLxU%CDTtz<yL*prL8qaT~`0X036KobQu#dqCM}~p`
znUgP))&6i$pPU9d>FvYi+X=N76qJg_%33BY{J>N8a66<^;L4_C4N_S(xEmTR3MMxD
zvCD)~u>4uy*5MFw*Zf6e&_D%FM$LXCk0;R7yUnG+t9yH+*9&s@rO;LJ4QQ-;)S}15
zpU~al*U<5emYy@<Nyt|1BO5+zBpJHMgf~k6aD+iJW?jV4H7X#++V+yT)w7k7cI_bY
zLzLUg#8D@i%_mFxPoO&{^uGwp0aSpDGJXWQ6Bre<ZtxBF_RJsSLQ(bO;4c5`&#<0r
znxF(K6@?FB!%^^`lrKfN;k3c-VQ=xux}{sr^p|jg00@OY?__rbv0icE<MQ8<u_Ylz
zY9sPW;4-rV^)0)W1QLsj@`>9@g=Q>dH#sJ8q(`$h(Pq)Y6OE=;-5;pb3;@x3zpGwT
zyegL|QW8mN+Df0JHj{X>-%DrNJi#OSSE{-G%v!4kcS+)16wE98GO3^nM7y0O%}oeC
z5cCJBjQTjQ2k@yNjoOc2rpqWNFmdZIiWBtzP5%w>g?^CrDBRQp&icnOmf~Ha7Ln&L
z+8_Tyv*?0JJWeAr_-!rx)W?_*2>4%kSm<6Ux+A?=H%rCM-0yHR_rJaSe}231IZw6p
zmwus*l}?iKPGv)jXh=-DL|`{pvnjUwub?<wR-2*v$V??Q=H?X2n6Z5xIV+d(N$;4D
zsGoH2lnUFb*I%+5(l2Cl^LM@Y42N2|z;Scf5(RKo<Dg-yDW)*Kr(r9eo8(>WU6w?u
z#}YlVBGx?BZ8pbLK~)ro#(w8iz+ab&s=2H<*i;ypQ4e*nYv=Z?O$()@rgURyP_Rm`
zJyN-NZcq_BV}Dnh7({519+~gtT6!jA=~m!3cQOp_rqR2rt6Tk2rv(~Xv%<nHRJa}@
zYhfq6*nf7~)4fb)v&EI8IX66cYo#G^#jRLiFhyvDX^ycb!p786QZBACaz?Z&$KY@Z
z5QvZ%WC`w@guR;qKL|xL!#ujtFvj5W2|}N$E!~3{9E>%|IdK(<+RFEj=0eq12&<qJ
zT6AQfI+{G8NJ-2ZMB;GAaAFa1da@!}Z(3<6@6VqVJys*iK$v^Q2%ZRGk2MAS$r9pZ
z;sa4zt{Lcv&ay{r#mtsplMrdLD=WUnP_4`vp^7OZtx&z-4%`O4EKKpkOM>&hF>7O{
zt@}sgA<i#~(4{@Vf9gvA`(%91=%@JAB6H0XJoQiGS83oBBb5G95qJ_G0Ol+2;T+)>
zhS;q)fOx<)2d5`vIbfm%$fda}%;YeCCQ8Hx3XLAT6+TMxYop0;jD~B>aO}x`_$Jj8
z484u^v+2!5iL&R#8;BXfjPe)Lj)tkOPFG26tT_e}W5^d)Bol6NC{Tnp8U=;$f{1BO
z)(}!m`OP8xD{#FT6!SouKqJ_Y89m>(G8s+CHpR_i1Bg^`G0~U=-+m}|SW|=ra&?v%
z^W2amD|OIzV2#Vty^m6|MPl`ofCUngm8<g%uW_*N7BByOouw4i5>V$)F2y{#G#k4(
zD}^GU`cLJ}Lp=Z_X#fF}@=Lh>Qr#BWAf9}hJ?_jPDUuO18HJqB<WqU+z?-zp!49pc
zxoHwu@`PI<hDCoY2L{r1*j;^(s|Lat(n&|Y#RBy`Gxc?*3?S=Ym2`e&KS}4{NKtM#
zRRr`FhLiLdeR#ZolpU|94930*KJ=B9G~2>JkX~TN%U7UH)5#7G#en}EUN++H5kd!6
zBml9{P#q*wo2&?Y7ZYl|#j#XxRKI`nyz(X2r0tvW^pWDQ@?BWEd4*Eog`7I=j*rQf
zajp`WQDFMaiA1?5eG|Bd<e0v=JVOcUjnoxjfoH1p&9~SIppL}a`s((RXB#i>KYhCW
z@cTzt?mmoP9J_>%@91c-d%TBjyZ^IKo^QNpd~-TI@BRqmyj**9|HW!LjNUKPK1%Tg
z2kH6*V_+sPUT&mgF!}Ie+CE0{1;^<61Y_KP`QrK0=l36NKYIE7nqv$HL{Dk+4E8RL
z&+0g9e#gP)_;CIC^B3D|_aFYp{qMqR$+-NU2i#FIOiU<Wf@FMOXD-^q^;Isw+WPZ<
zUww$5w%+Uowa{R%53m>dAJiDGV~bN?fB5*x57`<HC%cC@om>Beqn6*uq^logR9TxZ
zD09A<);{LAv%kh9{bQv*Uts@K39tpjl~2=QTfa?5eP;aF7#Mfi3s=q>JL@1DKIvgj
zy+1nIe|d(Do~bsKS^qJ5OF?2r1-WIs+Y6gN9AY~dO`t+gupdQo=pi-G0)qA|%EVL1
zxK;`YXp4M;a(suwJAe5K&OL~ogBV?Xw>1^NAN~2TUALh%cnXeZLXdVZy`ml+M3box
zW`aTiFZO%*LQUjr?w901;+6bYRNc)V{_D+N^Wp8UYESQcsZSYVxE6e#+V>ZCYF~Vb
zI(W_AzpRe=7aLOxyk22(GG!!(Ez3h<MMCbwTHb$g5R`&nm;48foXWDh6*SpLqG@6(
zX%26EfaH>s5l)mlp#%y^%@75{Gw7;k5%;G;L24jiP>|7rj2&zhIqxWzioyUpptN{t
zb|^q%1Bjo|Ob&g_ZCYWA_yo}10*~roHau;>@!eYCc(OlQ>>r-O(SvB#va2Zl0^G{M
zgo>>OW!jngb*y*vEG0WCB%h2=8ex~?lWnDR@6Rc}50-=@0E9|q{(3xm!#}4d!!sO`
zJLhd!^gznU)fsbwHlFU)>WD!zA4yEF66HLSuQ<T@C;fv9sXnO>0DJUncO*TT%Nv)n
z$@<YxXn!*PbbJ9D2xrf6TPM;~Rv?~;I1xrf8_v@Qq{$Th+bfZQ6OBQVB3Tw3xBVqD
z&Y<K-FwY(4wa07QtIwZ;0_rsW6rvh_Kw%g<Hcxj?nMB)hB#HqVxvWHjH2ez`>+Uh~
zTJ=O+U`8Q+FtJMJKG@7n_$6F}0u=K+tb)Nqa?8s_i)C}f@x~jC+)}DX$A-W54APX3
zUx{nSJ$c|?;cEN6**_ki49*VE;C%fWw+#SsKdi2AJbC_X`^lq}*{!`%dO}pZ(P207
zO~1e?2ZX<Q*I&p#falqFGri$!lm#LE&@W?m{5@U^O_3dt01+Zg^T~Qjbl@eE@>L05
z?vkSjzkhLl`2Vr@?G0^QNxuL4Q*;<S61D}{Ci5VPZ45YO2iq**WU}Ds3M4=ikXT9h
zk;i`a`}@`7Ji1!~CduyJz3bfx(0%$m>eTC0)v0>fJBR3{><%%Dh;Sws4|g7p{VqnC
zoKvoS&BdLBUsukG+P|<E8>t49$H-vN<Aj^@=|#ElSO$>4(2<#ZEeKn!*6hE+quHSX
z%}m?INasVp;!+jn$f!0SAnO@;A~9rwYx#%Zwm8{>!SdtIVEmz%l(nUwr!)RxZO}DG
zUJq9pHipC#xHpi54B?nMV3L<uFbOOx00B`Z@PLqs3Gj3siMbN*pB4w-4c<Wyx)TI0
ztk#`#aD2|IKLDQrctjIBA7YM8(}^@&l(sZ4Zbj0_*FnypJZocsEI27SNedEFLP@(<
z2td2&CA737!E<os2#}he7LdIaMAIBB48<|?;Er4%xzqU6i1~dSq|6@nNufM#i!grN
zAX&Gx#1?eBh623_BY3b0<aw>5&bWV!{IC(!p_CnW4s3_m1N9RYYpH|~8!mWql7F6*
z?S6krMJv<t_0)gO&9=x4!fh4d1!nD1P7q9AC4COa%`{1g%2&SVn!K6}f9MT>Wzkv<
zgoUhqgG2DkNA0!s_07G#_Wql{ZN6^5-Q7k@6+#dY;mQpK$PyCA^p=Jl!2Z!v=a>^{
zOW)}s=vQ#6-#R;nF`XN5J!@%DCBO1t>k~&nD4Wo&=H~dVb4HW_3hdwKhSr$RWHUq?
zA#zpqZtfho*N?q7q@RSHyDAvzNCFBQ+4c*On~oYidm5T}O$}+>_(I^HoD5=@ib0A%
zP)#hipYR6-E@>Tl>r`5Y9{sT;-_vk&2Bjs>u@_II>g0!Pt4^O_vO`p0(WBnKUqN$Q
z6BJiqN5~LX>N%bN2U0;3S?X~i*T!N8P-JnEUtOI-ex`akUuSt63@0mWG1nPtpCfqa
ziP@i(W6A?NX*Vu<*yXM~YV5%1!%JoWK9xd!cnVMOq|NI;UmZ)Ct9Byg(3<sdsQB5s
z1j1M*SKIFPj(N0~rQ6YkLfB#*XNQt#oWwf6*N~)@<>k_Li!W0{zZ+lnq~5n58U$yF
zMJR6vSq=hy4+nd9F%6ry8JEL2_7RMLDGqSO0;%gJABMPK(ZOoQ$Ql<&x%|<?#o{}t
z1F+T?#4eXMh%^0wJC6ur2$)<z7L4dbSh_^sqz}VUmr<@q$R9-EG-HTIafmCqIDjeU
z%Y0BAK|$#DyMx1qlMoO9qc>sOy!f<$Wu2cN!QxmTT%SXgQ9S5J@r0W0gQMBX^5X}^
z%JRdkus|6kzuk@yvfL2V`WO~a16Z6J8SJjGm`?@kMURT3Pk<}KdguHQQ(+-3kwuGP
zBPgT_+)_{!X`#}@ljP5F4`@Dlf(q1CC?OR8Y&67uHi^pzqX_@X?3cI@d8qh0H6`LX
zGK&P&5d7={;t{31K{(W&0*AH7)A8Ov#4IrnUXMl>K_HWhLVHBW?_vkHXmJI!GhiLd
zD5+IIw~((fz@>$1=~vy>qsM<)e)Q*+#}60JCKu;CSw=pa5MHRy_+oetofO5>HO`nh
zuYk|c4F_7Tk#i%XxTYKE(rVEL*81Ko7F!e0(fK(K{Gts?E%v#)<2g~YPp*(HvPZ%q
z9{8z_RstK0a4>5R<->ZSfj1Wt9en!y8k;~9j6t0r#0`w=6m#oNP0)kps00HqWtuE0
z&%AiNP|H(gvZz6!f}#*fQ8WqEj4ZPru%s^6<VY>Cjm&zZNDFf=wp-;`+#Nluyac~n
z6D7Tg8^6W}`9brn-$kbA1q5~q;})w<+U0a?YlP@MM0@!p3Q$gQ3P-{K?BPhG>%nI|
z&Tj@s`|n5B0StRW{<j0JeKQ-r+UT|*R&Mk^2FejvXKQR>JAk(!X2&>#VN^P1t6*ry
zi(nFCh%THHfGF^*+Y?$SIl?Fodum6_sMDJCNROG63}H?{6;r7ofucy6lj{logv+=S
z-|MdCO>}KiW3W2qI_6b57SEh>0!2>01}t@B)Pg2GeB*9k6lVgLxW$REXy|Q1z*yBX
z<K^Qk(Vb5q`96Vj=#4}6VhVUr$X(8s#jNdrS@}*DTZ=$$(zZ-rI%QE{GrLIgT%~MW
z+pj}k+u&=l*T7)SWrCc@t&A1|x)A;Y$<!OY>QiN=8A0P+%I-ir)xKr1vkPg9Ujb9|
zr!vY8f0+>$zlB;v_?_p4pv8n2!uG;nAQ^ET%Du2=0|jI}IpMvZ?}k=)7qr4ccwng@
zPGHi7VuSk$7v0ZUM93ZhZx9}Bv%(OtmPxMYFi%Mr%L^ECEM;B^Z){!Jp9un@kKo|E
zqh#K=YhbtwHKTFYzR^FzT=Y-E;x2qQ_T|mt$GrFscePYqvsK8Qa?Nf*2`lYRx&?n=
zne<pG9S3n%O2-dmZ8;Rh@n`(;EP1fXg7ZtLV6vpuf1a)$F7mUd^M9cFrNhYvtGf&j
zuqTbvvlck1W^#XsI~z;kuUO)TdPzPxadxD3rb96_O_{;cHFG#Rx_%BcyXur>CL_N>
z{L{hroz{=bt-rQ$?Nje%@;b3<VzOvrnV_OGvFmYY;N=CoZ%5}hK^+-LN$u9C^70ge
z%%bb0bFc_wL28RI#_;uRu<h0iD>IqyXq3$1HsW;tTjbjc?sjtt;HVR9<hQW}hkocS
zEPc3l+5LFP|7`b<dxJ5nINLwx<5Xa50?NqU!7pKsWhpts4MbLL;w2hDlY?!!E4Uh%
z#|)+&`aePF!W4CPB1B$TJflC!vYxH3uD{$||63dJYcjgnZ?fV-^D^rAOe{S48#`QN
z52?+ZcRtEs;4x-&%#Cl%UO|u(+CF#9cNJC-q6K3b^iN>ZLpNf(qH@Tfcq)7}Y_*;^
z03?BBr|EsJu5PdWy}gH(y;@^O$$V6assEge0*64k0nDx(U<p!Ls)^`cRlAJB1{N~L
zjVIOuaiRF=B6c%iYRtE_=it3_9G-y#x(nskR5pBml^e?0H>KN6QHN8hX>EZp*>v!?
z?<he5EHC6#YKoa}$3f#<DZwtmcxnL;rkK_)0<ETnK#oUPSV>V@MJSB`;{@myguS;n
z>5M+X8Va2uZDD0*YWh5wM~ClZDOA&d2imXgD=Y<?5Gf^YVpnxc*3t^-n5>_-2Nk3X
zdYivy7HsobW$1{z;Qof^NyjRteK!OV+{AOIVT=d{9d)KpU<DCMDE4+X{$AKBi+GiZ
z*6Uv3a(fJ3CKB5kj+eO2G^vo;!@xJZgzL#(E;jBg&K(Z6PQ;`%9-d4vDRJV%J_WN#
zYdpLf9fSAq`-=|kf4$;xF4?ieIqt1@hw39F-B66NZtdm{4AhW;fj8WIh2zWKF)Om8
z;sI<vu;7ef>1MhU*qCYU>B1vCTr6H-r?c7OsjVZObvo&6&ak=xdx#Zw6D&^J6i&Ym
zmd_ON4k35}vqYr^Oc5H)7^bT!Glo|e)^02Yo)zv2?Rci)%Evq+jL)z-*z;5x<JQ_@
zoS<jZi8G{)8`C2erfS?z2Wzc=dnX0&Ds6>Pjjv8m!FzeU0GX8ZF6yT^CH(jfdMEo?
z_}PBy4}B%&^iV0~;o;%q*5To!!^0ISqm|*np!2~J&wBa>l!@!W3IS!fCAyHgiaD}7
zggcz(KTUQ)C>~LeTuLb%99NpX#nZ(?y!R+(3_M<<gP_$#q$4VmE*9Uy!l?jogEp<4
zj30#rS}fMac#T96M4!~uD8Va?L%2G+AVx}1Lb6ZP4+F95&frn9dOKeb*Q#5|6O3(T
z(v3a^TP#9OZx3Vcjnv7Xq2(OW(s7C(*M3KALnV3;6!Z(M*y0U%b8^&&e?x4rAS=Xw
z4KMUQ)cHAL97*e!=U1lyRUtP;2e3vD2aCmC!hy;mOx@w}QV01!aEo@lWM3~{bj7Ny
ze}a^c;X)cPR2IF>CE}?<lRU`x;c!!=pY~cV^oQ~a^2Hla^9Wos=re7(1qJ)igcYz;
zMjgh;5Zazv*Fe{FAMIrrL|HtRH$usZuZkuv+%n*|(yVw=AR7RJ%)oag#H=rg7PCDK
z(YrVVn|FvaMx<}0J0SNQDnaSEzkp-P(qY~n!a;`67<&LY(DmQB9-OM&AqO5M-1TbE
z`v?XM^Z;?eOV=NcWM4#a@PV<iX*<9T!m&RRgbj^&GtX5DlJ|~q0cA#C2J+he^j+4c
zp?}<HcqdPxirNgYsrDO2#T=sM5<aA&=R`!6#jO>lvZSnb=?_cfd!Ypt80;w2ylfAn
zd`}5f*crvf5V$A_bKrZUBLe)T1q+Am3tAjuR|0?I?$K9w+@U<aLh>+*9f);o2HfOu
z%7n}c9(-fkUS+r7Jd=wVjdur)%kikS%wMo~)fLQbAfW|9<WK2*m2Vd11F%RsUk8OK
zl9FKyNsqyW0VQU@z~teX#q>dY9VL3noC?f1?g>iO%3dGDC$@6iiLLzliFs53Z>~tN
z8jm)9ZpaQ=&2q{<iy}sv1*su3<LqWCso%Mc1XI#JYF>b38V?tj7nemv5O!4AVABv}
zKVujAl;4ZfdP)8(y;$y3&$8s7V3y%@G9^!jou2IYfxbm|ZX5CGp$C+bpQ+d{1i#Zt
z`73{p?{pKyk@#&);PFo$er{#f&8#Hls57+|sIfrwmY0$=Qlm&USj=p`5=!dU1^yIm
z=NQJ*{DueuppkX5Ie*@J-=iVXPLTM`>;O`v62=5Oz%>Q5w4JJ75C>J1yOpnnjMc0Z
zwcJ6U_eGkGkKNwU)#>7xuHXw8(#W%HtYY!20_TFmfkApz0xu+1Bl`+&S}H}*(5Zyk
zX3;?jHJ8VQqu$)&9dy(w&U>&}_Mx!#5Kh1(u;5tqyUN`0b!VdL7GsF1+DV>RHlV)M
z>zPhK6xWr>fo!wAv56~s>$tYHK@~WA^>%M_H+j$HB{7$>Rl%AB;jgg)64)6apTax9
zPt9|zQ-mJFA68^Wx@|@m?uIE@P-}|UL+C&#M@=)VjuygIInZKb6>qq=<&<Asbuy6T
zTh?65H<`lTgs~_S=*GR(++&>dnfQ;CYB3`?;I*Ju`=WE%453)f`3{Fom^Lzda@vh|
zKvDe8#h3&!i)Mre7ayAFio@60vIX4sIaw(FcGYLh17cH<x6L>kSsHWkJVL09@Ds5I
z2&GLSF<hpCs*ClNBDKMOJTwGer?wK7(xbtm*>10IuOT?I4P{K<hz(ksMjFIr!0qkg
z=Yn@GcemE}F|Pn_j3T_?mfj_@q1_&m+m8_kfWm!H&QQp13(M%>7qx4K_x{*_>g&Bv
z7bnBPCsfA6ed+-e?xf~iv^gDay$O^t{W~1PNBEpSz=Zivncxd-|M(N2dqJ<48%xcc
z?Nvnag>6m1tJHG3lh+)D(D;ajW0lDUnK7}{F<~1T*N2K)l7{m5>VH*(rF+%-<94@?
z++9TJgqONUS3e0`VeJB^;8$(G)HJ1ws{u&ohxAKB`XruHfLop}E^YTmn_4~za}@I~
z^fT2{bX&Y08sbus)B6;zj~-GPH>KzJXk#cvV#e_~PKJm@U6v*Tq*X-kQOw#wMu*a%
z!DU%-d4i%2Ca^&_;195|Bg7{>-uwtUq@pgj3V<+VD;eB4I(a*`LLtQmWI(q64nRH*
zkgp#m`KoprzSLk2(#;P8{;uGSrw|)P&mg!T8rNzl+T<yHaJs-V3f^<ok>k}OO9fG}
zCS@Ov>28`IgGCU|z3zrL5wKWmPR0;4ga7QW;=Y(yfBY|30|{N5-M3D=u5IY?vof;S
zM46TbDs|Yw>ik2%15qwL(ZNUlbXAO8s|!yLmJadn-F;0tOIk>JO7w~99gZ~%)8f&n
zGXXoi<XO6lyvQ_iFzqvBVd#&6baDn~d0OzaBj{D1L=&Qwf+kPc6i>_$+QHc!W(5ia
znzvw*VXkF7qy~sKmc2_26wY9s6Vcv&+@=*+b-sK%OPI2AcA0Kb0K3<tWP!_X!F7fo
zESuguaw347{kYp0U0{jD(q6IjuNtCQV(qJ|z43A9ve)E_h|cM7BF%%LxYBdc8lBAU
zG{V>X#5nnnug=Biq^Jl@*k;5a#aP?;0`T?#cyUUA$wA>}Cb$t5cBw2V&MKmhcS~o$
z6ogDDCt%*KK`+F86ACZw8yKtk=wP%|w@A<w0-d|6kn8|298Xg?R#*4_wzbpVc(cDp
z78G1pZ-e0oC*`Yns|mcHFt~&3;Z-PJjn8|%%jUyn{zG>J>dnY^{&!muj8pwGupCDI
z6Dj3j><>Wl$NrQ4*RAc1_KU6EJ*t3-tqFaT!;4w~um~3M57tw`mS|Q$<FP_Ssw2qx
zcL#LB6FeFb^6pnujCO&joe{sY);w;P@Qx&!Y&!d_s%O(BrZGel-cxps>zhFtvp6a8
zFgn-8fl2i#F<`>7#G_&d(fG!d;m?xU3;H~}_#s~y4tIb4^Pj<I{d)J!M-46q8D2(@
zDHIWftl`vMA3qhxWxdGvM`X7H?O7&Gkw9CkJKJ44X{y8{E34(clz2^}qsN~M9Z_%k
z>?U2aZ?)S1;q@p3#Azd%9harQh-LxUi`GL?auH?lFEADP<P3+of)xy&X+OAMtY_TD
zJEi8@Im@`Y&0*gS(SU0AO2WD+Kw%zz-ydDcHAvbTOz6Pzzzi$J79`|HnCQ-j$3KW2
z6-|h4h30m{7oYUr2r5W_`mxuY6C9tl|9(ane2O#^c(<|Q^6na`=9`;7{HM1u>*Rko
ztfFw;aKoyc?Ra?k2}b$BuEZuc=iTRxwDoWsOk6__j7{Im8HAr!KLEkxtk}K?hbTe5
zn+YzTOb{^}&g-I#N#*cwA+z>2w_gO2WlNdD{tZO4B%6V82Rs7I#r^`L*w}*QZvV~h
zckR8+oweOH{K^RDQC~#d-qjJmcr8*+If%&0I`KGjZtR6L>{3#<tVTFO-_zCj7nFi7
zfCV)I=ta8<n_43*5jD$+PeUa#5m73iO>0Q{Rw*sGB3%D=iu`q09r~Thm=WqlPZXU^
zi9-L89Wyuu|G%_ZENxM<U-}NmmuZ^4K>-H3sVxkRC~C?I`ekkkN_BLdT-4Kng>|~6
z=vH<eR(lujsU)iB9p5QTAOU;88S1Qcu()`yfcSrI@e#G6WHVI15#EGzNLl>RcCm-s
zcuccGI)c)Qqz@#mA^zifT3?*-W`|PU_cI&5KVCZ;56=;T5I>h3v(p+(hQf&xQEu+3
zwb7}>6a-UdGaO1~QbqIb7$JSYZcI-9@O(ENn*stUCTDb~LQnT)_(4*_AHFKik5BqI
z5PFXI5MG?jo>+|$u4~JM0R2enDWC*bZLpEXKI>d0#2!gw)OXh=M~jg{fV6TTwpA-#
zcw`W>LqKB@ap6esU>plCRXUOMS<B40HnN~tnp_y4B-WojGpHQEiEa5ngXNZ~eBv*S
zoQo|)C(X?*YHL9#xdeG9y{Vlq`Bw9CjGAFB@g3L;qLDBg&LA+iWf?x5ayDHp0)a_N
z*kRIHV829;8i3eN#4mJQ2rvwf{Z`>X=)nUdAk-`($Y8`c5KAdyScU1GBYQx$a9>CA
zkk_=yiS?cY5)vY14mJ-9Dvd@ovfWKPmG2iBy@W1+u%O7gkrB2*9~GAY<`Dl~N9uBm
zJkklKRYUHathhUEG5^a0P+jjF`5%;Dq&z@ro}sCDn&4v!Rm2Uo+fbz1ZHXOfLu=dK
z#2+;nD|q~0K*GWXaBi0%+ek0M*5aRv*FekPo3f?q3#K!j;NN#fFqP7hsx%_jf*OEc
z!ve_1L0-ZtMff?tG@nVbrM`#c9*fEFb(AXn+=5Gl&6SCtkQ55PC95g!X9R01cO=%*
zh8g|N8TxVRbBif!T3QsjS!PX(a}A!oT^pax-W(a~c-*2xt&t78P}i2b2gUMyUk-Eu
zLeqI*WT52<qC}%J&CsF5?*XS45vLt+hoX<i86VZmPEzVe<&(13GFvSxK$nmqJc1fS
zcbwp3%R(i%%;8`F%3nWxz)&=zI${e1g`J(3M;(I&#dM^4+(^BbG7F+WqGwf3;B#UK
zinZv(R<<}Nt7Vt&jtQOFnoRX`#`bA1PDUne&8Xod^Gw%MUQS63a~NWNBk?v`R6hhY
zUj^nPf8~0F7hnPiDc6H$81BAU6i%+uG>Ze=TVn|IPmdn_a(}geN2yp^k{_z=6bn!J
z|34T0=P!M$;lzDK*d*?$sQQDRwUCrPJa`C$T9~%Ua;_pubN#{;RpCQfGvr-$te~>?
zQ?@F#Z`JqzQ1N#D>Jkt$|8#I$f@NxoS<y%Y;$q-FKD!$Hz)?66C@ju;`Ixc+;UunY
zvy#NsrL0<mP%ek?KFOY8WF7Xirudcpv9fdP6mBwq&Y232(NSIe_@IsSfq+0p(JMJJ
zV;@gutg?$A6X}v8lR_GG|H2v|kwJ-2ppfe{#O)eLfO`tTjPjMm#stCr@%B*4e(?0<
znCt4?(Qt?ouq9+8Vc^DQ^XX?xOA9~CPyBeQm~}w^<<h&K4;K#?{dc~xdzJU?nHIj(
z@r1IBq{_3SKLOcbu<aK+K)N_X0@~kHy@Ff^kVYmRFL=_?YLj2G8waaTm#hY-y_kQv
z8Zk;{X0RytW$j}DPl27TCFE~(8%lH;D6<K?N8%`cFPGA9VVo;Dl;+710a>^w11WNt
zslKHbKo)Z@$*1?BU=nm{btC5$98Hj%&8C5o=%5CgH2E2WVN13i>A<B=2n!0^p-z!n
zzlXU<YH^DUB$F!kKqEy20Z843F{uST#iYz5lTNfORv^sbk$K;<Xb0#IqELWeb3j3z
z;}5i!0WA1M+{4rW<Rg*`UYW3~8cNs&dNNZOYLT4*DWg6tBajUS1KG2&hHB9J9Eb49
zK!wpB<}nj|=RjyG3$boyt+sznEQtHAd;=m4B``NJ0d(Ss!}Irufe1jbC)MDQ0N~hX
zTmi<@{q+tLrt0x#CB&+Z5N#}<WZ%56UpqP5UPmeKHkl$E5Y_B<0-5S8100CTH0W@y
z<g~#V5nd!P7cS1$vb!nV?abmIxsgG;jC!r^lL%rXoxDqb=&MVxs7x~d2!TF<R{T*|
zOD$=0J&Cee$~GKKc`vMg8e-ia91;y9%mx*rTJ9^nG{J=<LWts2Ehz)3QD8oSBqTE0
z5!aD48xCk+5M!0)MhBQLM2Jr(${INlS@)v6`?Y9wG!cyD_?e3zi0j4DzqqCOPKmLG
z02(`dY58@E%o+y^n1fGt0l6}!broZA92US|!pN{Y<@%{XJ`0+m(j>_Q<1W?$@dYv{
zYMi&_B=Z?Ya83lQ3G~u+OaWvm#)Nup`qd1S?I+P`K9}op)o3E{y9}j5H`G@H{|ZkN
z+6~_aS;@HCEa<4Z6tSs+bDdLLLncV0iG=Ha1`h|`!7C3VBCcTlu~`1NuAw%;$X6!a
za`av3)_KrWx-G*Z%m{N-7Ia%=t;J?T%IF#Rm%wboTyH?d3FnBUOuqm!vPhCj(m0|K
z6dov+$$Wb~+8N>~b)+lm#!W)svVN}J>>$W*LyWWmFs_QHmz$Ix`wbGRBXbE5#buqH
zBmABmq&Z%Ntm>LdBmRwCUdXzEMoAm|stK0Uhtym(*qn`LloTrk`m?wghmd?{_*wBy
zBDPk?dr(kg)-j+dr^Pm677yDBoi<enUx2-^zW63A!(PP|@GWbGfFR2&8C?z5)!V5C
zePAc^BJ%DH&``6;K+VJNxS==cch1+(S+l`T>11A&EM%C+Ss(c*k)h`KCj~~+dQ#6Q
zF1P@u!u8WmNl=o8Y+W>6YL`$-@n4K)7d7;&{E#{xpW%pl@cpxQ_n#enzxwX}>cRIo
zg1(db)f`u}jVLN+)T}_P2Cn!sFG0cF*oZ&E=P&_4-u4HQDFLPsl+(dWT;x8JcqO^I
zgetd;j~!fH95J`#2{TO2t<E8>Ij$(-$6LhE>O65*=~?e%hhbRE$h7+C5faHlea8ia
zf3PM-`MtaZsCZ)>X$i<MJZU}P%&fSGQVXeQnXo`_R&zdCux_@01a;Uq!<#5>8<>Rf
zB;3|+y(c@;e9ia;yJ5UB-vk`Du^P!s_Om8NT{s-ym&_4QX+T)Kzp%>xJhN<1zw|G%
zeiN&d<4A2jXlZxw825p&GoGJVA3l5s<8|ffBV0ID<^<Cg*Z}Kt@K!)ZS8_saa@oj~
z#adk2jNU#MwrbalU6hfkVzOk`PO<@@4MH;J%6-p;5-5y88;v^1ka&vCZ7f0s;_Z6a
zKYSi$Z<C}d6|gIenb?cTzF@MJeADRU>;B~B)zQZAqJum2yS>xiM_00xg|K2ybMc;J
zTQKrWcELjD!cAwMIHMULOiio-{ui-|@M-|n%@~)y(gnSh6k>$OvpJfvee>`;C?-CM
znZ(Ezq#yk>z@=E6!-|tn9=>BI=@aQhna1p_<;<+yM%4SL&GTIyEA$xv$>#OOOZUyy
z1SK7eD!6#nPDElm{KW9rXh{HLTm_~HzJS^0rlvo9{|Ayqqu_1)oWTo<${j!*RRVPD
zS*Q|F<o=!;bHn4iG<Pw$tW!o{1;Qm`SVJSxpoQ&=;{h~QfGNK3cYv+I5{YSJ&c5Hs
zh0)+;10|#+vn+d%`=4CBe@tNDV}}_*j6q~}WK&0<$l&(IXD@~$#cB|6P6$M&+u=Uh
z(bEM<cW2Q79FkU^&qDTWLF9B*=4yfT67P)$FkO#3x)@l)k-azwGugbFr9qF^*n%<w
z#47|LJZTyt>iyl{1w`DZ;6f1YX!B$Oaa3hnOY)}=5qjYCEevxW`vj&$9guK<F;Dl?
zIa8mpJc_@{q{5%TCF*Fv2}cuj(4aEHpj_t4$4M_FR;Q3=QYdh&ETQ`lNse(h5z-m*
zbYz1w#2OiusoactGSpfHKrp>y&ATtmoS~v1SpjOAWck69wGxdIu=K}C?r<}aMo5(V
zN!l{cxC4-Y*T82dTfv|xF0Tc?p<gmFh>5BD1xWLQx0>?NlR%MvN1YGAXr%9y>HutX
z&>-)mVTj2&;IUnu*3MNGYxhhziZ790s&iAHJR(DfMC?!KA)`<KrW9v02c!t3*}UI7
z27ss0Q)KKo<^=CgPG1h8W=O>@`7N1R^WH+!^J@O`mTVH><90e7F$gexl@Q1C{^aQD
z_y=664||5suAvXZ(_7$$p%n>gGZIG_{C4z_t)q<E5u1h6hE*~4HbHwMk(pg3&8Jcc
zB2%;)P7Ym>rJ8M+{3JrECLge)Cd6B*{)l#d&`@?^#c?caoCP6pUx|r;Y%<DhqJzK_
zx-OHIgt_E%kOR}SB{C5WLM3&XDHADo42;r832B<%tBF^Dd<3I1@ZSMgJeL0d#mfG>
z@UjF=Au)~9gWEm}l*fF3!pjYVa*iC4A<6Op?X^Hgilk%;(Tnm`3wW{e01igDtBv$v
zFQ9i*fJ;>q$X-lCPOZ+9K{?l}S339Km)AS}mh>Ac$XJQnH&TvHl!h@<!n9nRfr~uT
zyRk^e2d+-Y>@ae~rNIHtt5|agd3Z)?5x;&UqTdSSttn?QyY5XXhzDYxs5tZu`4T?C
znI?!m2L<%KI<DRAopf*!7A6l-2rz~kZ#Y5AArzLaQXu(L50J8?EWS200R&OVhbvGB
zmPUI}WZ+vR_>yl0eunk9{8yQ2Z0k0I=_tOIW3?sz>K0ASlHXWF;K#VvhNw5}U=d1K
zj@mC#rwPX+%ap&Mb%{7hfJ)5_ccMBac}?RaP#grMmgS!V7!e^@WN9sM!88ItulBJv
z=%ximw8I@MXoL<~6$+A368v+W5mA(Q&0r$$rCQZEiHsH?1{_na5(LoVHOix8ZmDY%
zR`U+9!&xs-h5z;#(M$iOj5VQ)sfxrS$z%hFrW&*1DAJ_a4nR2u^obH$H<UDlvl4eg
zA57H;^Y36ni}gGs7m53d=84fl0+57{RLru9fj71(7$~Nj6YMOjW)WBH6U%^gV5h#S
z@dF?OVPX0*=dw|I{AqBk<q@1Q!`rX6YUM&9d4n*Z*~bI3hQ3D)d(d;yYC%A@&>yKT
z*hSet4}E%#LLXavB;*;mm0gmzrMih78yn&YWU6dxKCa-1#&~Xx1zxqVZdi-^$dpOP
zEyhgVYXxk~@e+1@R92Q|nCOn4o;NMmP4Mgr5DCSg2H(UVz{OG^Sjy{|-p4NDmz{LJ
zB8>o~SJ2hrnFyuB1d^TsGaNqkYRfe6A&W1-(Q0Wu26uSR2Fnbnab^lNKRw(KY|(yX
zFbW`XYePT>MC*dwquP{f8h=#w7o5l0QM4KC?q1%3N6={H0^$Zm#rGz9hlTIhXwm$8
zr*YDmbk0|W3B_5N<%MMj>BNM?K<;?2;@Y(MOPEPS)c{yxoV~%_ey3sWP=>B}J0PnA
zvf}!*R-KKOX1)oNVdz=%fV)k4b}dVp)GVHneatWdS$2Xoz8^r+3KkoeDaS3+tvc8o
zyq7RFVK1HrZ`?WQwZ$ziyg{`Z`b%xX;If)8FEngJ9Y$r~F4A1TZzJRzVigL{*IWDT
zwY|UXAUT+-K8Z7{>URj8(QRc}^yD&Hcu^T823AA7Pv9>ok|7^}&Kd23%z3>DI(*Mq
zbS-Ac7s%`@*^5K(njCYk)eTR+TAssgc_a}9m58~{T}7re=!qA{VszfjSF5Mq-tpN`
zxG2o+<u+gGOJoy6jRmOGxbwS<dgBLKKYA>;qjfqEB3TrqY0vVozC*s!&bZTQPv5@T
z{abr~cXPAaU-GKH*Vv$TqT6e)zka*+HF{QXtY@(Bw_o8g<snvSi<#T_4jDqWx1R5=
z;V$&f+Wt#M%$MO*X=Dp@S8*BNvFNX>uBG+N=}o?5^wU6Sw%hp|KSLI205%i<Azx7M
zwEgc_U>g*snQ954MqgZ4-B*Aqw7vyc0kOU*bT~$W><L&16i%3#%u*WXA!Av9aGPgJ
z9yw9Ah~CMkJ71DdJrO7JQ+O)zz?sz(x(S{!;dx-q=-ydR#&^F*^z#W_&p$mxY`)e#
zx<FLD99JA181LS_d#m{0GhI&df?6%wLkITPWYvk$Hv~O|xs+5aZ;kJ$ABlUu6!!rX
zFr%mhpO=`XI~hdlX|+64Ec(lAqq3w~Y)IXZ>7G=bN#<UM;}aSVvs01DoZuQG!VQkk
zddw}!?426$#vJO{HCXv+0|tu{3qFyWJPF>WTWF8wP`%y!9klkbL{XI$Pte9UNM?>K
zTu~Yz87Z!1idWWSR8oY<-w?scS!R2Hs+(gPXOv{#QHr2T-xJWbrA70wdmnClMDKU+
zFQ5WcjYA$O?PHHt$=KJxBcW&!y|G*zp(6gzM*#h&*I`!{2Ks=yFhQOn;FV;~O#W1U
z^hMgjbc^hn>OC()C=bZX|CP}rO+@)mhYj&CD4CykMd6AA^GX<1u;ht0B^wBbQ|=ps
zqU?b#k`H~k`O%ASY}LHITN{6GU(}3Voe}!5z4jcE<z-}NTb-22wOW!3s^hde=h2-T
zYaD`|Dc{rn&74t{Iq@_W^<&jmZg6Ha&}y%X>Lw6NI)z<Aur?L?Pr)kIGOa%=Gn^eR
z#UUXzkz2T9-#@1CcbV&2tpA-%ei~}#U@SRT^c)Gniit;O5wXw@ApV%_B3Vyj77i^$
zhR=i5Gl~pRG8#1mSxCM&V_SH0cHC>74aXC4H0tr1OXVT>n*9i0l|V1kKuYQc3!_km
zWH9Dbb*9pU@1<;3<Ga>Ht6RM63(zTPxfFK*LW&1~A;t!K$^`u5CE5G&5=iLj7`ZTd
z<FDvi{v+x}9FO{!Ox^euN5}$L?><@vwlwbJKe<ti6Q?Aq2X?R>ytJANcgI9Jssz`6
z12a5y64XsXs_hOO;_&_ar<QQWn#^sdo1v6S@~D<Z-j|m3HG8F(=?EE9N5fg6BmU=?
zK{l47watxF;B-`cDHEDjzdYIW*&IV5S<5QSAl13-gRX(T9{$FtL<`;uSgd74F8lPv
zLo1h3lDvu-@f>*DSVCV*?;kV0gQI0xBnYEj#~BkA1QZzw17L=cdT_><NkQXAL<Yi8
z-6&41{3Pb))`|LEFwrO0>W=7TZj;&{bp~U`K!#u@vrXv%-Aj(?-89KBfrGY=v$Fj7
zK>@P_jvP~&S>sDFUp#RuhoTSUBm*+LD1xi6$sUruO1NR;JH&JH_>2`I;-m|WmPLmK
zS5R?36}U};5ET>-*MJt399AZ2t~2V(JRIE;YgFCUs`t?Zmq<Wt3><TN5&+AbrQE$r
zN#$(pl-0$xCS6jRmAkE$Jf7Moc$QLeL@Dxt6p-p|*%v+_*6uE@s^6?2@jtJwmP<9z
z0O^tpDbNW<9JJGi^$PCB@WVi>PaIY;w$uYF3m$r<LRB&~E&A8}$#&=He^psUJ1VUD
zTcADHU$0eV8RTm8TIO8rw({&(E3<CP2*>$c4NEwrD_XbGW)a38EhtHLQZqRuYN5oO
zT%$vQN)rZzzJCv$R(RT@T4mOPX@t~ggm81w!Zf54;rDU`7t67Jd=Mp%dBdZQag9Lb
z230$O{4xW6GPFVY2Ff^;^72tJM}(3-K8AHfs&^*>)b(p1>aSAFBP0?o8}!LF)0Cum
zUCd6a_GS7S#WcgKic;Az+Wf8D9dt?vHe7~~w$w>0=de1F0lUg;rahD@98)CfKGoLu
zkOLmT2|bpb+l+aEf&%w_e+)cM1s8O5a0t4hput?C1c>Cu6^}NYu$o^TxoIUbPOigy
zzOvHG`}d14i_fVvm~Mqjtf++nDFEJK%f$W1Nl)+x*XAm_2=)`aYE^lVq`AJW7Q^=R
zv`B8EROy+a;yg!Ttt2v5FO!7WiCQ#iyy|@HUtC@IMHpnFKOfRjX;6ee;^T1-!M3q5
zAVJkRrOQE-9vvm<k7Z$>X-z?0vfG`w2Ls#N5P5iQ%V9gjJ}a1EQ|0Z5Ncb~$eF}oD
zsg^cfCnM)koh%<H*XALda2~wUE7X}zmW+E17jn!hM<0R@B@8sUSydj2@?A$=wF+7N
zVnDFd#s>1by&<r?Sr*oYE6=LGq_|dmUZA2AQaKWJiqY^xdVzpM0)AP(>8?mHowR5J
zqmiGsgGi8es%-Lxmb_x{3zPu2{#RO-B4>7zm>KG*XE<-Onx#kb@D>mcJFR_O=|OeE
z$X}K|wmy9L(1NG)f-ZfHrtXT&vhxqe4h$Z1g6W}yXFMpikFNUXUARtqqfc5%sacL0
z^0ih1fqZFraf(yYq%DO&d60_nCkX!`6oz7zh^wKKch(V14smiO^CR0u502fgJZkJ9
z?unjXg^@gSc;P9@0)W!TTq@=uQAk>|J8nc-3!8qXERnvRDna)`B5IC-1hPY>FP3PW
z5Mm+(8-iTf>4HehI;W|=;wt*;ssxPiPh6bFe+6-Gb>=W*ovQg;gT4m;0FwbM+!fq>
z{Z0GL?$*~^uh+H#pT_!|SAa2UlljB;*6Xc39>@LP6xSKV>)QEwa9N<n4Qpn?5CtE;
zJ2}-)kq)T`WYcmyVv!D+n06WL0)5HCHM3O2CF}k2>=M;d`)xS|btTuRN_jO0qJqa@
z0Hp%=eXOB9k(OQ=gqI{k&t4W#9%k-Fnv<~MKGntp>L1phXcp1yFIg${#;m|@I^bTZ
zdrp9-UlDYE2~WU^4WIxH>7_8)gXP>SxW@G@0uY*8J}(NO&UR6`<hCgk8xwhmYOx~D
zwhawlVfC)S0#IW0lX}WvL*VU+7Q$_N1ws)K9ROaZ8TL6Ha@E@+ry}O-;m&CI@zV~4
z{Ad8|#6Kuh?o}CG;gU08qAY>r#rHEVR1=EqDREnPAL7j#Lr0r5=avo*#}D4!U*gR%
z|3XBA3{V|{iGwx9ilTr9VPe&+7){CnUhAFSZ0F6c6yHd<7hj0wXpHKms2|XrHpJfg
z%gtAt(o{<ADDzwmWBvt+myW<*kQpq#A2(uym6gY#g@nPys2SDH6lx5b;cry+I8XTl
znqbkmCvgPjV~Tq9uxL9TTA(evcnP3Y<7=h!FRU5!{bo%h#a8&(NCC<=c9;<^&3S%-
ztMFUf``}#PtZi>?py*<IYbQK=vAy;+{{Ch~U}g8PERwjdN!{lAmzXfv7`OH46J>l;
zwpdvYmu)dOvU;uqsuFY5Q6rXxiY6=$P4p!F;Z0~w#oUUaFlD#Ekz=wA$IQO=b9MY8
zz*HGSf3Uu_vHP4QC3HJ<LKxL#L>vs5+Cj-X7FxuPiJM!k;v5xgV00nxNP)}G51?A3
zFom>pPliL@n`%p6CxFstfI5ES>;#b6STKhABkdHR=UTNMo2_3vf$t>+Eyh7Wv&5+c
zuu7yhHAkS(4#@*k3`|%;v<-y_Rxe*38)F2e!_RC(v=0!mFT(*&#Zb0JUT{xJ$Q>Dc
z_{6w{FeTF|%MmZxb{+S-tTu<OBBv`f3SDuSw~UbZJ6W6{Kfe&RXN^1f{0RR$#{arm
zbIC&^6htnGLC_a22^3#`Sv*=<Kps_43;%fdC;j7?fBc2N&ofoR1^grULZ(3iUq)H$
zWti!k|FO)5&Bu?5mcFHS#{m#*#t-@|>&mb7=~@3&BEI`bBH8>8?Urd>Uf91|EJiSz
zBWXOxZI*+t?$8l+H5&NmPcT<bnHZrReu*cX?yaR4WMj10)|qpWAgff+5>k#=P!H`V
zVda%6cckw0aStewNIq&HS*XC}3UYQ2C#7fW=Y1%?rH4F^lpgt6qx={YReH2Hf!w$#
zJ=9_LP<g%dLgvKtN$L4>oDPml3s4Yb4X*}$oI@M}Itq<7S)&hH$?e5?2Zr$kM5dMO
zO6@d2G<j9d+(HA*jS{lpgnBu|xeG>kP#7gRe_$^v5xsb<3qpfXZSFx)3e8CLkYpJ`
zH(sF*#HS+na)r!^?>KIF48@MuixHi+gLznL%kb7FAzwLw{Rk31WG+5m76{PaGyA`l
ze08O&%)V4tOuxT5SZY9FBq@;~hW4}ijv%g0>O{y}VSp(*wh;PWZv@Ms*7Wr2Gzuz@
zX%Bnp4`kak%mOgdJ4h}M%+hUu_b-wRD3>cm>r!tg(7O)_3tPm=uxN=N-GVnn+bnCc
zajR3PBsesUUbvBpM5nz0EARH+PN6KiDMVDJC=rAmXT`UsAw{ggl@sZ(4Ko>3#p7Q4
zSU69c>JoxJMTgND%$nnkI~4FtnF$KP4k!7T#vBQTbO>uXV^v4Or^=lv_bT`1ZD(|M
zxkA%PxjXXXa%bKWow=$P*nqe99i8l~K-=vzRJz&NSz!rr8V#my)#ZxYY05KkF;JaN
z{DK|RNLD&!Mcfpg4_lT(2`MYNAD~5r53x1bIC+_(N3vD&uCxy8B*CZ46&;9k*0)n;
zcs=T4YQ}*~-4|t_O7L8pIYnqVB!Sc{Un2FdE~6mSL%UJ|rxrF#o2WZbm8yodLf8+3
zbmsL>@L?d!&Wg527p7Jw*H=TQO4Deo2XTZtp=W?R=^*lT3&Ra<O-{0bj!Ze8P1z^G
z7cz0m%RQ-<owa)r#AvCN`}>jnM%V&7GB}3AeDG6Fn;&+^7Mw|T%wTQ~b4T9x391K3
zwGk&3W#f_eI_v-%-H)s%%*gjY_Ex*wKo8xPVAn!wQ`pr##_ZB|O%YpgL-i&qFPJb+
zHEl2=&3eguxDEU-dr0lR(XKp|$9Nicexm)-t`&FKhsQ{;eQJD1<jjI&*YIKpB^2in
zT*M8OICp`DE6&=752k>sZYN+RiU|q7^rh&qHN+HoNutuk?Wev)yyH)7{0S{U5g=tG
z9Ao?I%6A7%i4&-wHq<QH;L@~3539uspn&ZM?m<#dPGs}tioW7x^(E#G)wTB#Pw+r*
zbL$)Iz0?D8Yk;(RQasAop72@f*7&}m-{GWQ&z#45GM#iaJeVI}AXU9@W9Sc~TX#4?
z@~mg1TH^TF*rZoc8cIZ<uZ0Ns+D1Z#W}8g59DT^Oz?nwW%jSg#mcGRwE7!+@EAnx6
zPX?m$p6W-nyBZeyEn=gfpKW)JdbkHNjys%equj~YSAZjK<XiBmH32w-?ZT&^tz~F?
z<OMOjpJh<IbTD-%au!Wmxm=07mN;i{Bkh!6nQ$|{TD)9Y9D_mAxbIU;JDUb0swF9c
z`G_lgQTq|dMLrVeSw<>+%t>EO`&8Uh&~<!F8w^X)M+$YoPij;-&hXtr@kBHH>e@K5
zY^D)b=&W9|A~OHFu7huIQ(XT$D+Q+}*Yy)hNsfd<DkiN+BRE6=dWC9x65E?C0;>vS
zk$lBCXB;7CeC4p1CWLh#>)B=0Lg%SCxu$0myLK@&GwMc79~|ntRYFhWcgs?ot=5?L
z5Rh_XATuOvzS#yLX&0cRXwweuY0#z@L%@B~@11ukT)dCQR}{VSKH!(IP%am5^aJ`|
z#6%{Ylv#tdd5FvU4e<xde%Ek^dCl$!A_$+PyXtS9RfI;8Svuhsp&x~qe22vc(v}C@
zR>(mkcRkS7V-T)O7W2ZNx#-_)%Y10t9h*=2kYQBs$CRLbmE?}U7s=x+%z1#n<PcCH
z#^VpeQIg%iLiD^E`|XDFaVLBKaG`*iP!fjM^kr0D^6M+fdk2z7&MjPWrCARFI^Dp}
z$e1;LAYoU&6tXPuFWK4ns7A-~(+~uQ9~z8|fXG+lp6a*gnE@BVo1y4wOFSh1A-$&0
zl0Pun*45N?@Jqq^FpN;jjJ}RbeYx^YsIWKgWHP(t56;AlEdio5qcz2~F?$VuT1Y|a
zqj1dX85l)yPEUK?Exn@W-*bC>@bXEmp+QVy`%%RJP~}m^N|jy95e1O`fxuDX!{12$
z&^WT0qa>hVb2J)`w1J${$}B4B0Jz{YH^DkXpGmb@QG>FMfHF=*-mpd)3ciyhyeGuj
z<5uqq!bCMyU%1f8NFE*qKpwycE#$7N6K$!!BM%S)3XTeLGW)u^+Q=3|=uw$1A!K5J
zdw!0Us(r0X_=Y}<gcHD7Mg&Bu$kDiH^%3gFm4MZRL*iTkSM=QBA%BD}pq*&<dM;m(
zygIsovIGTnRHq~0sUR^EoDlh*!*}WxhIe}>JrsRDHbww{-bSDL8Q2rDHWP`gYKxA}
z-M}-D;huEnxfR3A%ZqPt(4ut`1R^DnIQpzmV4MJKk=1_lfk5-m1Qj&jcKBBK#5DLM
zGmXnZOaNC}4=4%Wi5sv#<jE5dqiq8ya3b&ZZ^(LWSjIAaD$K|f6l}cxhEYp>yuC#5
zl3^-;Yc{@Hx>&+xvDN-6VoKwu@waLxYjS}>Co;2>>Zfl$VB=PiK^Xmqy)?cMhm5~f
zd)SB<(j@iU)TI`MBCG_NOc}3HNVQQg1m4~oi39rvP~063C*^(E0mj~R7v5;(8sIP(
z;!2HdGk!%Lm?LIr$8VSrG|F*-dOC2!BTxp3I%qpAq__fCbAxH_T^(KYCvh;KVsA7-
z2I;fT05ax=y=+67-5hi=SxlvKKJBaI(2*a{OBlE(XWr|!F5fTxCO<nHpD>e1g1JDM
zpm*BX$Q1^XLKxpS{oaQP_G#}b&QZN{u|UPbeEfvQ8qy*G0$FCQ#YA>B=dWbz$H9CV
zMXP9xI*2l2a6Zl$ElzM1lO>YFS6~DmC<QHwTfs1kkq{iyav_}n<%CC*<DkzoM6*Kv
zaYdz9)K#iyHb?}P9_(N8?2?|?PI|)=eUWI;>{iY`To=}iJ9hjO16zGweVJvxsc#=7
zD!Z)6A3G<qbP_=4I;?1*F8XKbQx+B8>JZHCAjFiD+vOnz`vbe}PH6)Wgt1bE&rQh4
zFGD$3GvUh>ovTbycQmt3sInXDy+j>R)!xbxq;kqJb=j~5>e}$bnC@=*K6_e*8YWz8
z2jD4AsKnO{o`h1YIcKj(=?Gg3Z6%0Z_lyKlRz7Uo*jUfrXgcfoUQualRk%4(V>u-t
z$|;mU&G4__n5yo|7Lwv?HWEm^Zn9(x4q4Q==EdpeJnmg0tR4q!$@507UgQnMJwgE(
zSsd4nK9c9nV18d@lrF6YNoPPaF7>mz9=WyAK;@@ncGC9J6gBe*Il$^ym(C1U1u4Ul
z%t!QA+6NQ*lGTWGtoZ;7xX~Gh`OX0(Lpaq9b)Rm6DyC@|7VUaChk#DeNU<v6Rpych
z6#VqBO={`awrtuW36=5*DFW>}8!woO0yKUCN>d5N9Y|RXTBh`BU%G)Oc92?lNVl;!
z{8iUHSc7i1Tb$Q8?p8wUzjo7fB9DMzINZn)QTGK<93VmO6=GUBc#|a_K&iqA1orqO
zJuGTJkL&tT^(g=PaUL@`e|a{-;Ras|`Z&aVD%QF9)6}Ylmcaxl8Bw(Z%+S!wySU*V
z+3_Tp?1+}ak*qtjzg1-;P*vR?J%pzVxemwUKK%av6f+I6(8^nM%W|a-0Yv#{2S#J=
zOMi!cqJr-#R*IZPjcL1zxWFF61vbtVGWYKjZ=tS<l>XIdg{}Jy?7isaIif5pyPX((
zk9k**!8g*9t^~wRf<b8D0lt(d^5EJ<>8Q2VziHj#;$3jw+UEx=l8d#v8QPm{f;VOT
zY(<hA(Xdhhh&Vl{`92EIAnboj%Em-*um$tyoiT+hPBugx|48l#B=68ubr-^yR3u_Q
zuW=_3UBw6ZSh7a>t|T`E8VE#1aJXIss={6IFcJZ9^cf)-`^I1x-O1jzy;ooU48*$M
zIW<DP3BBN29OA(LGbsXbBB#Xn4xJ@bZ?AKSlm#J41I$}IqRw+swb7admMt_gg_jz2
z$y5%aucMv?5p|!59#LYRVN19ol1k-Vp|O7Ec_#0?$(Be6665lMZT7zjubQ$E;=})S
z8$q!F&Rq=+fbtrE?tua1eNHtYet_g?EhTZb>tNb2iX!J|zIVa)E78w22Ubwjx>!H9
zCBywFvr3jB65d{?qLqX$s2ya8HMmeuI#R?Y|G5eBAkiRAHRGan@le$%>6)Kki$(rv
zki$0MuBpGaOQJ(rTdY%|V|OIPD^*_H6e6@PWMu)u$v#s2_%<Otp3CAM2%+h2Rf9v;
zh0Wtc3QGYEb-@k`M4Al4U!uqGy-uFpa$to}X~(l}Fj-L$%in_QP=SgksBjv<us0Ya
zotvM$v+rgMP)5RR!j6cz^#GSeM6gn$80--Q^$ih@x--thifb}RWS1F$@%Vvo(G(yz
z!b?bO6kv!OR6@&uBigQ(5o1qZ@Pr}pgTaMsHyCx><pd({<xpqH5T9av+H;a_oC~SB
z21`lKe&s%EXN72rq`UYlSXo*<HJgJKKE3VL#&EO4Dtkj1=|A%;m4obY_Q35*J)iPW
zNQxomi>@ixk(!M^eE%N@hlfk={y-apG)nM4Tp4AM3H%xGRSj{un`T#Ma{S~$@lXI~
zJR?<QldQIX_`Z2K|FfjA?IF;6FnMx!VS#@l2KmY1hx_;OM-yS-e6?^_-mvp0NE6P<
zd6=h_YeE`O$PkTls}15Wcp*m?iDaM+jDb4PAi&1UH?Nqq*sUZnVY0nI{v05hWOROu
zfasmwH@F5U36!DGh5k6RhN#xIX!|C%kP!)~_$Ykjpbl(s$_Ap-669?MMdHcflvfHR
zG}I@p%Sde+(n5w3<}cTP!_5+GCIB^~zZ?h&;PP&Ld)hGTjbY<$K*IDtGL`sMA6L<3
z5UW>RZF2cR8nJUoN?d8cdB{JPcg^r88X{&j9H6E7K?BL}bvMkIO(H+6rXfehl=H4+
z^zCAcM;E+yERI}!ueePW2pj<arff+9IP8UOJ_?<nt1g1*5iIoW*6WQo-|n@M@Oy9T
z&Fl8_w_Dp12BQ5=>lmt&gq-0g>C|YM5f!hFZ~&vJ(r=7#8^zHnS3Q#!0>I>Nd{Nn#
zi=(y<rJI6~S<Ux;afHVhWszUBkz;Md>+c2+k(Wuw2^xg{B(M>p>Myu)XrCE)bUb-4
z;Jp>!!>Gj#?!Dcp8Ozu~$!A+A8NjLD=&*dxj(e|n$anI~JlK?|TKe|TJBP?l0Q&JV
zE}|wa=8rp<xF^`th=S~bSrXomJ=fc1C5324eqKWeQG`@h#Twm)Y+Q|FoTCD4QoUm&
ztxwz;f3)1M)+6{+EOG5p!z3k!AZPXQSZ`^gcXV|MswDX`!9Eo@2p)7cMHG}0oUbk&
zj^D8ubDH8DdCw2WOL+Y3mwv+j5e<BsMCVep$Zo3`HZtkX>}==uQJ(VL6c!+GvUc&n
zi9ey)*c7OePOT}@S1tm^s1**G(pYCK7|d4q75RT6l#Jon1@1`J{0@MgD%&E<26+LM
zlig%0Zl!LpUx~Ajt<f!oY3-bXh&eR$op3x`RED(JK($*5OF(<1sH9zj=BtMR4hc3J
z`rwuMQTPC?*eEHPBEgb-(UhWAS24`K5|!|IoS5J%mr`tBAw7iX6LMk6vlZh$u_G^X
z*@!(t)&foh!bH`9mq5jH*lxIUHZ5E$fmbf{j5m5BP^8XbfIYw+jq11pQXK)Sj6Gxi
zO7iCQGr;Ch%9;!Q#BW@v<ZUdnYl<sT(ofK3&jhBjYGDdtdg|`5H(0HaLX!t}VC+Hw
zAtg!!xbxIaJGiPXM63xe8b2M5o*oWZdBDeUd0AyMfJC;mC@-rpkzUH-Q957SKimSq
zl?9A@bajHDq7qli_AE*#={6Ua&bZi!WjegW0G05KVeRVb=I*Yz5o89k@7_t4_S0q-
zZ9dV3T569g^RkGfX<~UQtMMsCL>M}~%$o>GAz1DKXB4L?Z%+7?yC5E?y@~J$`&#Z4
z5jh=}ga=Q!x}eS;=Vp)-aqD%R)~s3b0jGRxT_6Z<y-j_MG|xo+Z8Dp#cb=o>!+Gz}
zf6;+<#-l@6CiuV8-bhj)o^+1sjgx<>IP8{E-UBe68qD(fYZYf0gET8^c&mBgs1#U>
zqfbZ?gA8`bqa^)SXvWheQYI&@_~f+&T8Bm;`q8JP37Fw2j@wX(LK7OSv3cdk<(yk!
znBkMlUT?IH03-Oh5pU9a)nV3?U>6R}tR&Aq%30e@mZT4v8MDAyRi)$=XT*fzNMM#c
z4JG7~XWbrBharvME4|8RM!_%vq=6JAKkeqC^KsW7V_J`>!TG;3Gn%1?J%BdJjfB>u
zKUGn;JWt@1rIcl?C0~VyaBOE3Bt=mNLuQxs12Ab>Uc!sw7{*-5HO`@$N#RS11)T)d
zjS$(PqoqAd)V8pakf<;AEW7HVrTBS#g|Z^=!2+}&Ye>-$#<EudE&=6J{WP$cprIla
zJ*tccOh=c7;r1Ib8^1aR16qu+jU(fMp<u$5B4DO<9arGwlMF)0Idbr489@MWpaG*g
z2yLly2a>c2>u<l?e5p3pPqk*2Z*+IWf@X?9zJqb0<*y+~{A+=4+-mT+yc6Go+u;i!
zTpqJOCLU`=FP#0Ol(F$r^aWvrU~Ac{a%b42?Z{DTddv8Y?{~Quoc6oQl)mYlU-dNG
zA(Fl*pTb6XP+Yh}!wpKkWY)Ccy&?Ubv`kGUHO~F?J<ba33rCT|@x3O?KH;ZT&C~BN
znpmau_~o~I^qCgH5pu6~{c?YQXGzXhJ_AmH{H8SPG{Qh%1q7)B%-*Ckxf)CUdczT0
ztS~sA;6Doxnf2iw(hjjiP&)k#5ZN-C2qJ=tD_{zBfm{bJtte7Bh}F`?1-1n2nMT^k
zSBi8hO!R65;`rYrw|91ehb2<Q6Tc|KE%LKm!ao{Gg#&2j)X>8v!q991S(F~4JOGd2
z;UW6wQP8R*$aNPdQdKn2Ajt$5SS+nrEDCJd;BanI2JgxUK7*uma(7OLr*Y`;+B)S+
z`!#z-M(*$Udw7|!%E0Es%T!Il*NRIvpRz82_sSCNeL~=PWl|#a9JB<yd-z2GY>>Qa
zIZ=jEmLN!2E%2Xb&4cf0t5AFlfJ`aVCHr*+b&%S{1N9tXfkBn0ESW82I6r|01ZjXP
z*f(E6Np(v3owvHr4(gJrW?5Kjz}q|;4&gPcpGbX=(6g-MU+wJcKeF}q$F8$sk4(!W
z>0GxiQ~&bRCFy=oUqJX2c#;jTn04)1!uPfW(=#xCLrO3fV;e7qBj^e_m#Kvjx|#(-
zyp~JQ=dgaf7_x#VMl%rE@CTaR@!dh=L>eI9siqkRh-iBH%=Ne#4Y83z^e^-P!W9KF
z7jae|Qu&c(q*u(EO=75D_#^5Uw%*-8TwJ)PpC3xK7qLRvXv+EQj{=pEnra^-w{)Ir
zZ($gEDi#j>fOVtsNBjZ5rq;r0Pgu0$D~R0yq<qGtIfMI@7P`j?MyBoH$j{d9(E81%
zoD38YW!gl9kJL>Vmi(hDTtt~<bJqQeap^l^u|@>kb7||RY{lUhiWs-qpBros-D+Vq
zm08y9QKixJZ6bSjE)fRNhX)O<IN26+-k2`MsBjX3M$g$<6r|+d(tv{S)jxWu!d$2j
zHt9>`)9jqWdA2wloiYt5m9c;y5apZ#6UJll;1|rbX<0>)BRlz2(77_;BPQcON+mrb
z@AwDl{O6A+OewM_p-ys?`8Wx6@i3x2LQsb8DfAixkC!PzFs|7Fpb1H^oXS^$3h2&F
zxnK8SpQijmEra6;>I+%dZ^A(tG>Z^Wq~*bZ>f?kSs<w><v+d*~*CILZcbMfDllBaq
zb=^hw1*PKSCvvx^2zAh$`SM5uA8zz5IR-B<M{%C`0GbYVs$hos5_FAVPqC~D7IAbZ
z61oe73+>|UAb*I<IQgEa9+#?t`dK=_)zyv7=Wo9TNSrUFgVzF>kXI@WGY)%7%zSJ`
zeGTI>^jf+E2L6mBoU<70JdhA*KdZ^zN4VAtwt!|9!cjp^d1;UaCcwfX`mKUqAGDre
z1JC-PZb;4NpMXbbm13yx6PAqATtcGwPRv}Kz-Nh6PlhN}dYCJcU0W@_q??Cf3>cKR
zc(Hqkjs0@TALGWYtY1gg>&2Fs8we{%BwWNUY*V8SK{E2;7k6;nR!-AK<P@D4&2SW<
zFQrS1$Hf>>p_#M!$7~F)LzXLp*ub*5^a170+~Hu3(8)gMi`09Es>I(DM}!msd$DsI
zqK5m^L%6gkikXT^W`n-4X1l2tmsI{Z*(F&nzE;X+keu|t-5Z?BB?zZ(;v?({TEV2P
z$4beX0UwZVbopdHU-t<)k&e)U%N;`iTTw<-j_3lie0(5C^umELiJM;6g)9zY^dmHM
zwuJB<)Og{QI=a59B~zNA{(`OVqCXZ*Kk>Em1m^JrTN^ClvizszWkFRVmLQnM<MReo
zIGSORe;)C5BbYo<v%qyL1oey<OfPy+nWyTzz)}$~1Vrw2chK33*cZNaPJhstIv((s
zlgPh==|&3#h+{9Y6DQ4hS&vgZXxO5-o^+VwXkNy$0`I~Nvij*t4e8QcxG%GJg^R{5
zk^n(ek=Iy}WOHB&nR=^Ii4FJ3;3&b3chs4w?tQw0b(l$HRvMLK;!8CsA;zZ$@d1c=
zl+6K;NGvW18Kh5ZzZb@I(^)%j%CBka_`61ta=!sle$ugj5YA|tRx)QY<h>yR1%BS)
zhc-^SZMpucHS6>C;G3mZ#l;okK2g1?H=tWsN<De40r=mu&*6O+BM4}hrJhQx6|*KV
z?7)IdJ|Y;Dt;ngD8^ac^`MSIkhXWd`v?GB0#GzG<LNx~Poi_+@liPuk6S7;t#cAu*
zfwDebx2#t-;HI6Xhp5(d$16e-43fYj%w(|<%0c3Ov=xrrrGE(ckFr?cx;_)$L5G~t
z)YinKF?3MqV-Rx^KqJ~}cX+(ife#273sF;cG{lLZH$KCCd33jU(fxyW)?#Opx#j{J
zAn9w9iHJrBP)l}gg$D+5ojRDcASj&Ki`NT(S4<sQ8yc3v1DE@Pq}knP1g`|sC0L$W
zl}r{`zPu<Tu9HbZ##x79)`RY~w27U~2Tnd#uAr9ZTF#KjH~P#?uBgHnO?0Y}hI0}3
zv2iEC`Tlt7!OoP$1vYn;Hz<%TS9TYF>|aWcx%G?+{nFXw;{5PP&1K0#k%|OFR$(2&
zq!LO@T!qqJUtwo2E~QrW+Mp}Q%=%)o&ot!2(B2uVDWr$jYLme}o315eK#v+F785Kl
z#R;T(N0b-&Li!YWoF4!6ujQv?1CN(i<gw4(%s|u`^ik@6l!esK(CnA0a!?rS*Qyix
zO6rBg6xuNDMRa1riv(3m3&gTlYM1@_U;oNkr#vxZlr~SNxC%qn2hpIG>?Kui#vmq#
ztFrnnYU~qwjC);YB4wSMWIRO)V^Nwpa7KltVeevNPK(_j?#UuWAzHd#5w**Zu3D0V
z;HGo|xFa0No>f5JP>$Eg(HaS#4Upz;sF2O91CQe*)pYFlj1$o8ZJ0xdn<FMrUS%K-
zxF%Eb?U9piI5$09UcMWTOAT1L<!fckhLzSeW(cl<bgZx|SXriK<Hz&<k)AUvJP3za
z=Ry=&4P9albqUI;y&bsHI3>#x#t0EU8A2bynJb`3?)0n0gM{@;^)$#o$ako$!d`SB
zj;W_3J}jyFAhyq!L2fcJ2W-KmhWq>alwfj4K6qPbS~;X?E&;qBFR~j*Z)ET3tfpn(
zhEVz^Ew#CP!CD`GT6xe|)L+_K1<rTbyben9HWRLA=7C4C02L6SRq)s<>!Fr}^N32P
zYt(ARR@Wk^a0w8T#3gQ#N!wsYGuKb)Jnvc(>%V2_noJZclyaF8wfVnFl#3o`=|ZdG
zA69yeDzGCYD3>~u=Hw!7ipQ^p;}D}7N_+Ng+~R3Gyq13hv3sbXRUD0I%7<wZk;tl4
z;&gywt8#!wtj{u@>gJk)l${mG4RBca?`VNROsg735ot2Z0^?f%XNY>}gf*mJs|m)`
zV<To>t(b8<(jBHT^TxY$3rS0H4>nH%grIeFLG%OvOzl>NcxJ_{Op5$I;$k?CF@K?*
zfxxKcz@OR{<j7az$7XsVE;6QI*8dTMJz8*ewLivLZvZpK4$HaiA>i90<T>I-ZvwrA
zMi~<>IOinBk@W#I*cTziHsJvCCX@>^kOYB}^d$mA`6NvrNpA?r^E?e01|3m-#%XGT
z5UWi4S&+NaNuQ8g)<X2Uz)xugvC3oizQkH*oK3xKbKtqCpF1j`>QS~Au$o{49u#-S
z58$mVItcZpL5dXv*T30pJz9SFxb>%%N6SLhtiw2*6bz<6|HMMO%)L`^{uYc_Aljr4
zX@{O46dzdOc{CiHqV4z!;hV4<umJPbrT9c(xmAr4;vm5RDLN_~j@z3;BLTQWM^KJH
zQ(DTl*T4PkkFly-=uU>fNK$~!7UwHBop-SFYLCO*$W@67`oQPcABLl{eF!x+eg}gS
zZ72NBzo@P#Gy$U8B*G&MY+!<Vs(jg5i5y1=<_bLvtQjaWO%mLK$3$(u!d0+x!2z1&
zF?CMmM#ub{Y5As(|F8Mzf2jHB#^|D?ng=4)o+gAKYG@<|sUZ%#5~Y@SS)22zN0w;K
z6Dkuzg8Spw3!Dl%r{YG^->iv5dWrcOS1k}DgQgz*#V9^6BU<U5bSi~G_x5);*Iu<>
zzuDh<@m>4H+Sc~l-OW{@DB2Q4f3U%Y*i5|6JVwls$Z<f2Lo^dxcre!}5U|GEO7YaL
z0UbT0u#`_~mnv!mWP07C*WRLP@b2cnzTMj0-0-Pex%$e~W$#nzKn7{i>;#GxO$7_E
zMm5GN-woqFzxYQpCb#!?d*7$_cnVa1>#0Sw5$M^<>L$sH)9t;je?#TeJ(P&mO>;b=
zaP;3p$=JGHa0C16-OWApz?GDHK~084)P-{iKpw&dsaJee;IeNM7KaBY-HwYLxPil)
z6$y$LOO94PtKwdP55noa84I+D<CxvvEj|~e5$*JZ;}hUZO5pTwaZ5$o-!TVtqDPeC
z8q=?XAiaTQ%stTQ_lcz?Gm=Ft!zJbpoHwhEcB1rP{^cN8JSv-6g?Xt>8`yzUJ<<(Q
zSmWSFc9HCxsuToGe}LPB%pPPrTGb-7tvD{47RTD^4w;?;xsEJjp#aaQ5hK!-y$JuR
zvw;)%Db9Fy$QT9_SY4iDmOZao-6_8APhMUf6>mtlIDX*dh|kofstXh9T7g9u#w{FA
zl)J!1t`~J!9d0b<ExbD&7Brf{3*0N#cD4%E%|o44<gOBz4FbwA6fZ{+Mo_|ZYzIU|
z0n>Yuu8+*e&s$HZmjQ4;_GI4PLja{1=!3tCnA0E2oh5|2@Z!b-135mM(PPs1ymqGy
z^b2aZpUJ+HZQ-=xF_#NNMdy?|iY$-5OJVvj;<^1k#IV~={u(@DXxwjxM_2}S8s~_S
z1`#Qug(W~XHHc)A3ABR~AELKvu<GG&f>qm{5?0}X<HQZH%0B-Mv1&?>wOFMQ>0lov
z2qxm@zYfXDlM-go$bDzRicmfe5xhc;`@%%r1ex$a(jiRI<TOSEkQTmA0dq7CfRq9*
zVdL{l9O19~Qco^&XhVHXxU>dvfFf3r0;us(V>w&w8P;>i$*_!sB|Z1ZPoGx>f54BB
z7Kwb4g4~<43l{A|D&p}(AmAN@g<(TEb~>~0$g>Jdn1?7X9vmh^DNGbW2n$%}u7L&2
zJlTSYU!3$o7Vj&=&DZQ*IH?AE$FE~?QFl$i5ty8L3Ci=hnT(5Fo4JUrrz+>dPGzEw
z_BOQteK-?_)X3o5;+G--kt=?GQ|^?KOs&wH*=sF_O18{Z_1{G9#hpt=h)5Ao&s<`3
zXyqyi5kyaki9^@T3^h`p?S&szx<W`Y7(${?U3=srspDQPEbAFNTRa@Vt`~xHG}e~-
zP@nQr)ejuk2IJv*kJjGg>)sq!Q9nNfrrvUXvyOzSs0z$xlC=W_%p*d~j*c?V$dPWy
z*Jlnl=QU<hy#c1psb^pKIF3-7h_N;-^HQE;B0NgI+!<!tMuC4hbA`Xe_PY5cvot+H
z#Xs7%8M5VEKA5U$LggK@mv8p=k@n=vCHZ1OD%Huw6w0JrvVre4W!OTB+OYvFbd)nL
z7AoywCVz?C$!mi#9kurHjwQyokK^XtL|lZq0gyhdh`O^r3WBCqOap6nN|E;>S|uH8
zJnp-NSpAVRx4*Hwxxf3J`J`&jUr+$vXq<fAU7<=Ol-#<UoWmPIGdNuidhEiY%zI)B
zDcz%08!QfLi7~ugibYiqZX4-Qq)12!2~=A1o&pBFA*Ntbq^Q!E;uJ?T4XK|V7}=yG
zPREt3gjI^8#<+;gfVvNkPJS7ZCm&=m<J|bBbJ@?YS%S^j%p3R8H{=$wtpTDTR(1V#
zTSG;~jl!>ix)JiB-QMf1Hk2})5Ut8N(OR41Rzu>RiRgZf-^QTkH{b+S>zqJo=OI3r
z{u0gFV2E(-ml!VLDM^S9L%@mQQ|>3!e-s@;Z^w*2RNiuTY%GIUq}N|d%pk%%{sDK;
z+WCYo%MkG?`&9Z0PMmpL*g@lX(0GTx)79TG1HzP<*6v%rzgh3~8%;pwG1=VF@Zkw=
z896#>EPGC!B!p>EFCqu+qXXhwgXNhnC#}uF@sJlukj0R(EF#?8ZHQ(<`!~6gAMqB9
zu*y`-IpVA%Nxczh)O?lM?VAZ|9G$n#*(bNB9xgJIRJM|JyH7D3tQJzKMpr>_7>4Iz
zddg=i!&NRdV)?jaVP2R77i(m8P{eL=3X9(;a58Qq2Jg^<_Ai`EhSYjm2(#%QTi^p9
zAeP?*7>*@Wcf!h)efeEH;*z$HM?5U=CYF*g5$vss$4rW0obxF*5>a=`vs4x;t9hhH
z%4|Q^uqcf9qxfN0i|2^WfXM{ro|@U;5x4S;QOw|GK&X<z&hFMXYx|q6{WpKxd@T;X
zx=|vON{4t^N{o?ZfU7~sPgOibKlkJA>gpT12_CiA*4JU8Zp*0Ddu)g0tf@zB?YTS|
zjdu+@j6+F&;WBP=L<-+z;s*QMn=t8pDO#&M5Kn_re#@X9yQ6k9bH9k6zFioLBxFQX
zzT8XAlb1KNJEAY($Lwil=?&|wG|Umb{2I&iIvR6_BiMK;X8ISnVX8_^<S{-niypF}
zhO0q&?sZ@Y#`#unRFv{=Oz%f{t9>QB!@Y4fvOIGmaJZaRip-VOt6pBdy%cf;Rah)Y
z(;{-*g6`L|U&<PY8A`jjvyx>asdXj75aQqKg#*%9qLled?fijgeMt@o<?6__KT=hM
zqvxMER^G5*ITzIRs%uhfAOe5DCZ><W3O2%7^Cog7jG7`v{>$6ypZF|YX895v(mgbv
zB8c+mXA3z0jC!Z-f4_o{r0HYfu!Hxj%uezYNsLzymJXl3Lk{Do^Kwb<@L^>|fe9On
z0YPSp>~+&wT0ruNr|<4zu*0Xf9?5Ch10BNJQ)3$r_?kTH{{7!a`D<ukKJ`Bzt7zW6
zT0kHWgQ^uN{0(A*B22DR<SK;4<PsqvL!_ec5DRo7*I|M2ezM5~r_%yvQyC=8nR`lY
zH9WwHPTYswWj*Y;rAE$(!U%u!NS6ojNlhYTdBF!akBV(e_;6iVJ4M9JVqC1*ZuLf3
ze$~bx9PFza(NJNe06lIHh9qXjNR9W($@aK^d4||A>|_7p@*DvUP4EoAkAA1IxxVpo
zv$eao)>_-#d-&)tt*_T#wf0`FJ^K7l4ExpB`0wln_B}rG?;rnVrRMu;gPEiA<;LD#
z{nUQg+<18(lg5PONXbHGE~5N0|1S2ItsEc4H_4LmeEM?BFZ1tWe_<rT$VMsw2TyHV
zsxR1430$?!t@orK191JY$<tc=z(Y1Mq%kO1%Q(KZ${rBc@PPA<ukA1XSYLa!xxKw5
z4K_CSTI<i(ANvQNbzZ;S-e#-Mdf~Utzwf-+-Is@GFF$wJ1qnZUweh+9y|ec>`Pm^_
z%WuFT!LZ5*e8H`%y!hMh^UupqB?Lk#6d5Emp&$rOmAX4}Rk!%+N%-4H5~BGDPP<B<
zSPO%WPCm7HD3(`ZC<u>HFbDo9e)4CurHRBYbf#@B%d7O!<~cffx$uBBS;gX0R!ikB
zv5GUFz8PbD6?G@#Qd2PSplxc0!oZP4SQCY6M$jlvkSCjNZMzCw$`=ZEugNqt^Q-3q
zVX9Bim|?1H23QBn(Q$|3B<j`ET4|vHU2zCSjiJfls(Vr-W4*(g#ybBU=7cV-#sQjN
z;JZBQjGzEX`W;hIMQY(3O<yH{!0Lf&e2wv0M<_9<vXCRHUvM<E9f=x`iXM_yfO(GE
z5Y0@P)<97ToF#E;VL}DDY1%qQrq$7;B?U-iFuYy7>|MY<Ap8CqNA>Cg$v_h3z}tUk
z;sULmV4PN}_o#t$U9wT?QYptp@N)%#&+!?qv_d@y&yerqPk;Io-q1=|!XcQVYzGBT
z!Z)t)ar1FnA_!dh7s%Fhqg%{wbY9*C4}~?&6>qZUu72DI*@%-Ml1^81Z%-XTBT6Ew
zSpar#z4L+?F@0))Yfeeb4RYPRB6_FPa)CtqUEIv)GK&we7&0zg#xCGD$$UUugj5gA
z&#*NsQV`1-R+cZ@GNT=1X|98k_4&`Nji26PZQv$Er#ege%0u;|Rn}rI0NgIoLga85
z<SC9^G4?M7%ctmWDea6AA9E}^u;9FCZu$%;X~nL+O+!<)D@DcE`ZpNEhx_!6AbB-*
zk{N<7KOt!sYg*t>)R-uZgru8Z?ONX8Q-7d6vM)=VQ~ee=BLC_X({=D7yJryYYv8;Y
z<%Ea6)olT?dF1GQ_`K77#jH-<b!0f${77Ri!$P8U{O(xYsYGpDN`=u6=>rBls2B8)
zEK;Fuw_k1UzkIXNZri;DyGH(zE|@(55v;4cfnA%}DVx44(IG2GD&r{u)4yj)=8EJO
zt^Wc1IN<XX+uQ>BYlyGvc9KKAT2V0ZXNDm%y~gMW`n`3bm*hWAgq{>R6(tMbWbFha
zfd4av4!YAj{PZf`ATf$Ui!&6-Y}r@z(;Fr%5f)x@s$d$o%6eks<1pmuR3srof`hbk
zAd^ad^#}1E(L*cXBQgfRESWBn7!3zJp(Zels+?x=N?JiVgWrN%pUedcMKY5-wM#xB
z)zWA{1ZmY#)!W4XB{mh|6~=FN;N^}!63m2Gdoh(q4>2t<8uG8dEZbiQ$i$y3E06K_
zczpI~aU4E)6doO07kI*5p0CS{nn5mPMOBQ*#iz{hz&cSETXgmwnqz;>QsLK8d5m2J
zP;p&Wv~5&F8coCnb~X56fZOPW&T>`EBR=Oq@(HwvXUU)AFf|Z_beef)vvu;gflO(u
zSqFfS$$*)NrG`5WkVuH?e$PlTfvM~dVQ1IKA1V)m2qFZ_b1pTE#><_x{g=*MKyH(>
z9V96s0=YFA21L*R-=G-@9I68A^oF)5b%N{%G-JnS2agy^S2X-05nTPyR17TY;0JQs
z-@(FBEp)&DhcR-iCB=vsAq5c2KdmLOhm|9n&<I9g{Bw%IxdPF%l5F7k2dqbrVzN~E
z*77o2vtXAVUtL0#A0t0G>e=$UG(y`?9egFcz#DD0zg^pXz4iL*;^$&hu5_$#%CB!x
zA1!+)4|wOp8f<Lt>~5~Z9kUVJm!9FG$kzHkv1ERTyBpXVUee@wOe0!Y@I+SEzPuWr
zHMJrglPnznKuLSv#ryIn($gP{|BwX`D7F`T?)W$bLVymhw@G5~gyE;d3T(<l+7ySE
z*$mbD;7WvSpjLM(Zw#;O@bGmGzZ+x}QC%|x=)hs$uo%<#0Vy_b)A!fr2D^{NJ2COy
z9-sG*dy+R!&REdDP{D^4`_Py69@d;dDjq735|*Bm!@#U?M~?V@r>K0boNgXToRz_U
z`4pL$1Rw1`(=i^)h$Xr9F1sHO`JZh}eK2O3WPqd3iW{1<so9XYG@K6G?bk4*?Cq?r
zZ?@Y!EvDbWdkjx1?~%f~fN}*0!^cgg*+a>pU5*G3fb4?M5o3u<_-B+EC#cCOy5l09
zH1@Dc^LK6tf^|Fd>VOy%PM*%WEvQbBQ~n^+^9=fBB%_p8bw&hha6OrPcvB7#DmMW0
z)I@T|ZG<74O(gK3D2BBa@T8Jnm{OS%8|iOA=x&P_K@QPv>L96)%E}&ZP)Ll<@p)eo
zd4xNeZSWst$vbE;DSR9Eg#P{=Z6iE~nVuF}P&sJB^e*W?mAlDbn0FLpv%jLvrnWyA
zaN>(>D#~m`udVdG{`FQnYpiOb<ui?wid|Q}NV?B0G+ZsybZLGih*CSmmZ8G&_pS%r
z_@9GZ{Hy}H@Tvm4IGtp572pkqVM>(E3A8d5CO8=>yb*UQmy&Sd=JjxgaNf~}%?RS8
zU)2Zv!`2hr7K36ySaPH6#3(oft`I!pK2v=Wx0E3PA^*0#gi#MFJmd12j+R%oK7NYN
z#9hx^zX)M~c}rij6Qt!djnlqh^i$r4KCv%FtI5BhWW^Wx<d_F^RW3?ezuF2*3jrlD
z0Wm@#JmrqG#>6De?Vsy9{mb%Pr57S6l=CUw{H3lNsPXIv9L|8HAjP@gjkj8cqZy_N
zTY(b-?~5rvYD|tLi85Z|49jWYRRAv;JQ<H(^+VrMSTC+ozJa>Ou!d@A)=(2ZsDC|C
z`>TaWLu|Afx#gyqeH{}WF)o_Is)Ji*Kn}fACM&Yw&}fk6Q^^9#hGMb_OQHlR{KD=u
zLa{RqQ9`coDi>d6s3n4aU*I0!+fjdW@E$k}DI&#jBkz(+vYUJR?HBM`yxra6><Gm2
zr`T>i-81yE)Ggj2`3qjKv_(j~lr^%P#L|Vz0joQZw-U4tmU;&a+VDmSG^O+Nt5H9*
zYS|6exbQLZja;*GV^jY_mM&DqEmMLaWu|hGTrVD~5?qN0C?1i6%x@EOqq``tX0**W
zE9s;hA3F_kg~J>FRRdDa>!=<(uVY9Q5EzT`0vpEatdU}E{yoE_vhTqsHIAy<jkV=$
zB*+S5)dXdU4k;>ujg+fj)lep3h6EvO%ODPfNCY#wsccie@q_==R+ykN#cNxNh`Y9x
z1WV2BkO#HaN<n`TMD~-w<(D82z-}T6IjL%hmQbk9?j}9T1VsvahrT$M4)_`Bl^J^g
z<IjQ5y-`SuXRX3Td}=D6roP6|dNw-ge|!nftT9zD`n_{_il%-gOOW9)DJ>`c^5r+L
zJP)*`iN%TP8_SggOk@MOxs|qbX-nwl&1oy9_l$b=tF?c5v+JEx-B=87wq8$pv%9&y
zxwg0IUx_#-yxep!eQjjz%KXV%Ot2=bJ_K*j9y7fg$ZEOJ*h10@<hZ_O9|Rwx0sSp_
zL2Hd4u7<Pb60GXYyW=llNaO$Rj*-Lhjvu~OSB+!o`Qw@9!S8we^0NjMw}v_msHjyw
zdvFKfZ(s4q2VW=@jr4DLIS`W`_vgVKm#piMpzQgJ0>6nGN6Tb1DFJM09&g+d!ofBm
zRz&I@Ij`xbdNAaFOD3dX8>+)q#dn_bY?s~0U@wGy>+isMH>!DvriD$)1g?-0U<CbH
z8^eBYwcw#8^l&-}=Ck}91K%DqtGAfNR?jkQ{IqxosI(9_&OGzMxT9SYP`~Eb%~VOK
zq>!2;>MDf04dDF%z`-<ekL1<b-`oGX|6N=1hQO+|xd)hhPW6jG_cRLg$)&E!J9wta
zH!IJz5@gm5#80Ie{Nrc6b)2Yqt(ce1{a#FI3Yg6?tbmp8|9$d_!|#v!J$`G|Fw?n9
zCMo@ntC~@&Ty=)6H*%AbB~n?>vRo(=N=05{T{Lmd4a2)_(gG#0U_4~EI1XuGUaWR*
z53mu~J~q?eLQ7jQ-}tXEY%}*lX#zrLQ*47sZrvF`2^QnvodSvUs0hz(e7C*P-rjn?
zySDpXn-=FeV5OZ8gT0{A2ug*F#jOed$^fkvYFDr()!BJ?p<`I#g})GHbA-b7cp&_i
z#LI|Qxf7&NnNl+Ykj3W?SyI;g`27G(&1$*ZE3qQUGnSx`U+|fJ%$eXrugC4KHtakh
z?_HyT`r;V2o-l&EzO&b~&?3+nLT;ej>c}z3lB6QD-fNdk>Vc6EwrtSYHTx1HOL3m!
zp;6UaKfrn=W@LtW_*W4qYAt3Dz7QyVR<YSmHpw(FuWPD6-}wbArZMh{Xl-CHG;BQy
zdmMp{#TtO{8w#ZZq+E;TpV6KYo@uibbkh7~Z=?ix&0yHPP1@JB0d4E^JQ0QVE{2&$
z#ZS!EJ*FZ?Z*7|x=`yk2mjiR`6Y@HGw5WiSHPH~u%}^yDS3-Hs;<|45akxc-*6@pT
zZ}uesyIO2M{PXhi^1~Yrrk^}|`1$hk3O*sQXPxI0AOfV2)FLizr;Vvw_D3ax?vRXj
zgTZ0kgmC_{Pf|vF|331Ct2$YgFr3&O#Py~r^8-SasS}ulO(c!thhwk|K5K1v8XuT6
zF-u|s<b988x0((+ZZ<Puo2mkqR%T{Usn<&aiY|{_WQk4Lm{sW&(Yf4QK_~nZb1G*l
z%AhQKFx{6a7vpEc0x2gYwutUCUdc{S8qpeX9^fs1#7pk1c;HY8EHL93lGkkS>IhpR
zIWUzhisnz&9K7|H1#GfdFc{_(CHyL7jf+D=G{8xnkoI}fJk7ZgxdQ>gyVrA6kg2R9
zV;VVwZ9^O_y_r_|Jim%Pw-YX5r&BENU!|9IO7Ib`(K7aBZVyrFa0(`{kq&4sgM>>K
z_OxhzLgk6UaG{X9L03%Q7&JpJB*7HoW))d7xpqNHi82k-4BE)?1-+Dhs%YT+z+;0s
z;_(uGjaCat&`Oa2hYABHhY1kf)8jl20z0`Lw7S|GBcw)el_2d67DLn(;w8<`a~3p3
z=y7h^$0m>C(ShHkkeC98$^=0?Dd}V^&A1QE)Nvq(&z%M!(`wPeLd?P+MXTj^&-hmc
z88XFnl07_>(?Ozm2|zl&g0<eHxuYbSM=muL8DeW&e3c&LSSW~*%LfGe_K$<Vv^n=0
zqVD}bcm+I>%B2=b77<=%3z|uC;wPBd!fgH&C-Sd;dn4HXGgejwTgrtZGS<vBpqb5+
zpcZA%`n30{Dx~1C7t%TLHtH5}Hqb$cZZD9{PMFSVCZNGvk_K;);3C_ud;&8U0X$96
z8Y{`Ab%i71;4g=MG7-5s%CKi9zD!5aq;o-c&S&e(F{t_VB3J$a9JTY>&ErJpb>-CH
z2R<i|laa-Ar(H*`?R{dAd7qR@^h2R(V!GAc5x$zmp{agrG(eH6;~YpA=BcVdeCI0a
z@Th#3<5}E!KOUFGSxpDdsd!Ppv~dwe<&d8oIWiy`RnFW#K{o6i<WGMJ_<)N<YjfwK
zoi5c3)j5L2c_w>#EijWlbu4#m$_s+Ut>QIZMW_)|k}m|2rcKYmcx&s;-lvO`;o#Gu
z*xv>?U>uq>J={6Fv|mdm9S=}H^k1;O_?HG*gz!deMagEy^do6<cl}s;aILo-djSeQ
z{vnBx)M!n_b$JLe^bOJrh2(?ckMqZA{>*tk<=Ih>0aPL4hp+h#igbM9+Jd2ne4G+_
zf(JMv9MRM)onQblxVJBQxaEt~!R#3>=?>FpxTnA#GNlI3fML>{UE;a(wQw%+_H{lS
z!t`a@99)zd$)1AL0KQPJi7#I<1c`-M{DH3!0bcS(;`8ix*cc6tvSx*(uZ7c&lpHb6
zX*+^OcXQfyUb_Xxd*QM%4@<f~vRLLr!b;PP19=2zyCdd|LIT!<`HkBjKVSZ9TbONz
z4OvwkQR0lx@^kF2=2W5cnCC?e1<n}>#jc8g*rK%Ev-<*?*f5s5g}AJ6@tw@(6g=x-
zGGT`Hsk5DGhhjP&U!M0NihepA-+O@n$sHw~+%=s}pCj)w;W($!1<3GonHFIsIdD4(
z*gR-}Z(#k{c)e!^1mk-Nm7GKUzX%n0^@A0uiyRe013uV8X~jZ)rDqP-Hm<(F-HqE2
zS;m|(pdEYG4Y<SXI1IhYW4uZ9^0d^01W5<|(uidGk13YaoP4r)@sv&|1ku9}y-(1Z
ztzL#>`z4-AK%>2C`U~`pP|+hqJ)%WM2?8<g50}PBds*iP4SXQ%LE>r=aE<d-i~pW)
ze!ca&*m?VWduzS;+va!jkWo23l$P;Eo3A(OnqVaEGMWLB*n6DeWJP%XK!P1K!X!SU
zrEoM^DDIP=BK8-XC=cNJVr+yw30*7{EqA7+{c>X9e)o!IOX57?=1fwA1n;*_jU?!}
zdDeMI=kiDq#rGA%0h|~}f{ugMSy$l%TV{RDW?+Jd&SNcya=Cx!_Eq*z#Q4xZ?;dwX
z2)cC8;x3Q`Z~J$kzI$?2C=zG<ibNYMXJJW+@KF166gg<`nu`k<Isb(gSdy%Yqgk5x
zFh}9|;GU=@1^t9mc4ryfuApt7va`d-#S`9IV<>w3O5~Z(S%yM|oY5FToz^e9O`0PM
z26k41A^-?IIOQK4I6~!*iB%5~mgXzgv3xPRz0)4+#0$T*+!x9G1BksH#&92e6Sq3=
zVZDumPV2|z)?Zui?sN57&j3OG07K3F^!-oy!(ZkZMc_L!WB>BL!;;`alo%K0?-0eI
znfITlW26N_;D&xh%nIsZa0VYTq#qYQ`g(6i_0X!uiUSW*VIHb=-F^h$8TZk{j{n~6
z9bKKil*>%K6@Vicj#-<pziF>+Z@>8#&}yu|dA0L~vH9|({dcJk4}KM!2ETUFSev<T
zKQu?`{r$DweWw0H*J9K{Br;wHtvL7}TunyCmJY1LU&I!0l<mQEy3t4WpwP<foyn8Q
z#U&u6t2L$*wzkchSHPvK<4IGY6u^Us0daUF#$$be8HdT)sZ6Ja7eOBg^#nf(WXUnC
zZJagQL$1)>*Wkf*3)j>4*4JLQH@0>Ihsg#Ig2xc}kY2{a-QLMLYs3!+n_kDT*{BTY
z7GTk(A;aMu|2UD9&)=wR7&Zz;ML~%0pE9iEv>|TFdB*$8YArd&&mcHJk)7L$^iUBH
z2)4B15k^siqn_>ty<T^`i}*Q6rlxcm3RP<Z@B(O_z0ib;A}~|}B@HgTWgwsXyDyA-
zg}W%5(y62vhhUK~9*Ea`uH`A)Y(;xd4McGRmEDsyyOjZOIRKO63+uJAwO=tunYF_#
zMK_<(oS2|&!WE<%n^px3AYq=D;dZzZW>2_*`}98c!Pun?k(yrZG^g-ZLXpWoyBeJL
z2S0d*8hax^WgkSL(38(A(rzHH(iXE6x5s@D<0T4T@`p<lK11w7loH=R-+B%Aw|dG<
zOvba6J+Kwv)(ZNTBbzS9xka3vAhz>~FvSv*;gD=tMs*;C3-DcnrcqgZ2i`8?e_+5;
zhkgyWZAj+4F3z|1H3;I{;phjriuag)!9T}t0KLx<!%!j%$$YslJWDI-F)@)h6%JrC
z%!XLkS#tueQ0V00Aln?g@59@60VN@hFKFd$BUA(R2aXybl}=35e0a|XeO{tVX9%#7
zAZ%N_L8fJxCkZ>lmTpN@9K)iRumQ)GW^Dq8Dae3blb{}a&P|HO0iS^6X5-En1hq9?
z@CqjE($=8Aff}X?^aSRD^#;^u>;fl&SSv9n5qxGdQTRGID`=NvMFgjy7Hu`Y<3vOI
zov=Y3RG=8>9Ys~}cYfZDK$mWuo!&7E%=oN-Qo(=>REpR%p*X0xL!y(ZKGRrI@v;cG
z&dW7NX{d}anWrO^b2po@Ccfh*%G4bH?6p987RbUD*|sLTL^v7fNqn+T05at}qrGr!
zwQsO7^Q={iM&vP@Z?;`5Imp3{B~V*6(#F#FIP*7u#OViFK+qT6sd)0$;yr`Ky@mOu
zi$2g!`vp=7#H>3;tfMFBr(euJ6t&!DL<|Y=kr@#keZDEyNxu;|b;uOxpKw6@A8rE^
z42mC-mtq#JV@@LFY=j5|ZE_9Ejdj*hXb*ty9gq5`hmP)(8!nJ%@6~kYJL)kX@Fkj+
zPmCNo9~m7}QFo-?mC&%ljA|<DEb+1u)mTtc%$kPv+i?e5FQ|=+w*?~nR0~NO^5(^K
zc`eMVB7ao5@Xvsmeuth)&zs#XobK1Q7aI`w4hx2HA$DtLZFg<|%`Si%rgf7MqIzO_
zw)AZqX*%Ateq3Qw@Rt=L09_z*t)`=TizK2}{E4biimvtgR0fI$HAd-9bx$XH{H*MH
zj&!!^)Sz6I2t}X7N8ajyLPPHi5l-J{FUhuCTCzh~1Ua5TRrP(=H`+jjX%<}Snm=!N
zasp(G2qx1Zav_fN(JFCF=Gms@5YafYQY=_MlUq%gtIS0)zA}qh0C6uq21TD2uwr2%
zlnf4#T#Nd4>-EN)Z}<ER)_(qWYkMP;quSy4tcRPzBU}W)e{@>&auP@HQ0^~}M$iXk
z%(>!lGN(6Z6lVZrzR`$gGD7(yXM>HU<{|#wTbQHVXv{yHe|UH}XB+hZk+)b&SH`17
zQAw0*6M=1u^akA3Oe)<y`2Oy@`*%em>FiCnkPk`Tx)kZGeo>9@|M>HlPnrt|lcq}Q
z5{$M@K?(UvzMr)?5sI=nTN&13&}gu*G`?NLFA>wi66<?Cn6qi!eYYTR`(A~JE)sJ~
zFo&Ji__=}t*B>08Uv+z1Cz~JpV@gDk-4Gbg+o;<>v5)V7&GwOM1PBEPa1p3LdlN;Q
zhQM<7Z60LX?Zw8@;^NYKh|l3@3Et2lKUjohAqrw=ZrlZW!T*<f;D+>$&roS5+<qfr
zUR}!VHyo9i+nqPHF+NfrFa}x6hWsCB(UJfR^TrwS{QUnk7EK>mM5?%6ui(`YUMU3s
z@wSV@(dm+BXjqb1Eq+b&0hEF<Gv`=uj0}ZLmJve7NGia?)5DzR6@Xvdv5W0X_wK;P
zx{v50QDS(}9Kw0RBF5g~B3iIf^LU|n&5F-!u&&~WHz>Xwz(&GfpZPC~-QE+n<-dO!
zj=KHg`SA3#-x&;xFaNQ4WWPUyOz18iNmI6dfhaV@Lm`kB!4b@cgp2ZL6RD3(hKB3&
zSSVCSXoXmKC<yev53>ZngH3z$_1-4CB9h@*XCi5)kPHfz>mx)tcMB#A#TzV8uRCC%
zZjWnQx>I+fea&zO%wg%Ctbd(1@waIf8cCu~TzsV+v{UTt?Y0nl%O5LR)R`^eWwE>N
z3UUV{K@-tfEGrJ_Lm%`o8H<N3-2i^Mw1g<SQ!ERUkURW9F<BX1FOrF4Ofc}D4T8Y9
zZnxQ}*z3SCBFzZO(XLAaz^w1_Rs0vfKY>2i`+$R|e(zmiwo)fYf@QH$PEN=LQXa%t
zRaq{AZ0&~W7+Kek&m;;OeL1VR#PJv>z|%z`ke$u7U10F>54c?ekzF1I)QZrdEoDtq
z(wZX49Ku^D+fEGWee56;D|}zz#i)x1arY_SWYyFO2$Ll0Rlv$2!(K3R8_eaTran*w
zlRxi{P;RVQ`3#<aEhGdqSE@!_EkeT38;6O2!)}8}#MoV&4~H7U$tF52)CC1*$lq>K
zUt%Ykp<p;d#atL?FdOX$4F_bh=SKh;m&h-}W-i6g2nl5EVpk_DK8o~RJj)fQ{rB9E
zt4l|3VI>BlF=q%`|0I_PCDj-Ex5XE42Poh;8h6guBV~Cv&#+>|N*6CWM}16j2Z6EV
zr=#M_6M6D%tPvB7z+B-t{=v<-P5~CwBO;USg6n?x@PR_2g{v3BbuY)GCH$|oEWo>A
z;}teuvTHTTnjrHy6o^4sg#6)pC+>3KrrgK>%ih~S$#q<3Vh#RS6eLPKyMisP<n`lt
z;2DtU!TbOuh~F8^41l5Mk23>+AO;-uO!v$*+0#AhAN+`9B8Qe^24*ZLyV^~TqcsmV
zv6CZjHnBIRch@#0Dw6D#cV&B5kz@OqNfaA8QqF3Rbu8_g>&f@s`g!&0z1RH)06|$+
z13lecuj<yVTeoiAx^?T;WU0CaE6;4Bz8`a&Qk*#w@$z!Q{@gyB>CL2!E*||!&`X4k
zP2dz1JlK(pQE=Ela7Gyb>X|&!B4HDCZ%cBMO$NwCI?i+R?vgzg;*n<qR#C8`&|NHI
z$sh}xE3H|fIV8Ja<DCG)1gxH8^BDpJFV%hgM%B!C5m8X^7z(>jkxkY^%w}F%(@6|+
z)1y>&@l&DlmH<vQkiejMc`zf5(~o@<y3-OeR;)EfnRBZ)T%+nv2Ygb?4Ce0u*k_vU
zfX;xo2Y!Di@WLQqx(-+y){B4>O@KY|K!B5!nDf*)fgObsjGQdcDM+b^a2^OY3Y)6)
zAR1WI?<LI67A5J`5Ly!ZX+|*>QX?f)^YR{>k!(yXs;f<5u$}p;g$==2Gm(MG&p2FL
zR8feTo)c6w(>S!6(;5qvCF;wV4n=3O(+d@4PJ9K<ClL2)RP7B)(AlUf@by4Y*rZ;~
z=>}Pm#r)FK7A9IKul3TUa;??CoWq(3lx@}!F}nyT15tM(l|(>QFD;P+D8(gR%*!sl
z^sFu<{_G7DWp>r7Isy{AP4kK~<b$pYbk*$$&aESICa3P2X`0N%MY*#8vabZaj<0ks
zf?I|NBlIeDBAUq!a4mk5`00LK%4Z~5h3fMDGvc8c!W(^dM$inWaAtnzrQDGA7&H^}
zq=1j*b@-MwB=RmFq`>pWM+|(MqR`C}UeUJ2viRtRa-Fc0Kz7HI1)$s{fUfQ?@r*F_
z%e1671zlmTrlkw2?8OVZsM;_zW%;Rzqu3{<5UP&0O;4W&P8@ghq#Y+^tl60>do4)}
z{3u%SZpl(d1gO(ViTsoWU?;}P0wdm(Xj4;>mIs-|@P1cTMZ~aO$J|OuY=PdOFy3bY
z?A2@fI{fM=zu}qFqi7PHo>N8jep#2e&G(Amq^P#g<Un2{NDZE<Go~sCq6X>Vduzly
z79(a>cH6Hf%u*n8n5DGtDm~)b4QOW3yzNa5c4718Obz=L4UJI1r!$P1nmCmC;3m*b
zDH<QRtf$T3<zpXR+a85@nv6FJd%ARQ*VP4a?Mdn9ZfLMrl|qz3*HKQJqF{KK5;G1x
z<+Y+Fkqf4*@xo&Z-IsGjH8rAe*9cFd=4GE*Th|nH6+|4mpb`LRgfV_W>7!nb_@{C>
z&8<->rzRyLf1TC0AtUD3aKhxYt7s9<%>)zy66=VB8Pn7rb}0un^;loUqOm|di49L2
z-z`*dwySv6ngU_ov?Pd*)s86Mx~Vlr0L#H_w!<`uBM}WA5&izHY;SvN#fNp9A?>zT
z2$6X{?vGO-x-3$3eE_{rj~>o^g(^bpartm77eL!v|AfmQ1`KnXpwUg9+^|UV*Dc#w
zrRWaxrW&q7qFob?+eGGz^t(}jwIuTdK`Pwsb#fM^>xTN-mjl?k;LB=6iEG`kt8ZSr
zGK`R_Tv>!}N-F~6BxyxA&pgyqQajmr&<IWSA2Zc`Z<@U2(`3q+T4>eP_Gg<NAzFUK
zxFwdjM#^GC6nvhAbd1%SKU3wEw}Ir9_8Cx%mS`1JD55D!K*i^XQOx${rWBZl#4DO+
z=q9X9&ZZE6H6$&vD`pCq5oc*qiqk_IgGrFK1b|W5BY@xHs98B%Gj)%yu0nxxH~F-_
z$1N)(<K3x6mMhZB);tC(VDopqfPDgW%M8qkW^Xav1vcAdFwVB+3}YURY(TW759HFQ
zHU>uu!7(~1*(aW6m&Cl}n#2cb%%fwENEz8IgrfqZ4(ZK2P>(x#dM6W+C9yTx)Pn|u
zIj-AjxMw&ki<BFqNlndYaF0+etz{##qb+TmH5;>Jc8M||twJ1?W?r>VnoqA`J#4I7
zC7m7A)^ah03DjiytORZ74tI_^F~Er3V}c8l{hrd*VFvWC@t!e`F+F^`v<jb)brrH1
zPJBEv*vkif8hHt|=u`y18Fl266I5KJaDsN}hyKSP-AkjoG!$f#$D}5eOHhd;DUDK^
zD;!d_B@d7vu_Woh5+P~Q_`ujODExq@xWi!Moc`BbCTSmzQaz=AmH5|7urw2V2t`Q5
zO>n+a4Dm+Rh*ry$9eWjl4B;-vQqQQu_y%XaB?B6V%al)tQaqGLAl%mIIU-|;UaRy}
z2yOa$eG%{8GgSjDD^ZeXx|W3#>!k?lPCi*nCLy1SEZu3;#HBWVl&AtH${{VrTukXX
znk4j|jgpD7|3c-LyGxuM1(fcpz?EHUI4^*jlnCOSgVQR)`m%<)TyB)wcCqahEz|v;
z1ytqjMuDyD@rcS$Jc%7aQ^H^F>uFnM;npE2W;um?qemHeK#c_^rtP=kG_t0uILAOT
zk{|>dT*e}i4pL?U#GuiKU{VOJ>70ScKCQJx5aw5#Y1ow_$we&dGyQ7W+L7cw{iLpN
zEWvo7I0w^~p6D#~*c*|(5y?4K*u0?H6yFYsDlmR_pZx3y>2cN5CVONW6Te_}A*#w1
zbynzviI6a@USVANkZM*7g{IGB;(-aCWDYf_7}w*lOrIhO<ji`x>DVxAE1B#_F{bln
zXhtRrG3Arc?xtlG6Adcf@HmJ4-tcgtK?N72(tpz#5$NxE(Gx`hJWbyvR*VT?HMk7v
z4;HVts#OXII7u?Fbc7ugOY-on!UKMx_9U1ag<82S6JPV%Cf@6qwL@w+?I*)z?>Ol&
zdhx}&IP5MhGz_9%U0#d*9t1t-;Lf<QEWTsVpJg^xL#kK<T}1()udbF7x*Cm7GB<<M
zd~s!aA<ixuVdL#&6e0AhlAOt`!NaA#_(fxsD<<9JTH5&n2*V!|G0h4Uy8Y2~0Q?}L
z7|uW%is=?c?7acph{s29V&Z&wc!GEMvL=TE!3a3EP6024W81M2iKA9A5i7G4kBc6x
z&`MCd?3sYnl5)aKuZ|e`LYb%ieq}e2)nE_<v!k^VsX;ea?<d8)mUQ28XeHnhv+@ci
z#d1BVZUc%qkU@=FYljiafwl=GN2&=n;dL8ClWuG(=BC;>mzgK-XDFcBr$+OIznr0F
zSr6k)xm2Qo+HnKGeTte~hNfwxydm=W?4C%ibPBa=Uw_gZt3-@UMBp+txFY<<;T1yy
z6tfmfgazfgQ4*}AV4rwRp?NozIpBzmTOk{WJu;vKgWIWKDjD&qrEO72gQXN<;0W|2
zb0B#DPRRL6`DvqMv4tHQ6ct5jm0L+<n&ogG*(5>k7<=}%L{7AjFKx)xN#p$%$Qpu9
z8waM%{Ja%oqCdutZVmjNK0e%>;aF*(8qF8}a{5q9kp76dE3*XYPovzMN3C=UwQrwm
zs5I6{P>+aets-hU<l!ANZD0`d_6VFyu2n9TmNX~?4ue0xFDqhQoQ$y31v-PsGd0$E
z(8-(vnaBAQEF4iY59nc?DyN*F7Gb010q)ttf03s`R^$o*EWM;P`*;cxj`g(-QvwD}
zYOE}{Z>QCE;_E=|JZ{J(B;e7jXGEZMscU)*U5IzAUD=;%V9#9cWh<xyMD@aIGvB{|
zpGy1TF13x>Rm3?w9<IfS8kQabuS1_`0DC!Q!)3HNM-lMY7|#s%nZk^7&r-Ezo_I2;
z@W7l|Me+qJ?4Aj_81xxVj76pgT?S-Va4~o;+)$0@Q{lPmjjS0Ft5|xXf#AtqIElsJ
z-t|TvfjR|(V02M89gjB|Yp~P+Qi_lqxFK?Gsyxx{y<i(0bcwN}OM_%4zT#9x?8<en
zN<(t02@iarR^>R>#D_=JnYCm4aHwjdbj(m%fd%03{P1i9`=ARazDe}1AKoX6n6exW
z!x1=%#0~m6<1i$+N20U_cVVbu6{S1aBIEJSc(u4$gT;oXSd1PxtHSzj`XG?Q=45?%
zk@c<NmY#i^`*v$S*xkLG1lrw4uI=gTvW(;KhD9~(;#71a{FsLB2(|~pPSLCqRIs{1
zQyVu;dDHNYl#ehUk?l$xR@s`Njy&BS-s;pstQO0wR}q^RX6y7uZkx4ujv|0C!y<J9
z0V&Ejp%scK+3{H}VYf%0iR3-Gn%~SB%ErDl7E`-Uw+vTqL1`l4xS_ICAQn1KE+Mod
z#E!*GJrnLf{-LA05gxyOb@z{u4UX>S-`!~lUoO;X-ndDOM<Cf74IGt&DB_Vx9xcOe
zqmDeaeAS)-rCw3MkxHqQMnrv3P9LR&6%mmn7$u}sbO*5vXFZf58nKpF(kX&=mlajF
za9A6da@q~%POLNiNz9I`8x*Z+<6Bo9X=8@b9iKc6f|wkWWxN)mKqM6G9Se~RVL>Z!
zS?;3!<p4CZB13I+&JX8H7NKxXnoIeNuP)AJEENhjJ?@i`HUcnXZimv@!0Q4_Rh}iT
zls_z5Nwh%<ZATU-iYc4Rn+c~a;6T@feKpasaPu|JA!CcO8Z@{$%@emRrE1|(v&@Z)
zDi**v?br+&P+pvIWNLGI#Uyo^Eu<)o^*oqJW37;Fr+`F`%3c7nU=;_qo7qcfVVS)E
z2^P(5*~&68Y2=YCO^u?V(w1||80KeH_Br8pVJ`A?T_*F@X}GF^<YOi!BUK16{OL5r
zY(AcLkJ6+K8`x4xgr}5CL?~1!4G@Q7CJ9tXmuSp{4AG^5ND3(LW>n*=#x1&5Fg0Wa
zB4i<mSY!OOWclX`%)bh3pq0qD3<W_;)z${O`a;m<wknZ0ByU2$GP!fd2@cMD!cu%i
z`yiR#R=bQ5Ter+E!Xh2oX_$01)(e9N9*T`*h1iV>{L@%T1QM!RsZf;fj=vV3l(GI%
zJFc`>wCdQ>4wpHm@+e)xU+Np}4@&&p#Cl&(#JJT+f+LtUdm>HK#H83n^ddTW@VfJ1
zaIwRB+vo<-el!>*PQ3mCTen8mJ@D`m8vwWcdem|{&nj?DyV@yn2=8#VRXU<*xGllz
z+EhgB%yB3<Xq1=hh5FSV*@wiGj^OilNdgnHgw?>kLHW%}YpslM<n+K(-?0YhamJu}
zq>egh?F>1PyAZ6j8co=ZSaHxBMjhT`_ACSF>6f9tW6{lHu_^|QY~X1?JU9!z0)7^?
z*y7eBwk1J?J_!F)YUW{?jSU{ZJI2Omv3oYMFg|vOQC(mI6C;N~1oi<84!c0ZH|?9C
zygy);M<p;QAx;Iz*j(=m7R0TGVRT?@BlV2F!%kMmZ4J5904&yIvyy@B3d+O+)n5Y3
zD1c|+oDMix$`peJwi|E_m+HKFt<gX~6`FXuA?g|IV6tGldP7u;@{gxIcw~>N?><rC
z#)pVJhK3H@6n;hwc!XjpMvYsPG=|f)?aHgQj8K^IGLdqLLsD`$Sb<4kO>7p{_Gv^^
z$3!@?UZVL_airxr0&+xOh+fxvZ3E|e))JN<+aly?4MwieBs0+uMvs_SACnCV)Tg>|
z`%5UX-GN~=mDGtUkzRiH&T7+*T`({-NqRV_m3{sDjeU2c_`|xC@bCsBvM^E-ml87d
z1Y-1snK`)uMvIPgf#s5IVUHdy$-xiwD+Mt^WZ{*cn#zxjVI$4?1%He~Q&WfL=X)sM
zoCgc$!4o14LH!i8bx^3dG}7E~!I~{InrSC>lqt>@eZdT*44en+^#V@Sz(dS=m_|rQ
zB@EK|=&`g^*5Z(*lLpNs^mkUMfvI6bL1sc&X<?})K`0#2Y&td|<^dUcgL035C;(vv
zhHo3(H~B=wAnlwnX3sG86HTkqj1>Ecqbx9+c_k^2T2U}r_}NLFlRSwEL<CA(`4GwJ
zNwAZQMMn7S5fV@Jio=m^X-6u!$z3J%LnX$ikt{ci;W%LjAU^a;1cQrW8RfJ{g`r?m
z*hL1v$(h!mR7q?K^3!sCq6CTL^r)UFp$+9TwJ9R3-Kx@BRL7kKZ&DA9h*aol;50Q%
z)q2Cn&Ns<)kEMn)B3l#wa@u2tRIC}cBN7L%KM@&St04O$<S3(a$T;Z5bdFtrp2aYe
z4voG|9KYv|XRoY+kU>nz;)17l{$yLIr7gxrc==mODMBQwH<1FiO`5qP1WO~2HDE<@
zTTr*1JbIa^PH;lAg$nQLM=|;QW>Uws)_tdtM;j1zfCsJt-U`0amWc15VdXMOxWfd=
zVi<&D&CLYiBTCKf9d5GYpD&g!eW#L68!yz-4|tX;wN0)l5f`x+>k@7RgOW7_JEgd5
zv{u&l+}1JKXsscBrC4)s%aC-YVPBHF+yqJpUWhr0Sf6mUhK?lBM}=-2re+IKXM||e
zN-Gxd^`MB^nN(IWaK+E$_R^KIuH0tVG}A;dYGV1Sot?PH0OUs7Juy6|cTk?oFYL9(
z1_`G*zU=&j2^LZP>)s$%nUKxtJ;0K?*fnS>9SuYi=Xnq=cy-MPa9UbviKEC~nvn|F
zx#eO!80zm2dJk-%gU~M#&SPbc%3<q9zZ$;`T#DU{@AuoBM$zS=t5Ttwu`v_SYvte*
zb#P00sDu*P89XH%hXYnMaY|LT9J-Ohx}Y^K1X{*d@eWc}rED#u&9H<skwvsmVm5c|
zCwJYcIq71u(wfv|x)$zJuh)uUTcv>t8Fmo}9_=&ILd29-DkK;L4$_LP8CkL3L(o@b
z()L6;zfie*W)ZtvO?XpO7tu2OkoF?XYH6|Sxs1K9Jw4>A_-r)l+FtjEJ<cn4ph=hD
z+*^i$P}wuJ4HQ;V!33*dGBxcxxFU)(p`2wXUFFqRItxS+{TIYqS}<yba?O+MwuY-b
z`KUVuZVk#5sfIiPun0C18@NUi7nV)j>E#k!Q4oT#9_=MEaYmx9;S@}3Ighd0tgT=U
z>d1GA_B=c8Jn1kZo=&3On+h~r$=FeqGpsH)qZv|Bj5#=#!T77Nht%O#7j4;v#V!<D
zgx}qw_#igIuZvw5{b|#pm2PsdS&YO;()3NHP~*2L@gYY9NzxjIK?VtxI13o?{?Jw4
zWTa=YCZ-ozEa6OssuO2vCjzM#RDK$<%yE#@nrkowLwaKrE_NM}Ovhvd#@j(0()MQJ
zD<zBZEu~|naQo?qIy2A_xyZMFh3F`q-FbN5GJYf-%?YbCZoD;Aj5t&nMoiKlwCv0`
zv&gya%^hQoV)H?<Mt-m_fIh*%u+&N+ktxW7(+dl;^PDq6^_4!XI=UF2$~1t8TQbe!
zE`w^RDGG3fD+ip6U#VXeCm+rj(EMm&<BTz0#k3$x3B8NO@VAJwKtWnBH=Y)7C^9{>
zfQb#PiyH-S=_IF`5;n>>H^b*n<{{v#4&Y->!=A<Z<`PuQYNLVL%`H(>ckqChWo4Xb
z1esEvdkvkL@Fr+aO2<iGtHNOI(4HnEnOD`K0UL$JZxacZ$_1+tNhfmZ#VG(9y$j3m
z#>X8G_{u8)1*YSG-@vX97BO7W(QlCjsVSLByA17!o>}Umj>WFYQWJ~jan+}bUE~xh
z4+W?Q*q51~pA2|YY<ZQs1`h(@DXi=Y5z}b}-nLzZE^;aBoWD}wae&Uf(AJ_FpPI-$
zXn=4$`34HQus*&U{uv#SlL#Ox!t6s+KG65jgF>XndaYF{28gD|xWozsn#-QhW&0G?
z4&y=5tU!Bht*wW!G#vD&Qj(xA;T}@llY*WA+^e|DVk6L4h@Q%8K4d#1BOBARHp{fW
zE!sg&nAWM|uZ_*p3VI#<g>@8;sc;;6oWOk+6?23ngPP7f&W<*H<TZ0&1$5~RP*gbq
z;{jJ_5D_f|x~DqnK<_izH{cAwJ#@GQ)UY;LX@cpg(4(}IJ*+e%2`K{TmsVcs%!0RO
zHiGIylK_<^wfyMo>r;v&uM*|{wUdnH$Sqff+BQQJE!V@vA}hyBG=~1^4>Qk6=?+GK
zNM1vURN_Z+d+J;=Wg;cKCBOpM(o`9GR>(wmf;=F5UGPKI%c=Ns*#(|3sk@J{##G&T
z8{1~8y!p1pOdzId@nr{}#8VTfgUxB|)91Mq>AbXaO$ufP&S+COCyQT5TUZC=&;&Sg
z18NE;LJ>5mKkQ@J2IQWaLrN{`6&w+CM%1RDS9{f{R0=?v3kVGpWN!hQ442cGYGBYp
zA=jd$vSk!Q#{DLhSFA;>?dbLA!3C>+8QVvvjjiZrmN9B#%!oa6qsGf}Fo+Dk#*83$
za*4roLxLw^_D>$2Z|NdVyT&E?a~bxwbl%pSNx4#~X&a1+G(-Jb6C1X4aZeFD5l`p1
zY_bTqU*6fa*p-o0mdn*dkF+<X1*ohb7ALlVS5>E>zYjSxH-QJ^ZciI((Bn4O3vh&!
zC@E}(4CPoSbaQ_nW4d%@rL+lWlxpz!*r@=MLMa%to_ScPz@M;ya|*30@4m^{V-lV1
z_OzE=C)FvQ6P}EB*?+Q+<^tFgOZ9z{8x2Q`A3&pIWwzM0dva!UWO8J7c5Gx}WcMO2
z=~CHM#>)sz0-NNvnlBx2dx)U=y7$B7O}mid_C5Mu8XN2TXS!_<J-1H^XXa#8ke1Wi
z(=^r;^ulzuZ+%Tn!rE4uxp@Sz&DLF7Y22R1p(3oPRuRL3EwolxZB8s~meGHe3hao8
z3fQO4;$a~y_u{aq*sn7a0wuOdhh{U~p~+4g+uyL#lY!z*Q6%!s8iHtC@^(804jecz
z*w1aMhXQz8Lx;+`Q>B<x4Zt_45mH42c~P~b&ggtc?Li$IfZ<qS2Apm{lx!5PG;olu
zqsHValBRbuDw3SmRm)Hr&Zd1>Jg7whQZhy!M%%a8g{{P?@`=pM#rd4lSX^$CZQ1tp
ziX3|5LW~ZLX^XJ95{Onv1=_2dCVO^+ncV#baLY{8TYOrRM6lMv&TR?t`so;;cia{Q
z5#}`&VD&Eh;>(CA7@Sk{gp|OYhgakP7`zI^@{yT+VXs<&(|Ueoy-<a3z_vPkn=1!)
zxEYU)=2bKXq(ZetF=j$<g#EI$$#X4`Lar94RtpYg^j;(lfMzWriwEd{jrGjD!Jd#f
zjrDw^wTk=pIyT-q04FZseAbvTCp49|OA8rHETZkpxo1_W3}rHmGWgMwWAp(rTPY0t
z7)G?J!buijJcEjW_zqZGnMH&xopG>p$8iH=jtfRjh0e`49iFoK@azF271#-}cF{Gd
zIi=Xz5SJLXIIKD&(%4I8?9(?9XV~YIgRzWZjyvjc4yKCWG|8baH9l(jCYMfg85}mt
z<74t`#+FLRDg^kD^}M~K@>cf{TC2tod!ub*AtZEQ5vuyixXR_e;3>y(uq93<Ff>+b
zHwhJd3r$HF(;11uK!tB^pNCXP7<;PARg84DS;81o8eu5YwF~1~&0{Gcq7~zl#LlLQ
zjS+q0WKpVHUhNNemfSbEov*>OmD1j-*@52RMc3F<f~gt*MXZrJr^?OKt!3<VV6Ot3
zf*=uW{H;xID5%|iY@VyF)$gqR{L&SU`uuvW(WD{yR*$+dX9H10g0p+*2%nh|n9@LY
z?(PRG(8aZp-^768lp+2HvWG46fJ7@PFf<PR!m$H_y+y5#Xv<Mx8kKSAp$6>Sz9l@y
zUF8%W(`OgKYLm#6&R=%Z`8$Na<PL&MB}9<dhrm)a(oCj!U)@*$Lm2f+RitgQ&@z<>
zqHKoYlrgO2)@+1@h)SQuu4W;-t<1FM&9V4Jo1!-vB}7byW?;`JiNPakTWD2xtm$14
z(RIc>i}s9Jfo?xiSADq*fG}bUU=_7W9Ky7cFLD!p`u6xPWvIO<-WZfQh<Q-g%+dPf
zFV16epAK<93OE%^_r#t6$0^7L*siB462>hu_m*tOZ=2G|(#)ljQE^fUm2#Ng(-6nP
zU~#e*s9jd}Z+&!8lgL!YAV5*gv(ss=u1#LH9z!x_%)LpoKS|m^S3wVQPP1^r$Iy^-
z$KqDFj-Js?Qro%Eko1ngUh_Z>^vSU+ZsWqRCPz4;s&bSDwlT(6AEAibI||zpTFkg>
z_{8vYrT{;?Iy=}&Hcy0EE^=Y<dpHW+C-GE=`%FgZ-TiN2@2YsviG9r2e9)M_*S(1y
zpRIe%8mMy*Cmn;eT;j<bqp(6WN-+waXcuC-gloH;^S8?z9brJcy~De0C~@=bYQ6fV
zlQ1S&df_KcKk<RvC6=x?zk$l2W{=Pn)XLm=B5->`?X)rl2ShR=kL{dFs|a-ld$#%8
z*IST`^n;R<1sVnAXl*76wJMGFOJC3{Vfz*_!zW-(uMn(B_VJX?KIJeX-y7z~0v_<k
z#kku-l$*m}j>xj{v~I!aC2!IU#lAH>;Fbd3T0l)gA3}l$AXL8=17{=JM`^7O3Im%e
zUO}b!VHq%S`=ch`;q&VEMDO=zRu#6u$OK&0<SCDKf(S>@Ul+yc_Q~q+&8(eg)`*If
zGJVu>Y1*Z@B`UTqy%^Ood>et<k}#o(U{c~D(JV6cdR|W)INv1!LZH+5Bd6oCOHZ-4
zR<ChR7#RhCGn}lGsB0Tl0BPn}9{g!}{0!bC=1y-YI+*?o+<1f1vr(u&ouunz=n>fu
z*C)ne?TmE`HWd?BR(}7Ch(k0q(oiO;op*CKO}Rf=*C%&58SAAGMntZqF6wbUv%JyT
z$YU2x1*mG%F73e8f!%15(oo?Uq@lV?{D!)7T}&A0LTst%lW}>vD)$Amh@MrZ)3S;J
z!5xM})M@At4$mXF*dy%h-e7JN=kmn6IYtDnETSo0Pom(nQtnw6d$zvv?A++Hs$6%d
z*12k%I=m<pNpPbNyV-1vb)X0KTg2gAof0uPn*kV}6H(RTj41gUSUWkm3kWSvr~+RN
zxVIUv@mTk%V-GOR^PDM(XxPSKdAG6Hv2N4j+v8tn!4W#@^L_;yv9bGUcE@G+x^Z<w
zdB36e$py*>4|n@uUBRUo&Q`}&)O0(9=@zqFn_)C-fsuUF*ucAxZ4A4?>Oz_2PK1}&
z%eah5(?w2!6lsw~({WM)r$mhSqG$_uSma;}ID>a7FmI!^xycA_mC{;a1qMVFq7VBl
z+?t}RTD5c;)GJ~GMjg3OTm??RHanI9Ml-?`5F0hOl(_0bi1-FoNLOut>o9ue<zk73
z95m(O7zP2PX<LUESBK9*$K?{{h0;Fdoy=o8GEAB`a-g1v&@p%_mHf&EubPNjg{=l@
z4QbjCfe^EpJP}9TYmqJB*(7FGaC{4cLJ$}EsH>&HcZ2<j9X8WG*b0?(f^Na^p+E?7
z251|lG2+l6eaxG_TpaVrgTpZEjHsYNWwlH|$TV+j>s5czYWf|}czDt)SCre`@v2tS
zS#MRp-^R9w-Tu}$9legx-fT5)cdx5j-|Ak!jcpHm{jG00dR=QqM&8{nNgoD^pc93f
zY3NXX`vm;$Zn*0Ax3=9*e=CvmmUOqO`Yq|~+tliWzP`<kc6%E8;Se=&?7dA|T*%`%
z9R}473pJc;c`L`bt^Q3Xf#36QB?Z6(W~Gy-^XZ%3ct?vj6~luR0-m+sUg>V?zda2P
z=X6u=Th;D*t=7Elb9&f(w|aK`TQcgE#~{N!@h-*NqR@sEh*o~P^x6!D9&I#<2?1k^
zNby?&72EEIwhzc`forQk)JC6X6o>|5rm0&a?^|mu0OWyT8*S48q9gQRF2~wu3*l;C
zZ289sVxN1CJlSwZ)Ek<pWHC?pa5INbSDQV!5cR}J%Sni%h}=5U-X>!#G26ylz-J(J
zW_&M~pDw?7WG3`miEX)_A!f(6_<3S(G~g+tDq<UfxQ3fyuY`oGuN>O94<{{f-hZFY
zbSqmE-bNK06J+!o8=#uWm+LsKD|R>&SV;I@RT(on+HrLQ9AB)hyom!9&|CBqk*%>?
zN&I@_O~iU`em+=5c%~ahIKmLgyRTIQqFlw+MGLwlF6OD?C`i0ThKDll64;$SHHK)W
zts@xg74By$?<S5B){MKgN_9j}Q8nwUSVPOVK8hsg=+xedMpYN!LHelhYvaOc8eybn
z2I3T!c^IvlaOa@`b|DPTnPwRG_dzGBMG1t3EwOb(*@7M^;V-ncf}x4jipo}waFXg4
zL~J4I%{Y=YT!hg_cOOTrTq(?LZ_H0PNf*XPV~i;qUMp}9X~?0BluffGn$<CCVv;><
zeTPi}zwIO9!p#jH26PeK3Gk^!I(wATG0p{ELkAzpV`m%D6|k=jTrSw2>%t=^%EG+J
z*v0C&%`7?fMn%^MP6KksEd-Vy?4t*@1JLB*R=TP%V4hBJhBClmo(P;1xDve@L1r+D
zjAV#lG8ly*`y$*)i$L+3WHUt|dAN*DPR!4?g~AN&AdHpQ%5(6|%b%Q?pAR8LTVyf?
z%oLbAWK#O8G+JwffZ~0{=j|{$S-AXksga*+AuQ8IDTEY7N#zQ3t{?@MlamWSnhF8W
zLOh~9bQ0THVYzr4_&h8d=QT#qt2f)z;Ux0REzHb^xX0m~8Xawi1|C}qOsWgo+F=!i
zHq<u|C24zT%pMsBYt>f%slp{Vu9Pbwk{-zEGr^<CGsO#B?AVOLDGzSu*ETBxD-90v
z1gB;vA0(`hUQsw800Y^lfzXdrOFuC+l3Enc(y-qsJAx_N_^HAQu3D45?};kB2};4q
zR&_;Uk*8rXU3<D*$Uo-t-hkgtk|%hq8BBsAAzA@5MmlI_baWwXQhvY}b~8A*jBteW
zqa$PEAwq!T2^q#fytG-(^AmA6Ly*zt65QF@>42XM9+@&EAG;iFohSD2IMOB$msTEA
zsS6-B6`aYRKn@2+Y$7;!MaL-!Tau3r2Ac`uPHsL7+Y^LCVdrydsn^wiVty=24kQFm
zl<GC^g44G_OHD`dv0hJhrW=`z!Wd~<hqqt^i&iRn*&r=TGy|%H?kuTO6L1lIBEmf*
z#oBTyj{qdRNYn+w^k80vbU-LgL}!H0hd9y9OCWZ{3V?%$nIuNmAm&6H2{pd7%1%|T
zZZc?ftTuT7knj$VlQ99>*t}23jlzfFT*r9~2KRs(2kts+WP^5c5@W}cl_M*KVrc^^
zFr~Nw6I^1xu$}P^7EqqLkP2$B6Q~&Pr+{L<4p<FlY#{S>02NoNl4-?!?Le&}L`q#Q
zVt4sLry7zw5EA$R0x~Z4MpkjzK`l6?g4qVMEkrI}G4R-hE?Ox-V{JTziLL|SPVKGG
zInq;<M^vKXvM8qQTRt^5J2^YzjisqEcyeu4HVaEQh|*`<+yRn5lyJprDVUmAFeK|x
ze0F5sLqpg>E*zYyiO=gSZcxWUHW;a6O+nY9`4wCa&+$pA8uCa&XG||l&gEx4(L2J_
zV7|86#O0@@V6MEj4#3l>Y^Bs599DDdr399dbkks&A=_eQ!1zML2h)BmDJbzl9);3r
zaL<mMobtO+L^vE-o5~L^Ko--u_Vh^{PsIY3csTGI<xh`FaalM!pENjyyEG&{pO~p6
zIx^D$#%NA<O#|4z0~Ch7xNeJKdznpu3H_OlVW-YTcy3(bw35KF`ibe*j=kYP!nwJF
z-jGdz7F^!9qBNU>4$s7U#RWWDs#h{V?q$9`nV`n0>7&|#geMB?F{MML8PYj;#0Ag^
zSW==bVoJ86SRztBdBt^VHW20J0)t~?u^Q9vPuORokwYBvMs@?X^E?-p2W)Rph0oGp
zy8|nMYP>MJ#o$)X)P6!Gnp<ewlYM?HKbTjVwFd=i_aIMxQs-XmnlCegCxe+IXjxo$
zjIk>w91A#4-7ZnDw-x9d36K*S?Sh^zHQPg088QPNN+rfnNQ~*yWybSlOyjxICB(tm
z22o4}d1p-$r@VxMgSFTk)9KiW7VIc)%JN4Ny3%4V&$msE{N#ygZ}OWgT*iRGNH|fd
zKaH!LuLjeyy?|u_BDaXjfukLDiTVn~5KG?nIs-Yzwqm-LpKh(NDS5+eQ685j=@sF0
zEzp^_->!H?SgH*$@_;*{+Se9HiFS%iTHDUieq3!{%~qYw-KnbmsZ6{-Wa61R7ID==
z#c=XWC7Fwp2cW4Ue`Yj>eBL|^dJ()(ZhqxGP^-ac0V1SUufjAlukJ=zac8IGnsC60
zcR$2z-=Jv_KFi$mN_^+JwvHigWV<Z9-SOF;;0fv%Oea`lTg^G6=CzE-yMvs8`VK^K
zlLw;FSIW!khXKiT7P7a)Fa+zFAUc74s4ZCZCT{j(KpkhO=)_FjF**XPZGuf;6ruTW
zISZn(H5#>*GE5?kV4;p{EU*VgqnPO?LxK**0+@bJoR81zqdc)hbsN<oENmliI^5Ip
z;*6<a#+0VEFrx}oI!v$cJHTw;^$O3}D!VPr==wSh&;3~5M|-{<J(;^5iz+;)N4MEc
z)38Oc`$~1NNhM>ip_c7+oL5V>ew13a)0MGWvh*x!*<R;*wPfpS)Uv(q_G-!8`_p4L
z#{kd-#;~D&Nd|T!J3pY_n_>yC-pm{#_1*-N#Olq+GpP5b*v6|jGZ&%Wn_?ud-pqV8
zKXP+?HIJJRU{<S?pDtx*DpYz?OyyRZnWLoAo8YKur5V`?mEIIPxs_(-B~*G-yyRAz
zk&(79oo+-NZC@5eYRSkdtYv$A;?<IgH?}WlBDG{>hV4t0NG;iVAN&@x4Yk#yTn1yw
z&iaoG-tDMWZWY=3U8>klzelUc(&?yTd(yzIB3qB6itUYkw~Ef)-JhYSWlNfiuRyPH
z3zhagJT{iFPE>0Ta3gL7t!+vuFYeonIT2dSQp%fnVivbiRe@(=++GFkt&OAOqZ5-!
zjVLi)?XAU@ymsHZv!&F+_uG&eekT0mlX6D*9Y>g=PizZ#G&6h`|9A&0@F%=wmjKV7
z=tTbZ%&9<orW}QLe)Q%*SFjr?n=2mP71##la!m8wFnW7Tr{^Z}k6y)%q23PfC>{jr
zL3Tj1X@Jy&1VHg#Awnv9p4;RYaPUZEU>mQ!#7~2DjTrPj3d@)p13PTmQ=`Tk&>*e+
z5hLklx6ar}Jg<s`mHq5U((o4hhymtWNetlV>=q|c(r%{~OxxS*9G#sJ1ih_9vDd*C
z$kR~DLgZlP?j$BO8J=Zwcc|A}CfRgI`);Vii>FYTZ^}U!ACmhl8@Mwg`I$tTof!#c
zaGME??jGGdZ;N;rw3dB-L`o)X5Wd9QYh+5MqGS1!Nqf^{r4}~K5ex*oK4rK-8UKPw
z+*wlna9c_j%d4x5>=mW4NzJt7H8$gkIV5CjO|97|eaI6#V=6Hlfcdf6DB@#C$yBs>
zsk~WS-H4XfiJ1V5k0y-l@nTC|y$IVG$0JS=!%%O=PA8rs$k<%k0c3Ovjj`#t_;q<2
z@HqM$=Za49Ojvg004a`KBBTyDmEhnc_XjyvZk5LSE&PE9QQSe*AII$(jfQ77*NHx~
zj$M!>CB_l^e-oRwo`{Iz6(@bYBEY<vWV~L%<xxJYjx)*lnK4{dxmGVTvXVo%@iPG*
zoWcWl-21zOaqP|{wzS2wn~3%d6&#k}whfMp{&5A<Z*T{gZB?{!6h_1|ugxmaux?xV
zcCt|RHgB1dP&7uNlXaYe!6n#N;%gWCC^k1bxFi59l_oX{Yu>!Uoq9?3$?frKoZctr
zPmXx<4@t19RRZ{_Y~ZjecVo2laWG4SbjI6_CZscP4&Oj{vto82lXKC&N(2-1%67bw
zi*vY(9!Dx_)%JUrCl`DaK{=kd<*ub@Dp{I(>ZB|iJtn1>Bs(G)a}V~JX8Z25aRO?n
zo*LB|YI|KuLmi5ev@_J9ww-;dtlSShs5;fOxjl%<!CB9F@)VP@6|Ue;m1yNd#nmgp
zDW+$H(5hVVK%A;DEmLuv5815LcnZxCq&hJZfJE<Ic1s)AUBsWOTy0NONaWZI+YTx+
z`%q&-hp5JkN9}+jb4Q?S^D`r}5Q{_09u1kEAjXidovrRPZZAjBy*XS6TlacJKY{~H
zb*%-_8K_K|dg=7+wC8?%8p~$f54U2)Bjz~ed@HU1FcQH%3rD-^4!SXlPMny^Z<d$6
z6AVa20M^T;`s^r<9+aEdJuQD&+;;@Docu6)xdS~eNhU_mM2U)oOwFBm^psbyT9g|A
zAzOn-XU9(kr|PxVW;R%yrltyO<rQyXAQP2(kO@$E3N-NC=~P;P#a*?C+cfRjBLGzA
zS+XtFClCj9*%P=bEfZ!FxZXGa#Dq70n$(OSnk8=~HvwI=lH^n2^m&}4SSUX-Iy>Qa
zAb~MGrrKpRXAqtGGOiAf>oE3F!gy_OTA%RU_mP0b1h_0d|Hj*zq;|^SX?xls4rkt5
zbWY4CEjnZ69xXJW8^z_UN>6A(e1twaKQkTeD5hj8Iw|HLk7<Q!ut30U2O!JX7aizp
zuNNpVo?4IuroAFC2}htSgWf!<6SFl3Cz@+j96pbdW6~sN1HvgPao#iXAZ=D1>^M>J
zAR=HVD0`E<LKqh?rku4wnQXul=hhCuvrMtX+}EYsgT-<l^u{zPwqZ<^4e_gPIFcvM
z%OLS4LhnX2A&6C^W$HU*ZNdZ)E6>DC04B?m<;U{WCA>Z|+)N(8)nKjQvDRX%zrXaz
za%m-4Y{5Zv$j;(e<cLosX7di7YpS}r8>?<kA+S>3#H~f;ruwzbnsIYo<5t$b3XLgy
z>yu>fWDB-}rauXsMZ{=Y06N$>CKW_qpe2Ija@G}VG0{9Zn}}awq4c<2;dvH`j!R?>
z(`T*@FkO>G!kAHb`i<@Y(`|Yx%pqTrb^wa}?G$U>`k|HR0G;dvb;W@i>!zu~(<Ou)
zDS29*P0lp#r^@viZviPus&M2mTpWRkg=l+{kNhQ-%NL4}6xcmuyzz+88;7X_hw(~P
z*)rCO3~qlKc&NQm0(Iv{r#zJ%3E8NC!li1}>vX1MF1}jw&LtxuYvJjWBT*udkPlXC
zdEw8zIxW%2v<HCmlC=ij9z30zfKaS=aMSt9x;Lg&T0-gg0GdjY9aFUrBi?<%7j7Md
zngZpCj8ssrtcW#=pkk^`3Ml3a2Xh*zJi2G4wt2NuT6KaA_+(FPHq8aSRouU=i-(a2
zTo#X4JfENS3~y6q#*e@*lSkRM_|NC(Qo+oX)U3**F0qp>af}uweMkiy?70#wlyIgB
zN-cn>m1G=P0_U_UIH>3?=ciO+CTv=p9{DpavZqJjbps<xv@}yA3rM>f#$@d3^D(ka
z7Z{SvgC=R)=%(otNdpOIz5<vqaC|7XDR9Xh0CDP0kIs8SCVGNpqJsD%wQ^%0Ge+7S
z@4I-5OU==-lU|D@EtAl~sk26<)U5jeRBA%sCo~Ugx?ddO##WwASy{9jA=8sH3I34N
zwfg1K8V-Fae6UW0&Zh`<s)or-O<3Fma$qMQ3RTTRv<8T<?x!<H0u)Ap@xdTz+Z~5v
zDjt|plbh8Y03~5`x#U?Pr(uR$&enS=yxv$Ta95$NMIr$ef#%H{7&&RsIt-jy9_5KG
zFL4kckL`|C^qY2i(Xj-K%EA(Drjco12qPgItrk2}+4MrvPJ4Weau3|=YbIr?R00M0
zj4SOn(Sa69E9=!-rM7m}>o120@vfQ~olH2h!c?xCaYzzZ4td?*VS1h!TTbu?m>C;Y
z-Nf-pJDh2CXWzY<q@DZuiAe>|HMe<-w1n0QD33AIK)GBQ>&%%`3;w{B-cXihUi%u%
zx3FPon(1^x9M<HRY$AAg2|PaXJD}qh`HTkWDUs~}wqKMPfZhz<4xl?hCv=3yaDQ|>
zv;(-`Az>Q8n1-1SP;@}dANp3bs|@TmB0CiLa>CUE+f?S;YT8zZ%%M0+2b!4+qH%uG
z0cPfc!4Ae&#RbT<on~jo5|u~Rp{cY)Pi4h4GrtZ-6v}j3X6DzyzHI#J`OBv^`Vx;+
zWMLog9Bw;+?a$zv-hMb^w*bS3GhyfjE@_9M8@42&Q~aPg0%{6Ad{ENkN^=4L6$UTf
zK*q+)1RzOm5-hwqv!i)#W`2ADtIN2Fo)2OZfVN3S$5wn^N%D{s?9BeFX{0xCra(%P
zObL)Wh&dpYET&_im@Q=aM5K~MQ350vsYFOI{W6(Q9hA!1#G@Egel$aj)ewslQSM3-
zshEc4A4%ziIJ@GS$G94dL<`_jC+rjeuOEE~E~yfbsFbd>JR8BRq^Y|E;U`upW{cX#
z=vD|lIRBWp0t(Yj%i6KQW@EKp@NAeWEunP0J)4bIvt02Sq!Tj%&<^@Bo{swYQ@v={
zipSPla><b6ha*XuP<Sj!3_WI8&a8(z82jdmaQcXfLP^Sm!Q5#HDID9pnuEvOY23Cc
z0fpl-u|2EJB`qTz1DOY}!;Y7N=FPY5pt<vHOk~U@hDMt!6)ITgVhlXpFywC^)F}=a
zrm!i`TW1aip(%Ryu9fnLkIBVFmf2};e$-zv&Xrec>n6yBI@=Ig+qYxUxz@55O@=9%
zFqt2DJQe|PzHr5mNDIXb(f?40LAj%BsuQzeGCJ;A;pUMN$RR%Z4Z8uj9Ad=f#(b$-
zWLRYn6`h(5ms1n@6R>;sdoWH-q#TPn(Rr~vuVH#J9?9)swjYNIW=z1fg&7rh3FIqp
zh_>Umow2}T2B#>n*pQi@fWGF@uvLZgSrLI<>Q}9QRm1CF-<dTTFgnOvnapPe-hNR%
zUn!L~ui})1hq_5l;GsB%Q<Kw=dgYBb+0-NhBf+#AJPLggLA0{SjQJ_->3RT5IF*BN
zq9TBJ1p=&Y4EGP;>NMGr(BIP&{<uSm2LWnm2Tp{-z5=JEgWf6!A)lCTS{UGYs5mAd
zqrp7lE$4hNc%BK&BObv#Y#w=}J(#4~n>-dJuRr$VywF-J)JNy0=v9H`Y-BM#ub(Br
zjP9i*5bK=J4>(_i+p*7;i@b!~Y%ulCpPSrsquzN7qh1|I*oV;nSvYn9j+LJ0mLwUg
zaAt|VFf;eqnQ>AtN;{Pva0UKgyn!8!aSTM|0YtzlM4Zxl7u%R~uwk@<;9~iO8E>0i
zKftzMeFJxQuLfs2M8xBH26Zwf!X^tiITt=^$NPzgW`G9WkFa+cKXuWOeb1J?!!1b2
z!~kb!<`?E>rlK^~i9WyE9V12cw&U__((?Ii?NYf|NMwVy`u({f&r1^tAr!~;TC;r6
zi?Hqp*fy9S8gPw~u~82=`GH3sIoSX3pf{IxtbxVy{SPB5p<6>-0-P%}aU3l{QMoFH
zd|4=~JU+QQEIH0KTIEe|7fq#Q>iBaDC;ho<tQMSG$9jW|eDVQoKAx|&f^&$27%bF0
z<l6TA;{%=$Ktd**9!K1A&mTyVvK8jRn}ZL<MbP8wfdSHSp1gvaESlc=IF)8Vv~N91
zM~BnneOT5cl#Q|DmCa4B<gB$%w6AmiX<rk1oJk3l#yRwS(iD2WuwJX(@AbMbhSFep
z-Jc4py}t5<+;8Qq>5-7BK~G>E5jT@D)fDG^<Vu`)L!Kya!1n{p<mmz`J%KLmw`?pp
zPiaa8*Is#g0v@wbPWdoWOzRw0iLn^-<>6RRtzE8^po7p)Ee>fPayibsft<4&@k+Iu
z;hspcK0Hy2Z-;kMS_vq&cuE2k?a)}e6&#N-li{Rr<<94(p7>bq&Ri~c7v5W2x!f!G
z<ZD>|w*PY-kKgq}|MXqCUw-><-~Y;v$=|+zVI9}h!UuQ_2SkGv+|h_jU2ycj4&9YD
zPTZTc0d=MC-g|d;YtaAcXZQa0;h#V83G*KO<NxCmhwyvwk3aQ^ynKH46OYK}?|kAo
zK0i=+<P!(+dHCl?KXH$we_rSRlK%WX`Gj3#oiI%cX2)~6$sKp+J~>sJF=?;o?%45;
z9dFO&b|U5bRqDaNiO>7+4)}XmeZM1@yPH4pGM~8*;s=O!{AY8Z-|&w~MwVk?`LrKQ
z%zdmYmz(96VSxTVe(_#-xVZyq_vdmS@E*B8590Z-n-Ju3Uv__c-}`Wmp^3D2pqYXf
z+se*%UO_JR!M=JC`|!Ekzhpin2u<FFPv<p^s{2&7+{ef?{dj{mx%cAjyoPhRSNrOX
z`U=vdtw_k-gZJOU+j$KWULC=RknsyU@PGs&ye_<**KjU(!uk2VueXo_Pj$cYC+{83
zUB80<dFs!%cHI*U=bkTpF?a17FOGk4_UzWK91^aLfARU(|AN2n!mk&n{`J<b$0gqz
zxfjR(_3YMx)A+TuYZ8CYpWiyL$V`&{K}jEz^lP8r+BJ)B&riLvB_I~?W9ZfEr&053
zhwshx59eBU{Ml<?{p20F-1<j$0jHP2r!Q58hI600t9w5lZyi|2kC*v%y_<h}`6u9$
zT|Y<2pLx&A{rL6KSD(A*se7*fDp>UVYfs&`eoq&G{SXG-&>y^T@r~k3WY=q(cfD}&
zE5Gp(0GvO2ZfjRR${&8M`M$%SZQV2U=O<riedTNStzQR(tzG-^=xhHc5?^{2xshS#
zQ$KWlk-)!pSdu>aYHRNc2l#OP8_2u<75usW^N<wh&wupQ<_~;s{Effd+S+;r(lz(h
zKf<3E4jciN2HN}AUuU5&vrt6}eeQ0;x&t{Mx%idVJ@Y{CAf7(*gFl7j=KI$9as7D^
zhv0Y4KXu3a)-I;KM50`~_n#y03p=(B@YmJ>(rD`diFExwpz-`?-ud!D{CMOC_u_Bs
z;n}TSuOk1o@vmRI=i1v}n0kHovFraO+jjo7Yo+Vg#$T7*Lw}(2zYY1<1s67<^?h5r
zeipxU3tuAO>wf{nUVCNgiCk{=5i|!F-!=5=*QQ=Ng0C-rgb6=(kkYvI)p!5O>G%HF
zzkdfte)IkKyaQ?f=J{8C9PNJBr7sNq<|9A(ZG8ROyPm(&eAlIa&flDH-u-t!b?sA6
z?Rc5!Z|&L)hI3DCZSAVy^Zfaje;ibJ`6uybYu6W$a_!zGzQ1tK)~=tFUz_slrC;EW
z+{;(+>-yO{a=GiD1j^4JeGee!TJK!{CqKaQpG2nf*Vp99eXOhXc2@L{nLPCBYuBH_
zldt}R|G2evZR^anum9NBdqK39PY<BFd&l^9*D(M7;1T{kaFBnG?#JKNpW5|7`M7sM
zK7O!YKJMNtAG>?yBj1gWpZ)2(AIHbF@!ZRk@_E<Gv-o^*{GOK=`TL`<zp=ITxx4>9
zaNB`@FY)2allb+yyNB`j?RVu~{oLJS@)W~}!|R2Qei=nxx=X(PA--<y`T&~q^2d<y
z)Q;=N(W3KDq3Hn<*FSJSsnCtD&mU!mT<ad>c|Y^46ZY1wUOWRk?8V1#z9v+-Hh#}F
z^zKSy<Mq}9&tJLk&Q|xe;rAR3?*6ae*W7jVetx{`Xcs@;b~HEq-|T4Jef^6d+t8~Q
zKYdqj{q%cSBP$1ew|4(Se0l0U>%Cp*rR6(vU}ARcwF5te^sO(Qe_{M9&wqQX_0Fvp
zy7x0kJO8yc4Eq0yy!(9V4V``KOUSo%fP8^6DD?V^<FC)YIR2Gcl)HAIqF~KlKZ9Z~
zjQ`<Q>nmGdQVi63ed|jUC@iY^`#R^wYvX@-{k=fvHB|OKR;HQ`aM!0n(QkeE{jYxO
z%jZvDMDp2J-nf4__v`#~?%KBxf41@NYj4Ady<_W(&wuW&!+%--cS}!7uDj60U+16H
zTwA+84@|#+m!k93J$iuc;NU`CBS-H1lpa?9cIbD1m1Ao4k<;&e_uu*SwNGDsjRp0~
z_S5MjcYglezcTWk`Lpl-l@ER2(C?mmaqOP=e)P{Tz5m*C@4R;UornK@{q5Htzvs^1
zKl~T<ohP1uwfFFs>eq)}U3&6D1f0U>slQx54fNMb7&*n4K&0I4^IyLQa#h%FVEo%d
z|2t^@jpECE^g`)6=*e&E6b9EH0M(xV%v~VN8Sqb@X|w9dtUS2~PYz!j?*7#EANm%i
zp&hz2u1$TT_`<yea~?=t`+q?MkzqTAe)lNE&a2XN$}pXO=+$_MFKq381$5oo{mVjK
zh5yz80=+iYecua}?)zTE`@z@4ro4fspaom4Z(kdKLxX-{{M)v5uer4-#2zyo@}8~M
zeTrk(|MUAmE{@6Te~QYU|IB-ZU(oD3NtGSBp+7%+{rBa`JJI}m&+>Ks(~|I>Ctt((
zkbZpr>-PzTpD%qIllB|A`E$Q9{U^i!?4Mz<Y#shw^WKlX)_nW7zW6c+{MSBs_QmmU
zW4IncQ?Iq&7;1fcYu7vRedrroyYADU?~zZm1^M2~eB<978vn-7Hx7TR^|oK10pzc}
z<D-ApdfT<ne(Lr6-tmnW4zK~=`qB#rSizG^7k(3?wuxd}yRXRKXXWol<nNEm-;c@P
zzaxMDzWn_n{>~N^&0iere&>V5?stOPFZ5o!E=?o3$G@@FBCWv=TQ81(Q?U^U{Y}dN
zme5n*82at&|MR!Dw!ZpD%nVIs_X29Tw$OdgiznWB?a?m`eg68#@c7!JZ{XWc;M*&e
zuHoEQ8^{gn{`Gsk^252@Z{huic>m;{T&|AyGTx{0-jDb9<NY6cbGbjk`*-pFWxPL$
z_s8-61fKErA5ix7Zm9RW0SoWH^&`1l5$}J5_di5BU%frK+-LCq65iMG{!6_7HemGP
z{Se;&1Ad>z=M#80@%{<C|2ERzg?A3`kK_HH;a$P|NxbLqK8g1+zLB>dpZ^=)uV|cD
z=g03G=7+q_{f*(=e}M@m>1Ec0ci>?K?~mM%_q!p8z}B2zmkY(9P{A@qf>tO9r;`9>
zO#{qt+^;mex%sjA-}$Luy6=~NddH)`aQ~lw;cJsmVEE^9D~A?w8B)Em*eKzyl%fv3
z(^$j}KrNWl8_ia+T;pXh{fpRrV~C%|;(EDQMEtPo=HeXg$bn0Bn9{ddTo&v~KY+gU
z)@yt~4V9l?)SvwQHqh+-c=KhRf5Caq^g~e8`7+PH>^%P$-#?uDW4w{H^+#A_aOP**
zie1C_8-MZH>s?{`*TGl0AeZarr+$4CJT(@8SDduJf&O1dTGH#gmB0V)j#D_zgR4Yv
z@@NV+m0a#u?wt2h4&Q-1^UdOEL=&kX$B&NBO^;6w4fYi)VDs<K9Y5ZE{GqWk$GcY$
za(<~<FO-{&;j!UHvp76__IS5u`C1)m1H&r${K&`%9>L6CURi>-7VhEL7)}9!;e*Tr
z_<O<@K?!Uar<L$39UKOT;U{uG^L@uxYOSWiV93JiRQ<s4NG|s%%9g8sQHDHjhdGOW
zn)JbjCNF``j<66fLZwc1St^Buy&)?7;GM^hKa{m=1kJ}r#?e7ofXqI`Zp9b%clZeq
zJ-2kn@#6yq-gYrE@P{7<q-S;<Ki+>!ApO$Yj>F<zxdC33whe-ffzQ<Z+0ofY`ehjG
z+ObsFXsj)ju9RUW`v<wDC3HDGG7I&!MsDwpr4{w$*&Qm&YHsQAmCfeKas~Ufxi8$M
zp8g!lm2q+hcW&qIx<kI=@kZ|M@cU9JcLdOC)z$J^a~<4Q<b;A@`TwA7qtv9d-+(7m
zsa}V*4p<eRg0srfYO#DNm-`pUv$=%QxIKtt{a>lytLoQNzztVA7fVZ<wKBun<X%@W
zmkOJkOAG^tv5&gz&81>#xwW=dsxM(D<mp^)G<TeMqP5M^@bLMO$I)G+&YY&spWM0B
zsshiaa~}sph>coR8hq)FB}x6jol839gF65_f<uE8TS2^&rFD@%_5BL}YVIAR%B9-V
zrKQV-W}{Yx6;&wHtk){5n@f#c?$_>+7(A?_RK%U+fUs7|-G7&O$mIU1`*f5skn1Iz
zDVM)%!88doBQB%`y#G<8WlL^dG>OOVI#X4cu?FA?-&@_Sxm|ZbmZm=bojWEPW15h&
zbsUPSHggMiKGrJLuVR}UJIWip@_V$_lD=4_6kb5MC`riWe$l~YK^S48n9Kc7PO8Qb
z$ulCs<=XBA1m7BI0KpQ!<r?UZ(JNOZ+@g~2^`WHsuTvV=3;)aeeq)StG+&&@IS+S3
zKl$G4uMY)a*TL%!v-R;2s)z@n1~Zi}^_+$#Ebdd%*AvoVDhh!(aG3rv{QkLNtO4%K
z<-VdS$bEh^m!nP2{O-MTKXRST&F7ZzcRV+TZxgwh+%$eq;O|L%=j+$*`nP|{g|m7y
zh2L{IsJrqac$K0P@bMF<X9wUlbM;&~SIw;fS{W;V5+GM|tGOET(DJkc8y)-cendW(
zW!vMoa*rY1<R3$sMs7vmYyw^xWg*t^n!}Squ85}rAl8w-fo~1uDN3C|?lL}`SoPO(
zEqn%2e<Am@JYB<YVn$p7JXL?!0h7N<fVUyf3n&$!o+_XQ8q%h~W4+6GvI2+(f~sBB
zM;sr@y~AnCS%KRCT+@bu++LLJkNjo1gSmI29BHL+siK|=`lJA?Fy)|4W#Ci9_o`s2
zFe3kaKgvxaeN9S?qNdH<RiWM*N?{_-J&Lrt&?vy>PYpTO0D~<g4f1#o0;2)^#aaMd
zy90c{USuwXhv^&sYKKoB@VFxQ9?IPdIJ1CLL(MJFtm(9d^@QySq}`@PSO*CWqh)<c
zrqOW>ZP)?sE}&+Eb;yB`#WMO9@RaRufW`%27WU8E@A}7g9VsAa*T!<?YH($vQf(aV
zYSpWU82Egn(8zC?pcIIYv2m!-*yuxOf&jBrc@?hXI-t?fE;{ENyLV?09Klv|qsf4O
z8KDiSawyrLZzv*r3!DJ2BlXgUTBxN|oW*5nn7-hxZXddj$FHC$C*et6P!d%5_h?t4
zF(Il#y(?&yM^-qUAMFy?hOT2r_F34`!WO=hBUby7eGy!b?6ZwlG%yh&S4>-un3N{&
ziWFLh^%&|;U;*A_U~_6+N4qYUt3!ibK^5VWk9OJ6Abn>?p<j*8Ep(X>Sx37b!Ofb3
z1O0=8eFwX^gU*K5>LWO94Xtve34hz8U02}68f=P-?T?nOE<k~*qSIDa4lWNZKT;Sd
z4IC^XoG8fRW25mrvM;#>VY;+6EPrfvWOih1Y-0M<<B#hv-~0XDMGsE+!#iMi?!T}9
z{{DUa2m4R<PxdeLKiOaGul7IF|3d%Y>wl^L7yAEE|1b5w(*JAypYH#B{}=ne(!bSz
z&%h54<OhZa9vzq&SQxl8@XWxE54<q&@qwQo_{D)=9{BvgZx8(5z#k0!Uju(S@Xdk$
zFmUH!*Wd$#LxYbDjtou>E)704xH<T%gTFEOxxs%q_<MuDKlsOke>(X84SsX*F9zQ+
z^f!mzH?(`Ge`tDWVd#mW4-TykRfZZvA0GO#p`RT3>7kDg{dYtE{m?HC{ga_j4gJ>8
zzZ&{CLw`JUedy1IwubK7|DOHt+y6uRcke&E|HS@l`+sKtf4l#s{h!?bzwZC+{y*6N
zC;R_$|9cMHf8fajD+hl3zzYX{;lSyKrXE^)=(&e}=Apm$Q10RHd-&dmA9(ok!_Pka
zsfWMv@V|TbKR$f-!FvzB@8Hpc;|FICo;|pDaQWc+!Ro>0!4Dt&$ibgI_<tX~;}Hn%
zfbtpk)BfxIU+@3GKrql8xH9m8!C-KBaB=X1gX@D$PE?pF;U<C4@3hyE<D2W7JHr=6
zRb#WTf~zvFZmia-R~Na!!XmHQ#5XZF9p1T<_q3-zUmPnnR_bNMzpO7V!22Zpc{qE)
z$2Th3%h~nYqWvc8$5phWQ=O+vmCaHeIAdXal1sNHR>UeXV^O_a+g2GY78{)FvK5-B
z;*R1)`J1_2s?11z?UH=!)E1MyUbwoA{E~I7v`X6tT9d63DKsPN?QIl5OBlncxt$`H
zVHn6vu$fkqR6fzd4IuSWQT^;xn=Yh2wp*1RPur>~QYLeS7{k!^iX6-Q?9?igo!q(2
zB9kz8W-Ow3e^GtF-qHyxo&UxL=@_Dp*VYt>j1^&A<1ligv8_T21>DA!t;F~xN2qF=
zNm?iTA_c9)+kR1NC(BmU7|9G5;f%sb*tkOLSx3ArgvlD>r(4X^%85)4Mf&1Ncqmq2
zGOJxCk;m)xT0O0(v<N}#aW)idp!%Zs+Zj0F)Th27%zXp*vu|$VJdEu3?R;kErd;z}
z3c6R+_-O=3gF+pmvUktn@Zn{gDqzp^sjN_(XYhc<#@w1685SbvTPGh+%ZMPg8Gyvf
zx+N81(+}7wQfM`6<FH=GpA;RzY~iiTd?&-rrs1HsCX@qL8-^=mFgI3-RKs})&;`T+
zwC&MUREv{fwN+i=d=SuNUB%u(Yo*!S&87m$o-@-wIz4l4x@T|Dt()tn?g72KoA4vf
zAJYY-;9-*(7BQvmqbO1YT7fj`vCnPO@kmzzV?U-ZHbsqD(1m@fpN0fXD&b9jEVmQJ
z-%<e)vU+3iHK`iCdxAp4a?+mQ8B<xISnM?n22ju8-%g*N4WMiLsyB_PgYK|H9dd?=
zY)$x!bvZA;POxsTG}R88pu0?k1W>AoQLs1gX1}1jCR3??P7KPc2(O8}!O^3rT1rdH
zEgYOlwW<mU35=p;6i6u)LPx2{e5gxm+Jp75*MgPv#7wWM6C9`EpsRQg7{up(hMar1
zJW;I8p~>+Q&%vV|V=?FrMN-gRB1%N{ep737yz()&duo@F)i%ozsEo0}tpc3Si?qX1
zTH31yLmeAmrMeXaz1=0-n*1@av53PNA$;#1QZOi0pru>(YExd(J3@()z$9HrQ?{?%
zSgJw+!xbT@TGMQg;AeiJzkAEo<{q?EKUXFDRj->wzsvy15@Ktz0h)TEI6+g_2s;8H
zR<^3(STK0tfDao<LUaNR^Q!jNvFBE=y2v`iz#jyPClZH2D*qeCr@VF;KYwf;7cCTc
zg%(7t(t~@|Y|@j@Yw-r**polovqw=_F>U_XWmwiq74W8F3H178t-5wr%^A|`J9Vr2
zrP)U0hWWJ=nnvZC!*%y=3rPx6!Ejw;s0IM#we*0I^pX$KG&;=962mllMNO(b!O>$u
zj}%GbUw3v@GoVPw<eGJZqZ7im5J`j}ctKQJ@eL-`8L<R{_!bI4w&=v@tk2B4tw6=$
zb5j@Pj_x2k6JU@F_3U6rf2BdEghl1)jO}wZrju~mv)_Vlc0#F+xzL$CY#N(ke+Zk&
z<l(~xb6SYKgk#WByqf_6Cp)Xbs3|D?jS0*?bFd5(iDvk#EuJCN1Vl5!U|SPP`$IGx
z543hPI(77RFRQr1PW$58U@tR4=IXp{D;UwLViIH{SwXkVAkjLJB=PR$Kxm7KlvCp+
zP2n@*IDxaQ$cyonmY;3wHl>XM>CjgfC-ZcK<oB)CbF6+F+)ri4ifTnEGjXNmGPsH0
z*F&ZOMZMZw?d>^I#+J;n-Hjvr<crxo2#!#!U@<9=%y;V7Ugd5|Sdq}GLsbdDBD1JQ
zO=~eYTV1PYB%ywrj*RQGb%bwb-~%l=W4%e8xGGwn*e!H6RkK!EeeOUJEkhe>g0Vuk
z!F~=EbkYg7kj6ID@lZ=nsI^yDYwHuX(zWD|VSGD_N;QR}AzF8|*~*RaDp$8fOy^YI
z>;tbnWj%|leiIsVDh!ouTM%sO3OGVl3MmO2U5NKuP8(Ik3Dj!1Yj>k7z{yS?=Ul15
zi)N*XjYcv06>yhZsn{pS>alIL4BW3W4EGwcHW5%CPJ{uRngaUh;LtsEa<v?+)%(Cx
zz->vYu=5=>9%+nbUEPYtg^&<d=v?F|)fqsESkH<IJ5njb5!S4_fNh~6bX6lG1TRmb
z5W%))&a$<HtQ==%AQtkm=q0#jlxcol6pk0Ad&7<xRRoXORKhM{1OtsnARSVyl~k8N
zm1s0cP9;w@0Zm>jgLDFDDF|rZ)7V+{%At-;1EIA+W2a308d=icN|;`)x<NhR+^(|-
z-|O#TA6vdvnuLDt%Av4dLyG%j4r+8rZ{1UEDc(*JX|9s34{!@4g-}WTMnkQ+CuB3#
zbTD|rlAaWfmR-WyX$?wu_{2(M?2*(9E6tHgW!e_A#1bxMIs^g-d|3aAVwtvV9@b{s
zux&ilm=UH5M}&}L=~AJBeGDaL;V4+fM!){CD*lmfYt6v6GSu%*YS6a*PD$wtP}x4C
z4J_olFqtb=sMVWwJm193vQ#hO%8QlK5{1w*;sf;dJkxvO$wL<(e0I-s7%kRH)!U6-
zAD9U#83N5n0^mcfS`%HSpI#WaxTmj&03PT$96YCooAfekj*wc4!+N}lI-l(66SVqz
zSS#JCL!8;86cxCgqj}Y7bVEshFcD+?p+QzMLkuAH7NjYpv5^OjK(thZlE5i@N?VZ+
z79lGd46bVFT%ZdpJU<bl#~~@ZVZ?djAQeNNvVzzTV0jB%;P2Rr=7mtBjp#cXbS+jF
zt7Ev2`zoTaZ1Bnq4x;0WReFkaVSHMng=u+WA#6u7rV(1&C^?0(?sAGTa2)Lys8gT|
z_6}|$4KfH|12eHaIu;`r{Zp<kDG?F!0SeeIXzUfUExMcZ(xeKX?t%GYAo!6V3GP<_
zN!SdixlAEWCep#G_c2`;HkM)ks<j$H@9qXz*F^~5t>_+E#zk_M%3|$CTrh@{q`6lQ
zWj)B9wqPjlRRYmSfXKs$99Q%??vNrlgfS-bqSJm?G$$@6o%J<kGX{~KHORV38a-XQ
ziU~orO3TK^TOtxzsX@TVqEuO6n&C+TQ+`kI?6Z1kCCij<5h^zOp5RC@V8*3P+G-hs
zsP>d1&3YZ*)R=~e=@9HUU^N&t6xpIOxY1fh0h75OEQ7K}KQklE5ly-&mj9?OYQbcg
z&onb4CxDrQKwSEuW~aji<?kf1V<bcXVRc)i&d$J*BDU)(L)e>!_qY`sQr2XP@}?R&
zX0|qYB6A~0=}cYp=sf`)KdfrP3eYvClw0>S3bAhej5(FssUcl-)EY+{)Y(;OuAZIB
zIS}p-#jKOn#Q}l@ECxj-MKU{ilSFdZmbgIfZiKQiq$?E1Z1U5B<`CG9a39rcF|!Iw
z5$kA~gJ4WTyVZw!TsR5PJkF_1i|SBHdj(VVH;hn6R{()zi3(v&UJci!>`|qMfK|=<
zDk`*9|CxFLNzJUFTd4d-OUN3YkZD6UOlPEEI5{aRyeg`dZnB!7W5k>idf2xVZN_4&
zgk|s`hM3FeW-;aPI<a>MY?>jB(!|rMFE)R0y;`V;daR_yw3scr3o=LB$&R{iNDRT!
z(t!Bk8OCV^w?>K3#~0PUJ%<BHJr?vI4)S?^m*yz&%p8;X%mSgI5xAhgT)ZHQW!CI9
zSZ8$Gi)Q=UexX6vusFkji|d8UKt$#)>i)wxGAKuua5Ab^@9p6_xTkMvX=;4o^vu}O
zQeO`=$>w?u$}ja`%#tvYLKVhEmhfUT20h&8A$Pf4#;u1ElB%?UE5bx!MvQz;Pvc8x
z=Evuj#>Qvo#z#jM#>Y5LVL~>(Y`}>Ylc-0VtT`eZl2QMm-ssOmht7`AotT-QSUA5l
zJu^2oGPxHj9~w*whgIc|sqf2Jd??@`3)xnyUA;Y)FJ;#x<Z!Slpe?CF@{D^WvT!tG
z&1@#Emayg_{$SSd74Ii_;w5%dp@Asf-@zRf@gQV7^{R(xljjXRT(WZnhMpf|`qZ^j
zD$xa&>WXS+8&@$#w5NqJT+lb#qZ#|o@3(f_y;PtL7nm)@9*lq)**&(K&n~k*c4lIH
zfix`EsA0gx9YMJr)h`Gx>~35P0zd%`jZIH6o)$xTs)0rcM;OEkZmxr_ghR01mze;F
zIv0Q}s8f6um30p%CMCHOp5?K6IuzjEilvF^g{7&H#{p7wB&~k-_MAR_c4`TOZ+>EC
zx@QlUTDnTk^GA*Z2&&4488mbn2S+z#sqzXXdWLR5;-pEJzUDFEWbrI|djtMN&EB4U
zi*+nm_i-tvw%DiC3ZK?qx`hymmfsOS5-C0}q&sm{6OVZTN1>b#_D3N=r)P%AQWeKd
zL&Cb`7sRxY1gO1#@xp~CH!eP~LG;KEVaxN!6cI3`5J4dOV6Yz|i@zE&6X?aS#wn~>
zIu}z2FxV(uQRNj*fUl+&nJ$eE9U-AjRFjPuNhVPxn`3*$;$~OFM`Wgu%@FksZ9%(f
z4=NSV3FK0#n`6tXBMVq=z*xYFhvOBnwOlf>?RmqxF7-9@gP6bq)$0Bhl16x8+ZYDH
z?grm3mIQmD|6&|ZgP>%9g^SpL>GkqXb#Aw%yiczlZ12Q{i7lNVQ)bRA%$`|Tnj1M6
zJex`4P<?D8nbd?j9vq1dPHb;2*BUs9qxVtFxV7@m?wMJa8dzvxEj)B&A5}K)rh!e`
zRBS~eSI;a-u2h41fj=n~)bgrTZLF79xe>s0Gl%c&F$PHG`Ovc$DL5acW^@IJ$9-X4
z=0>7u(2#+V2Zw^5>b?R?visbUllYxjQa=bw@}Rqv)LQo5oigQsBcVZ-(I|4UsKfZK
z)naJ~b;&Z!4eH_H;K2vG%OXV8rV-Y{di#T|zR+F1c%et0^juWJJeCX*PJZg<*(uZ&
zFU|+?(kj&yHbiA_2%4&9liYI=6_Ct{6$t%f=&4jzKw_gjiJe$pdLdR`rH19>1ICIj
zj9UKVjzTjq+r5YEjxPn5&=8+&Ha<+l_!Z5<2dl&KF9YjxnH(D!*|?rATbI{KM=Tsk
zYsHtdG$g{}h*1Z@U{`+5PI%NOM)<`?!BWEzz-%Q$dg-ljtqMeeiA`btYPGi6C<6o1
zNc>1KS5IS|!VQ#6WVs2NXhj-3p**#R^@ZskszshgLp7{uooj_(4Z0?0=zA*)tW{ZU
zDTwq?fp1^y=~8)Z9p0_Cq}l<uT6O71QcGR@Q7`#H_X!N~3_Ia?7n2oC169Ap^GI6>
zrDt&wG7DmHYyzt_T*z>KX?}cmWNrjs2}o&wqTtBx4b+<_dl!oj?zw_L!6Lt2k>87Z
zx~c9!v)brI7a@R0Z_mi|{KP2LM6!_C8gkH6f>QfBhMPlsjr*fX8oVfTX*w=~fz%&z
zs&=7ZOdnz4o39_ETnOK2Hm@obGQw-~NFJ?KS{tzVi)|wTU-+*!`UKm<eYod^dG*hI
zWy+)w^fo9q2yG-r#59T>R`vA}6Z2hT@}{5=T5wcvbNbz@Oj_Z9lhN+4ys)*AFnpDo
z4(hDSFqBc4#^vZ9RjFSHBGnxs4k8IdAFdYG{rNUHt`=}c!1q@LDaH^KrcQfIuNWeL
zvTX5@A-CmnTCL?f)}TyARz_?WK76c7&`G%W1)lE7S50%o0aH8~WNpkWmrA|1HrpU0
z0U0T%(QNegV7sJzWvSf2nZm1j(P?KS$2n3wf_r=9m^>?&QK;NLF<ud;Dl%Rp1tGHB
zf-(b2XRCsoHAw3Fhh0>(=da2lYhkrfM0sjt;QZ<F$w{uJ$LB^y7beb*yI7^ug2tq6
z_w9#T<x*2f{vkthqY{9FLQOt{Mx&^fj-T0I7J;byHNhRbCx7YECG;4!91O(7Ty)pQ
zGNW{KD4H2m^gX<lZ#Od|D=9q`(S5QubZq>@nNt*!mh0LgxV=^kZod_SH`EF&+RoD&
z15@buHX0K}uYHdUY$J0tYlqIe(FU8gVu)anL|+NwTTUE|s;!@xYvEqpWrq_u)>LPm
z#JwS6>{Xm@LKB1Dk`Q6xNE0zS(08j;XB%aRsGsm|um@G7;*rS7TA_%_sYf1g3y&8_
zdkHPv1X5BkJc2|BW@fY4LAlpJFl{P<ptOUXP}k(_j?F5*Rj%dM`i6^bP7m~D@zJ-1
z(9aRaNDeA}!pGVK914qFZfAhL>n`!+CWNwbO_f9OqpGLmNS56o1>RD%i@(<N!nVWM
zua9)0n+28tX&gE<Ix{sjGChXUQXv|zIz@ivkLiZYuh&|YB90Cy?adr(B6@Nj0Q-93
z2`Ah56yh2JxJR6>NdL+r?}|gC#RYQ`Gw2SpkR%D`G<%VS6SHLlSrMw!BGQ~FH7)&x
z6zTLNY4{daR;QIv%{E9p{`lzl>;i1UaUv_f2~BJ8ip5FMx%&!dWwrDrI|vF+9=0}-
zQ?zo}EM;ih4Yv!^_SD2QcCp-csZHl@cA}m?Fzj6R(ER7}@}Wa>;|pizrk6%%#>SV%
zCgw*@OpcG;lE#@sWZUYn#D*Es;dYkdc@F<j+9rrubgXW!m<STZDy65XLW>d<J6Wzb
znkFiNTTgN=5H&nzDiye5Cxp$R<3CpGfSvY{t*=LqWT;J2#yntu!7@2F4zVFyr8*y-
zu}czR3Hrz0ogqYh01fI1CAer|>jtQLhJB!s$ur~Fu$%@v^^8GURSJl31Y<?OGT43<
zFGa$&h=yQY?M@?`47Vo!<x0?Ful8yOxLWlRj4wTViT^LPMSP;V0%aNwEcB36r5ULS
zj)cb8eF%F&w0l;$vr`EaDtm)LN!);mUKfM+p*rD;^8>+<Je9zjDHK3mJU=a%o^6X_
zgtmonC(>+Vr&J&-BeiEE%#cXmleIc*)U<7A>{Ec5l~yNsz?BCTvO$cmtunAeq4Biv
zfh~Ng6oiW;Wv4cSyX~OUBOLaPaJ1=VJ;vPW&|RGt7qT$@m2@FU`%r|mt%Y#p)SO!l
zp@k2e;f(HpX$LD#^%OPB;Bn(r=q09lFdK<R5qadqoQR|+#k4W?031V%MJta+rD4`+
z<I9(@l8=})BgP{Tm-GAm@pUjNTf#+YdTFa<shUH=E%L@@X$4cNA`v`_kLj;Hn`K!2
z!Ht)}uo#Pp4>;?ibJ!%n;5VtSw}+OgAYZ^<6Fo4oe<8*zfOFg}SThn>4?HYFOOM&9
z>1nh!VM=T?mSopY?Uoq73bq($qw!d-3hy;&m_<Bd=!pvsmhlWN%zy~6GQv8b>Z4OR
zw*OWtSQxfR5>uiy%j8szMOkNRTCI_&;gE>;m!9F3b@4l~zZlUtpz-y)sfWlT5nFU#
zzaA8KfW8NhD_b;u10p}fQGjRtK#c@Lut~cxc(JeN3XT~V%d;}ihpYhqD;>hbj7=JP
zMpFE~jjxv@EzQ15&u)wm?ohFRj7u1|mF<rK{Ls6E`Z=B32jRY9VTdEv?jr62taD)d
zx$^;x3fu~G1aG3HHO++&J760U4~0XSjp)YdvnB-tPKU^u060DaWxDgA4v{zroxO?i
z$9&G1GTywBKU}^l_}Gk9a1SX-wbbAh${NUC**2Aot_)>6VqKAZ<N^pnT2-)(1l>Ve
z0sXzySg%!zXvsin2qMotY=J~2JBhzj6ox2iyM5>y0t|dv`y;r3ct-h!v4e^y^!JO>
zPx6Zxw;-b#gun{4IA+4RP$fmJ5e5l`sfvzb(x{b7S=BXq!2;_lu+baDE)jT6&@D>g
zDsz&~D^R#2tXzPUp;p$nZJ`7YayQwK|L{%<@p6@6ctShe{jn+t*U5=qK~PiR&CsTB
z322teFmPy^ptnQZ6wZ<xZVioQ9Zn=R=bm66bWT_%dwSFoRYM5i)1>UUTw9}%IQm<l
zsKAgvLwmWj6(cs?t~!v0J}#FqVEN$^L_^3>n@3&BElgsTshmEI@@WD&$zN!yW=(dz
zPe#Ronuc+Lu_ORUt_0N0NJ!*8c__{eQIe@b@B)euifxS0){NjKY~({RjGm>6_~ZkQ
z@x`}X31rRTHPvJY-3MH*4d@wPKvY@(lvb!iC{d^m(zs1PL911GyS=i<;|7DW)*237
zEw>oW39fif$SAZG-6|)rnNQ%y1S6O@^c7S;2#~!R?pSoo-3Q1AkQ>Xmg@8Nn>Our0
zV%P`}qt_|5Ei^acsTy38OSMz4^#Ke|#l0?sVrx@b<0#Y5Z(e{|M?Hgq+RkGOb_V=+
zi`%iJd<<=ab#+48?pAa4ph8QL333U`(j+?dFe|4=Lszs^Ir<s?6q~B{fuX~W9jKz?
z7*EH2EdmMa)p4)Z3u0YXt&fnH5c}jV8YugM1P(KSO{528C`z3;WFwg(QkX}W_*D6;
zf#6G<1;k@v5|B_3bTfqWfCcHo9x!5*d2KBUj@C>BFo$(AToHXJEQCgj3~4VF>SeBw
zq3Rn_nuY-<1vDIbDnVfAy@42SC7%-(NhLhhM$Oq9<OkS~%q+(pB7T-(4{8vcgE@1M
z9;A^bL|9s_Zo=Q<pos`oj;A<J93FyBN;qhi*?DtNP+pz-IR%U{vW0L-kLPyYK#*D!
zq!0xDE(%sF)*TfHHnbt8#XuZYy9p}0`l}2>+bCenirXk+w@f;fnGqY>Z3(#QYz=k7
zN?Am*M<ul;e^3}lWy?iFGW(!YjW*+AG#Uq3S6<4UQk$bN(2ZQ^b_|40Xhsem=cl%c
zEYQ>5NtRRnQHOO47c;dD=xY<#Vcl}d8@mnGCQfR34W`r5S&~6c6OZ%A4n>Vj*G@(U
z^^lRL7$Jd+n{DJiS8lF@;z@PzHX^6WhIR~eXEEEl7LVCSWTJh<hKpIx^hf+KQ&N~q
zyY|$3otCDk-&)DTwsTGuQ7fxV*&Sm?%7mmKK+&2;iaEzcd8s$LFySaq9}C0;BQzkO
zBq9Q#ST-Y(SCgZQNdK#|QF7*RrBLNj-t_|PuQ;+V&fs#~oX4ibrp1`4%Ck@VkEdyi
z$qb^L5>+#kbAV{n+`{QG!RyKy<h{;!tX~{My4n?W)Y9lfH?LlWIT@n8p(MpTBfL0y
z{9SeGiFWQ&O|%?ER**f~&{C7ouCN~a&56L~G{uE`dqT9)h!hgnwR>yUbO;mn9i*Dy
zOegFSB%Ia*+#ZJ*^F3-EwnrJmaaw4~M?xXJ2bLVsH0B*_{SHt;>KN)d&tMj~FeSi=
z!g{?7115KvOig?2@kAAy188F(3U+k8{82HM`r_8Ec#F+O$PHlck|%*IVJxScgssT$
zcQ3sSM0c^uf%S|~T7Z+Tw${`tNplg{%(`{U0uGw28cOaHnKdXzc_2ePsoT4SAfI-Y
zDASoHe@yv@sG{K<C<C~YUkH_CBXpA`7-ffIAr%NV6yFZ*<L)BfoS=|%=X9f@8^f-3
zK3u7kd9UD-&b%aN?UWZ^;v?gk9RY5Hmn>CvRA*ZQpfj@r*iGpO&+B>jxS{JGrVec@
zt`OmRHuKS*3A!zr0AvQ!V0L_TlH1|Sn9aC>$uMPd!JrIV-TXjW)#n@rA|Fu7kQBz}
zu8iA>w9_a&Le-*3b&(ya@h;_1wCjHZ3C?UraqP8w`(mUF&CZ%lDRlK^2u@a+n{*Cz
zLp%r8oC-CR7ezwbr40t6ykdXPDiJz6)hGeZ<La#jy=*XVF{aWI@2p%R<C(r@UeqIK
z;8PD}!$cSXFito?bPYyQCX};lLO1+y2!o4q?Lr<VVm+?H-zr&T{`xWrL}FWLu0jJL
z;AIS>9yX1Pp`U~USE+5uN`7yTR{w=E!dKv=OPCCbWH7D0A3CIKHR7Ii^4KaEo+ne6
zCdPteM}q_X`}_N8&lTKU@4LlOY!o7-wR8vrT@fc-_q?k22~O%Us&Q8;OvpiX<xr$x
zEI>wTo+SwY%s@YM$hgW$cam)k_d8610)ihTu35M<Cx?l7`801Mka1=Z+P_xiMab(e
z39(n~>!=j8f;(>HvFD+Bt(O)csbN(IjGxV}DHw4i#0#pewHrz!H7&xaB7Y3gC)FBT
ztHVeM^&=sESXu8LNig|29F=gfl#pJ=eQFGyXa+PFDG3!jLIIs>=SLSeM=E5;&Q>da
z5D68d#bKUqA6a1=Gg&*N2&tfZ#_0DVU84l(*sZp|43JQrj%%8rLRfjP6C&7%04K>E
zc@xwT!jjT83bT;L?3(bGEB5@H;wKwIi{Rn%R(q1^b#lt^^9LVT#Ho?Zb%(7O=<TvU
z9KsmN34o?0Y+PG#UUjZmNPy|Eky_zm`Ve0LwM4=%$Kn7w%lUNBnp+C^Ng=@&wP#R?
zB`Y0iST<5uv=aFx@MUDWwDQ^#BaoE>7Cg3X>rGs5s8PL@jZ^X?Y?}Tpc2?A5YSvXs
zt5ir6m~P}(wl*g88HSQ$kqD`8GzsS+M4Rp!nj|<*_}OOWkQu~hJXAQV{9eSI1yL^X
zBf^@s@+MDNi9AoO#RqQ{qu9_;$cBT$FDXLBQkA3G{)lj;O(k+prtP3O7Q#+ga7Tc)
zMVQ8fUNI9mDAI&L-1xa2qy$)8q$n^J4WDcvxbDwV(LxZ;kj|4eoaU2m3FS%W2RKvW
zsq$vn*|S@(qWhR{H&4RblvT)tC6ml4QHQ$V*?tFv8Ynxvae#CVYS@Jy>SOcclP7!N
zRD(!E`Z;H3Ed#g>WWr`d@WOC%>%dPX_t8$(yx=&t$!%JsC)G?TY|lSFslZ_)I+QW?
zj3uTq$!&C2MGpIfy5s%kmph9sA~XTrQK_JE3JZ-IY1ZrB9t%fo99%V375M|W%Hq2@
z970c87rwS=*xf-3hr3JC24@nD$_({1G7+jalP`4h!~PJ8hBDbF9jHH|Nf+l(3(PWU
zjPxjPtq)3^0t;nqpya`BIk6Hsi{(g<eCeSBQw5iEFg}LvBy^!sK8-E<0!yn<(3&lx
zPXHKO$s3`W7J)sK@0DUAa<}C&%!YdY7QNELSsP{z4<}X^(lE||nA6VCNZ#8mwsOOK
z{o;dy5h30t^2j1Y?9`fA($N^}0+!qk1=;O(Nzp+_<okAd{frp(6NIFK>b5pv#OXon
z2B&97`pXfDIN2VE4yWi=PS}B?%9ajEQZK^|NF3-(brmvm8IjAlSqbYQT+JokdgwcB
zBWVUEel7+UaOCO|eZz>lb!$>Y8cmRn_Q<h{RPbQ@iLpzHHwHpH(wBn8DRaWr3u02J
zIBUlx*gG}_f<mPcf~15KLIlOKju0OKWuIF>+&&QLsx1+&iiI<=2A*)UyX~Qcq(9CJ
zDx@4aa~Rb9ZjKpejG$(=uVG=YzNd$byQ#Qvaav1QxbM;!aN1fEjsK>XvdX3fSw_pt
zjm3O(R0$huBrzJF*@HB;-5@$kihMOnK*YcJ$qP>|BF_1Py^DwRw>=N2I1^STXPEh|
zwAS>HJwF@Wx2(3iy+QnDyK5*Z;#@JT5<2a%%c4*DK$xZ9Hj1bv)iw&)?Tk)klns1Q
zeK@6*h*LZNY@z7^xB$V+reW-f=zPN?#+wckBwknuwcjq&^MCSQg_wXP$4)U+Ok*N9
z_dBQ*XXURoab+BZjl|H`+NMPZhp3M%ybm3kk#m=ffEhWbMg~P@eVatWiLxsj6+GuO
z-0ld`Ld~4f;ygmP(Z{DI6Hh#08im_@JCNRqIxPWIIoKbsAP6#$+8aF8;EEa(?M87A
zn}8>YH9``%t4^P@3tNEvc4lX52z1XL(-C$}*=vsG>0^07FesRUJvu*`srSf{BR%6Y
zCwq1-V3yZGmLlsYbHR{K#EdRy{vv?a5Co%Yz#qbvhK7Xu#Dj!2SWv;4Dh+q%PHvf6
z4MG!?Du;9yE-|R#N1X@j&^3)<`KqIp*uZW1ym}t)(F8TQRXPJQrtDF4Tof~qUAI<r
zAY#phP2p0ZEV}}x#W)ACS&`#}T4mGCcbaLk_U#0*ckbC4DxCgw3QbFo@Yqp*sxx}9
zn%=-_Lw_|86NbClRFP3<C5zpg@g$DjtI^YmujT6coCTYMj9t)1v_-^al=JpwwYkbw
zn7?bKB!kxe#PS<v)@H084tH5ZB<XbdJo``edT<$o+=>4SF5*fTR<2;M?_oJ}k`VmY
zWML5C;EX;TYPiEd<_fn|C~~SxjuNcVMmHL?6>CxoNCVWGhTILdR=R^({nzXXK$s|0
zj#OpN9{M*&(-1OHi7z=L5vQJh3V|VUft#2=RFuGYR!Sv(;xEo{<OjCl&Gp5kaKdHH
zK4m&hq7&q>ze1Bc1;+f<jn!K9D&15!ah*xEiEk48LA68u4QE+3tMA;YGy>;%D%?zb
zjUe;%?@sO%TV{?3t2Oot9-MBLob<%$hiEm7&}$fI?9}bgIK&4xm|*i>3z|k}sh&K$
zxYFFfk|Wy%m?KIn<yC}Wr_v#A(VZ&^nz$Y%xh7{G<C#6qU6ekzc{;PWO*DH*K`hZ>
zn`xA_s{>62R47PbPHdJWy<kh2OocS1jpb6YNIQUj9KeI8inuTt=#$R_yg?#5@%hsp
zaLKTd6lcCSZQ3)w5J1cb^6$+*O-NU3)8jPm+JxS`iO7#}AI?~z9%09=nYok|*Q|-`
zx;f55&~vo8$u3+R@s_@(32W@Mps_hcx+E?H$XnZ4%E*{)vy|Z!z8!d!17?5|b|R8i
zlSy52-Q@a6>3=t*S(s@NnkM3P7)N<0h!VQleZlt<QcuTKf*uA>E6I$YmSX&*wcdHD
z>^9_#TL4~Pj9=O;c|~~O6IIx_LT;JTm6eh#P*CQEn(JsrK(xIY;f&Hw`qQXte0XH0
zJ>>Abf-V@lw9-&fb%=wB(d;lSqGuGTPEF37`2X2^_r|D>BX9VBKE;?7J0c)%cI-sP
zcx4bV8*>T5&gK<FJ4ge}V@adcNXF#y-JjpDF5T6A`pk?3ILY&_c4Ow8(_LL%UAL~T
z?%w!r=h4Q)&8;UJcQ;9~o!9%gN_5|7>*5^KthI^sB?(*Kd!!}tFIBJsuV1I-Amf7C
zV#Q4V(61uL+=ANTPoOgtn=pwRW$lTNWtVK8mpa1&=^+`tY6GSiQQy@iiyf_HiUcNJ
ze~NhxJfKpcOEYC8KG&UtgYkl%vC_i8DIVsKOSbPdy>%JLE)@&|TPd)KtnbR!IjTdl
zD^MlBvMgZ?;04M>9c!V(?_qgIMG1>^gZbC98ij2eaSwLs%8Lqb?l{ycCAvZ&XYpkG
zT2uyg%Ai7C5Jz;n=ELHnT)s!nGz>C7BiW62K<WF$G~7YnD9Yf#BX_$ym3RPz2u|2t
zYEZa=*#-h>Y-mw8mge@t6;R6T@(CELRcG`)Y<P6i%Fc3@l$o@><?xf2Kk<NztVI!R
zO`L=V^K$?!ch4$@a+<$P$bs{{RpT>~(_Uc=tL-A?7?(wnRwKRBt(5#JJW!8;T~j40
ziRU;dC5Dqd4F(8^Be|QvbC30q8wD;aAHF``VXtC51aRt$UR{--@7-1DDA;L5RekFd
zJP*4R!^zb<NWKc+U^Lq``v|m={(DXq_rvPgVhvt~ro?1BS)mv6N2%LCUDj{bg5m#B
zx8QfHQS!QPeYpJM-V)1bY!ZFR;+c32knkb@-5GASHI{w8hL)3>Z-tizxW0m!3*`s*
zo*M|W*1xcO^M28!Jc(H3b%l0xuDPg!hHSfQMaz0?eHlLL8C|$A;%%S20#MyH_tmh9
zz+-U6nl_c+rkJw&*sd3BMbq4^%1ZA^?i$vpd&1A)=IP7-1DC|6m3S0~IcSXFN?E)G
zx>J9l&Dq-tZZ*@SaLV*9&-*i&DNzYg4<WwKBiG&FIeNucT)}22IWmw7T$nrh5hoDL
z1Fxgn25!-X^McM?)&`%yp9g!>#U0T0{yZRquiOC}zRd%@r^&wS%jB%3mFvecbc}P!
zr80hnrru1;m9v-3gsb+k>UdW>SzAl0a+we{qa5i5=G`0HY4=dtIYUg1lJe-<ow|~g
zE9!pTy}QM3P0~99;Tu)*gVoe@t&(bOo4;hCZY&YC*}Y(XOi*mmW{yO<0zsQehb(H&
z1V9?o-!AJPxS=2l(pY}xaCyeuhilY@#Qsgb(1&Xw%1iiN%q8cx_R)>Aa%b04DN|0k
zW$qQc%|Wj*VZnOp?cDl4wu7$8c@7iFh=SMW+uJZAiwwH4$g*)4Vc&Bvb7j6?gjQ%r
zB{R#ndItZAJcmq?v;vi1xku2+LGdw^Cfxk$&R~C9O5^0|t7y}Eqmz+OArTHKpzAl+
zV!Y93sxC#+lOIJ@ErFpDSMgJv&VyYZSEXZqUQ)j;foX=R*zP60c}9%Sp&aL<X8;wg
zU7cQ@5vL%qy~jq(QK!Vu**XU_VE-+&W8AhKmNWE%xYh@YdSD)?Aca!_i)$1CPnR@=
zQH!%D#ETHGW}Hz0pQpK-&m|?nE6)%^V-mP9sv?#r6K<y~{OJTq0}krhK;)X^prMF8
z2q)TY5b46#9*D3OEa>1*NS;V@R}EBj34Qs3oxYjU&E^2nPZFcpK`P}bkLIfT{>bA5
z+#3tp`P?8V8uqB6N5s1MG$=*iay^TUXl;8iCJ2)nbq|?Jrf1k8;a+Rh{FIAeoz`we
zOgvS61Yg~-G@CY-O|4nz(X5w=tc5S77poU!m~@0((S`sRa!=%dWE=#4Jhq5aE;gAx
z9qSjKH_eg)w1zgp_W0uwZu3N$Fl16pQeS;T4zAdp{c8L9@$<tM_#s!>%6+f7MfCMv
z+(V@xWfxrCsuYzkau~KS^n3z77LmP2gCykBXE!?YNQV>nvDrvRu%w=E;Zy8Fi7nez
z2sy>Z6I4&<)&6Y_3xVk{*1zE)LeBM&S;dofLDbPM7pGJ#%7*mI50n=lA`u?!fr>-w
zcui(fv>w$1Y}A|}j8MXU?_u>;7Y=Gs@JvHmSw9)p{4QEq5#tH7kv44B6ap79lAR5x
zjC3k(3YS_>!qCpN5}qsphv;I`!h)Iu6CO;{H}!Tjzu|EmQ<;OBm8}s|YG$;oST9og
zT~V^SB20xI8G3h3g_j+MHPD*ztF@baxxB)c6fig=S||IR=4CQwW^rUN(!Qx+IM?pz
zmh`;Vkr}{_+=rzoIun&*1W)@SVkDc8L>fflS$J19s+-vgA$m3umTz#XUT^qrx0bRj
z02F!<N`UEyDxk5MC%Wg>_h^a+_KCBWE;%sMMh85?>gM|ZNg2NXotfbzz4>!Xk#C&=
zD&yBPvxlSOANQtj|G8#o@DG0VE0{Ky?qJ+-LMT67e17=+@ZyDQ!(R<39<WH);MbZ8
z4aK>Ac6j>``I2{rmsv~eJec9&V>Lx})#zQ&M!p+Ed!W$av;rV<*R5?%M}O5IL9D&N
z6#>eF!9|E`(YXMFmK<=peS}Q$sDt&ve-E$4{l_E%G$RZ=j+@2FOh<aWbEKVBh0?)h
z(o^inyf*yrp~@V6nO|c1>nxxOJd9``1N?W}86xI++}(}QVGFSRgZf4PeLlyCZIM;F
zW3eHw5;tR=8n1fWg{#!hG+<b%T(PY~>3C3cA-Q8!QXx53{dEJ|vF@_0>sZzv*9T$T
z0=Ro==VIGzgw7uq;qovycE8;S?wp;4yq97i!p^A#69evGw#CV^nnj(_4gc-m{%w#-
zt#fmHoPfJMt%1n|+_gmqs47l6DzPbh*t5m9!Yj;r)zj`|5xm{0$Lc7CE_Er=j}tNm
z3bcd*26BG?Q@hf8<%#}wibh&}S8!NclJ0oFJyo<fNO@PoNu3=z3xYw>+d024S#&Ce
zU83ex^JlgI>G`Xe0A5X(MMNyr<%Hf>b`d*U34dDNmlnduUV|jQC(ieFnCir0f`5uT
zQ1n53X8m2l&t$1R+T^-cF0FgID4!gAz(hOmP*PlJGKhKL`T?hR-~t2~4E>&L2PPhu
zY5s)qWycM}Q;}em!`ouLiyfh(?~IpVD-n_iui^C8>Y<k8LU@BNEAVW<B?1dMJA@Md
zUk!Pdswbvd)-vSol!(w*L4-zmwioJXP%x_Qv;i}fk~|}io^moTFUUq@jX?b%XzS~p
zMA&o+v{c__x)wsmm0hVB$E=Nb+^K1;tIn44XrW))Fg%!quW&I<q(?`qQ2<p%Qo#iA
zgBuT_GXdZ?FeRikU}N8A=#PaZ`w4p2lEy3qVO5u;LuC#OR?<e<7X0t3;QP)!pCQrC
z`58GwzAr7mLfCkT>4v*V=UWw|d@*lSZ$@&z$&8x$lRO-06<NO;XwD8%SxPQ{H@Bz_
z2dqnBgd=E&(50-Cv_*NaZC}@s5&R-=RAxL@JvQ(?wuwQ(6NUs%dl91Grqs~fDsNyB
zS8Iy1ZvQyhJ9z_hRxcKu@Z*l$gHkUQ6rV8Rfkz=X;uM{x6vGXKqe)$a1qW|vMXKx5
z+T6N>D}!4wx>50vWpeJcZy>)kv)yC7XO5>GA>%mCXBC&L7ozJUA^wOOqDmKV59IiD
zI;pF>FhIptXn%ncI&@`l^9y<?Mf1+kTztEXW}E>ou%x<}dRpo-+-|O^xaj)ZVa)e3
z#l_IZkaaEdQDvTdWUyEdFA^6rYW{DPiQM1({;_z!-PnB2*SR#$gw%}QwA4`$MuasL
z{Yt_pCsW+)$rGMJvx(K!kMl1!DSs?;h|xahL=gT2ZyOd*9<B{vh(LZwku;gX!Nj|#
z``B~zEQ$L4>9Jxob3a@0RZ-11UL@7z66M_jV-(SnRB(+qJyK6KOUkm*0Dq9qTm^OX
zbh1##q;;UwQOA2ZB{DPgFP~R)sP1quzofzl{qIv3P#j*Wh}C4<vFec>4Hn66=;_<>
z-sNCNWcBV`ng=GwPK!H#d?YG=Org}G&F4~T%XmuN;}l{zcRVBs1MJ?x=w~DXr8jv#
zT&Qpp@i1Gfu-fo=kBNDRu87SVrcp{I*TbkJsi7h07(ALCypJAlL6QX0JYGjVxV(Gw
zLp#Y^=aJg3+#t^4!><m7KIRO_GGiu~xZ`~o%gj<NPd~Zj)bi?c9^-2ABD6Ge_`G>;
zT`E6Uco=Q%4`|sXavmXTy?1<i$04CzMWILmcao;UzT!D%HOc_sO&P6-uV&0>h4SvY
zK6=~xVz3n{AJg|7=)VUhjHp%}{s-c>L@>kWboOtG<ji-Ve0r)wXO>3JlBdYpB(wae
z3ElDQGK<0;;$_a%5L%z!NPKBCZ;@r?_%?pJxC7l2v0oTald(<#v<|i=ot^iHt$ARr
z&~|_CilTuKN7+<=9dPRgvdoj~(*`{GzcwX7Zw6H<@sOx!tSg07-AhI(p2S{stV4lH
zZ!!E=cWXf$v4K^_HnRy2Q0diPeQ0&w%YfOGX&=egtHl560RPn$NeN#r+2j8WFk1Au
zo@V1fg`E>jYB|T|^&P71XWRMkB@N)oM4a5Bj)kPC?4KSS$~?N8lW3}q8Hs2(DFz5Y
z8GFp-M5;{>G=;&P_KTO+{;Kz*{+&k)b8R}RkL10_m-lkBoB^k9QghaxQj(Gsx}f=#
zJ6H9TaZ;g<w36d{Jg<uqu$O#+qG6eC$${Zngf1p8t=6={`aU{Mv<K;6jBOjl7lUp0
z#AYwL2zb}~^h)cPkxs{h9;z<~DSGRTg}$m&*+>5iI>zT^Idl3tsl_dS&N(Y9u#L{y
zd2+}Q&-;@NS%)4Wue6+#GUPHBR7fn<>Giy}t43^e!35N5M9lUa*TO|nkU)HIVE+>n
zNSvLN3R4Yy(18DZqcKP<59^%h>f!%BR|7rI{SLKE=n<wDeZntIWnx0!nbxbC%V9wQ
z&`=VIWtt&C1|e8eCUXqu$x2Nl;^#T$*(bkiOSn>BC2EfH5@)l+OId?4pqEY%Nvj}P
z2BA4^BcQlxNiRa`Y^EZ~hCxNHUO0yZK}HAc_~dBkWb_&d?A9nFGXDE&6Rd7P)bQ^o
zuKd(A`6s}j7nzfTnmq=ubYj+U)EL2L2dONoJFRZ(l~!aPTcVAgh!SV`T#~h);}7Q$
z`CQfB@lDQYXHLAer*hi>Iut+%VF6sb?$yD6ZsE$U?;mY$-Q9Syd2eUyyN#`HH@60V
zt3Q1I=)v8`_cnF4NVuTO29*sr4sZ#QffsQoGJ=EHdC89D)><-q^jW8smFdYNNQU?Y
zS%Tx|<o^rY5i>(Ygd0%xBlRHkfl*ZADO%RTaMY6g>uhvb5chFC%l$Gr-T=|SU2Pe4
zD<#P4Z(@rJ9jO4m8y~)w)s35o2of`=hnB2o;a52czcTm?|7#RXbSuNWJt5<w={eAF
zub&1wt-=*}H;!E=f(${v1&O=y<p2SU&_Py`Sh51)h?FGjM7cukLLfpl9oPR|=6v5}
z&i^{tcM~#al@sz&9en`(Agest#Y@M_SMTG*R`kZ+^Z;@~2lDy7#}8Ar;mvC|pxTDt
zKfAy78|IU(prBkhha<x8zM1Ur#mOv=zoz#mMf5nlO+C|L75iyC{)m%y8BEJBP^*HN
zeBe0%QyGDmr8~nFSq&3*)lL?54L$~D2Ix%ca$1R-p>?<gY5G*;xOI!0=APRzQ$2IK
zWhLf!>ahp3D;;LI)*Z^NcDkdGa%Z%tb;UECFrMnUicaqG++Fp65)^w*v%L3ZFtast
zsAeUMwd|>ktdV!qhoypQxmh`&e&YlV(aTd@UdkjHP%=S-*d<NxZ-H!Gi`uuhUP4YP
z>Ij;AN#Kmk*e6?BKs%oVL7ap0Z1_mlyy~LNZy6pCrjs^Bh!T305Vy2|lqC(cy4sO~
zWJS-z^k`1^Ln)zw&5%S!y3`D4ahmf8x02GvrzL#4;u^B5;}ouvbx~fmSmcocb%wg;
z;LOj8tQG7h-Wjp1<ADurOFk;KTEk%YsF(wIf|K}jDRBmSv`r=n#yM+a83{;=S%ld)
z@1Bibi`W#lVeQW3WPGqHJ4xL^q(R9FsAcswavELILg8Hz{Q*{#M44vO<oVHwXoQdF
zv4F?Z=?QFhpM1{r9Bvg*33Gr|_9wXS8?0pXrGqUAMZ@nO{r=J8KRz0Ya2e!2Rnmq7
zv1+!0ORafS{aG5aqw6WMbz2FVNHJXYm4qycX7|AOVV(l>B1)kszcwNpnO;!}BTjKx
zfss-m4YUx`n-NyTd1FDNr-JNhdQkn8RBA*_hVAa6v_U0AVo8^&+Ose?opNS}IhdaW
ztQZk%(>h^3kJ`DA(!0NZk6cbV;ixbQLfQ<ot8vq}4eqX<qZCs?)r{%U>(x@AW;hiK
zI9VQR5y)<rbDpTfPpNcnyv5aXdSj5aFZR$SPRkt0Q0v&od2&>Pi#Alb+k3tYN7L*Q
zb;?Xxnhb9te|L}ExK`%xy&E4wHk#t7o~X}c7c8uX@K$cW#BZ)=y992**v*+cy+5A0
zo!pabHJ75LsgJX6BAHJ_1DcUC7G4UYps!;pdSOJ~i*%uZj3ETnt?JcIc4`vnIyW_Z
zK%hk20oD7Q?KEOgss>W4oG#Oe9IPCTc!xTFUzJI;zD0&Rwo8upp2$~Dg|faazv1E)
zPN-Q8d04o=G(C!n&EP9vF@C_l-0`vF246}VCq%?>9k_4N)M$<NveC~{`eRf;X81m-
z!xcOds@3zjV6Pe6#at_J7eWLK8q9DjEEJ&lj=d2+4`FoFOELQfBl$d@vA$(pw^sB*
z58|7n$Gs(0<f(XVjL1gfCUw(@!A^Dvqv-$A^N7cm(925It8K+M8O2)!NhW<I7BTU!
z1w+}{rkdb9oLA^Gr#^o+lv1LE7fV{W+)WJlz1%981AQ#WRUs%#0Q1}%C{L~u;BFXF
z%b(Df#o72cMY=*vK*3>NAFQJK%jpznPq&NkcoZT3>hhbU3AP7Zg4Xk$C~0#-v8+>4
zhp9pALG*`zXpE3jY>o_A-GZbIIFxdY`=v>1LIH!3L;lf~X{a7o0+kI9K7;!Su7zx#
z5Vh25Cw6%*@z;pha>Czpybk#g*W&<31RCTHV+E=Ks&u{%!mNK|h?VZX<bml16I77#
z>(xBiJbZkzd&=ttc|OnVlxOUE_Fkur8&EO?Nu(Bn#Kws8p83P^PCH2!HbkA(Bnsq!
zo$C3BJ>_}>cz;)PvBl0D3au22FleQv8#|g>R!s7lL;+lDKmAf5JpnsX;Ga`xhw<rZ
zYza7ciGsF;@+s7?7Z(#?nwDxb4)#6OO_d7ntJzh+t&;8liTZ67$jZ0%wNky2_1?=?
ziQs=SJ~%pgm)4#kr<m>ki)i=OWc7R}?rh@P&lC6o@>g8BW785TkoN!G)f;x?CHz7B
zAz=YG<UHMc_WjdGJ6u=ax%Xge<D2g`?|}nv-u%>I`wFv1W{_Zgy1)1I^bp0=kyRY(
zl{Q@X#zUQb=0!8)t0n6A+H5Jp6P=#pm^H}p;h7;ShRK(%R%U7x8kf}9MRZDVo&{VV
zvH6D!8NJSN8f`iWe$n|wedhwV5filU;pW*9j&#Shklu|h<H?uAaq@&g&v6*%AaGV<
zstJ#2`!YXk3JgnBLlv+5F&d@yCA$MG5n!3?Rc-cV!wX){*x4sRjUNn0d#xDqz2n-)
zT~OMFX+K}}GrN0ujk<==oOX4H8&xyT@27U!ykc6$nb-ea#E=j2&7o-w7uC@o_8o3r
zD8zJAR>`21N=z)7?gtBOIsSQqvE0L~{i8A_q9O^|!Ae2|PLv}+HLB5!_4Ozc&b=rq
z+(EzVlvf7A$@EAAe6<99HAAPHs~XuG0due7c<{@)*|}_*fp(RA>O+RmC1HVBlL9h{
zaU|{fHi}>ae62VjN1HZ!4SDb`|39qO)jSsx@yxHyn=27gF^I)4Bmy6f*A-#1y8hyX
zOR3S3Vqw8fc?jG4iiW{#NrrxU56<(83?bt3=VL6FD0Z$#OO_c&7h;w++1s!{T9V&*
z!%TXaG_u(8$xqn(hZxX4b^sn?b@r#R-Q5QSYRN0~6PA-_qt}T0$^@i@07<q$4b;{y
z)GS_L5*{@H&Jg%%GCgI4z?Y+&YA`h;Z~)FL20a*&rxJ<J&LMSQphY%9pBtK_hH_5(
zT8}C8uwB#89FY&wit<e_SNIts=jWRbf~8-(ZFoMG7;GZS2jK1xf~C^(yXou6ZUeXK
zJ<s;B8=X?&C~qGSBTU#%CI{oxOr}@o%g4d$zN~?OnW3%bdyfgnUWO|bp1Ohwe$-;s
zr%2mm*wSr5!QY#myxYeTQ_@TIr>&FGE-s$(KhqYo6KKTUd^8$@_M;=LE1f`s)Punf
z0Id#0S>66Kv*cfVjUxpZO<cM8{P}GLyQCgkVdRJj^m2!x2QT^Y>R|1ro`A7G=6u{!
zqOGC;!0ThA<sB#*0MEpJx_rdtn$$_VIc`g{gq3%T2rRXnY5tC?t|%Aj{X|{hVRtN<
z1*mBpe7sm`l}ie6MbkSF#EdeG_7(?fbvQmgo*si_%uTAs+)BKCgRCi1mxtd=Xjk3I
z41<@&&js=NcsxFA*X1-4@tC>Q-;MY8r*GT!{|A>KFx6=;fS1@X`Xjrr`);Iogs_)U
zlS>+0XOja3<edZ;tG-GRFqjsQp>&dxp2FQ2uRdzZcV1pG$D6w6cW^OKXV-BpKf-;0
zC$kK<>r$nLGG-m8B#pPxpv5goO*n<JAE;#BhM=8HF`7KIhUC7?>pYd5q3urNZ$`5%
zVOXsMuqINr|Hy_(lux6LKyhlGhL2J~*R&{wTD48KIC*nCeJi8fXZ!Q9P~RV7g==y+
z-h+|>i|`~Zu%h;((hjo!vzJiw=z}OOwluA-m=`DfV!uIOCZXcYE9e4y$Mu<4uP~zc
zi2W-z<HZAyAX>q`tf%AG<DdDwJzD$A3;yTE+HZHh`f}~Xudbko=wURt`(}q%=3e}_
z+pEx$Rt;w_4X%gBuP@*Ex5>}Uc+Ebh)E(nUk0(zzf4B4a{{5}ZXFJ%%gqZ3kh?0ZS
zR&e`<ZsD7Vc*;u=PhI+2kr&%HU$}M#OX;f{cqLZg1{S<zaf%xn%_Jm*wq*%zZE#a|
zHzU5AFU7@vn5l#VsPRqI5sBY_clY#|DH+N)Zd!5&mAvRPF<b)>A-ArKIpyVB-rQrh
z2fKxA!9y(5@cK|t=U4PlmD1=n#zwn1p5g8G%H#z|1P)2dVi*mSG4XIJe(EJ<DWvlj
zixzeooAUn@sg|VICmdV03~B^P6K%G3=V*$}&fYXD>H_0eUjap7X5y;VbOBdrcK>*K
zuqA9s?o2v?4XI&Xg6wYRwbzMiN~0`Hj_2v@4UR*m5GHh)?%3{8E=XnlK1{$tuNGd>
zWAFvmQH8Z&nBYH1uT86wB3x#*&pO8g*WQ-v3WjtxScSqMUJ3h?_Pn8T$~c!gEKHT7
z;I!-<Eoeq6>(~y33QcsZGuyPEv^~Zkr>G!6<v6v}wiUsVni!SjG38;zCD00le6AGz
zq*6q~>u?FbdU56QSNQi!{JZ+)-&QWuHzY;f?ieamND`WNHAEfs`02gPrxAm^vZ5RT
zX701@l!A`tl58t!huxBq+?l<ZyrQyZ>&_g*SFbUiF4Yib;bP%dNe6<LK7M^XJw3#(
zD5*$g5T$PTaz21>-~;sn0BR@4?<ftqMV%*Bts$|T_tW-HFc4V`!<UFxEZ<2Jb9tz%
zVO7*SJXyFkR8<&Tc;Pm4;6dhG^X*DFR~S@*uVP!ZH0O1AH@&&(=bJ@hq%c4_U!%8i
zAx9z}i$JFW$z4Lf<ga#2V4<_y7TiH7lxF#Y9A2p7?wre$OVB#tD*vdzAc(#doVfmi
z5F$`8do$#f^DG=DLZhjVI&+q8&6z&S_c+=<ho0%gcV*;9OaNsQ^Se-2$9X1wVW_hh
z>U^^Y7)LPypCaTplv~dA@wdD;$FC}@%ie61)%jTq@~uW`n~t@Lf7&Si61PkUUNt74
zH3}af(s_6?YL@?Xqr6P$T9|y^D6BKtR_QMqr9CgLRsOf#^3cV-gPK+E@`L(WIjEn_
z8`RIrLDjOogZf!Hs9L&zP(LdNRg3ox>SyJkYT-G9`dK-sTE1^kKPv}S3-=D{XXT)3
z={bY?Svjaue%_#}c`O8Uy06y*R-9gr&aGP%PuOwnn8NhFDH|mfZ+*rZFZU;hKehy;
zB;l=JtMqesRTFkemRmR=_`(oVp55_izXkm_1*_ls4T{rcab916t~Q001z_c6@-@Os
z&3DQKBN+Ru?Ow3}l3Sqfg)Mb(+;Riyoe&*PuI^vPYaNBloD0+CWKGR_%mBs5GVfq~
zl6CKD=oZlYdj%l3UINH+z8Wx00PHkw!%$~Asc~Z`&#lxWm_3@)4Ou|<M|5{-EQ}OI
zhh-KfUjwR{09FMfp$@lrWuz9)7e^DAnlUiDzTR)3wy|D>w$*DtLtNy-&_XA2WRW_i
zIJ1h~m>rFGC$A>sy>&TX>DplPXS~5ahXHQR8cc!RrP*a3t1ZQGSPT{exDA1`B3mZg
z`EmSiwgmRz4o=TF!_$!{NV9zJ1w;=sgN~6gw2yf;F7ak_s4}aaeW-G)cg2@@|Dg)l
zeDtBJvF{93K!(Bino;Ze-uneZrE=(E=(5x+4kotSrgz>$3{)*F&-_~cOx$nCW=g80
zAKFa<T^iQCgXUE+(Z)4{w0va;{nVpTL*gj(Yd3LbwaWibMnYnNX0vu?<I(Z%n}+(T
zsn{8hz6^gMC}y}4O|ij~sHemb(n!>H7^=&f)Z^Kcd6fWj0WBL`7!zvI5#t2O9CLX^
z+YNRof%kHviu>6j&eLJlS5Tk1Wra(zB3<pgt}J-1>;)+Y7w-no_F#L6??Z-|a|?YX
zr3A~?o`bfVHCJ!DxyYh+W67#rb2y4|61O9dPw}lj@}(q*4qrby{yiXpp`?nrA?vS(
zjRTEAwiI#kxSyY>ck%g+Pe0whG&{I7^Iup<NCy8)hBXsFN&_M{_4hnnYU<xVmLP_{
zH>%T_>`U>6Xx&e)Gf%+_4xBs2jXK$?Eni1PILMefDm8SW(~@ODb1iB2Hn?!pZS}tz
zew1EH#y>|yN7GKKz?=->7E={8GM8pMGg)EOKBkwltwO4J=j*P=`An;+6dFSKlzTJ;
zSa9b}aTniWN!+<QFDPx*kH;9cy!u=tX`iWBRV!#`qhfc}Tvt8f(#nF43AA806uN3q
z`5hgz?u3V{Ov|(os7BReTV7&^O2s1<F%wphC|z<zJg-T}ymE*56m-&QMAEEU3dwhz
zhnFR7kC15F3E}`)XuZRwGO*itZ%fvv&>E5OXuOIJ_M?d+EsA>d8Xu!&6+ZQ$+nHhS
zdIt&a)|2Kbv%cOEu^=rl;!8~GFqcCp;~0~){deHR8Ru$3wgHzRUvbsqnr^0RtDYE+
zx9R%12F`mChkLuQf`Sz`J#w)(GhEHjkdw_Mhl|)QGf&{|{1Sr@L^Y6_NID8N0w730
zfN;yMW$-#k=h3A+rCO58Okt}gyf=P%`Wlu=moT(UN_o>qV3Ty8*sbh<S&$kNamJtV
zk47Q!!vc)>C9y;N>tR9sl!zbxO}R!Mrvq!Irofs!?7*3tiY$GTF~*m*o34V{shX++
zVRdD)2hPt+^;koEHIZb8r{Cjfq^SR-K`gHCF-t|`NyC@ZV}$M`RCeWxX#j2h65<y~
z7Ia}((#Yzs4PNqUa7Eb5*I$tkU%q<%mGt@!@<Il;uiqKm#v6XQ!+XT;=t<wo&tQsg
zq;YnE=awtxxihMQY&cs4z?LEm$P6pbLW7s5uU?IhQD3sa&_&BLGr)K;eEj`0`dCun
zNg~wd)2B)(qasTpEkt#el%+1B2CEi&GeNh9Q*0~N`LYiNb1Zm98n&<;(o`k=+O0;m
zCw!ZjH%kJMMgu+wLmdS0X?G0R<sO;`N`Tz`?Sdp}3tj9V7AVu|yq*W=2;pfCO3a^G
z0wt8vQ6(>UGRT5~XwzxoKxnmP2BJOE-ftTS(G+JI3T6MkQIJO8V-$MunKKTCPca19
zvYM!nUg58+qUcvt8ftcm8}eqeSEmepYb%tvr<+@*u8v0&tPZ^1e9Utk7m6G&ADL4~
z#}(`5<N!+vGH)c|LGkV?HD%u(!cniur$Vu1%=`IUdyMZ!LREAAgdRI5;WYj^5<~<b
z?H{7~h59LV#*CKOkZoq$LL&tqpu-ey(Na21sr0iW;C&v_l3bd>Zd|{U)t1_x?WTG%
z1NK5gKP@wXHXk83Gw#*EL5lDTcHrKsV_<g#zL(!Ie@WiY0hk^?&p*>dY0)mnT;=(H
z&ZdX?@eCCoj*fra!>LE-s{9O*h;vme4rpG4t<H$Ln>pu`_N@B^azFa~L2!|CF49!=
zpr4q%K%QrrIeU|n<8@)}4we)rm_!`ot~XppquJ>tL2;N*QSRr1eJfYcuP(+j_^n*Y
z7Inb`das;QcbR}iD=psWno{BA*k%ujaS<u8tUf#3A#Lo^_*wR)(F~^I)!X9KU1WLq
zg*}`?s`qwbs6r|ps?x8zV6ovig)#G`4BYx-fS@U;@0XYdCZ+WK;S4vE%Hd>O--FK3
zr92STdz8WZ3S344I^9%PeuwlwqeZUCXlyRAc1M4M=v^XvWi=kcjg3)DjK>9FV10FP
zFL84i!rW%!T5tU51bn}3)g`$-7Y>Sl!Q`)@eqE?OzUEy<JuItW>OE3RDm2?-uIeoj
ze#y#btaaaQnyJzBETI@xT*5ClL7t5Er>`qTC>!xN-r^j-?i5+Jm2aJ3MFb}eyOxAs
zz#UI3vp3k>(s+Gx@(wM})0ZnqsD1q!C-RNqAL)6&o7~I;+2O>4j42_co??{f|FCa6
z@JpspwimbbG+`b;aD=t?!xQ{EoK5$~&$V06?}g4^gcCW~v(_2#nIMT|V5j;HZe9EJ
zV05@QcsoJb#mn*FcziJZ38f~7gHNyBxHkBHhTD@Cw)Y3y-lNT<MI5~{9wC@?b+Ebf
z{nqBwoqL<uLU4EE+2%cw?83R0QLAY<8}B1dHAgM`mLjOC-8WMN&X6ArEhC2rfw-1|
z!VFJ@h^Vf8)+|(fsM~e%dicW#(|Eli3Y41Y-eGD$IhRTC@ak!})H+B<N1!LN9on+M
z@iQ$481_p}<Pt7W8L(h)x;%>)Z)jlhV05%P)MQy;r0*nW!uxE8<(Y%ZZfZ?zB5?x6
zY-gjC#54g%BARldmEh@6ba5CmaCq`+_2OW#eVcp7?u0ws?CSIvc9$DKfnEwoY;NkP
z93i=nOkamh7h?s%4V=$Mg6Ld%H2ZPma1R49*~9khY!-${oWGj)8?LS#O&EN=JfL3V
zd6quzaZJmRW5=)`LWgmT3oE1wnvOI4A)6U;%&h?+DN-3&KmkcG1CsoD2WAJ0zr*W)
z=C+%Swf~PhZ`NL1S^awbYWe2!uU44aUZX%XHLv}wW?REN#=#k95v;lubaec)=iM~0
zU?q<~H0o+FEmH+@8^t11Qwwr%VUcZ7xu!W6mlLy*kP;qE!?>*<&+Nt2<j^LR8e#kl
zS6#`FGQz2?VfG+yx3*OA&&n%!g)wkFgP*2@NDD8i{4SN<rQyG|&Vv{%r|@buC)5WL
zrzcbF)Wn`^oMBL=E}5W-k+Z!*vZNv}a7ye<toLY_Cmz6ahR`_f)>bCgM^SwWbnTAl
zi`AsNNJKsGMJ;ny8Npl+u&%cC?B0V%Cbg16C}~;+K@d*C4k_ui<iel4OE(5;kD->R
zlU`w7&Cn|k?a@}R4nDi_+s}aunMKTqF7Ti{DV06$+KpNrZeSXZXt*6ws13lJW4AIZ
zgQ!XP^qa(ISFnFhH~hV#MrX9#bp;6!dxbBDrw0#Jk>I0+wq3K3d^0#H50QU?uj0n+
zjDH^Q(yE-i8?3!1W>ApbFSwlGaEgu6V`y0`l(af1l-w`w%s>=ays*{b8=v2)WlN>(
zbK2soO9&Emzc9eVKJ@t~uV6<D!s}YuE--%urGhOr#h~0&)Hx^@pWpoAi;FrcyBrg~
z)O?0WAM8Bg4nxpz!a{3nD--lSn`Pxb&1SfHkJqO|;faR8h@jdH*1{&i_a2oz?8WF|
zlP5i+fpMbmR|IQZ<$)o<+Brr$jjW5<X;j>0aK1&XTbWoERNURZh22Xa3AhxqP*R1j
z7<L?f6JE$5u)VSvDz4qsY9L4;jCP_q04#+h=DcrHXo-u+O4!`?2-xcbGQLt9+mJJx
zU%`m6$xGOuo8FRiN>(<F1sE8X%Y)m4Pj9hRrPqm5i=C9Xsdn2-1aPt`N&6@a<9$>y
zb#cUG_tgy3+YS#ELx*t#^+l@HmrK&G+*M(c#LSf8E-JqS|KmUg+TYw|$b}L+&@}0j
zN~02puoXAR?tXazi((UMhZ>sE8*qaqj~YjO$7IHJ7pzJe`X<{Y<s-P28jk#hj+2Dm
zb+gZuBvVt{0gEeFWHz1*wB{eSMfCTSP?gWn4eloE4qRu1M9RMq2sQNK?Kk^(@-eyz
zOOCZj;kDuI>ny_N>&gQ6H>x{Cn1gCfffrG}y2ISnvxQecwA16mL@c49?T;htVcT0Y
z$n+xYui;E!O7d#L_z16jQH7i)>5ew~;}Zq^%#MXmPFF^UGaLf|RvZ2WF%47I{uXrp
zWReq8^X5&h#r=^sZCM?@qmRvV9Yj|ouz3oQ{r-d!2y7ZSm|a7XV2oM;9W(=_MEa$T
zq!|~1W_3RCO4N4?D>;?%Z-da^PWJZ3hfi^%5a!|YVu^v&=+His0zp})HI<YJ|A(Df
z60q&2|7vjW0cPUQ9zXrl&erCWji(#<DI)^60J!YnnAE2`JJ*KSuU)&29!*b=F}1$_
za&mb6jr3AX@HG_XO0tUo@aWCxcwJ3~DZZVZ`wza`+}RPk^=lpSlY^s8X=C<owgVRe
zcbFd>?d(mCS1+TvF7(~$5stw{OJGc`3wGc$(6kq-5i3>aL}qExBPg#&!~{_!9MQ@}
zWz1CNIB8KTzW63+OBYTLxw{7o-uf!CleU!Db-trnv{`w(t`C58K(}z$y6@ddOZ_z`
zR^8p;H6>p;`LqGpBK0Ix34*Aa9Adb%2b>2)onVE4aVjX1M#059S%ZLqSY<(8*g^by
zFx3_#b>`0(KnpU%HjVb~9xzKqV|(Jr40mU5hVVU@LOeKlcQl&K-olY!A?)JEsSB(%
zSUV2iAOuJXl>_NGU>-BMemXlAnHeAc#Eg-(T>uq_n<MolUVRA_1qWOo`QO3wmE7cz
z>(|mb{A;+*XJ<oV4okO*x#D^#s?!veadiYNf>c2z&l~aY<rw3^$#ejj9uIf{J0`%Q
z8g(8Nl&Sx0DAgDddMt}E3yc@Igz=*(Q)PTUjzJNI#j9-FG&VSY(hLE{Y6uk#I18E%
zQxB|k{WKY?F-!};IF?+4QOazAP`OFs6Q<^PB=C~pqZ6bEV*YQEunZg;M_}gTs)J@k
z7;jkMlXIqoIG`ppTht;%ACdTPV{}TB_o2tinGyrw#_@v;h{K&X1ZB_K1LAN#@(|Q?
zhhBP+!xGFpg>E<qi(66CQ?=Q0h7W2D5Bf5~BcPI-0z7Ap28`}9n5;j@5}2})ucsuj
z$(q_0K+?WXL=9hIC!;98AR_U#pb}rWltUWm>bycY0Z|UDP0tXZ4XRj^Z{f0^Ao&%Q
zF2;|pr$re*{{3)p>(1c%-uS2MMD1dTqHC;k_;Lhm0hjhK`p=L04^4>Pk2-rp^u_fe
zG|n_TUWgkEK?qEVI_e)+y~Bqv2b4LH4OA>Bbm@?~8|>;MISf*R;&>~7;V^mF=+>Qo
zzsX>7QYJismIs?c9-M<5F<LpWqL&alSezrMm=4wp^jN<^XzDfVu_63no$D2kwWisT
zSRCnSwnLa@q=&UN=ZwVldLE16XJJq-o2XKUwCK{>7Iy~7^3kd-`$M-6@di<M%=5n4
zpYFnhjHQGv5UgSbX6Fam6jP6oG5*=bG-L-B?0AQ=xk?8z8PQ-l;8B4I27azgT6@O8
zvM(V6eFcYz^YJ6*9HppYy%Y@YlRsSoPQ{dV-Sp_hf3;3iYHncxk~3$li^vZ!ysctU
zQkl9B#vaPCia~R_NDpr#-3xo7UfjOUyf2^ZqZVr?hyo`=t}JjHF&kz>hU=aVWgj<F
z5i8+9U#vv(t!*83{B1|513(4UY5BWI{ED8s329xyyr<m22NKybPlCer9eW^8_-{a0
zX-(s?zV0W7dU0=|Sb1F0-oPIpP&eZrxZwOn)e~#V32a<6HF5Dslb0D>!MxgmeFi8v
z+hM@YDYYFIqq?b>5(JRa;i1wjD}~!%#3}%|2JAguQ;5aG-Y$PRV4N*dW2OsGsmQff
z6FiDeCEvS}%Bu`Zq(FMqh1~+E9mSY+a5Wks7p4bYya!zhu15Ceb?`KY!5J@t@z>g@
z=^5V)$XI6{K0QBya#(XZR9?reP_d-O`^Tx6GzW2AGvZxaSU4M~(lK2V&KTA=sUA#G
zxH9C8?CyYC2aRbA9BNGtczz0p79jl^tkm^`FBm`Naw6EG`06G~BLg%`S1^Ig5Ynhw
znoRQUoNMtf541Gv8_ePu8<F<#3B5jexwD7$%H5M4e<^+~J|eSEjuDX=0QuZq9y>hD
zpbI|s3Bd67?PTvn&JbklU^WM=3~BwYaaQJfhBKr^;m?S*r~mL|Y1LlT5vm&WkhoQz
z|8pBxhe4wcap;PE<54ozE1u?qGCWlDW-9kNYu_9kOb`}Lb#L({Ua{!sH`4o0Kfj@+
z-@xE|;V*8<%Qwd?%o6F#Z$AIb%I}Td<q9{yxb+)(kGHHN#ZU>P3Gx2eUuyNd{Y|OF
z!4MV19u4k5ZR&eF509T#J$jZ&2NfgR=&r~Ko?IA^>bIfm+Jy-20v#u7KP_*D^0G(l
zf+KVbUYD<d8mfVD9823%W<2w1e~PRs#pQbH3D(B0AuZ%uS%;1dCR1}+=pj<5juE&#
zkbj-|?UBT~dE?e+gI~pe0**=1Ub=Sk)!=uNZ!kz+_qx_8{!yNm!O#Y)7J5NH><8To
zvk7c(@cRZRh2auwak1g0y+P=Zmgym;>X1K9PH-QWAFr3!F;p613X)#+k;6MNK(2XK
z!fAUDUK>x8V%9@Gwhy?(Ptdpdn|f1_5ZXJ!t>7r9x!MAxSWuFMj>kth+<Lrv@jCa%
zKi}T|^TCT>9bAVcl&qtq$$C+Uy<h`QSeYy<sY3hj?SG4b5cmHT&>pD4Nu-8mY+8_|
zjTAc21W=lwIi=J!pyvOG(+s3n9w0Yz*;}5{oT&HIP47_j6kyNeiG<G|fHN7x+<1*$
zIIocnr^NR8u^IAghQJKzo5g^=ncBd|?d@Wx+<gepsII`y=Gv>x*Q=;8FN7)J7HEA-
zUEoeCHv#Hf-nzX5#=)hwEYFpEnd4&JQeSY)JRIeEH2V$7xk<d{%|WIO0|7`pt#w0k
z+c1#1ukweiB<trd#a`v<wqe%=A|Hy1z&3#{I9%EX{Zn^3{Xyu(&QAxrCbDi9H4#bN
zKRDlq*vON2`t(XihDMDa(wXc72PKE*7oU3m46%77P)dM3oSq<ESdKzx{%$-0oox>H
zG`6SV4Wd1uCD++X8Vc!7470?HWFq<hUQV+lUI>MJ`>w|Pgn&!Y=!tFn{Z^Rbd{io7
z{1Um}-r_f-<Hd>JnMfzR>hi1<dd5ocL!JQOm!OR_x0ou)oHxZNE0|iLLmQcmTD5<X
z+LIO|sqBK9%pbf=KVaBl2TTa2xePl4<h5(0P}gl(+?mKtTF*C*k4Nv)*Rr>wNI$AQ
zm!%FEXXJGz^Rj5KN~k0IMnXkZ($(}|%7+~$RnRe-`dl;t1+AexI2pzk49t4u?ohT<
z6B#2f)mP=eslL?-qb@M$t{SlS_6)h$*}oq|1AV?7r3%t7qP10CvKI6XcU$L6)Sla$
z``BeY-b>zCU#ni<{#~-zH>Z>Ry?h*pdU6Ir6;n0h7}EGx59IObyv(fK1D=5)n;S5O
z8Vy-t#NHBz><u>ayqwM^vd~gdZ+-oT&8Ocy-g@xtPdoS}|H0!XSwtp{Q(0M$ruqYH
zWYGx`JXvzyoF$BUw=!oo#3~1@K><x3I?~|=K%=&e#QM74WPof>6cy*s3dpXSWN`R7
z_R<Q87n@<T6cn4(qJNveP=7gKX-f2;kT8?~HoB_>6-0jUacSqHBq||+nun&~GMfFJ
z>C6AaSQtD<M6j%RMyr*I8aLmPNoe{kZJECQ4sm+4W-C2mgmt!y&EbYi;39+l_IPx3
z*LW+edK(Tv!K<jt>&RgnFnBnARK{AGDi%5>t5{pabt@fWfD8*P0a*VA$0H<1rf>U5
z+*{AVX{-%+xxLBWfUYc%E6bxJsTtU+&2)Dvx|IgcGH;6!!if*_2FSV*!I!-VWAaHx
zp)5l`F;7xN@P$L2sRS>q3#KU1Z7Cy%CFn1@>@8kVScZcebR&eD@~Eg9M(5ykc7h!a
zb23RGJ!oE<>eW|boUGxj9)yRqmw*u9@X46cwL>rUNUA*hd7>JlG6zX&l#Az5qE^<0
zggov`NwRhmn}r@2QT}<AFIcC@dNT5)2|8BT-W_oP(@w9cbDwNnLznS0NfCZ~;?o>S
z0}O;daQe|aX(ODqJEb}_oN!x}j{N#M4nWd*DzPT-S5bh(SzsLgjtk-bjYt=LrnDgM
z%4hvGlax3iG{@04I^rpvDx}u~*-eNn<{)c;P*NJs$Z&>6=a0V*rd<AuICSQB1c$D?
zP1M6$7n0K@Y=U_weJor$X?OR&93B6W3qTOITfYg>F}%NvJmv64_8+1?U!(MJ4fIT4
zoDKcso%m5%I+UI;$?kBpe_!%2!Ur5>gVP`GACLbBE!eSVsSQ<lp)|HlAE6vp+*28>
zYlkOs6rN$Zq<A+zk{*Y5I5KV*8^a0fT7;&E(@VI^L(oDEU%^EcUBm)ICM2*Y{8(%c
zzH&J(e8LU`F$?KSs^zO^N&b;X=@83r%iXcut8uZ9ga^aj__%sit*Q<gS~Cp>r$Ikp
zO7?K{^A-jWF+U*W+Qx%)TCrMmpv7W3ngjzQk!8t&9Pb|i{%K;RQ><umg$do}Vvwp1
z$AhjUIfMg(fAUG9V%|-=82yuC2lyvkp?HJKF~-Nr#o*|B3<ORz3x+X0Q|t`GII%ym
zM6W!`>{RC^u+evf*J^5Jh<0jPaFiscPG;M#D%qo+4nb?STB$qqv3g+&UxU4UU7};#
z_YH0>9l_YszOG{IbinG)CEMHXrEJ<O+UG)mzThY+(B`;!#F~Sm10NCXACfkn9|FC8
zf;+AqYWs&4sS{sC33;f8Mx?_GKwk-f!*rPTjOS9iLJj<DuY^_&GoRqMpjrn_$}8ZR
z`VFvKt`TB9uGSWFAtc4dk*W}(d2(N<OErigRid<9$NahYFub<2^KkRow~y~}&(Yvu
zjOAX;)90R8^7+COOSt8L3cx+d%L(xOgGs8s6(Dji8zjyxbqGY0p6m8fNDO#$fRb@{
zD$i;EiJJWe9CHXVCN&gj5x7E%#>#B+7g6xWCBdayg{OnVMOdI876_MMby>!t_19Ni
z=y@Trigyt2;wJ%Re9SGAFa3ZNkqpQ+jm)uS@2{F(pu|G7Tpjji^x-Rp8LaW>HA>56
z4FlDW6FRraQ;R{Nj9A^n*oas@VAl^4-Ry$w5<UX$2L}9{o%`{tyyL~g&;{AWF9My5
zpPCW5{s~#nfict;fnaJoZNy6Fyk)Z*!qD-yu&e%mLQ{of4^IpkPa+vC#vHFX9gtw3
zIjuxAK!sn{@!qT#d2V1YDj&a^+tLLEJA37>cM5f4ANMAhYLDKPo?=as(cVuy>4;e-
zGDo=?vCb@$b8t}#8}(T_&?0zj(X#q2pIRJ({>$2m!19PTx)EAnD>7xBx0N=bw>1sX
z&SRfRA;juG10~ZtaOvvH>2x2uHoeO%)ioor3yF$8R<5+g6iHB629cR}GV_o@0<+}c
zSvmDy9sX&0isb*p=R-tPagiJxA=%gG5<7s_^P}ShzQW}MMe_W}GRPcBZv$tt47^C9
zApJvVbo>+S*aUt6XdpsWFcot26i${vkbGqq7sYm}NjG~c<gVxT4RV&Oj55r?DpCcx
zZ_i$qz}b&=&fKN{p%b!EHDJet#zY7WPdWYVeIbu(^=nSzv<xc*<Q%r=lgv`+jkLL=
zLf|ky(t}_ISpnna&=TWuxGrGec{G6zlELiCJuz@Tt&vXsGOVi!eO#;h!}kw1pIPiq
zC&v<QZ7r(=wAi)Uw5v%73<pVhidz)jhG<VJ6M++jd@PWge2i?|@i*8qW_2uz&MZIV
z-;sL=1U#670}552@xk47FS~Zf-{7N_b#1U3GX{4&47qyoua{PNu-Nmp9p0Gz_4@O5
z+@5_2zb}7%>F*Py5+;Sp$N&x}8)Jf}dsB%fCM${#et6@gim-zc2|X22kfcI7zY8J)
z)(Wq<R{G*FT>z*fsR=k*E-)dZqy;+5i*nSqy$^t1f=}2NMI01*<o=kWcLN8sK#JH=
zZ$!z%18^eZX;tMC@_xPcEgt#73yYMk9XQ_rEWZhVYA2mDB-G;b^)xQK`?xWzBeR8+
z{ll2K!-aAVLYi{a*T+{vWW9%6>xmGqK}?~DOv)KB&RX{+P8tIi3gX&u=WvQz#WgAV
z37n;|32^4(>dp^IB`npvu|pWqWm-*q4CX7?fzqVN3kMe8QVf<YoN^P%PH``V*1Ial
z6eW!cyWbdzheg~GgJ2Dj`q*Fr-xNCs%m=^|pjiRxe~*=iI$vAD@A^8h(tFy4BCHD=
z_nCA?SiFNxV&)ViaT8_mQvugK&MF>9=KB=&RE_1w!d-m+^9?zbYfgci-)!H~bAgl-
zty!gJra6w(QOTE%mCe{H%@}u^qeTF`xgq}mZYAysp2|Jeg_1f86y{M<eU1517ugwi
zI$-2H2<)UY6KIRmrYV&-Is?HyH66PJ)|kD;PF-I47cErFtX_Z~N3PG~a+x#QXw59y
zTu%yDL6Yk*WrjIJQVLT1uxZbJn`nQDwYG(jQ>IEk_;)*18|lT;4c*)P=KJ6Ev>?eT
zg1)Ays9yT=!k(Tn+f(b1iVasZqu2T~;Uam@K2CM~eNI#ygJwoY*JW_W^hhtZNrq+B
zXCR0v@YZGQfT+gJjlb|AGWg_-NR(#8*EYbQn()Cx*Cv<c479HWpnUOG;AV4i2~(wC
z<&OYnF)Z|*I22V-7N@oO+gz5q$lOBe!Mxovmd#-ZX2L}5QC<9(_zq{j#ma`1bld|V
z>~*6Iw_O&H)+$BV-Hnpgv6E3^yozXgU!7>76S31gj%`lSxIHwLYs7lcVCtM|1b0-7
zy7W~p49EMUqZy8)lym5@-<vBtgAVtQR#she+Sb=qe{xxARTq^=E5|zonAtYk!7<V!
zu=L^UC0gPDev)8i?jsic6ViOhjI$4ezNuYV`WgT<-$Elz4o(jc6gqeb7Z-;$)9~S}
z=b5Ak!Iz11*Nlq~ouy1=;Ck*}$`Xqoj(LmP>H;ER)h=~7Hv(zmh2$M643y4=&C~23
zCL9}NV_;)~`O5@O?NX!vA;%=O`!U8O1Fyy;et5qzX$(mrVGR*kNv#|0b!*-d!q6N~
z+fk@M<m9x|+2}PSkq5|I3k)zoCokUwb?p()RCSKaycM+9mV1bDj0Cs9Z#{du^JL@R
z&Uc&lpLy@#uP?1pQhQm_RcBbxWqC-I-E#b+b-IZiu>zg{&CYYn`relr306D1_SS7W
zHb3F~0i*P#>zANcE)BLXfq<8A61w5P{^{(}C0u)7ntzff)b&~@cfq|Q>3AqJ5B}hi
zY4Y{YXVO^dVfD0D*OI1_({%3kuU|^+zP^_6Nr24*S{f;Hyh@5KO^q6x!6lU&r&M(2
zmO-hiFO2k2w<)OS_WThR2J~lNYPu&(S%r4q_L7w=cX3gzlwJS43=}hC#sA7Qt-j>=
zU7VPCGPwWv>BEg@I~Mh$q?j~8lrUY0Wgf98FHl=7D&5L9Wm&TJU{l?t^&~hQYy{(-
zM~@$F{8tIIjdrVK(gjsV4cB>9+6ZEH?9&V?Tc6;!lHrK+S23~8ok)7C{CL)sIMiXV
zCntKr4XF~~5^npW*~vpqshE8?K74)hCR79B#Njb+PnjUEP8lmzoyx4#xh=gm)!cax
zMcsLqR@ypWY=Wl-;Rl@~`P9qJEc82oQq2QV{^QmQKi$dJQx!xU5{BwW6KJQ|w@g~(
z%%pCMTIw@v(0_Y#eEu}F$>+~o4by8k6PKINmUyhh*I$2~4~gJe71i+-zy2-WEp>i{
zom0_0z~a|_AVoaD!jGwFoM3VD)^GjDi{c22&u@IzdxS;3;t;nye}tQOo>>OV-S0NG
zwj_Z?mEQdPmBlyiaMIGh$x2lzw>x<Pterix^tC%RnJ@$T#;5P^oBN@#@@%)xY%swF
z6D~*^0HE$fw}w{ftHmIlYW)Zv8C^{UE(WP$H7^?>27y%{R5$FQuV>0P2)PF0*s9I#
z467s>RHoe{L2&Y+!AxJLAgwCaxKXO@R!BI-3S%VomqdIvc}bPFX$b`>6xy5GU|JQ(
zVExpJD{?t%O?~N$E!~ETqdX9P^rbjtQDWm(_N9oDHPAIbMf;Tj!nbsHvMgtWWm8q^
zBuR>{27>%h?gCpB`xK92-H}w#Xk3>*qMxLWfd#;F=(P+s&#6zwv1<hu;gll1OB4)&
z+eM`<`fB6RMz!#Q&ed6i5W1TpxN}=pNAo?XtN=p05&^b08SuqP3P__X({4J8hMl|{
z2(J26;0isN1~XQJ3O#aH0|6f3chRy`oK#sS#Hu%%*(3>$K}a=Yvy2nf(vTdh`2jp;
zlyFBGVsX$5ooo_ZMhAu%%VIK5+zCCy!Ay$1<b3EZ2h$UVD|7`hOW#wXi+x{=*@9db
zLx7FgEi)>B0e&7N-7vuwo{$w%r?yy4Mg|pB_M6dJA73v;<Ei7!Z~XbBoiXDT+F*r9
z#!O603HU9MwbJdfzDa2wWPM*j?(&!yThUyWe+*kW5}BuLSq<nE+990(Ywur=4|!A%
zGUBuYdg`yHZX?$T2dN@D2b<ve3V>+q8+sJfrN2t{_}?#aRZ&Xg-?*q}?W7e`7iqOE
zRpCBK&N&B*z=Z|>+L5TTjeZ%4_eNdXldcigR8}i=4GZ{cGQ0<&y%q~TXYnG>E{!<O
z4iZ@<``Y`|4hE6EPfmo}dEb~TM085aBsJA=S@5}FTD^;=AggjUhsE?0ihw$C6xMIM
zb5*@%Pg0pl5U~(`|BhVGqZCl^NG|{z{xmuU7j|<W*f?+HJLY*vuEikGqBax^y|mKf
z&}r<pOpoOpB`pZnuQCY|4ZVixs$r<*pU}*ZSv+aUr8$hi(nxOuo|@<|K%6&JYN}4K
zsO223WU~w6<S9GGXT>JwDCDVtT|(tDRYqY)Tch8^F6J&NxXc~f5`oY}rk;T*)-=<4
zuVaBV+(gYkG6q58>V#IAas~HSx$L5`cB}v#eXR8MLo8QV<A_P)HW+y|vlvt<(paXo
z_DUv;6FCkIm;k^`2Q#}PW68N56lx?CY6v;wJ1#9(J)?prV6Nlpr`W8uijUo9BMYK!
zf3zcUDR#Kq8=zGl@MAG2c;v~_+UyaX&TF=cW720pP#a2BjYP6Eq=ny{;uyT~Gp_xv
zSh0K}g-d9vk2q0lnP>qIC)rmmq&TIQU7YL_=%zS28mY?p&=5TnQIHpC-kMoT9!;=@
zHVWF8^fL6t7O=7)a=EJ0rBt)GOTN_4mV`;n8R%G=IFbzJaoR7rjX;XC+uFLspdsyw
zY8@I@5e+}SRbPk0UoDZ%i{ngIxrEP~GC0WW%$FLt!tpY@<H<hq>3O~^d{=9sbtr4P
z8Bgr1lfu$7lUGzL_3TtVE)$$Qa)4oQUyMY8?LvGyGlFXxFC)<j5=DHY;Gs044Eemg
z#z%OL+E&AL6q&RZ6Pt8lFlX@RTQ}fV`daPN2fzDP7HdHd?_ejXE(LKq`DN|}b(@+;
z5cT|PR}y(LkFb$AF0=A91BqmU138Ev@kr0SZq;`-;`vv;mO2@$;x@9)h3Y_q2ZwvO
zdlg53g6~{{ZEV8&NYLP^kV;7?WV{mI;3Q=Q{sX5c&=#q3`f}GNmW@wuMu)G*59C52
zuE}&ue1EiuDG_Z&_$s*KL^E@o&}pxTFDH{-mv5v_1&qpXX}zQqAc-;lioDM25Ei8k
zP@sU~Id`|{6AC~H$L~rgtEdVy<c#8d8XHlHd2Zy|585L&Zf2yJoSDpf29c0dBt#`t
z?|^`=;$W!`2IBd1{`YVE&vlQI^wijqIj7uoT<`=OAL+?0IQFK<D+ppBb81$fD6I}3
zGQRU>^wW5NQ)Q6+j;C9QEn9N~L13^8-%GZ6&U~h4R7zA^nD$NYPW^xc93EX@RSr|n
zr*73W;<<{xN=1|K0j|1M+3eN|4}q(P)xPn-MJA!dFA1yn#V4caU4Ik>`qaS_Y_flv
zu&oU*sK4cV4x>^wdZwN(fMH5Q=+pp{Xs?9a?u?e|9OQ@=Iq+4r%es+6bY_Rlvew~5
zBq_v23GVKAkMfh*w#-~u9X>HD6c}{0kJ<L#I}K5e_bUEuja?+0Xgr-Z$JbPjfp2P3
zWDQEDQ2a5-MqpNcaQJYtzdwO*G(Fry!V<Sx&gK5Zw~3wx4J}}n+RJY>jXf0Y)b(JC
zFv&T{i$m@t0rMTsz<xQ#EiMMg>C6%IHayT$Ec9?dmE(myJtk@tUW$08?24*smY3+2
z2c>SZB782$(c4wF7JL>D60VA=pxH3Z8KQ1j)xu`ABsd*E)>s+pb7Uf@Gb`+4+6tZ-
zo}i-bg6OC~l%VgB!e}W2`(AZuO=4HIBW8`ycKV6w)y8=)eu~JMju$b>s=EMNv1rd7
z<I8YQ{3M)q-{?q@QHu`qCCI7o=Il2RQR>tTr<i!#4U<rJ2xGELhdx#>`bUk#ur755
zE_}mQ26|)CQ}<aD!zG3`V^cYwajVSvwzF2-OI=Urx$DQ-DN4=}<Eai9$U$AE9I?cJ
z`o>Ay6>4L!Xj7;<IaWArzLc$-V0YDv=22_W?r{9Wc09QrD{NE$9;h8$NIqeH;F)$o
zEh^pI$}|tNl%MNcw9*SC>^hEVsoY0KN1_+5r%K1|PY%l$K=Sl2iBe|i+HO^*q98FD
zwfe5H3)v{>d-n3=z_IQGI8wN~1&!A*npqRhgCXkh-h2lw`{lY5;b5|hQ@$Bbm31b$
z3rG{yGw52A&QaJIyk`z3zojRc{z008<ta{fY{FDYGJ4DlG2b4#khzFDv-lwJAKLVY
z)U8q(6N98ZN()eLY{n2m^O~sKSLxPPp7vPFnkZiNDbS?^{!s6E(n{@J8Nx4#TUtYU
zf7x9JX-F?IA$`XEALT9sl``gkA~8CY=<F|OK^m;a4F#8((3*D?G$D6uw09FMZtZH!
zu=l3Bx~S?|*?tFxN$S5yC=sVBlS@-{a#0XcPW?aQO$Lu8&i#v?<f_aK|DnF8b&Hp$
zF6!hT5TE4!#OMBF=?Zt>OeeeJTulN77G?FfeftqywL@9p_Ju9GK-=G%2<z?CIl`G<
z%U#LZoc^~ng~k={GI|n3mCLoXrQRL8R|k&rVC8>|B<?SJmLn5?c{mexHokwlrp~&h
z`#h~w5Jh}$@)Hl1PZREpj3`sgxkjh^*b*$RvVw4Ir2wpl-YZ-JQFmf*x&gXaZ{Y5E
zO@#)dTpUUbf`O-gnvCCyXPRw<A9td5KEM$Vk|+0gxPTrd50IYMFlODQ)AN&dVZ3p2
z!o^yNoF+ujM5kFrKe0p23z+Q*D=3r6^=)>WyU-7pAmE&(=_&SFr*G^*j$ri%UIi}K
zoB>d?>|P=U%YZA6#doM4W-+HnMqF_bfBzqJcf60iECwao$&Z?VI(n+uf;{;+{K2Fp
z@Dp4m>#g{J2>D?k+l+T1Y5)?%V>3xD3~)-&*0XyL9wELqIz5?srT`0jFNF6u!pCEA
zqvTN4YQ3NUfowdF*mUo2Cl?^SkZ0{!a73nSxy{M!+sWSE_)uj{@B6k2xH{s%6CdEU
zYPqN5*%9_wfzO<FvGcOU6p5H}Zp8E2Rv~kN_|zirDBSlOCBmYCUl9&Zn$@Hade##<
z6*~t6MamM%uMR%D@!QWKSM7RPG-@nYhZ`@a$K3oZLRk=ljZ$spS%A_uk}3{Pppugu
z(jUaf_GVinkV46PlD;8VirMLG0H-8)^;(5W_f~W+g$!Yte-~vR$te^3RI1ZlWtDBH
z0>TfR+a-;Pb~ekDE*eS-G|I$TrOG_c4)6cLgpig&yX74zHW?HaTUY?87Mks&C)mi;
zX6ImZw5pRF5l5_YDRr@TNteNPejL98%Xk|?Sak;=a&j$7%Ex3dOo8D<x1vJC88HV-
zxGRbzm;>fTpi#P~NTu<dPsv9iF;X`$|4<F${D4ZJiBB!{UU1xWzI~fF%HP2WR45RS
ze~`4A6#<NF)!Wyl*b81!E6vK#n*7?@ewev3t0=YA86?w#$}T00a_(#T7z_4xNma=^
zviGC^evo{$zg-gIevGVmhNKB9CchXP>l-;?qWrLShZigkH7o>6Y==e#R#b}DweVcj
zTDzklzS*Dd!VQu8ON4VC`pguJCT~T1q*RsM(9rH-!C*9MbF}0&?+}2}D=8dCNe3Vq
zJ@+~jkqa$!1_`ay6Rltb)!~PYw!t!mCi`IQ%6s(X6$NPxR~vZLj^HAAe8I^Z5B1D7
ze6%G#ixUcC&fqvdobGX5Z1!f1i#WzVk9Sug+_=+f?KP>6#Fg$Bq%s`h0F$-x?whF<
zO0>lwr}@V!E{1})CDV@8;hTLdk8@j)QZSXW&o7al{UaC9-@cigjKNdVI3D)LSdMxH
zCwsM2_O!;kz#J#P@;rrL2=p9SpiXmj@%ha!zPPA%Zr4qw?H&m90khwbeBG4^3$3lK
zOn`uUmpN+(_Xuy_l+&=xjB-SfX{AR^eSlHP<}hF<3x;SA2+@XNxa8<Eu{Ku+k~jf+
zX;tks@Qj^C#a$*#wup60NzHn8P~R?(=%f6@ej_e}?yd0E0PfN7oA5%M3)U-(q2k(2
ztp<XKSy94Z)zhff832~T!s^N7kk?#@w1UI0z*ZEnxoQB|>%7y$WB_-tW;VxipEQYF
z*!-68GHs^E;1+Pg?ZKy#9;fs=acZ%X7oQ*cqjO9EC+9P1A4Pe3A9ZA09Jh%GR?a&-
zR16)s4b&H@b_10h2t};GB#D{DQg=}~W#U!xuEY*BO$t=KP3LLQ0vHf^Ce!YIc>xQe
z6W|?tEc|Em=F&A3x0&atal|MmW31w*1C72(61;qbT4W!^zDztvUVJsJD-7ey83Btc
zS0qI}nOzNRw?$OXq9X+ZfS!TtjL=2i7XHOWkfrl!c>B#hoCF`kgeAwC>Rwy~7HL%$
zxW6G>Lp`Y0bO@#>-%B%SLLAPk>REIlNY-#CX|y`_iiTGvB58urYqX-;bhm)nOd`Ga
z9z5N=`|L5UFWlOEvhj2SKV@uSC7^<KcJQ6MudiRbb{#FrlzDdj<>c`C8);oM@--BO
z`2{6KPaGlZXkDQj4R?0#KlpBQXGhf7uSKJ*oE#i&N*l9xGYRo>=D)Kyf!mGdy3lv0
zN4S>-ErDoS7c}NG(6kq-Q-pW$@>nJrf&45il32)w9}_4ZwCP5JPePm??oSTkxuhnc
zwtScpm#X)z9#PMn(#M&4y~hl9-@EZuHxyN0Vc2`7ZK3u;KI1TjsiBG_I=l2FkE!kO
z$AOW~<P-wcwS+>mlPsHddI+W%h<X*|0nPk)^m2<t(is>u$!PB`?wG`PD+p^f+?~A{
zawQPn$Kc@I(P%b%3$@Qew`JLlwFYa);Tw3A`w)2ZJc>D$E?z&K9Scj34}Tiqp2M|W
z&VA?}k^W!e)t89(;8`E}-@)^h+~knE%mgv~Yq*j(okU$mG$iIQW0%F|jZHqW4RSM<
z4WEcaN!-CBQB)%K*U=?FlmW!P=v%L1@yT4C<59g7mlT2wBa7*AWQvL2!_m(v-UoV_
z8<<lxQ$_Bd^oO-Ix7K-PBNG78S}KBKI5)XjKExBWkf;ca+8BxJ)$;<yA983FV;h2_
z;7ow>$W0v*vA~TEnS^A#WOtqhu@g(-g^9LcDWlaHt1KEIHRU*w!SYIk5lIZ|Uwkca
z;|q+5NFkKB=ma?8m>*1LP%^_P7CQZdr98pH0V*|RMg-&W?}vk1cLvw@#y?#rY8S)S
z2g{~?i3?ac3HXct^P~Plan^mFvo}7(Lqel#+7TfaUO`C2k0A*3y6PR|GI&oy=svo%
z>D^W5eRPM0XpkLmh3c?9KI^)5=ihJA^sBDK4^aGM7gIE_R6S}`JCGYMA;b`;?-$DD
zFEcq8EfYQ9!p`{?&nhLes3liD&ox@)kIPq^9^y1$xOP|tJ9Fh_E6A+(xt%t%)WTev
zFxSg^^se})hu|1U$3p`;>2~b$HP{t@b$oyH!4}JC=5+a~qu_(~gdO-F3s*|KAb&~`
zm{m2HzS5Myx-v+LF1N@}$;{^04he$i^C5LjKgqXW13dSTV~WR@ybi=`@m-qHR8y90
zgiFD|1lCgSaXt}lsX3is>%W=ahQk;H{b<CqCD+$EAnK?`&EPbnvEeXc9?%XCT9~nt
zqO4<_C$b3^Q_a0x7&fhzm~VYuZ(6N_y$(Zc--@L#bp|~wSEe&tG#3Yxf2L*0I7+-B
zWkC7PG}r3GfktXC%`6tXPJL-N1^8rGp(VB|V%`nu=D9E0+ioj^)pz6B^~339YYvfO
zx@Z^}x&GOIzes2R8}C0nP*k>4sbxxNll6J{6)}04B>PBnUHBoEdy#{_kK4*uF9UBl
zoLF>cEHL%9jJ`v<9G$PA29EY&Ojd_iNi=&;j>{8zFWe5W+t+PI$z?Wf?Qch4U-t|2
zlXQ7X_A%cCCuwUxxal=@L?+5-CGnC(>}8Pm?~M=Qws!<uSTQEv5u^D$Cj9_7I5Vg^
z=}gv=gwYpFtX2qC!MLJ+eK`iJF}ma<D@ne1wEWsp=FTh%aCX(cb1O!hS*O*6(}nvY
zmwQ}rFRiWoRq*^f6pFvZxw13P+`#4dui674j07el*wrDTN)+w+I#_2My(+lt`kN<d
z;}Gc_eFmYjcN4L7-i*K{zchR@vuUoapp-e3g1g(A1@ohi>RKv!%)LpO1&jQ)l)eS$
zQXGSg<JYIu&*=kt>VN7ZJ>zY*>ip#J*~%LP1vFV-CnKbg{P{#{<#_s*nzrxr?sWh3
z;P8*-fs$P)2u$NV^vAF6N^#p^ncC3N)TYzRnPzSgoAgdw6W6gs%1h+*D5QO{ncC4?
zVhg8w0`qoWe&Vw#q8PV72+qVz3ow&bk+p0kUzbe_M5ckVFtN9KVxM&-`GdZybg$ZT
za&;g>EA23T3Y0vmd2T$5{)PP*0^Q|aMy8D`XXQks#`V1sqb4qP#t24fe+6|zHJ_Cs
zvMvbb*=OyOomE$wL+Jz#u9;o*6Z|n?&^Pd9p(D9LM&cf%g#6Xq=3|D4Jj+7-oCAm!
z9T<(!AUZYO|7k4ESy}L|^+d{KqStLWts8qGg*dq;P^>EhK@~}BS_KT5&*PuHgG4&K
z2GWT8K<QvXNrYJnQ{t^J`c~j!Ttb;}*v?t{OTyM@(-2DIU9_FWJajE`+>et4Ng|=w
z?(`QHQkI1?_2pxYDCqmGs#tXf2Ci|17b@+Wwn^K25S9G?iJaH6nriwa&9YOhFwevg
z(^vt`^#nm1q+H0sw)m~lo2M`b-W8gLz|$Q|QBrEOw<ndwqM&PV9%}erM0$qUvBA%-
zkTgNZ2S22<Dn42dE5mr(8KXHZX$^NfP&@h@s#V8u5_m#FXpKm03p3+ai)=Dzd1_6p
zSbYh_gK0Cdki!a@MtP`qcQ8FhN@a;qLe(SMu1Xv`A!1@FBGJON9s)4z)9G86ZF5(Z
z-dTGASQ71dgb<x*|EU|oG#{RaC@ohUqy62}eZ<`#mCk}VcWg52%a^B<{XL-vki`O!
zX;J(}C!M1g70*(ZRF`9*%4H^QjeGIqW88z`u%R4@k<3}Vb?FfB*>avC;w=FtGi082
zQ6yXI@5pNo)LTZw+F8-<I_<^E(6^#w?#mVrjYqg)A(d;S828Nxup=^?zz;S}l-aw)
zVaW9^d9Q_F(d!gvs%ayDd-?z@LkEasBp$I;1vkHEX)JLF#NjEfgBST1CUhq;xm3!7
zXg45l+Q5=EB!MV1Qsu^y4yxp&5ka>zJl{@_Cx1bf$7uf>tjCSzta;5e^R(MAl*iNA
z3}=w?kT6~B3*eDuas1=yQO2gPo-8r&Oj0qCD1YNiBsu5<Hk?bJz8oOsLEp(}8tmrr
zVY!`lU($FgJe>@<>3?zgT8AWei*5?-tc=g=WJ-pmX!pcAAe5HQFTg-MklnGK!M>|*
zLl#ktx7n%X+o~lH=PqM^csam<@94nPr$Qyx3&s?^*o7=>r`(S*;}7-^u8NWj@305s
zM31F43Mt%9FT<*n6@=pg#@B+-poA|<K*vycBEV;wOq9F3T#aTW;w&UBk%tzZSGyhr
z_Q&AY91!3XB<-Q7AeQ}y9`tC+NNUK2`H8?bbFAb|hZo4bVkV+y#Af-^j#uoiDqo?8
z%`&1uiJU|?*|zHL>Hh3!gzR}lRBn>!7Y0c&AM;q5B*j+-H&bVL3`YaR8bQ#E93rr*
zBP-+hma}u_mF1z=uSxBxp$o@*lwHg4B+SXMkEFJQjD_(bR_u6sGIFS2?(o#1-IE=E
ziB6!y$b^&wktF(EBIbEuEXMtl0D)J6x7S^9V9EtL(amGgo)Ms|nll2PZ_jEakd<Do
zYJeHG%N)tEj=5B?YLs4>UNR2CQ0N*}zV_66UcPVK0;p2xQi5))x<F2{+yy9bz^crF
zQ^k<fTE_J*sPQx@kDx@Yc1I)fy%xL5tv-62mo)*dlrG}m+`<#M<jIcJU!J`?;s&oI
z&K*&PD(&$E%{u3`FD1u(yKEXNM~rw=R}wSzlMmFrQ*h>z0dLcJ^gYh*B85Ce%#Ky0
zZ04YFF?zMHD~dQ;#HwqQFGQ9dn$oaCz}i2CcGH>K>7jH~W?}#(eW!~AYN*rqd+sj~
zhXHl?aNCop1%|i{`o&E|aLZreu%JiKJx|e`Sce%%Xr&>ELAUWOrV#q=MKC<Fxk1>_
zJkd|UZ}@*#158sDLl}v>f}Av25AwAhmO|6U&S|Kv991N1NMCshB7U<s;`p*oM^oIz
zOX)9*f<8>?P8xkx0N7M%b%=4MN=KpwKoeT2=#}l+l)0FONg`*WO0q|2P?`W+`1<Wi
z9LP|^fW2tdg^iVjkChXKf;mPN2UvP5hsK%RNrZQTV5+JUU}a+tJ1Jytgjztet4%!3
zkFU|#aUt{%7{Yda3J9!_>FnGg5ghULTRvBlJ!SCe55$)G39~IDF_&Xy(3>n&+J25e
zVf$AOUw&tdoEayVf_aBxC%$>$oW;pcJ%W;Iw1p!yYP%(*1v(_@)7RCt;pA$&E}VPI
z3~flbi8Ub<EZC5j83I;kk(pC$8pQrdL5{I59qN)neAsB7dGbB}#kp8B?k83)PYzHR
zaG~f=8S0M6WuA!}86g^q;)4iCDCvR{UWvfSIvVgYaZvOlBv!<REg=~*{jG<6Mcfq)
z?Ld!)C%~IWdcnuLyPU2PQOi0sw54hyyy&5VdW+f9?qV#NCq>Atr;NvXdQ4}lRyq!?
zRVN_ds(U**&6OP%71kt*8Z$_U6NA!1UbtZU24_e6NMBI<*cm#rTBs|@;cB2&%7x{#
z1%OHhUzvx(Y;b6bWOPHfhz4f{!R3kkGS~Ln<WqX!2C6VEp`G+zm9q+5>}x4nRS(e{
zwTwZn;~kdpINQ*_j+;{%9%^xD8qmQNfw?x>u=VxU>Zw&@ci(dwmpJ46lQ?W1!M#1x
zLuH+aAm^fH%iRK6x3Nu}DmW0WL$jJwsCXoZk9y=M65Whk<QRUX-$dF(#{MDgLq!R&
zfPQ9Bm{=x{N{TgYVqmLWx3m{SEVij2bQP8WZh2TG9|ekZV=)cAAk7Iegf`~8qgQ4l
z(b+<v5_u%DA${^XoOkQ*X8)n`-J37FAwmmvuX~z&pbhGx{yeQhI}-2^+~_+fQ*sNP
zI&hV4>!^#=yK3;+&ospC-XVEVn4;=iigx%!23V=a*qgk<u1D+URYTSJ%B;3CVSgAz
zb!GMtJ7nw1h0i(fDw#ASBH}<6fIz2l0<L95wMoS~!m4MjXV<OfpX75Vs*>z=2q}eZ
zSSrqe@`RQ+i>zXorcFT-NK&o?^DbVpgycJY=f}^(os?`(RspUu&P)ii0)7>XMFEUz
zAPws6Q0>E6W7VqzJzqam#VPyzQnCcKf6|&5e+geWRB9gAhuSIqE?A?x+t^JAX7D1i
zCS@R8RYD)wargBUoZZi(O#`e#EHaktYff`Xh7O&Txwwri!)H!3<%k>)VH;075f_HK
zZYhF277nUOX^_TUF(pfS#7f60g?Z|x>+qHJDC5?QUW)*^;ccqSOt!2nGx2~QIVSsj
ziE29eBLjeb<kj>Avm3sa9^%^v3WYhHGHe-Gqn>J4bhN7wd;y8NQ;DjKA@~(?+@nVQ
z0w=JZL$8y1Vuldgc_~XoNGc-yWCgdFEka)=e7_)h<ISie?peO9pBqDqvmNYXu?i!v
zIfFww;1SMXYBz5m=e6z7&(O42X$r{@`$8{YAZEg`zdPlQt|o2r_j&lOd244tFp~`I
zSuC|C?2Mh%O}Q}iye=xMdKhuyT^?o)6Xj=oG$LC7NrwZkXYUSPO%LC#so=gQ9kfX9
zx7@}YxGgMncJhl%KqWD$V-dE<a8$1Db})TO#3l=MD@ROZ+x60T|JC}sjv72lLeM1X
zCGsF#nJt}w^-*&)Q&Q!EY|D^nNmd4hgI>{$TJUjvH$9?<pLNNFA@pI29^MpY$k9YR
zLkB2?>Q>*F)|hsHB)=K^9;${ei(9K&O+o??sf?j4?yS%g!|ULd>CYeK+#7>fora?j
z{hP4ES%#apT&K9|U~h^@*WncRcAg#`;YcOf$oXozzdwD;l_qOteefJo_<YEeyp$u)
zyD6Tt|I!OzYcM@s5hH~iHRV$&g__<A;rkw{pG>!~3*hi|Bl%9q+Fw^Ky9NNkDJC>_
zxoAPg?KJCA`Dvm8aPTjd1Lr(ui+LaM<jTi-5~TjKbau1VdW6+N{rN%Dp>5BYuvxWJ
zOdl~C@Ex(Q^Sg~dJ^udL&b`h18{dET459#Mq4@I6<EQsFpMJNwwS|ue&K4i<K7ROQ
z<1VYg#p9p*02;2wyE-MR5Mr>^>(}Gsdt;oujRVo3CM2I8;AH*O)pt)Z>hAxrZ#$Y0
zKm3PYAur@!_VRf5w`8>eJx&&MmA^7yNTw6)UHYpjijeVlqj%F&29D)ls=ah`w!Pv+
ziW{-ipxL13q2(SF(F}r3Jl!W-ZRT(L)|D<S<Ai=dZt>?rF*fXn;yy=!Vz1Vh$vsRq
z*V3W~N_{1g<UmW(vk?uAfbi}fkU;XHF#zf>rKC|?TMhOn)76W5mK_>dstIbokVg}f
zQZsNVr1SY@!1m;#mb~DF&I&dl+a*Mq&AKcBJZ55{vK+grnHO}(Wr=u`$|h2c@r+CR
zE4U}l*ZI30flU%e($2){t=Jll{`Q+wo|QL^EW3pmodML@aBR#iR(|Rvi7mu$Nib)_
zt(f+$;CT3yakvnlIaSYyLp_g@$$6YvEyiUrh_mAa*|@Jcw1QB2JSR4}{<%=Ot3j>`
zqvGvShyjhFk_P7Lw`WuX-$<mwaZ^*MPYZFHQ}t|gf<}>CBB~+&SlDu>!kqScG?3-3
zS1zsZEEf!zqDx2R`P1<$Zd!)#hW+TcT=fJeHU)KzGe6*(;D`vi4=YFFT5RlMXY*`=
z1P#5hA$^dAj;gwCdF04NxNCO|NqkW6s8-|9%#Sc1kF%DcSV~)2dao&g5H8GxQ|L0C
z$ncB_9$|_+XEJNC#cIk1#}AI2@867mv&{M8%<Uf@8JK@rXO1nDCn#V@^)C{VQ5p8z
zl8K}B1<N6Xm(PiB8blUai-o#2iADYvi$nv#k5`J1l@exw>ZpRcc+!b1i`ZU3nZe&$
zE3p(0p^D|AvgtxGU4Re8owSLdooN>)3D67f#;iuFlFJzJ`lNV?2}2HQl*6_vvn-$>
zxv;!%QaAH3ih&J1vYu|6JxyY*EHnOra%`tpH#MX=c7RAn$9(NhXu~xEnX85aLY%=P
zLMzsVmhZ*BwBDbkQ%z*Ly=ym<r-hAfF6mh1E;a?F;nJO@dkrMIo$EHdU<Bv*Xt_I;
zP#0d{S-Ma()<r%xvk2`<lc<l<;}KF(#7|_l`rVnjQNfFL(RjQIelCOERaNHh5o;&K
zM>++l>Am+Ny*O(ZQpFzp)AuFD0%e!+89QAmDuo^u@6XY(qJDq>ied}8ImZpn-KuPj
zEx_rxo~0W_V_iv!%`8A0s=nj&ijA5a?&$IpjR=c)){NW1?>6s0t8Qw~kKa6g_AP?>
zX>&%laJ}2)haM?1`!YB|qt^zTKOe#A#g(*E<Vs`j;Og@2S9@8lA7mr@c@~mCF+z2z
zoq#cm>O8)#9P9>Mmi6uG0})ZR_fH)@9O&_rw33`=c>-f5@9$*#<M<F^xAX%3@NB%7
zVApmpy2<8NkI)LJtNU`KfqUt62{~ej>(iLNSQ%21tR~wp<j5)lkR%L)w1Q)EpZ>7<
zbbH7bLtC)$#K$+2SBS()9bjoXP`kk~<_ElQx%E6H$Z*j~P2=D%xG&FyBpzK;$U)ua
zru3IfS_7ESQuGX14`h9u7SJdmn3d@~PDxJWZC+<%5)bdx+Xnan<#Y*crj&><WUY+E
zE!#u>)DX+OY=97i)}h*lo%@1`p?pS<WQ#x>(AruJs_L5^emMt*ZdyQ^i=1sOjba)L
zz?#+(Q2TCr_*$v9Byz;f`-u#1Cd*^Cq6gGBJHG|b@1)%XT=Y0X4Khgvr3#@7P+Be*
zsx+X}EN~qg<LIYh8wHqNE27~EKrEIv^~q=`X)+(nl@ij(F4remfqCt+N1$64Gj{DM
zOi#Dc+MN;N21h3vB#@!t;iR%`fw@T06>A6^x$HcBzTou%6s)wrBn)Jzy~PxK=^90+
zF5PNLH~ops28XHM4O<gDw?mR6zls#Md^BtRl_caL#Z7YYCfKLMa*}j6p_Gv3zQKAv
z(N`vVaz?gzwYpB*kY#pvagFD}OhqB&HEKB~HXRFtgfuzO41KnBWFTq-60w;xtc38T
zbi*Jt200ILlr}|8((4{!Y-L}d-q;|dL^3#%PDO{WS4<#jbixU~oYffP1x*pKjkk|@
zTfib5mG6G}lKUDsq+!b`G?21|s3PTJrat-TCshT$u){1_zTkSM3^J)y56e~MjGZQ3
z^&I%PN%RsdQ)>a4yGbMspdnfimPi0aKhVzogEsAIg+|KpVh~GDHpw^RAUE>p9*hPH
z;zZvJZCBO!==gOnzsAB}x|T%+Ie!d^KKL0}Pf{`E)dbBPkKSRc^Vezx-7Ihmk&Hh*
zmqD;ST`gGRsu|MEFz%LwU<E{_B#)9b6x?N5rjwQp?W5k?QQHwmGaNcXLxg+D7B(c;
zw2{GW1W`wZVkxCGogxwS@Eyo%qi~g{UaAwG-lZknlq7R;aM3gC?J_HARrg_p5LdrC
zd^F8%rY5frUL)GYL$gF3s%TdhAUUnc=GMkzM&l04m_!IHzLZY)ip!G^Nk@xxdIl;!
zqccM(!%Z%1SnUKX$Vuv^!0#3Qrq(bWGoL_;hgom(|5BVKq*=n=8P6<d>yis$#|YrW
zVfgN*qUi&Pc#7hL+)yV9Q8@ZywZx&;&k+7k`px5ec2RvQz0C4-oNT-GHK8M3YeJ`u
z4ZXTTH|xYfdU|01UyVmBptJqc19b(AQY`JxV)BvGOQ;+tp1=k5(xkw4bg^uc%<cfr
zHH8AzCAvWex4?-Oli3g4C-l@#g4UyL&Nn)CnlyGx=n2!UVaY5}YWzeM_mB)Pl?@ze
zCmS&RBTUFWziSpYO6PsD8IoNNeTzm#b%28KtXH)R(zp*ro}ZLOX(zm3|JvRC3AV&1
z^Tr7-ykXnqrU;1A)g2|Of6XISL+ewkV!p*e+r^4=x{WHWA*e}{o<lL!VFlRaCpaVR
z07rftD<MsiH4R64Lq|<#<+uP*-@ryLgLE!qfVdSzekLZ94(i-dKJYfH<ok(lC?#CO
zUUMJO`j^KQ?y9}>d9di=y2Id$o(`i4IQP7r82uF(5_~Ys8zj)-Aa%^5=(vG47C^X>
z&tK3_Dzw(+N?mZA#Mk?XL?i3>`rt<Aioos~+t=iBi&8g&mwL}TLYB*e;<S}|&5_yJ
zDe`ewn9=97rK>7%@j_eU4WG&VNX{rjdbk7`HJ!nbfcYASIO9&*)!|?Nyt=mi=dWH|
z`RelOwJVqZ&J1JkmlTolamkT4+c#fa6L@QP^ceYPL<5&y5+X01lm3WG5l9*UU>OEM
zXzQTiiHk!eZWS!4sDQkbyaH$<sOgJNRR@&(Ksirb%9DFY@?Rn*+80(nbF!~0owSfn
z=iC20mWp}Df?z7_v8pg08mxb{+&AK<xz^Rea^X`Oj9iiYnV--&9&m8a`g&}x6aEB-
zHF{7>i|s;gNp=V#m0z~gl?s(_E%+uD0?*=X^w<Zftb9Yl)p&h4`%ojVsJoc*-U^&d
zySnnDIX0y1k4d=l>c~%S3G_Tc6PRoUTt0p&Bhayf^ZQwLlX_$6t|c!p*Z8qXj1|P;
z-%s^QA86P_*OvS6RUcKolt*VqA5)c#f*GZpv7>&r{hadQ+A5;upQsasI0Gz^pFUXV
zUa;(CIZ1>1&RyODn&Q~0RFb9Cpn`yEP?3Y~)g&l^o6lSRn$)ESr9N9zE=!*VMY@6a
za(d{7SBl>12BXM3oIO-Ro+r0kA;4Y$yV(VnJz5}<eAD7K0VSRX1*t0PIPk{~0+`LI
zEXLWwyJ_~W_BK;1hVztt7Dsv$u!W(3Cb43Z8bLJQ_}sC8<tX&=A_C+|ndIkLby5a8
zRbm+Lhk3Kfvv*<9oKlHOds3mGjI`v$(PO0w!zYJ?nk*>whWxc8BMa3ThPtdImNCz}
z`&_U}Ah_Hyve07Rf`<oyl<~cEvy=ZS?qwbJOpc+9(C8%=q_|O9eN1&PX%NP|hhwJh
zVA}EqdBk5{XATY>7x4WOE90-iZicUktkJH9?{I&{#0dvpp@e!gf^5eri_Yb8j%z8H
zS+h|(O$X+eXq^er6=Zq=+1e_n;ObmlO`U6^h(Uu;$Ss^wi_<_7u*wvc+EpttZy~Jx
zH)&^0nM6jawz?G11TR<#v+1;L{7Y<+hb7H1Q&ASIyOO&LUI{G$sSYln(QAzH;fBZi
zbD`8tVt9aXAvUMRVmkH3!AU$YR3G3T;N-qM1Y(L(9mMLc4pJ8+dqY-w{9d|T51wR+
z6<Nc)>rYZsexD}Si3Cui8kcl3+yb9~01ZKoC}IG>JF-Mzf!4bBT#lD3>$3u^MLa9M
zM_9rbA5KqSze%cZHdr-~>$19mdzr(IV>n~zNvBQKi*4?m%tSw^z&3K)f;c;S)m($P
zd4#bNss!E-Z^0?6yxk9RxbU_Qd{{-p0u>n3hEqq=k870hLeF^J(|8zagF2OQPzS0Z
z3!zzCE12)kII}iaH$J?PGa^_lzUq{nbRd!B_9t2IoJTY3hVZ@7PnBRxZ-4{J?t{+a
zPjI-Xw0n{u%1ZwfjzOz?#fmTep1_42y(lR)RmbMAwt8=8WNfX}U~4lX^O#!eF2&MC
zQ7Y1`w9dfLiqit0t;NpPnJzPzA9`82K!0gQj<uF#<5=bVOsu_HKqq52E!*^Ldl<OD
zF)PtF`xb?rd0T*s*n|dvGwd>Ez5V>{+KVeIIa;mSi2-4ZpaPP6vPZ|SJrK^O|5ZHo
zat{u?0Qn1!x$6WF2uoZz+TrQuAHIL^bn_m}VLG?r@sno{9zWXn4)1VcgO+=+wX^Z`
z>BgV%?lZvkX_@4}h0NOnT+A6BVGFqkRJ@t(9#14?%+V7ZYkMC3a%qG<1%RV{&WiE0
zhv70vc&SI6q%|j&alM~s(F7oFx<S%|4!pY<=oA-BF;uO>i{_=yoxvAeRnv#tyk_8R
z3d!j|5NPo}HF0}o4Of~}?P-!hb&oz8vF)Wmz)Sv{<_aZIB8@<8s|Z8s#+1$_2Po<4
zt~(AF<Pn;WPEWup%Vj-)(Iw5|s1(mkhee&$>RrlW5<1LlU_;ZWYUr6&GxYSX4|=!b
z4Kf+?Ib&6b>msuFGcq=C-i7rfOT!6W@JlJKtHZmaLpkMVe1b)Nt6|qikAG%bEAPb6
z$)q#V`nt2Pu$1o`puc0oRIL1WCa*I3K*~7*T@;@-Pf^8n$|n+gM@r7Gd$LiXPrTD5
zUQn(w;ptrOl`qCN+?@tsbNkuP4`9@L%FPX4U9NUgC`}0ID{*)~i=cX0<{htaJhu($
zTP%1zVB~`Dz|~E~C=(rQm0p$Eek#sc4Bm<$f4oJ@d!}=z$IqW3aV9@5TC2<@pa0jv
zetue1Pj{C;|K{}7t1)g)*T+Sb(hBIDW6GA@uu?L1VX`jR0ZLWa1p}9?qjP%PoMJK7
zy}Q%X!)O`l;3JTM6<U{V$f7O1C4eZTe$fc2VCk)SV&RR4)6*hut)8QO6@ia|hO7jr
z>I$=AdrxARWRCGr9qM&D?FjTlqueUYnRO|?%gG#&X5Wv)ESSm3_<-t2lpY`AF@7>S
zN;5J223S*(SQyfr7QPZL#b*e6AcBt$LPe=KNRX+djp4I3HJ?5$aJkok{;9G2HMLss
z6u#qpF2g4<4Bi_%<!`+lXGOviR@GmP<aEtuPmPSUThDA8Q_na+d6d&@S)unQ8AOXG
z#PzlGUA42LHy1(Gp&M?F(s(+GeSr;cOXhRx9o0K0O1+W8X&b!B(#+r632vhj-%cFC
z(LPew_TC|VcrrVgi7V(uLqj3yvUeCsqdZNnP;oB<1EnU}uo=CA<>?yf!lTIeME6b;
zb;R)r?6y1-183h%nuLtX&}g4~D0@*6t*N<`y)*#>dS-(vigsU!H7*2H26MbjRcAe+
zpF?LObQc{PW=xQCQ^}k5?EvBkWv7NFzgE=?gMh$e!y5aA0)MGTT%hqh403YX(~$04
ztwMjyHmC^%t7@T_3^$&Ws*X^HF5Pa^SUIgfFlMF-5_As;<)NE1Y(-?P+k4klb07!@
zHoQ{OpmG|%9z6N>$<BjE&vqVe{1>QgivaaFEQR<}%Tw-az7~+4kqR7aF1k`g<Fmyg
zUsnGW!d)XE!QCaKs}Xf`c2f}jq=5fYTl3P8*<hNb&)<^$O^#UO8#HhYeU7itD4k)_
zVr9}fI&_ZocWE`XRd-ONP3OU``bq7owbZ%^F50ErRai=Ee%sm(#!Gi~=B#wE*NC@e
z=~&OJH`kh^6B}V&6^NMCJz1R^H0-Jt3W!Ck8a{5>-W#<oiNJn=6b=YNAgZT4iSMIN
z>^o%gk6>Z{xg>p*TwW>GKZ6@$gc2vjLG=>M3-~G?*k$pcfjx~lU#iLNp;Tbqb5YhX
zaU3H(G74d^DeschXX>iU^H5liB}Um^qDY4Sea!9J^0`B<t|~F1!3`%NOK)l$Qlkoa
zU1FJ|Un3^Uv}~G=FXo*6+64G)xlri_>K=^3Xi7hWW}Gu#_v5B;2SMdnk}&ncRgV{a
zHhsikXA(%$3r*4{AZN8{MkcXn`dieq+cSL?_*Hzx@YdS1rBLr#ZOe$WM+j2dLT^7!
zkNXa^ehqWJwWS5T?r0ggy-WRA%wsW(%htQY>CtQg%Oa4W1!U$+MpzkpVp527s(RA0
z%(>8p(*T3lWrVz{L<ff3HjXcFBwJWH&?Z?v*^36e=HLU*bXk;-cjr8Vl!ND1?*h+-
z<Npn>dfX(X20j)oW_fwT?IHX@tdv|p#@5LcXUQY=C7LhceWbZ80P{k@@C<(10Ux|d
z*KWC(b57v64#5cXjHcZ$T@X6e-M%!V(!4Z#L9^@4mKIpMfOjGe?%L4ynjh?-<M|R8
zXrDm!4E5TYjB<CYR#~;{q*`+v+1p6L_W47Q?rXz4U?5`nN%$P+*OB)KXd8eG?zLgb
zeFcO#EzWr4sz`bcfG(4b?Uv|_X0dwURR_2QeR%EKHLc-qE&beD5@2p>kPV1cs#P4!
zh>@`ZIYC@BRh7^-nlZpEGoJtGIllej=Cf}f-`m-_HXLBj-J9tiW};Yj&|~6p>Jl=(
z5L(B%aWfv=bL-lt2yW~R-s0v<o<Mj!KA8T5Qj^2Mr`K*U+OoOxJ<jaGDfdsFZr<H^
zws{ZS!*vmj`(pt^bNR}^H<hrRK5E~~SoLc8xW0}vfxdaX_2Ai`b{;)``f%erJ823~
za!YG>^!H0xB0U%tWTQ^<gRF+fqqjSDWGc4`FtS12narGiTf6gW%-vcd24chHKedlA
znjt2SXy)wR^Z=bms`8(V@WsLb`=nZ1(58sE)rfS2cvj*0d(&Nsxv1q=%FG==6j^3y
zh>v5?p#J-<7cIM5(U%dk{EA7P2#8+EU$s`1)}3Hc<*()`5Wd=k<&c`i;TAsejc01J
z(%#vkUMC2Ic6us8WUx+9>_K%<HL?q{XZtF`$fVTVUOEKR;&52DMj}`en^uI<bzuZ1
z9IO@&7EclBHUJuzb;z>vS3?IXItZ|aUWNsG%Poblk1O5z<<N~a5+XGcRr09gTNjgc
z)07py8k0T&`d>vn%UHhmH6-{b$sm=Qw)4Zrci(Tq(mdLPVowP^`Zi}_<TM}sl2QsJ
z*(X9uc@$p_euLJ1eo?0Zxo;L$BsEPXRWQ*E1zc^zSbM+P+!>cHJvGQ1qz%^eAW6YU
zPXRcJ%t;n1_m(NE=2PAHjQYL>Oh5BTCUczpPze-UcAhjWV^)J*%>nT<lH;0(LHEAu
zoW!53=Lwa(yyhinBQ*%+$77~1;$iv|Af^3$;;Z`y>1Xo08Mc+|I=$m^_}A6vf4jW)
zHFN9s;jta$Ni?_T;WR4PQHPX5zB`tHpTERny_m@6e)gqBlDzWOXAV~b(H*58!rQ<K
zH6_7EvXVea&qmWEpe$z?fF-iXZvOhc(#TZGKR^;$>ODer{u|?@^IsF8Gh~=?NUU<2
zIB<2)Q?EOd*=5G+!2!;$kedJoC)2^35%%;Tm1rt6TP1-~Kj|i5af0)DtVFsztv<y@
z#n#Z(Oa*!ilG?C1DQ+uGZ=r16@hbxsB8z<n_7USu-N+PGN`JkAew}tZ+X02g6sOJs
zFa3>lz~qJ^vuE>>)9fn4k|HT0Gz(i)u4b^J-C{ONb&KiC@-5orxZP4>C5pTyC{|3?
z*HYI@YOt0mYxa{O6UnN!x^7n9mQ7NwvXVD@N=2S!s`M7-<`=$wgB0jfZj#buG#v{}
zSF;1bL5$|87Ie2SSyu+6f-f#<Kr$yS6s5lQa+Ue4donk_&{e}J@|$5tp)3Tp;m(ki
z%0!+X*o#8E5Vdk-5H+*R8C32Q=zG6h@}VKo@yBgBEu=j80!)yoZotouo0S9(xc;V@
z7zb5KO0X*HyQ+BpaG|P_C{oo(?~4lAV1z{4^s*}YL%HucHqcdH5v`6MY*Y<;h&t;Z
zJvZ>t_I$%Fcvpw7WsuXE*qGSqDcR5{p4k7p&HviLiQ^9*{SNA8xVru4=d%}A{x5s)
z-WJ!9q!0ejr#Qx_3)z4!zV#qkHsWIOYXl4+d%VI;7gQl?TvX9jg=EI<`40Q+^Xv!Q
zPqObDm%N-yQAqCY+21xjW7Rn)FOiWEkr9!Rk>A{3{ATrd{I|t#o_=(Uzx(s?=W*+A
z3-kh}4gCA7J$S;>cD`AiM_co({L?3F41X)(#r;pv;<5DBDi=AnHz;%5Vn^YxY^-G}
zevg5zv!<014FrtsIkF)S>)cr6><~ie@mNUFPQ9VEa&mH;Z*htzu{cEdHU&(9W3E`u
z8C4~ltZZ964flh;5+^Va^3~*ggs?#ql^i%GKSLr`7iU;!jwH2_t6*{#ruScPA_@w{
zK}h^$t44w-lY%uTkxRR!<!lw@jzEN^98(nccI!Nd(K8q;Som^wJsOLH6@iPE@XOz6
zI^#E!3twK+6Fm=($Dck}qf^<4M3YG=YVcOum9->l<yiwD<*(;Kn93SogFRX+9@kF>
z%IJTk_38Z?yR-3Y9BE8{<+-j_fP&uD*|^`vaij4fo=!&Ht3Kks^(^vY3qWC9t>f0b
z{-clc^DUeo<rlX>wZ+eVLgH$?wA9f4@;-95KGsWr!&;jSf9hc0W2mKJS&Cr==V2X;
zO!O@<LekCxq6)N?)(698X~!N(+A--FqGAq=I~u;TGyCG*bC+2NDvz<8uLiuni$)39
z6xq^*22mFm7|mmMgts-}AI_gomQ$(HC8q3QxRKdd;M2$*jy1JYf{7t4uItkJ@ahJ)
z<6ca#yLG>K{P5xD%a0#E`m)$+pZ1{6?IWAQ;H=jj&CTt1M{*(?8zB%(7l@@ly(!*|
z+Bj{{!KwC9S1zdFHBE2eo5P5hMHkzHV^kQPPTG(X&<xuMCcC<s6FG@y#>4Xo&n(Ek
zN_#vWp7q+i?xJHZcDby#fQ66Ss~0K#<`1pLJct;bbh_>S95!J1%|D5(gaaKLG$bau
zG!CcB2WS234q>1u<>!9yvNzElSY6OCo<pPXl`jdLOEeC9=g1Bu=d+gzS!Ko-OGO6(
zwxizZ^#mm`HT~Wh@&)2Ir1}GhA@F}6i{MocaNC;T&`BAYaUyhLC>rz4y9?}|WFrBJ
z&aX#<9$36X$T~xyOnUrJq$Fk?7LWIMW8Lk@EfeF_xw#{JYM)|@O2(qIjKe_yRnV;q
zOq^BwG`?VtWhlLhFx}t6t1ujlj!cP(?;)s`mx;sQRvG3B`hT%m9PVu${jhehS!^E`
z`v-g9Z*SnB;QZPlp68c}AGWdo_UfoWiG#J>qaTaCtzvEW$Kp@hyBpX|`wwiR;XdfS
zgSqXO`#am4c(=W~zVm8>3dwWS+ub`VcD7$`L%%4F_6l}rO>J+Y-d6E)^I-i2T3&m;
zy@L(2rMa!`qg}SSwRcdg75i%kN89VKcGeDx{Z|M3dxx9oegn<!ZtreU?b&>}xqGyN
zUh%Hj{2mX*;fuAM9d<Rh_6pD+5WZr4Z~w=G?eAV36)*O7aF6w2@q81Itv%n_)UGh7
z^_{itmrKRQ+RL@?2;D(}YhpGJ=2%RjD}H!^4av;|_P2)rTjwE44r6_9_viqROBm(B
zQK<UE_TlDIv39V1NQ7)1?7du?BRWw7t)mgt+uhVwh-yJlm=P2q!mwi*8Z0(8*LKhr
zre`;szLob%V+fNA*KR{e<L1-;cyx>x$7`1C#XqLsn!Ihg`Q$C2=1%u*7spqoZ))v*
z$c^dipy^@1t8Ik6s;F?P!`w>gSfVPO4Ke7`ps79T(Y?X>FzCpe2M3r4yp@~9O`{Dy
zZip#puF`9GhwbRo;kY-s5hT29U!k@tbaRfG)m0Gg!FQXx>o{%utSAklW(PT`xwf+j
z61!`y^@DA25)b^|u{#_89czKh!J@EA?pO~3V|yJN8B?q2{`%%pf)t&)+iLxIQd4zf
z^ZBdW0F^RJjE2Eqm}mt5U;&?bmR08ze5Rm>`E2-}&Rj#0Fh>pFliAP`L{|z!M0@>r
zM{oVtfWh>$;)>`hg?*%kUy1PHjSzma7zjl8G_n%33*o6^p<x!})>Jchv{kU7wIAHy
zb?w)5`l*MDRmz8!Do~W+3T@;i-D6tDQh$i-1bs?T*f^KCCbvzk&4vsuVuH|_AXj9p
zH2d<{z6!LjwZ(D-ag$V!YA`hwKvW1o8bKU<?(s4h`D*cf*%ci5kR1wf%P7ex0n>@2
zZabT>9*pNUB7C5s)r))C)QYp4Ru+%e=4dn=?FmgqZzOki>g&j4h<ic5OB>%i4I*Ru
zGmwyW(u~{K{PMGRBOE{M4<aI;-g~q#e)js_<Aw20#l6oK#!nx}8yFDQa6PW@*@;2>
zmtc*nVISH!(knZCEEr{&$>FX{al*LR3W1{aP2?m-ry?@;o8cM>?nx5)=+ww)_#~O9
z8}rPwpvxZv_9P~cbb2y8rHU*uAosb8p#feYWfsc93XWN>vAZa<ZJ@jkA%nl(8%wee
zbbVjKXi-ybCqQl{6>NbC;vgfQtQE1l-p~1`RHgcc>Kb_%SWq^!An}+i9j9{I8Te!{
zsQ#5DfW>vu(IOZgkl&3+w0aOed24P)TCE^qLgu|d0iEFwB+wKwNM+aW4bt1it#QNy
zI{pB$Fo~BX#Vb|Dvh&A-d7Xt1MDKJJa!ZzA_}`8d&EdI5BUChJdZAVdQ*}7O@yttm
zndCQIuhlh6Q#zh>pd$KgPG9%p8iP4i5m*fxgah+zI1+Q)gB+R$>8l_eky@6FTjyD{
ziNX!-18q#lEGf%^OoOxuedD)^B(%*neE})laH3};fz8_Njl)av1u97>S_l=AB&G;<
z7bZawIXwd;+9WnG!aO5v+sHB9lQO)Cz%VeFI>KHf4ze_cxbi5t46csFbxBiUTIN#O
z`3voKyo(h5O+^CxdY{b&8~cuUn1M7(K*hS*4BKobKs6L>1S+nWx@eX9t$eOSL`|Xk
z{y+@bDi6AM0hkhCkb)&`6gW?-D;9(BKqNCY2o4}a3nj69@n%J~duG!hUwIN{(ivV)
z4iShCKTNvoExV2R&J+SO>4V!l{QTiVQCDnbfW^_!G+_APrh9K8Vq(@yM5Sk7aOh_j
z-Ls!VK+ocw56~Cu*44$;$u;8|aCyTwMR;0ljqyUyBVeAogSU&Vy~Cp)_c!6<p5qzT
zy+gICY|0cV%%(G@^L#j73}H%7s3IvKbdL*?>~?uUerXt00mp%0Dqyawpy3=Ug;b}B
z>C`cb)dvs22B;*bA@qed!ByK}{`Q=Gc@(#{4FKN}71&gkkJgyg)vr!E6}_w?5DB?Z
zTRq?@5oCbl!!*DR?xf`i*at4!O1Pq6O7fff1vIR?donG7>o8vVE6dt4t78so%gx5q
z#Q*!G(NSR0_j_t4w|T{aw4FtzQEg@y@sJ9Zg!Nfg=31<laSB|84i~jXcn%YissNmv
zqEQrwx+3a={>`<&)OZwdH6X>cH!30&!c9GAt2N4{;w11{Rq|IBnJHa>_+oLwVttgg
zw|{@Ny?NBsXK(!P*F9YSmeda!D?rL+Odyg$jL$5JEDKYqt#NbrS{vQd>o+M_Q+si<
z)(rw2TCq41E)t3n___CA9oY$(X>%uy1rvl#nO%XV1d3T3|L3rCV3POdYZqfC&0<<@
zW$sxR+t-k2S0qu_=WZj#j>=?{`s898VJ>g6O>CA*3?2}ZNDs6~HJ#*OYwuh4uGlBl
zh~itEE5sF0RgXBdiU{ZvZ!HWJtwp7l>u`evGAKg5<!Ap0CA{2g25pn<fB?#+s|lcT
zV@>s&1{UYy1F+B_fXlFz?5ZiCstCg-ig`gBNCx)-O_1JOd}B?mHA!v*F3?(2{TiWN
zrb!X{sWLx9qdLM&w3nN;-HPIAP*rDZw{8KbX*+C(+^vJkT2XFV_XF+#^n0x*Hhq=l
zg9~b0GQ`72$yye9m3r^9?1Tjn931X^G%q9b@-0kmt`K2RyCxQY%U8;SwI9l__OT<q
zOW}rAC?mAB5VSd_U{mC0t1UZrLScY|cOS*L(yIdy-|KnR07m@alvu5TFOiO_pM@4M
z1cwKAutmX5;Oa&b`&EE2AaOXxMxMFtmNPdUJ}MT4nIX{Zrcj>~IlXNUEIdZ6$=L-@
zGyuhXan{CGxUOVj0Yd`TRGC@D!4P%0%bZ9|Z6Z4LXHaNBfTw^0m$OZ68neGTQ$eAL
z3<NQiPbTd*A<NDbXd*b!;f;l+xPggARrbL`)9(AYrHy}vp%v!O^&YGhns~K5!B_&}
zEJrALBs`Po#iK7hST=q$F8R9nEac9PAN3c{;(hsfwJN?wZozTA$2ekF?%K9fi^pG}
zZ~K}NAleS+{OkjTN0Q3h-U_YYJpl2)4-<nA>1@@%+J>;z<-C#G1hHbd9@LWWT?h2F
z-a9Fm&6&NaRFk69R6f&PnhK%OE!U%GvXh${%GEgX*WZyhLPX>5kS4c!Z2UbPAQirJ
zl2Us|om^)1V3)f)=!}hF!KcB}fz-oZjbmAwD3eQO)o6Jr25}Lou8W?pA$lFaJ+xSs
z%I=FWh}6!A){Sk(BjR^FydELscR~A?oz9m2=H-oF7odkw>V&IMD9Yt%hFH}WH|G!2
zoh!xTpxXuF+VrFXQC#?XxT@=%*J&e;Kr@2+Ssz;ntSDA!{RQlDAWb_fVYdTFvKHV<
zT_5(5p~)TIvJTg7M1^_4oP?I}J1ky7b*n0GK$?y?by;MCkbXfGHLmk1OBD!=0~Ui;
z)cJ4EO;7Z_1mY4v1{DGu*Ekuko2CV}k#4U`xA*4W4z{y7!D%>hnZp)aG5lD-D2C!F
z^;tVV?_ZBEd>>7xEJ2#`MD5l=K1Vj{4O|kjj`+nNPYyQ`c)5n3f&(1sNbY2wYdw01
zWGasye(}dI^*T)s=m!LZ?)`Ac_^Lz1L?SQc|JplPY2kmhU$?<%jbER?+TPii17dX=
zj7enkU~Tts8!?lu`~2L(4*?L=AYb8S-n_hK87lzcF}lpl9Bm%Fv~2@-5?&7^0y%DG
zAb9VWlis*JncQIDnAC5v6A@n3;QRe#_`hEx{8^8}PnV>#NO6^+%K7UTL&Ok!^TEsR
zFH$LbCPmZS<(D`g+k;WT@a2$AI!Ilx(<SaGK->Z@7{ew=_#nFoFN0d#`ktq^G*SpL
z(!BwKGJ2g5>rmzBOjiHZFcSlw6iks6p6|mQNwvZywgB1U_g}caYlEiL3gdU-0c}JR
zehd*WqD&9+vg2&=6Sst8m9^{1@C=CQ^CXar9ICiAoiJp1eL}<QI<Ois79b1iH>FwC
zhy60n7j&=qkT^_|w;Xfwm<}Qi?v`=8=Ua-ooUOvv8>gTVoZMlRf#HKJ4-hEY7m};N
zkkVoVpb7$%Hmkj$g%hOc3*y7u$>2+yt!;Rd8^cEMi@H5$Pe~^wuE>i0uDCBB*Q%vi
zOo3A+$F0;I=qv*+eou!0kds-9Ws;XDf&-mAC=x&0i<BiC*S?dCMO0-@x8s*B8eri0
z6a=olk;w66jq6$(8Zo{C3|JP!xhqLdCYeD(XQtw$jcQY(!DR=j{`|VV{9kL!|Hs4S
ze>=eyQ#cp7l)t(E$-)EfQhVbD1VL`Y%h1(3*k@4bJhrVoa&K4zQp){cupzhVQVkrG
zM1BgYrmn^4g+7D>QoKZ*YEiTYUCkUUUd;<J81OCE0AdsIgJVouj3CIWo~aBJ*Nn#x
z>O0ouJQ&rggSFDC9>7CQLJUE|8kPyr`sHnycfQAoYBfZ|+g=ApY}$kIJH%FDDQjPz
z_TF3%ug5EeNcsXfHLm*|9>_uddLAaaCyD(-;6j1pMw|_YMJLa3f=Lr>o{Vv(40+rX
zX((=30j6X?AEksIbyT@{NLx-2P1dveudQDukqw9MER}O8A}H0UU@M-rpq~Z1p^2y>
zdkyGGv$ZSD33bJB$a(NMjt}4wY29(0h>CTkG9eZo#zdeEHV(AQ08C2m#HX)V)pn$c
zea5*y9^uWFU+a!M$D$vsA&P!y>fLYvD-I$Bx})oKQ<Xe=#Mgv?54vx<zks6bue{ne
zC}FNbfhf$mUTP5^S`Yqw^PmU4yCLi_Dr9~oYgI3+wz^Pyh1JCjG2}3yJEo7b&ozo#
z^36y=M(o;ZSasqunbzPsY0!}m%37cVH;#zO__YG3lTPRgL5DI0zPwonA{fw=i#h~G
z^UB&#I;BEwKq$`|+a4cbZUKfNHCeAf+GrNdTenY2maUJl+5s(|to@YMk-$_%Aq=o`
zq2GnjKim|WwFMl;>YWAK%DF;$%=wDsY`WEglW=17z(izvs?>hE=yf_cm|HcS^urcT
zYM!>wentV(Oxjk}h)BLdtT7GY5CO&kAS0pV%^>iE52s=>+|J`}SQ+8_uu1IxJMdw`
zCIOl$y1b*cMw^dKPpQB0{s~xm;}<eBHC^k>2$j2Jte_31gw<cr;*@lvo2YnbS7(o$
z6&JFc+K`^C&~^$RxMwFL`A&E;IBn5gJ~k&-HSoQ`gh}@mi%`W8l?A;c!>O?mDPABp
zRcj0>YP|5`!nsj-VhSn9UL=aFP<CeEA++lZZuXUqe)(C3!SAVzCMeMbp=wXpHLXh7
z*rz-@(1J$wV2UUnRB>&Ng4LrR;<Y(?mAGns&|2XtuoC9(2>0sPYevv&X?+pO%cPJ7
z#X&2opq8Kn&yU?Wlq~&`^#`v^I9_#pG(ys~(O9w_K(yDEkt<?YBB$)e<72H1mwL<P
zm}E|6xr;;o$z@ba#a}pzzffCd>NhzY6r&lL7T|P<yR`lifcT3XlK)Er_+Kz0+u?%>
zPSwlJfsAc(=zmwftf4p(gw9GMf!DIs_b%{dx>fck?1U!uowe_aL&gG;Ys&%vJQO>q
za!zPHwxE)++z!Oz%l0p8whTgRv0u(Akuv>@Y8O_lnEig;7W(O3USDE6=JFJT14An`
zdFg4uknI2muY1U-fDv4WRAx5)p**I<f8SwJfU|V0RjkD}eF-DdYEYIan+tKtPG&d0
z<n+d;yFr_DQw26{6yU<roxtS_Q#WX=L4b+yNm?*vQ*CUOOqSy_6>3@wA>|O4YHvw`
z(D$^3F!EXTy7D&4ZGoBFOg?pdC?=u&zO91pO&(`K?M-#|N5eCC#gFaZ=hwY{hbb58
z2NOzPTy6f+J-ePz+HUm*QbjpX|EpT+;RVx}{2DdF_SF?Korq<aXoXM(pg6)k5;RBM
zM@n8`-|hfus(AD-c?+T-@Y_hVplMiwRvTa`Q+KLfu0?3TaPfhpRF|=fS-Fse<M2=t
z{_7CoajE3{KfJj-A0m0p<CQ<k8gSQq|EjSiLk5#M)rVamIJN{GTt{4f#37I^lM#g?
zhH89<w2)1y7ke*p+6CE+R_xme?}&&kCD%lRR&;vM2%}arGHZ^SA6*Ybi>tyaF|AyR
zH-G6(bbqpTOj2&0p2}bpjTdR8pmeJ0CWuqxQ4s|%YrAt|gY>-<dqfM4&lB4yf%D?T
zk!}gG1at_DY!x-3;d7v<i6^UVxWM)cbPrzu?fnHD)gktvP^j5!rONb7@E4FI4)qM#
z`&Gshb6*>@vbb{rb>zL0QW>bHPoK6n_qJMdPnk;O83dB#R+1EP1_OL$##Bqi&G6d2
zh>Qh-!*K))pf_H9`atT>6~!lo_*sFwOXj5mS9C=_&^3uHO<in)@f4Rez9coQFc@e!
zXibn5NY0?Cj)Vh{y|;+2h02L!Laco`4tI55A<+{WrZc&B)|>QiO6?hW0tjAZe#ZTN
zxM4cTD1<@Vl`)p$O~{8>cF4guiZiDNqOt6nu);Du^%&;yVpg=!IxJv`E!IS!xaKa>
zr}vRirtGoD654a5;JlMsTDm%QI#S7L;$FtceI`m%lvubror8JO=`4~`@B44;yB}Z!
z{dEiB8E;!Zy+-0R|K!l(xi^5`+?#9;-u6br!6h7z-?v9SE)>SCC<Aub*WMr#fTqTL
z2M*NoGs+L=M~jDq^9U}7wtK8}+GSG81h3_wEOJiZd6J{*HeOA{0MWC~`(aQZZbDc=
zMRX<;2@Qo`vMOU2K&c<c|BQwVB2iK~{|_aihW$?=9p7AOJ(Yc>XY})1%LU^FdILFf
zEzAl_L<(USsXe(yinDv;`%fQeJ@x5N)W0Z}hAfI0by<;0gj2LpAtT5wb3b(Z=v`9G
z(dt);@v99m!wH(vysf{PT!7HP`d=(QCHSC&XHC@*V8;t`7Vo2`m%R^mU1JtYks5Gf
zAzd58j6-8+`zjX4K}O^iBcOc0G>4~9vJd?tDT$<K02Pr(d@p+X+jg%nmZjl&%;aN3
z2pFVSBLFgva{qtch;6E9iN;7hl0otqhpd9(<XJ{3LW(hr?>+>%#Ps=URgZ35K7?2Y
z5^k-UjIr)nhLJeCS+`Sau0=(l#s`zY6UR;<N{3yu1rvio)SMU-Ts<IeT-2WMS(h!%
z9uspGo)2HQsQjI128R~%I6efMvBK*XBJf(wp{O!e-(3x1Yqx$<hD(dZT@&?h0%@8H
z20kHpwz$y)Yk)_41x_sYz%BECAb~vLj`dSxqDDku;gZE(I+%=ZYFC*AgSKk>nAw>O
zvI>o`H2f428Gam;A#6k^(1p=02$_3|FI*%+1|hk`%~2zJAwRe<Eg$P5=#Rv6CHeV~
zU50z6nvPk-7u(fq=$4nEGbV|eRSXD8xW|iyBaKuj#7RJita+)#Sg?l~DaIKCU^;VU
z0=#B%uCL+^pQEMm;^7f81Tr<|<ImBYe@jB=^c7>gG<r*(=|NK>S}l;yh1X9=s*(~O
zXu*(EORhZ~LT1j?a2y0nR;0Em4J*mYVIeQo6b4lMls^H%z6n<T<V@C3k$yd~i-MVA
zw$L%GQQV<0z$$R;wr2PN6&lPDUD41-C%wyVAnY{GP}S0)N#cM8sFE_;jjnyuVnTmZ
zMyUsiH?iM`9_X&O_uKS5oZR#=g$)+6SwqJ`icdKuU2kK13opQ?$Qy(4yumK;_<6h0
zD)$;H=B>jJ01Qy&FPx50!&M8*V4@q-1DQkRC1^MpZg@_OdDU*T`7$-9X14Yen(!oN
z2?1(LR6M;+3e75yK;znh0HCG?5<&S#l}60+@^r@C_GHllTCrS6wQR$+#Dr&%Lz8ZT
za_pJ6xsCde9hoen-j@|gr3`vO4GN*!5zKO`OVk{X{mu?nG>+EeeUsD4R6==XXeuP7
z^ALCMxu$`}9wC;X=xORnKS=?sW~zwT_?1#kV}Q(zWs?~yHJKM{Y7F2t9qTW(*Hp5p
z<?H}Qq@z-+S*fO`&~=b3Gad7&MR07QCYrs>%D`SMKNZPIF~a|0;#JtE=ldm9F;vT5
z6L<q+x-d|p{*{L`h9{Z!hO<p$L2517+EM=x^eel!x3N5xw%c1BU!QStzP+Fra<>;l
z0&MzkZZC*sa0cw0L0zxKPDN)C_pc-8Li}5hK%y5TbZPib?#Qw^s7mwp_~*s2V-Il^
z_v+~w!;7%X0|Cre0qtnAJf&z%H+i!r@6rNXTMS#u>PL{=&)7H`4OUBQ8y+l*tunwc
zn(|O;R=FlZzB1+DaMDKhrm$AsR(!4VmT4<_lBY{Y<RgQn42d@n4?i-9i*Mw;PJ_*+
zG9|WV)Bb~NIzcr&F6)Z7)eJJKte4$Z(bO^hXU4*bmM~)t(_%di<xagF(S>naKf0MD
z{|y@1ZFN+EJAN)XA-f%YoUxJ?p1CU@__4+(`r~b#c{%QRRbmfkuI#Ni``9Ly?&heb
z4Edg%_mg*6VIex;NqV{VA16qDzIn2<x%(ZGrSteKE)xmv5WOHI*}nR#9V78scB`1S
zTGaW&TN;x)-2v_)N!n0<NSqXrPh~z%wefuF!>uKyd2&rW(KvhDhQy$J7^c4D08mmx
zJ<dtRbB~ETAyL?H>jQl*xGJH~&}38o%T6%LCU>Yt6rdw`LK9>DNL{1-A9|e$L=|rI
z(A*n~Q;&_|Dg;U2QkpyVS$l40D4sNYR`p(2gEG2L`@=IVZFUBaI=Ly<wyuq7QV~<S
z%poZ9yhkHL?<T*3-<{zkwwEb0;7#FW4DRulF>*(O3Zh;y4v{$#7>>+lj$XNaB$@?5
zKNpDc%|3~lLfu|X8#RlxP<dDwos0+n>OlBIhX9W$cjscWzOn`Ul`&V+-=gw51FW*#
zF3v^wIlWr+I^RPQ<7f1S8^Vy<uelJsZ9I-ayRr2(Dk*QSuHt>aJNQ9w{xCa1lqAJa
zt|TXY-A8YC!9Ybd{U_sK`JaG;Cat>`6-QLLNn);?gGbaed)dB9cZ}?N;vx-HfbGkB
zV^WBh0e{CQ(PGPt;l163xHIEm>R>_!5AMcQIKGkLtNae|&chiP7Z~Emb!MPjT4~ss
zKB7gD;w6P^UG9}R*O)B`np?%z-3H+>&L9V-%Pf_-`rp>m^EY3civQ#^{CrkCcu-Je
z+=*|?HeH<Gd^x~|!rq@iYc2GBircn+X3tpdy92)DNgOjZ>K#ZWlW<BelC$1A!fyAZ
z%Z^6f&P?4@Ov))8+3gMQ+f{CqPU#1iKx|qkH`rGkzLSxB<Xc}gMq$O0oznU`ZsWnp
z@TO6qn_Fh)c4L^m``WK6kEract5`iml}z?tT4Pl^m!-n;n`=Z}NT`oKGY#B2kf1>&
zYSb*YbaR&L9e2Dyz_QrO&Zg>R(x0yCBDnZTTyEM9R0p%-^REXm<rm$cUyy-HEdiQV
zB=nis7xwC*QZUpd=o7at{CD>S5Sgw(Lr7BVbp1Uw!riMtSiHD889hKXjb8=6e8Wm*
zBH*gFini*Rb4Dk`OrQ%sGl<Q<VdCNMKbjdFA{a?CzBnay_E_!oG~)E`eiZ=49Yo4_
z-TpNZFPoOCA_pDeyx+U><C|RPom^1BqM~2>E1poCMjkR;ImB0`7$_4JH&EP%D6ge%
z{*}BsBpanlFMvHE0;(_T_bxc5N+mFWIE)&R?j2MeMG*)Ht^H`Ha<jOdx|MEi+nq|&
zTI5!=I-YD0E=V5Nj*D^tv);s>^%z1W+1*L4ZoG5zuf7tx7b0Ws5h1b~D^uvrF^Q=X
zv9pN?kQ+R#Um&Ip@yw&e*8lT=|DTZ51&rKsVaIP$Sk{1YxpI5|m$sLXYf{6pQ|m(w
zr0YN8nJQ{@$<DmE1DG#Ud^-}ahv+F8vx&|c^AZrT34~H0Y;F0Oi#Ij0lDhc^PAg<d
zt#3@7lDPFjPyGdM$Xfty(%G!N3gqBOKH(oP=~J4Pnl?Ts#Yl;{E5vsqEfg@BjO3hT
zjnd+3xSOkq|3G~el`TOh)ux)aKp5?OXouHlU3K0BO@1u#Y`1PD#`3@^a6?xOCeA!O
z?g5!)1a&QOx)cPPo_bu{$L4VuV6iz&OU)ek+~yn2j_3UAE!a(WqDczqy0i`L4!#iG
zSeC5Dc+NEmFOy}a#Z4R`Fv5ih7Yhr85a}NRAxcYl(!7X@I`B%)Hj3~hlF9fr@#l*w
z*g;GOqKDxtMS--uc>KJ{jZlh}0&KA|!cJpe(URY;9^k6cNPv0MEFrOgD|7tF{l_6<
zz$Q%PMw#U70wl1<GM6KjY$3$}8fSnkXDC;ZoO(p$gZzB`_V<tBx>T4DijFV`swZe{
zsSrVej!>sEC&>`iA;J*~fC!WhLa|{4fW4qtW(TsEc?_C0fB`;ACA&i^u+%}*z<+>g
z4$GXOqt;0y>REy_heX~34#s1_HDh(VxGGq;A6>RLj1Y6A<&aXry&$H4RZuz2*-6S)
zr>)pL@o~Bj7d+n`pS2}Pt_Vc&jxer&fa@9Q(ML<zDliXfc=f){M}-Eck^nRkxjIMj
zMyNEZ#v)Y31;p4|Fayvrn4B-RUf;t7a4>m;*eyY;{!pOZq%y_*whYKm;-sEbX(??%
zNC(u=;GFU_EADy%gra&V7DZ98Pk>psF2cs%>Z+Cs%ZT=sqvAvlx$4U$IZD9j@4)!)
zhNI5A5gdq4F2scBEXfwcvb(%(!tk7qaj6K+sUO_YZ)_b1k=bA5*dd4g6v5uY_3`tv
zCWldAHR2MP$|R&4Cq9rbamSD^kxF)=#J3TcEWz}i?8+!kdZjYbwkm5~55^b0bI2YJ
z7qX`%dZ#=j03^Oe`6Q0T9&bnegH+^1ao%d^jpZ?S6%Y`9yU;@*r4ZJNw8B*SES`{;
zeh^WpMNmtL-oNN;Tnbkq!<uP)Up^yL6eCKU>INZRZE%5^p57bBT?i=#a+1n6jlM$g
zoty!Gzu1{0B<u_KC7ZgO1YW`z=at0A$-;cw-l*&r9<Pl??Hl<yTX`uu^PaErZ#8Uj
zdU&MOI!qFaVj|+7Zr&zCRpej|a*3a+AjyuZm<P$TZuuoIpgrNN1y753hh+mk-EU1G
z$dn#g0+bfpG02`Qkvs~8$1(9&xi6kX5uOcOxye;0CiauQWOV2SYj_c-brFIKySpG|
zva9JZ8qhs{jr*oSRchfO&R(RiNfcHRvo<vXz5jAWWmm1KNHsXpIb6&xq`K^0;$SF!
zvUK+9(%ArHNYR>us2vH6NpxQrT3U<n)eZlx2Q4<X9QU;2*wb1va_4bHKqMkgd4&%2
zyps!VqB$%;IwD{}O_w^FfB~`58(;O?xSb-LqeeeL_ZW5GVWo1V$#EoZXhgtX>ZOGX
z9c=>@Aw==|1O=dzf*J-0k~iQQD$8;m`WBkWnhQDa!WwD1ww2QHInL~59CMiq0GSA#
znvi+pXUA4e&Al0p#|RVF)IAKspr-3M$%OpxD$^0}&Xbpv_WXA?IUNPSZRHh3JcJH9
z(1+Wq$b2BLH)!{Rmjc25oQiZCVO5-S3sQz21iVdscvf0{Z&7YN(}0(r60vG;$jbo?
zr#J;3|1Fb`9g!`GZxiMwqR?U@#A!Z}v-}=y8Q|-dbqV=22X!w|x00oVpsAR#(qeTd
zs}q=>wbSKCiX@v-@bTj8>Y63zUJU)-pA!ZbXdGbEuod`FlDc6H7<yocjwWeCnLm|&
zX!Tz!7Sy-bt^ZLgRv;y6A?bYmYqe@ix=-c(GUNa6w_3@aOHhZYv+s{b$H-cAymkt;
z0cs<DzY(iMkfB2Sxx|tp%uPSS0N1$6yiTgD_lIwCM@AW`C}5{bKzwFAhhTKkT7F^p
zVj@kAVA--5wfZSsWFOjqNQcgYv!9}Wl>0sFciRIDV=M~5g29=z4L6JmUi)Ma09&r!
z@Wb6T1jl)TR(X|RjS5Kb4Bn>!clzlG6cy0zmW(>BZ16&B9ZMa3(B=KWoUiWNZlBUv
z-C(AeZ8F(3r$-ze0Dy~>Hn_U#y>pGw5epv6pRG1vW98^FkcgMOf3_sAsYNPWwQzF6
zJfm>AJY1Rr20Lq=^?SOFM{V?QZGW3RVDAxGZxIJ*gFKNY7$?TtVnPV|3+NpE?r3!t
z2`zD4MB^S}E?d{4Hx)%(pFd*DkO7M=OuV-$XN<!Kg%+O?ghW`d>^DNA1K~LGzF1Fs
ztwPAPH2oyG3`#r+eIUhOT09-=@-x|hSeq>3Kz0mW5=va}BU8uDka<twl@l7&{kprU
z!Wp3}&8ieBcal*4f|Ss{6NRrg#<nxSNnX8cSKc|>@J6h4-Yp%A`==!7Pmn3qQple1
z)j@aM9laIiGi3<&G)>y`{;(}Sm{653ATGgS{6#!U5ynUJ2ibuxI9Fu@80`@(x%D=(
zcY}Yd{RcRd{47405Y+k1SU&K2>aa*LmuR>FP3Q*^iJAZpdlPYr<tmEH?&M<F@#DYg
zXO0-S-|F=xkV&}Q1Y>BMB>OF}zy(7YV^OY;VdkZmn@2D9HYD|uc2Y=QrNOX3^f=0g
z1Tv%USr^G;5JH_K>Kh>I1MinfAp(l@ni()x@fj`+0j4)kULniL$;ReBE*n`p+S~xr
z4a#Jw0oQ4B7$9b}vrdeNyHusVlU$H*K_u0-o3<FbPTGjkpb0#;pdK4hCG5jCp};UZ
zLU&~s0=cV?2RYMZ2N(bNe4d<aZSQQJoG^IPf5Sbir`=&1_OIFqzQ?sHQ040)A}}Dh
zhv`nx3=8a{v1F&m=oa4r(xQ+tkBU!WCU9B|z`R8KJb^b<JDUuVTL^+#YTVZ}?NS-H
zw@yYABnU$CbpMJg3;8as!FkZe!WbcHM=SCckO(eiW!1K&FtP(AuX-xQ%a{gy*!PL6
zg^z_TV7m*DAnM#8O8{?d3=aI*n0C*+n`lRFtLO5DEA5L`N92Ycyq~i(5OMwi)nji+
z7M}*N#i>y*QLwxTBd7$f@r?wY7iYh0tt14^O<uVD&0vi(MwV>!XiAug!63(cjln`N
zhf^hBM5o+V)g3OY$8ox@k!$l>uFLKWRr6s9=bRJM7rl4cOup)$a4(jEAfyk2W+)t$
zC_i!EE7oRu4S(VU!+{sux*i1eA{G*C%+i$cdKqHNYvP>}=}FF2PCDTT{ediW$W$K`
z^&m~Vo(LSDi$92%Yy=JPtCyr<KB0M0aTW4u7{mW?tVeMj5G7lE{xk#51By{rJTe($
z2@TAS=z4DTh1u3~F8i}D!w4!#w4reiCrb->zZvO)6o_BEPYOg)W|jGaoFFN>!Bmk8
zf|~Luh?>+1SGL^3YMl%_6$d=^g8+Sb1s(?HYZ{*fI$xLxoV>O*sVcc@A=EMb^}09e
zo(zdMS)l}Ow_DDTN9Y7*^+?2tJ6RG@32RrL0H3djX5?Y?jah>7abml-JkkrC--h7`
zSV9{|d?(T5nrY|`Z6Ng??Vn(YR6ATiwym*T!IDW}9h>j{Fb|XLMxD3Mf>Kf~5{~@@
z=?_OlZ(>h+XIlEk=HGj9885+|a`T6Pgc%fZC=XODuIb<ibrXBi#v-M0xKHe5h|J>y
z{MOV1>s*_h>@abIHZqsNL(9^_<DNoEsY*Ue0&VRaR^&rZRYk4s>=Ka*EEI@=eaFbK
z$}#E1us6NV3xzLxW4QtdiLWQ`a1;e9HH<c1E7JjQqfbQi=z3bPm+>mjOBHX&c4|ss
zLeKnUQ8na+-oVxy(Wgwu3Ija6fZ@K28dx0t6ehC`275`C%#sueP@{YB1h0=oZX7kQ
z6f=ZD(%>N}P<x^#f+S(Y?Kh0aD87<GUG&}{OCj`q$65nib>=`+un=<a=0p&;*!uW*
zP#k|!e4B{E#g$LKxqm$RM!w)WrsCVbJy^K^v79su{cQ}-*0G&{Ewx4Zzi}D{cjL*}
zP3ejaz7}AA0-Z?5@O?<M8oC~T;@uRVeoT8&0&cXe*N>PREf^h~na7AYft!oBKu}#~
zh@vDsBl1ugG0m!GSKk9L-szo}KQSkTHOW9GiScVOO2y?^xXD^RIqztXOyve}WYi?*
zB4}&RN4?IQE&y89>#5_1YkFO3J^-1XX9|@J5Lf<=0g*w8F^FnCVAy|1)A<`>ut4mR
zTnH-ci9VbNZAErf4bV$MCZVx`@*tyiwNf%BSsAQdu&(Z}*G#V#fs{d^TbgWQ3`Tr{
zIg}aPtBM!G4f#r~r~7mU0?bpOsyB|Yz*^<7%a{3SaeM;Qiez=&PJMv{ITTt!os}V+
z@S3VFxdYNt(W80j!YUv_R=>pdUt=E$o;a=&WEWaR)C5A?QO<@syP$n6@*2YAmD#w6
zG9;XxEY{#Ac2hHH;bTGywdfG<a|D^t22iVL2+h_2TbYVmF;P1`G)0nb6bZiRI1SlT
zg|^WqsBs2@Q~_zn3pkOOn{SY;-KWO43|g=g;lIUlnxrwKU=Jn3>Jo>VVg*vSR>}*>
z2-ygQQn7P#_$XQ6`7txJMUo0Y5Ty^)#GYv-J{sN7X$41>TaZ}ZbY!0F)$S>kMOZE`
zT`8U;J1ivD`E_4;oVoDG)8i}P%rh?|8!C8z{&=3AF06t}k`ZtI!8~1Frap3!fQBo2
z9}T=IIs+%tVT9uiiI_DtzXCgdZ#>UY;1Ug-yh219@<9)9u`|3${IpWAx;#R+M+7J&
zd2x!=5`^4_i}zT(Sgpsrui*m+#v3@|01>Tz3YGMxz{WJtg>cki923SS(<VdXL>UK1
zM;G`TvY<^3E!<5g%by};1yUHd-kEO*^9j!u8p9FDURtHb)diwUFLva;i?yW!fa)2u
z*iW|46c!L%3;zk&>@^{^*K?M@+G4d>0YnWqpA6^q!hh;2B9QWfI;Plwe*1N9d3mAt
zg#Z0dnRv16Phdzw{eYE$_AJFJ6ok0LP)4BGvI;FQ$sd5WW}qv6-C)f(z$`5oUUx(<
z0rq5h_AfR`QE+d8o+z`$F5}!jR)tv7c|=zhSr#i4y^d6MIG@6xY)|<U-3l>Ag<ij@
zkJ0*deMLV}U!e!GW^IW6c;0N-?pwUv&l}>*lvhBvD@6<T9$*D8Bx~K?HY<i35ruwz
zbxI+SQm|XnF4e6H<>u_2<cL4Jz#E+I)o$fqAj`kwVgTq&Rt5zL8GWB9VuM-WU12g1
zVA8z1GIcDiG;Fk*dz;k-DO(q>7k<{U)Q*8GxVhTUWZcboGFMj|MOfhvb>DslBAjtB
zE;}9C8@(nkh;aw~kHhil{#kh7(@&KablCn)U=d1KUItdUQcxGdYm%n8rIZ%)fo=dk
zsN@;(gi5c46rE<+4slEJOlR<C`;bx%%Q)5eLhYacy{w_l2E_sg5(X%tGrv?kR&7E&
zeq<+zI>pbq+r6>5b7FYm{D~T;=PFwS0?JbdPYHvs9eB38aIVb?niH}=7u6xb#LwP|
zVzzcZ4Vi?-q0^Z#sgC&X@WHXb0s^>M&|<~Qtq)j69UCS9m*rnbB$H9{OZT|N1Q*}>
zof{B!L1U(7T5ju2bX_WGp_6rWL6vne8s}U<May@Zr8LF1f{#i>J60-lWU!svzJlQM
z>T2)<rRJL>c61KjZ<9BTAaSp|h8iZmcx?A=yCT>WhiLBoU(ULcz?!MS7{nodsIYlI
zYvnE6GtmElsz>fRlQw4U6@iv}=PxA&dPayNoaSmfeueFBQNp&&QC}%C9hyq<*y2ms
z>#Y>HYkTWd-gk${G73%bgIlXKkPwPGzbpS$PD&k=+&`MLv~EW~?r(<uU}Ok^uvvce
zFiY&#XHZPEdgo&K%?)nFw>E+e+*-od(iXE6c?CCCQ7VUwC)?|r@)`JnVis#BpQL8;
z{qX2ud;LgiS$4!C`D)o(<ns@jI5f7Jew1d@66<>}_qqF<7NWNW0=~%09h3`=@)9zB
z$xjo9F6b9Xu<^c!Yev?dqpw&<F8&ecv&^uDuSj`^D=BMG#WwfX4h|KO`#4M4O=`zl
zvGn>LMi4)&BZE??!9~`iFc|CwQ8Z=dZNE%{rCge=u5Ryc?IAB%Eg{@eBhGt$L7SY|
z`dunS%DKJqgbHpiEE9gG8a}CatdvvUjDUrit}nlSunjg`+Yy}B66Xx{d^~sQD-i7r
z?cb%{oV;&sJfnC~SO&r>B>RWIJlfjYV)BZ>6XUbtRkt&(1fxQ;5?_6_^}N&ubdk%f
z%okr|C4}>mUMgJKIbtu7H8F|+Vh@+e;ZpQ5FJhDcjP%P8j)xE@89sDfkFa$$X$|39
z6R6^dwP$>!2fKPr3sHcvr27=QoX6_^T{hs75$m)hP}dcdXw2F%dg;mVP@>9${(!MD
zm<;%mKL-<PGHn0$O?lp|82)M4qL1797a9m@9f&7FHYo5yRc3V~N@S;8v!dovju_hO
zrt0S(&<c<oax?zIEehPv!B9*G;beSqJ?RYJB^S3P7xt~LDu7WzU@!+wwy=CJ!{4Gd
zjg4h{ybDj**SPz|2Uimxp+_=ez^WDHNMh5Ne?<n4wt8Eua^~AV_=VdrLA`(;deEuV
zz)68FV322C-x4v{spxWl^~7rWp?G_QqeFkm?AmS((2H9av5T5j1WFhWbhq{}*VdCo
zgSvpg9qi#l%CYDregHQr&$;PPGX=71;^8YCwYb}5{4h^p=Kx9AKgXr3BpyM%bqw@)
zF?1S{9rnn@DE0T)%lSx(r%h<$o{|87FOeI{W?kHhcrQrggX-J?)Hn&G%YZr&x`?oK
za>6N{{F~uB1mL<c=Ea20%i#}c&}S+3w^*nVYQQ8ZkHH~=vzj7;H4V<F`O2xNd5vk&
zD!?cX8?~?dc4T0&uC(whuw#GaK~>sNfgFqgY&4352FV7t=cqDqWd=I=x1`{aJvjIn
z)}yV6^CGxbylHNuFhK&fTs9@*nmA$;U1oyvcn<65wG@||^%&|9$d$hpN0ox=d}+kF
z%t&4d)1Zq>f(nPtEO#eQ!WpHF4K<T%n<r~KJH_9yg`&zM8;E$wV3w`56djHUAc05%
zReKL$cgGK47Y9@s9s2&Z(fjZzBtU(N2oPKaBYb0n2j>NP+|l_5=jjijKBv8~CVDUz
z1*g{#6nl?yHM__a`JAL=k#<dw0bC}~^U22h^ZxaiNkW2FkRaJWYecWR1KG9_n2>AY
zduqNS{ng;-0ajGIW{~rQ?KT9yYT6GyFpUmeGy@j(<KagDm&gl3k0kYx4C4gpE<8Yo
zepIhJ6I{Ct`=;^5J4N3jIc7&!^Ag1x@qm*PQgpTP4__DmEhE9i3&pcC!6+>$JJL}Q
zAwp9%&HW7oM!a}y;FFn#V#u!_PYO3O3P}e}%5dN5(xz`+cTv!41g&h=Bvc?w!Gs%)
zHDstldr8kWMliZn9CZ);=u1lDL1qaSY>9FKsua!k%hI9-@a1R8AuFMC(hB7eVKK&j
z{rIQXEk3v4To)>{y|EO2CC<k2E<Z!VP6A4j@YUOo$1#pZ2qhtbFf$^ma%xMzS^Kp$
zLqFk%b_hI2@2cRJaw`7Stbjq;f1xTl(z@yoU%N%7OFWr<AZd_|8Eq&Lx1k2noRBei
zp?4PA*V7q63&JL>oP`QE^|pD>eXk>dSKhQ;zg7};p$!8WGEcEo{=nI`Zb0xJ5K5}f
z&{D>V3?Zse4a<Elep9*)@T4yROVG)<1Xi5`d>g%8P+E)su6ljsWjt|W)a#4C(zyi4
zXMI!L_^=xo5$m#j0a?Fs<B}{GvW}I_J}aO0bQd&>@qb4&rw)UF=JX}e+#BHPP16sW
z5tP+R5p)Maq1)h&BS;+@VFc;R*+-DodcP5jnZfMxjziE&P9k9t`Rmz-QL1<AK->)x
zG^z5M;*v|nhKMee4=JTK&^s3jmrQlWa_en-bke!Llv`1gl2NU(+KfQ*?_4}r&6Uz<
z#mPDmy7r}(96h~?4owYXmKIwxv(rt3XTh>{57{;|Y7$-LPJM_TY+K%ODbinM(W~PZ
z*Mn;M05G1d=Ke4jSi}jHPiu$gQ$*nlV}JnN1XWpp7wW%&0dEI^aj0bmQ7na`%$3K|
zwhWj6ViP^o$WR(9st4gCBjrP>j2^92Fm1?J)L5W-8vH>*=Nx0Y98!L=IBXx}%q1{f
zrnN`WWPJJ4MrlB)6=OCkrW8S|L)rIy-TH24@A=x!A+m(98Miz$k!kR%mL!Um)u!(i
zG5c(<o3w<93uog|!XJnY*#s6iqOuX^5dat_FjwNcCYdw6Nm!YbY@)Yr5y>VH(}q*i
z94*{`Eh8fSBo*;rk`ez6JxRQlOaI)#Bs0Zs$)y2Eb3Ru5rYsodx~WhINvTu^O4x<-
z8@rr31Dqn+mvjp$L_Ws==b3i6fdIj5fSs{@wR_<Xj~d~o@UOJQ-V6LnLyh>yKN;ux
z)CW5+MVEvViU5e3Bj8~Q==bPMWRFZ5k27MY@DRJQC)*pv)31v!9)AAtD<SXFw`cn8
zA$h6vNveJJ=qp1MNYEyF`I*R;7pl~*w+C(9;;lYe8BF#e@z16zp9=~hpFovlkobUz
zdYv6P`Ef)AM55YY8G%hDbnS@+jmKIH>O%@yiA^#ZHhU*D%p$K_zNGTb^t`&8OQrXW
z*~;okUZB>=i4!%Bmvbp=>}L+N+gg5BUrd<yCl+zPMB0w^EtK!+jBpx0IS^jcY6W0U
zatnbZc7}YGDjJTX0!aRw*rg`o#PAF;zz~F~+c~+0ot~=v+mT0qsMi`-ZbW1vtXZ&W
z9Om-qk-qJfn3rNT99C$OLMDF8QV&KW#m*z77)3mr3k5Qvt<dhHrXawYrgnlYF~MDK
zi@uaFvQ@*2gT|$7oi7&We-#n&+k;<aS?9l$6~7(-^6=rWzV!cwOxUt5#5Iv5kIE{a
z1<H?H8f6NxuTPv{k9<`zBKD224xA(w>K58{;N7k1vZquD5XeaZ+$DvdmF1?q#)#+&
z2ZDV$({<ga(R!*~WGNxBn98>ksG#GPICgx+6*TA&*qRjOAw4Y>n1<(ZI^EOjHz(oM
z;>PClSKpmHUt9mv(ZSmK=E?rS-uK%Zn<snE|JUaFk?L;I;UyS$fMp%b*}K~NZT)S3
zx625MwS(^t)Ab;FdIG5AsXr{7kFeRxZcN3H=hrQyX8=yP(#Us`SrZocP5dOHaR>F;
zcX?$9s?6ob?R5(hn@49qo9(kBu{^9`?3I4Z@&wreCeA_k%)u%^s_8N<Fr95A)94U-
zL=$4ieDABPGHn=cSZs03(iH~4qlZterF;up8K=Gz9va5bscwksX_}|lGVc#nEf|5%
zaNOtYqF79XqMjI?R?ECh(~q3a+5kOdjmohbEDeTA6obL{6f6-@2p}BDCT^1QG>B~j
z)~^W4K%tCk5Te+Ig+cTQEGb?{cvL5hy0pJe+mD*@X7~(33uQ%_I0QA1de~>;)o`Y1
zn{D9&7PJvuLs46iMJ%+N?620qt>Xh$)F_=(1zk41L{*$q^;}-$;B8@9z_G_U)!S<d
z>_DH_0%S`p<j<=19;yqE%Gc8PWkp~EK&Xg@(Elx{<d|-$kWZc>fj0UyM7YT!L$wE#
zFG2khw}ZEl+Mbw^RYx%cU5;kXPdUC`EAsF*6&5Z;(0EIxjJq~r@Io7*7ZILuJqV>z
z6ltD}O{WMMZG^w|YJj`e-ds$k_R@^cT6;t>%Iub6!0NgefRiCuJ=ML#rg}Q|09XKB
z?3uio922(iMu8pni`4WHxN0S*i`sfYEPj2%8Kk_(_S<Mo<U#<QhE0RCsS6cm@dc4s
z1Uat03Oi1CL=j@9j30nhg%HO#oC!#y_A;o8q;*TX#^D6g(0TJmq3%R!<tj}=$hN;j
zWrIUZF;g+aH~JucvsJJ@D?%u0&17J4C9PSPY*d>LFQJbG1CNK{=LKabrioTs5=1Dn
z%a8#p>JBhc(Ym4SuwfJ4$bt~JeQfE1++%(}H}P!$B~GiSmaSX1f~IGB4m^@_kX{4S
z2)izzln$;7<z>Pk{#ou-!T!|PtYiozx-jO*7WxZA5gbX1Ys90=m9iMO>RfhysYW48
zv_+YU)zwO73lMxA-35vura{=DOhR(yG8UYsvl*D%&3Jkhq3cVeGhLt>63Smmz^6WH
zAa~+-i_9%ths^^xddlXR(YbX0I&rbVp-9Dp@*sO1{NplDAz_7_zq8UN4yon0)_A8U
z4vSjScS%lkYbD6N0#RRHM_4YPjRXpBSDkDO92Q+r29i1`elo{b6;67<vZC?Xmgn$Y
zypb5%W6Lapl4@*G7U9G|7gHYvr$HOJH!%M{8dy$z3add&4sxeWz_Ezxgw4!KF<}@M
z3m316j_Y_6wZ$ZJ&ESDeV-X6X<gY;niY~ZGV#I2OShYs3l~eaJlSvxU^0V$H+?lCF
zXad#-7)wF|P=@Uc4UM@PUio3=)c!PaV;a0jGua9gSA^1jkYsU3#9R`t0n7P2+qaf|
zk2{Y}=c9}&(`YbBaeZ$@zL1Ub5#yI@${{0e-CGn^GXrFk=rjThPJkogPRIq0f}o9c
zHHM>$2ociU>(3;R1Jo4h2|+{?J;NT>(gytqb7h95Sv!<4w;!^iE}xZbI6UR$rXPwj
zizb2p`V<GZqauKv8le@NX7a(enHzOlUd=pBNI9X`m(vn@6*x5{=k<LVxew9IjG$sm
zblaV#9Gj5=;u<YfjisebXtZUOrk~JLPY<Sdr?11(H5EY{X~cUuckcgugptYesC^Lt
zC4A^(7UdDK6V{;YJ2=d+n!T>zI@J(l60+iKC3``U80Vu&GLmq`qqeC3gs3~&+eEw{
zI2iZg4F+9i8bCIDgur0!K(4vgj6`lavy<i2<aM*rdVrbc6YPIH=W)!=1V~YWK=r9y
zVCyEvhE__k`G4Oyg&Fl|l3j*&-KelR5fTh*e+3cTpO!hyPDoNGE$zJr8jXsnZjeJW
zy{&F=g9xZv&9;E<3}qhN><Qw)3UQ1cZBvef&cQ$qBV1NEIJ<dYNbb5wPsiY0TjOAi
zd+WWXw$Ok(qlbg)Fq!mPV~DXY?>D1Shc>qk3fOv1{>{x{@i5fT#+01UzsK68bGs?M
zJL{&L>zR+2kOO&Vdl&hT)1(xx9@GrQSEmw=3@^zYm4+L^x@x@jSQm9!wA_E}>aG}y
z$~xqp%hOu~oJ!y7SuAtm1@*pqCR_L4K_eb3W`aClWk60;AfOwTF_s>8A*oS11YN^4
z&1}J<EUSBuE^bT4k9DrveFq1wM%M^n$$FO*5|$_=RY=%-b2oKxfm-Tb2aJ$`-y<hO
z+{6J=<xl4((@=w{>S?)(NF3oC=dPrurjru6qU3DWIM?3rT8>d@`mxp3wCY@_t4AnU
z)*QsR)f1|>gbjwbaaW}LWy>}s^76h5k-yzX%rTq9d;;~x?pX44&mk{s`w%8ZpMHZx
zmx0vh=3+W4PgK|#<Wrghb&ZRt3TO^fP(owEI#uC&3&RBbQf_zX92ncq5`I$7a|kBK
z21ER)mF{&@cej=!qGJc%SIOp14Y_s2lW?9{Z#G}nj4p2<HZM&gcz5mP=HWi~;ZIgt
z#~as|mp5^)l$WNjTJM!YIpNe&m!1`}N*;HU4lgE?t3P!|(8Q}Eg1zg|YUIEa0)^nb
z8#5rvr>HVb9cagOkicKC(|yxEyQ%q@vsz7oFpkD1W|nv}a+P=w_s-8DWlXGt)#N2E
zWuqTOJ$Ir|*#|4fHRpF>zqoE{5H>@1WUZoeFv7xD$&Ql*_{<*b0hcx`%2MnHXrm%z
zW$cuyVf!*V#S&yWfG!h4P~|Tg$^u_c$ku-%4mS;()bRj?jpeAfP?LDbsH#$rG7T}M
z9&sSqllVg?QPzW7@Wj^L>uP~9>Y^4GnjStk&%AvW#`DWcvnm)#Yju@rc@vDaIyd0C
z-kDh@Sp_0a!*9Ya1DZf=Uy5>WzhfI}_WA|_wmoEUIo0GbK%m+7WkLxa6_m|+K76kt
z!FDCgX-JVeA8nxe3YhV6dVS<E(uij(=<j9HK#b{EybH9V(+=eTnPHepQ8*;USA`14
zH<#zb!Ocl2uBt?X4W;f(fi%8SN{!UO$(Ap`h8bo2IEC%_wZ?ZEwICfmgpW3Spc&95
zNJ655?FIZd`y=EH86i}q6;?+^a!Uh@vMV(phw|6Xkm6(9;}sFvW>x4$N`2q%UvnZL
zQhew(haa}!CrhlX4q@j`^s%ry)T|u0tP97XC3>?_m~2vtDcxrtBk4$Z)U0kYxO`Ek
zsm3~#qdkJ|hSIiSn389yqa#F()%L>=vv*TD8|Ph;=pytB2#$p`e(4b~n0MDS1zOp&
z6rwDtFU<ylw#fUs#S-x5p!(~Vp!4gaD*IIFt8}j=j>yqSTTd{47nP}AiP>OOXH>0@
zxBBe~Zb=N>!OadSQ~uzJOQM98s*?Ujy|+lp9pW0I0bq8uODQ-x7;SMVGJX+LMRz{<
zh;q|u^CNv5mes08cH3a+9MVQaIEz1DHgAugw>vMT&vif%2A^JF(C;(dFp|evwqb1t
znGaW2@4-4l(_G&<h;GMHxii(icPG?jRzV(DVyD_w6wYt}qdXDGpd4Y+S&bD8;F@r2
ztDAcA&F`oySoR9+L5W&R5$atLF;bc;)2OW31S??6zsDr#y%z-hJ*J=u*c7=~kp6xa
zM2U^afu{SHzBJ`PtTjydG7xsy@Q7S&hE%DGu)fqokg~Xm*C&zbX~ZXwJV#7dM)FG5
z-Q-z-NS-MruE&(B$;%7uZ9wZu-q1IYJTfApS<=vzKI!IR!yY(*IgVYD)+K3+uqI@f
z_KVl1woouh6hB!)1sZvP5+AyWuP!n6B42c3Nm(^Y#dIa%`k>3QLntm{%R*E?rLt0I
zqbN0f7?3K^mWE|lYWrw`k*@bG@V`(j*vcf(1qWw>yebr5JY-Zn);mQ{0!3OapwIhp
zKq9(iW!xGKF9SIz@gA|*x<~;L$|?`SIJl@96tXOrMyN}5EF)`msf@1QC=;E2po>zW
z=)3c2bQVZwF)qG>GsRA*s${jgYE&B>elfw5&_`{<LBX4$VUlo-eu?`qjB7QeR1A0r
zhfysrxmqW~wb7`JNN`~dC!8#KTrk*$i?SQsMf6f+>zaePTkH1&DzF?u6h<8Ds6rpf
z?aFiE)MibzxAP_Ap4Eu0DWt^1B5E4c6fp%&Bcc<$1b}K0gfS(QPc-GOLe95Kyc+yM
zt|ZmPp(ChPgvH7vMisv`jQ|Gct||Z~4r+l%?_M#Ni3to=p8!m23NRRsD=3gjY3g4q
z*=-UeG99_pkwKH!Yb&Ei<?Kp_x_s0i3gcA)!4k<B`N+;1E21yc@XPb7Yg2t+Dri_D
z>5Gr-tFa>b3M5?_T3D3pd-RHK8zuc(O?Ex45zeC+J(aE-$q!hSDrM?r1(`xw?rQ)>
zUr`uT`g1{5JNf7{R4pse`Oc3Kl($BJ{C92304ftwfGdiqOz?@SAk9DV`vLTo8himh
zRlCD#rN_bK#L``)jEScTnna$}js!T8O8L?nm#0+pa+D@>!d_H>So8!Jk4Y_gFAPN)
zpBER}S8KPjs5oHe=0aUE9{wq@iVO`CX6hLhc?D74XUjuS-r-I}=4@nmZjE_WoEmQN
zrHGaIC@#@Z!)@32onXn60+^mSBB<y^{D5r1V3gG(tcd0plT~hj7*|>0T(t3`Gr5~h
zFWrF;1~0hyd&D$r;2;YHp!;RvO_bXXOV`fermM6l+VxI!O=e&yFL|2q+On)qzhOnM
zZwqC>tTW;|wpN341p^-<-s=Y^q2Spd5UnTE7^Wz%D6cJqT<hr@ua?baUNwBms>in$
zfDWj`^p_eSRpgcBO{_@b5W2}L%ImD6zKInBIpr1A+ngM}JXf=B4c-*!%8CFlw3^W(
zolU`l^FbC_eO4s%@vt1Y4W8aVK$Kt0b#@PB1;Ggp<kw6&F_JiIiynu>e4<q$dDnH=
z+(D)RZR%O^sQWqTKZ}C}|3JHh$D{Sh;h(nm_mTO22oE@t^y07Jz<S;BCoxIif;66K
z+b1W?%&<M+$?JZ1(mh!x!>}hL1yR6piAVQ_$E1}>aUGZ!Ue+7w>2gug!sU<kcd;`~
zxf$DC)V^~)9m#(DPJ11;aCm*Mp)2&^cLQ1_i0`l!74rwh7R4mc`2ijuPU$4#b7OOJ
zzu4V8#BKD&+QGrvkBY_+y;%>W*Ee)zbRzHsLcs-3_$`yGVm=Nb+K)N}I&15@V@V*N
z|3EUu7Jvj~^n-IN;-H}g7VEYLoIx#HY6krduSdL{>ei;bDB9u_`Qr&COQ`P6H;)PN
zI!Nh_MtDOeejx){fQ-!O_JLvq?bYz-6f!)nCwT9K3%N`p$X)hI*2rOdr~ipSbsq$*
zR|i6hzDAmkIwm?H7cep6T|L$%zb*Ppc0KA2)&#dy{kF&5{N+=7uLRw1x_x&G3}wqW
zjU;l7dZ{SwFGO9Fauh2tL|XHngX^WDSE&|W4mJAUCI~6agSzSeH47$jqi&p0gK6Is
zlK9`9cOb!m_1i^sIm#6}5)F^LrpLGkcMM-peen`&Rd$zVtyd=lTuNlZBGAPP&8I{z
z@;#^yx4IHHG2*nbyE#eT*ak)NMh~28;_}LCT-5bua*@2I0=Ux~{Jb?B;lgGBI8_J;
zJP^Bf5YE7%92T#$*buQbdr!H*<z2F1$RR7N8hRV34hOo?J>AnUhl*kE8##=5`)9dn
z7#z6Xdx%@Upq*XwB3$luwN_3}UTz+}*xO)C8SnP=gAh1BSw67y$Lzokq86^Nm@7}O
z<@{{r;YzWObr*3|*vMxlW!$+*Ef8ANDE5j7=X2bs3pGK<(s+&ss@hfWt=yqe8Vl@|
zd=`K=Zw}uglC=TyhjEfm2QJSQI<ePD4gZN`I1HM_qRT@Hl1u)?zTw(;d~Llt+FJf2
z?KRaBs}Pmmq%CR`&<abOK;QrQ1iCY*Nm~|O@ca!k`7E{?ONx?Ud=ZF{&cXEbmS1Rr
z^F18!`w*fZP#|r_Ef8&w%AQkXX-b0I9QD9}*wCHvf&<XCo4lS2oKjB{Uhx<6WhQAv
zB%;71f!Y#w8%3o<v@32NInM`}%5r2;6rXTJ7EGqHLLf?%v>-w7&i=|W0IeHHp*d%2
zQG_Mp&Dtq6gV{eaWAd<{B#sZmF+z5*z*FH<&<#;DCA^w(Z)Nfkw=@c6i^-UEn{&rB
zHgZAh`e_M<iU3xq>o@<hS&-~h%Dq15#%wiptCQGV#{IYt@vwc~eIZLyzE@?)6TIO*
zS!OrURfS?ptC!yLPPpMI;|E6*r%DzCNy_nN0CRN>X+es<y?;wOAcU6;IbHThgXpk3
z=$uHBbL{+yv8)+3%<+(ds;eb9NeVW>N-HdZwR?6k%t!2tZW%3TkWvuQ;8=V$<{-nA
z)}u(mG74nrS@AsP!qXDj)IwGT85HqWLVN}XCM@Jq2(kjL5H(z$Kgvog!Zgw$_QXjS
zcZ*AkpsGZ_OHMkCzF2lCKC>yoF%7C43{TM|B>!T#$w)K{S#0RtP;^NN-d%d$hjCaJ
zU8lOa2f?*;&(d{7kR%f_=3T_rn0_B!%?abc))l=n;R4yU_qK8-qNy*%x$i#c_!bps
zGXEkG`U6S3QBsb^&=Loe^Tl~y8_t-X9QJVVs742Rykfc!PKQo-AGyW=K_9xg<eh%+
zE^ybh5E@mlHg37_)dx#eLDdLJ3M+VkhrDto4Fk&Lg)$w7ynk_3kD9m)RYc*GdZuGG
zWC1GgpU8~GL49463ph6OSp#UoDYi>^0(dYpY)s$8;Hwu$NBa*Rtvu9gW+MZM-J5KO
z$`*C>GzO5QT%ZbZK5_#f_dZ&u;-9v(9O8y?_8_-fnU>Mf@)0~es|7E7eQ+_k>^~_?
z<N2DGmp~?ird0*P%wf2xjO#$(9x)EZ3WC?1I^zJdt?<Li4Q_nC=t_EICn0sE1p!i-
z<eZGt3BVkec;LMcJm$Lp2v^Czje0-iOD!R*lnir}s%>TZd4B>kwZM5J5~9uykp2Sm
zmZBwH3M%wf`51`mzJ&a&>umd+W|QBb4c370kK#VpnG_d)7lJJ7fC-hCCple$fr0V6
zo@957%utyPlEoQK%FiEuz6xchH_?;8hr`k61}Bpt$*+*$uDBUq!>;Y(>YL%u6v2=X
z$h&$CJ$C>L5(4|2cLJ}#T(hLNX49o@{%|oIUO~rV8{?wi`?*_;hm7rHgYRfCzZ(q)
zZ`cSJ9d2%9lamb*-=s4(7gLLXt5y>BO~fyr(4h=A+gBF7WQYr*xhCo!30&$=l5-t_
zmyoH#!*E2{!9CDp@&XZfnz*C~TR*h33~SnpNSk3h(4ba5T{b!ZW87oJ0vp2w2K~pU
zNY$p))WyZ`S6ySS*C8ASDk17Pc5rdJue(p5K5cF8ZMEi#hQFs5k2O{K*Yl_S;n~m8
zXvFI-u2wkj{WAZou{v9QijC8OHxiJAOWCn?)g8@0yVrSeuk-W)OTAY^LV##y4E^B)
znlX&Bq1#${`rzVmGpJ8Hy|+GG!P#y28TP%n?3k0z$BA{0>T{T*IY>c}5B@tLIpSYY
z-1sa@o)XJ`8_GFpMVe~ywz-g560DxThns^tq-eqh4=PF4@laT(SsIijA)%v|k<HV7
zkH?0%xaekjnwq>^$%(Jko<OzW{}%6!8OfcE#K>N{jik>)jL`QJOYzy2cR5&=qGU~W
zU#%9mLAgs$c|mC1j?y-10CUS2`f~okFjXE&2C$;u`q7pmbk+d7P5E&C=>uTycrc&K
zFL3C*u+D>FA77t707-eNW&=o1Tp3J3;spU6(7x$f3}iJ@fe)ss%9xo|Oj3xmSF@xW
zHKa`9i1>Wz$2$beVIwoF6L*s7VyM}Q31LM(Fp`shgx7kVujj|-%dk1R6Qj*}=O}+)
z!-G#R9z9d!fS(^#`Pt_w?@*?K{3tWltX(8(<1I^W99>luuQ>e0uUIazDOEgrboEQ|
zeRtHsjRi}^+NeiSFm4aV%gFH9JBKbj><>q)#mAq0_8Bs2rpStot%&SKoAU4K=@9Pp
z;-cGw8)Wt26X|Jr-21OCA4s{&@!Rw!xg4yjGx^uT5-zNY5KP-odwtu(5wBh!X^dAd
zdYw*p@Z=V#*##i?#(M(+(E4bC?vsLsLZRcf7Hmw9?%lD+>0K=q37VxSfW3s+=?xI^
zaSi`|6O8TC(-F#^4o3*HU&fqFhL@OcOm_#rq1k_oU<=GI?zZLxUv&Fd&5&QAB@&EI
z<D<{7e#vL@@k72*TKV{oU-ExXptg0mPvL#DOe3_|{($1;y}VP}Ii(-4u}T}It;a9V
zzI^n>zs13r>Nl-!?~=VKc2);N95ra}X!SzEQ+^d2aoJRCXktj<3EYxknk#oNeNbas
zeycq!w!O3F@)wA&h*J)A$vIWf_g#00sRyUdAN8=v7Q67*9}F+s12`Y#4VagF+c=h$
zipE4NAf^4A5!L5qXZQ2-ocF_Qt`=W%s?YnwHpEB2dp>z$^tg<xH^A_SRH+nj?kc7e
zdyzn=^4)2DnH{YbpLH)oZ%)w<SH9xg_vpg36d*tuD(z(E=464Ujuobvg6`2{;=mYS
zwZKzhJTnFlkr0*{U$P2X>Rq<sa!4UiW@_T=6_)~)S^IcP{$jtJiQ@4e35vCENRu8i
zVNSum{H}ZYb8iBEO+<#y>hh@F>0M*(e#F_Iu_ov2wu+a-|MLE{3zHU3OZD9~)lw>{
zhw<*C%4^IpnykVqMeTUVD=j}hJ3Ff)%}VP?XDloOyshjz%-{06QTuAO5L0-WFY^AU
z6IWuD)d+`=I~_=67ZlJ7BEii>SCiqDaN2F8ky9gU394?ZR;3M8okkeTwKb*^piSM3
zBS}inGJ;tApGQJm?;zseK`&4(s>1RgKxACFqwuoDx1>AJ8=J1Qv#!uh?SA%%|9kR2
zD>&@wU!y1&nF_tBiGwa4WuvnNwIZ5}M_)9^=3*u$3i6nXZDn$*ByG!brKAfX7w@&r
zX(3e}n<UzA!HDrl9k-wnZ{kjJGpAt3o*`Hl{=sOh&=h<6(Z=5T(U1F^qFz6ndrB=^
zEvct;+DD`6DV%og;HT#$QmLQ*xB22hxh%H|zMdZqu>y`^&Z%eb>-oVDPLS>|OM~J0
zkU7QknzB;Imbgeh_yjlNUEK&HW4G=;rt1-6r+;k2wfIqiz~Pkw4!=Uaj|&=+Be>C5
zPzlj8(EPfHv+BqJ^pW^BpC4`%JNg>ThPbTN2M;=o$!3O&l^Zs@g1F@eH`hJb_ELRC
zjE_$qWRuBvzZw%!u}>I*#P{U4(_!ak35Netr@z#hEOk0d*Zrkof2rSFg3fvIc<JJ^
zrHjv(F1}d0_;LwR7)w0JvqU%0(s{4j@4y;f>UWWsva{41U~hVfEDDEgd+BOPRFc15
zBQNsO<YH;k`PD3xhfiDudidnG2``KO)d&ud_S;wE?yCQIVjp-c0!9N&#ot1=(ABLT
zU&E>ZTE|OpP0}5@bk;++0&M4eu++U=Qd@$MAim~THCsgY9x%?J|3u!HiD9pP>-T=8
zn@n{x6-mE&2a}6mMI1qKe=}I+Qu5@t3d*b>QT`r2yjR=FUsq4NbcikeRgGHAmpGVL
z;a>iY_S#bW3}(>HWgM-*iedEHu+|+|L}mR70gX7Ic7Iu+5&*FtJ^5{XeOmibRlg^{
zl_n2q0I?0M(4Ql<Pyc4MH@M(MKsp4m+p{-@Ie*vC;3tW>NPPUJw3Cu3T<qj1{}Zat
zR)*r^ZWpGN(~wveCBd7SaYaQgf9wtq!!%^`yIyB<vHEXciIw7XDyDQ<s8dJ<2;syf
z4Vg505XI3`GFFrknVaY71N{h(kVrE3^HfBndRxs$DUMhPr)fC&qXoh?a&W)5o{G^g
zrqkE+_r~gc=DTNfhe|a|KYEfJ4ifKWP1FLpWuk7#EBJxY0!>{5Y;qBI^;4%EOLBLB
z9!r#fAbE3wX3j1aTOYsv^Mhmj{q)Jg$ArfkWl1^Y!XFI%LdI@>!uZ4ij7}{SGxo@$
zJK??fQCSQH!2FZi{fpQPyp0+M*1C7D=e!h>*1f(w`UT<~EQaEXRB(85!~bc!xIhSn
zV7V}ae>Y^jr5-#W>c!Vpa5JJU_$LGyHeq*Mir3E;!E*ay8m*nr7CQ3<qB~FfXIKb+
z{(2r_p*a7VE8j{FNv=mPj$ZCO$shHrw4`6O(eo#D@Fnw+!RKB{i@*)*x)QDU!v&u)
z&kF-KW<NRv`F+VL<ZQR>yo_CGi7$*m!bgiTgX8LZo;2c2Td+s#Y4`G(R(tvYKc{&A
z*%K1{M^CbJDZpdqc~Pm9n5jppm3A19@OU>o-&Wl^2wy?x{u^`<jshV?nj9s$13>ew
zd*c@4Pqep`AH4@MdF+BgRkMs(aLd9nP;QIHHvA0E3~)*_trhYn+{y2{qeajztW3rh
z<JMZ870W_>f2CE^83;-a^vn#jSIFC{C#1=Uw|exMsc9aYCy2p=cuFCxpfw)yar`>;
z!qG<ma#$5)ftaL49!66sCt@|w@oJ!BnM7{xLU3a$Q6B$Q9`n*FsZCylL!5vYC6(p0
zgm7%kwN{Q9zMh=_@dXwnC@YJ|+A*R?|GlC}zRDuVXN)4T5pj_mU%@5T?n}^x0EfVn
z-J_F#f3<hCd59}W7E3oWhs&D3k#dKxo*y1<AH6!-#4_>_lHP*2170diZ^|r&i@zf{
zxuFuFDkb1ZVn7!cPOs0;yD~{$TOv$DgrpUSD;1*^pj_dv9J#@*5{uVN+m&@zi98LL
zq}=6A?W`1fgMJA#1%|QCm1iVfpju){5G+Q?4r~iFY0ZO6tqm^k$ybg6qp-;cgXGIQ
zpM*&gx@!FH2-uqKu!|^VyTgf`mkE2FvmSY=IoUS7>IS3urO%xq6C`PA$?*(R@^}qf
zK5)Hvj*}wXPxk`J>!yB^BIQ;r7h6(dmPTF;day>u?S4{^Fj;GLpMib(JL3j$A@v-^
z5o0z}1qB`ML>S-<eC1<EfdvtUzcRgRWUw3{&?svJEihOAQqoeG(^Bk`M=>ChTlGrB
zDkBzx)zeTI=(-3%66!bvqhdD^XZnv@5|tz&NnplOiM++UwlPahjb|aQ0`<w7${E7y
z<gbT%j!4;)N1b7uSK-A<k$1>6v-uk~UuvILYn#{fn&K?;hE)fLsu6T^j+K)S*bC|L
zgbU_g?xW>FQqkRD1t?;!&(3=z+?W?X<4Dpm<rU9upiGV$eP<}K@!2`4@;qQ%R+>H1
z-0KalHM`NmITV<if4_#M43|PjsZAnB3l(OySPJmeBS>XhJiiq)A{=&s`8Fb}x9M3@
z($TYB4R8Q+$H$1Prg9>PCEi$1rV{5X>cQtsqM+oFDD{c(FCZ;`-go@p+9V}-L#u4j
zx6Bgldkve(msi|ShVetu1pTF;@mZYXBz+MYlXpA{7Umv5!*J7X6<T&TLEt>u346~Y
z=oN@7z_GH>;x|DbEn}_;W$!ftMj*cF4^I)Z7rJg7M+&vTXGZK9PL1?Bh!kcn8C%8~
zMoIaH+bZnaVnm3NgCRh;PEGdM5ESKkb|H&*@|wuV5Z7IKvr^2rR)WF^^>UsQkyJuk
zyp9N^q%MStDEgJQk^)$V13fUp=y<S!y)K4nVzw~0qFo%%V~stYm#lgs!RNhsU)Fix
zOc}Dg!ZO9BTx|?M!=^RRo1e9)x-A;9N-dYTLYikkcIKumCuMxqKVd)xNjN0<MoQ2`
z6q?&WfKbH<vnMmEVi6ybfVbE<<)J^3%NwAOo-)ZLIj=)i`bLstab-<G<cD`dD)Zl{
zCy`WwXVofr;07ZgAi#zi&lEI+GrClj_WTJ=>G0zBu;$d@vN8(ja6+QqUTGLLnoGem
zS`EdkD#{F{>r@b;D=ceiZ-Oh8L0#U`#o5!vKq7DHC6PoUv9i3<a`b+YX+*F^((rVQ
z_8KH@egW)0$h2$^_EDATdc@Ht<xood)%ZdZ{v8oH+Zgx9qtE|`jWg=~D*CPmy}vRq
zx8ct?P*ljO1UR1!sV)u#=OA)uIDMjDCt;=aKphrvCh9E6GJQfl=fMN6);R53Q>D`z
zah0HhfPTo@sx@_mYs}OM!US0G5eFv3(8YtuPRQhfk(Jg8Zi&HK5NJ}8fE;ug-38@*
z8%mmJB=+xO{&}N^i}9ea+;};sLu3Z97Ai$nSDCu!WMl8>5LR|9=uP;)tE)SkYu|4k
zp6u=J{HV;N4qp|6$yg|{;2IV7+rp1sqEQb8%OE*gvr&fwZhx@Q(QsX3NR4qq3uGs4
zAz`D0GsQQf*t0do4Naplk#S04RT6`go(4he5=53PvtbswdT<6sc~c31Z>e2vBA3Hj
z+1N%#^rO9lA5RW9aU0tjeu4^R%1Mo_=+Gv0^8cw%0tjLQ*)Ms{UNpd!a;NDxR;n$K
zf~6V5T4aNAThdWe*@pgSsZ$FVDsl-Ed8lox*bS<bqZ-kfeNa6_NBD%ZV7Aa9yga<V
zU(g@m9E;YB9`b9YL-9DpPku9TggfWwsQaj;6kFCrS)nEY6}XL<Nxz%FVDk|NG-Dx_
z3KZ#}2TJ|xHCXm!I5<O6c207pc!WlN?~Wml<gemG6P5{M;DD%X8Tj(iPqZq>`Dr{Z
z-NAxvF0$zJ<wexev?myx2(TC1)4%7v$n^YUokr155xfgjxcC>wo46p%(@}C}Nb%pz
zqe0sFNBYe|s-G=mwgeNidETZVQD|z0tCau0=sUtK1X$xME*V`P`t(8N-$33XPV%>O
zr2I2IMz-Pr>ehO@eCvYqbHS0S`2L+7KPq!%?S@XDx_z7ssT2^G=3Hs?BLSelm**!1
z#TIK>y&Nqi5Sz!5F2Wz?0?I-9KraxF&c`)&sxK%VaypuQ(N`$BgC8O-iV2)Dy&$UF
z992yqHadIK-gUi?i*=SCI4A&cMBtVt7$dkz#A~}f$>49XCau7h$w<z8$+)a5A+~1C
z2Ujib=9a8!D5c7p&(CpQm+EW6M>VC=`K+bRIJY&mlRl7&NYI#KNmyLuUd_i?mwWn-
zd?Hu)EGkH11*n2q(+v9%ca!uJYq1d=nt>ZTc?Y6{;uF%P-o;xNTcNnTqq~lsrfiZo
zMt_}^h)j`t>n^doN8hcm1RQj+0l~3pakLnlm-K1Qx__k7QL8e{rfr15l~NPB-d0jM
zv{dZxpS)RxMKr`&J@|8V#pNOQVW(JJ`Q-lUD*t=Dczl210nNCY6;J}UAc$16_-MhO
zgH0lop<1>rR$d0-H6ni5UHQ{ve2P@qw1*&c2$IBcBxQR~+d_%c(XEByxt_?3n+iHW
z_J|?LT=6$XP-}nVeI}>XM$tBChS7wP`l+PVDc9G(g4?@q!LH2DI~ZR~x;-dn84w12
zQp*%IR(vYfzG!o?(*5E=@fmHRD8FG(OIpB6V;E&pu}-YH)QkYAJX38ka{2%7rI{iq
zm?zU@SvmCk;pfYU|NauD+-VOA_dX(^2WP$RXl`!5JGw-~1M@W?<s$a2x~DhAn-Riz
zyEt@zKI(Q05;0PZyaB~R{36j%fspSpDhyAte=}fC3dFc0*nAEp=`28O;rZkplVHFh
z0LmN>5h}x-x6bfP@;ypA&w>ttaj{5MX8zD>%)=doPCD>$!UC1ZbN_^hSTO7Lgu&{F
zJd+)OC79)B{cA}x>7V<(OB`O53>0iyBi!fEDCFysz`0ah4m-Vb#4TNRWmsaojuAPE
z<8)~H^cs=c)VwW_9dVe0&I4>aAqcsTJ=-g!{U%n}yhA5t2s&_s{tZP)IXT2RbL=ms
zWr^wY>(QWx{5l;%)?qlN^vFcWti$3O=}+{+jn|Q!8spWuxg&gPpW@_;j70<dVN)UI
zUeGNQAVBD>+Nbdam0hGCQG{VKfL^=`!@=liYGC3K)>JU!4Kz_1<_h|Mu~{7MZ5{ov
zcCcA&9~S!ud*5$wAdmO_+995?Y4^i6(s;f)Dp2BJZTINMVsEQh+x@Zl)AsHL!leFV
z|6ud*u-H48+kUyfv%QIT+q>&KuQs-Kzbl@j-tOK}v9tYho5{hC_6l}rO>HB#3XQ$o
zJXn8$me-zd?`$9axHPx5eS~D<hsD<3L9tfA_jt6u{%U9KpxA$Pu)lY>iS9Sh?C$pN
z)&Y9ie2G}C74(XC#pd^TC=Oq&?d-6txwTh-{($fm>wEh@9&CU2;;4ABx3jT{m(Mo=
z+1m4+P3;PUTHjgQez{a^ti4?Oj?f(xduZigj>QzZ;)fT=Yr1*B{?_n+>qpyryBx;)
z-tN%>9&sji@8Bp@{bBoX6KC8HwhxJrt%JRnOLIggYM^yAf_l4fw6YbVS`ZXw1VxCj
zR~WRzQfzFl?Vv48&u%t-D|hm=Y>{7Y#@I%Wnc-aPh~A<gpHeT;-CR1;I`S*H0n2WA
zD>mfF-rfsPz`s{^P1WVb#tO;sQLf2603-Jd#8HZ0T5gkz<>>LYC*QGU3mRo=W*wI-
z>g4eQYsn%Q4ZjwL3tp~E5fm+tSU$uPOFgk#>ndX?(wQx$b=ke$`m|BAhF!mvSH6p8
zJaz7D@{;KZQ+9lSyf`>39wshzi6s?Ms&fCXhvCmM(Kgoim|@%0v3U=DVwo#vO>f4l
zTI6t;mO+IPz7Nuno6Ovuu6}p+r*rswO+)0>I=qG%2{VzRyw=wnfG3>q2PNc4z4=o|
zPsa(|O@V*4_4u+PHC%+L;S}pZa^!#IIit$6xG=;D+#9eONDBwefDC)!;na6Hpw!*%
zB9vSj2HA<u>=FaoMz?)=g4hORnQ9|7Tx?Bt3mCU>j;6B@lppaPQ?^?fQ_N<#h0r98
zc_O%Km7I1FJ&R2wwhY|^%FF<%$-3Ozk$|BDh$UXQVv;d=&58Deua`e#FL}}TlM@uk
zp_GNraGf1%;&C_B>QIBtksv1_t;XGx>rsZ0rGhcluMT!jUhEwnS-K)aPei;1g@j%?
z#!Cv8425txCm)i^GB9=52;9WA#-FIIBZn_E#2-WY_8g}lVNXTOz#%o#5V<%*3AIk2
z7wCN=&V7EYADO01#IsjSDojIe?Nj?L&09u_P^p_Iu`)PCMl!gF9SrUrEO8|VBVEP;
z;eh3d-59Cc!L9Sjl^jvB$1vUUQm!>eyi!ib?~Nq`cq*32h982Zh0_~^#DS0zmkHE_
zv0vSsOoqsgOfJSL06F>*a8Kj?tUnxs^z=*eLV?B$<gRCm&1zLq93ub6S@=M_eE_9=
zdv*cC2|=-mlq1?{Z7^$ZczvG-Ad)YUKftxc!$*%EiEJRB-6jDu%rjXz&I52R_C)Mr
zF4zqZf2LgcKD0%P3znLwVshG*7M`%9v*qHwH++Zx3IsI2Kn@6R#Ei!ghUwn|V0cI<
z$Db7TX>6GV50=**6}StKxV#t(Ga~xxS6_X_RVBEKOkJ4_J;DW<_fbPbTkwnA#|gj+
zgD3eYl$8bcu>jdsLbkt+9b3PCm<Ua&dk*s<ZaU4G4UaDa_}B;uIqSx<wn^!`2?FpM
z?s(_SFh!eSA1wo-vV$-~Fh@+`KD$_s{cwU=Kb522o_XF)Qz%cBtf^oPA=2BXM+=E3
zk3W5a==K+@q9(kC5e48N@y}_UXT{?${@9FABolT!<Lsh)_A^5(Sy5Mi_y<%o5G5C_
zBIh~2oOQdMG20bCQ2PvVDuVa8sz~2q?fi(3A;v+z5hBxTWg%tR7(7Q3liESyDkE|-
zyvEU7#Uo~e3p!}c6cadm!sB}&8JQ)~IGWRDzbU>=db7^(kMwdM$F<VWM&5x(w5?|R
zN8>H%W!gLGS!h|%l~P*3Z7LMX_|XGw389hc6kTxi(5-XUh3q;$8nxvN3{o5rNGtI*
z?nlwcJ1%Dy(#3uuU{qhl?1U4QX)rCzsKzD5fyA`7a9rp^zTiL>0f4xc3MdgJmNp?#
z$Y_`FaTHXJaUxkONM9;DKKgd?npyDt3mwwJD_kY&cm>}_m*Csb5}1cwRw;FgFq>I^
z=!_KO8=POniSJBau~$%3MPLcOBBEVpvbZ8yO+uS&369*tH+9NHIj_whS3j+;uD?1s
z;5o&$!;}5BqZhktFE_DP*A|U&lK@MF__rzkD0|q`-xYnBE)Ds7F9d${ZB8No2#AT2
zy8ufu82`$AZ5kIF^~D~9bA8z!Gl>B2p~NFlr17(O70k=XR!TIfA|TCGgIt3Yb48Ek
z(7vb`e<=Q2`xQ!@nGIwBy2Q*N;zVl2;b6Wp9w5n4`XpR0QcOpvxSGlpSi)sAjzd#m
zbTvvRqB;-A7;7?%m2t)q|3pYX4Dm@rcO5dpx^>73E~i%{OjW81FI#Mf^K%P_o6)*j
zv#b)yL%%3G3>$Z)3MSwUta97J=dm?86awBq4hYp)0;^Uj<jn~87(1sjPvzDB3=$z+
zw~~m#`yMoc<{M}vAOv`j_F)&oFtk15#ny`#-@iQh9@{F|$<XN7kC++$7g`T*csyaz
zs@;{i(D8>8R~7m@sdZ-;cY>lsT@?p0yv&Fba3dE8RcjkRf^#L6{-z=&A0b%GB<zh3
z;51^EE4SHa!S8x|Krf-f5ex-d%?gREdDu~e+0?mFb3%<X{hS$3_DWR#oTnRxd3ON0
z*l;5?EIMkT2gEhOlPv<}Ofo6v$mpCBcjey2Op|)>LB5}<hTUT#nG+c1;e8A?fkNiM
zEHkXE-mnU?SfWwvnSSepxgnOg&aeKa+j(#B1I{`P-!T9&C@7rxX-JcmU1tp~p(YGn
z9#smj;ErTq5Zl1AW(<)PP2fn2iZY9hz67vDjkxUb=zP{sY@4dEVoP;~sZMafqCW5;
zl{K?_Iqaaq1OSD=`H2&t>P`Fx?~8kw*d*<?%n-=Hi(7`$D+RiN%H}^>z*~rdxRy_!
zK!FSD3F84UhI;1?hGSNr77T-#Hoc;9wENhGpuL8<mG~iTAjw;R`IS!LZR>OwksI^#
z`_ugEtycr#3vj-Aq=Nz0ffoyJV`BvyTg9pVHL#N$ROuiMSnn<P8rQgh$tdoSZoC6`
zpTuJzofWqFaQLX$7|QwK3)s@~CDdQ*Tp}O_gW)MQ%>z6B1~eKyC2>}DR%TbEK^gFS
zVY-lR95-u<RVhc;8$=yQB}hAx%`_+)+558nbC(;F2i-pWA8)Z@@cu8nK-2IljAYl6
zB2X8cX%jAVYqgb|Lvj{YGj8_w$<d*MDaAt>(QJ8yEWBA8QEDK2Q=g<G0sz51^wT@w
z(kEHedu^-|M$Kbv6awf7B2Drp;EqJQ*w9kPg!a+gYb}=Du3?Rpzec!{FB(r!SL4^Y
zPe0YhMTYAHV4H|ZoKgtxOB0XCZ`pgEZ{~JZ3#u0+3U1q#fCh{viV%NU+bk~f5fmid
z13L<E0D!%qXFZ)}D2vb1z6vyTOd)RqAxdBXqQOq$GqOd1WCLDUY^|(FV5l}CQ20C^
zbT1K^){)jiIsq1L^S=3$F$22-DFfS}W*Lf4?a0WTln#I*1S^ZJmI$nBPe`I>(-eB=
z;aoTjf#I-h;6X8zb@2_3NlH<OT4gl$tcVS~$sc8_W9njs^R9`%PqMg)V8V!eOIK2p
zCg%dk%5gbCK`gmsnIN{vGPdA`g*X-quVOKi47OsqLp-!++qkj`HdJU{<C^1scj8yl
zEY!RfdSw}RdU2B6hp9UbVJ{h!ot6}8JMXvOaMaTsUyuuVUDOam>cAZ?-o5{;c2@Wd
zv0FZ{90m@O6DS~;r=3b(f3dx@ak902fE{nrLv-tuKug`GrXvAdEJZ=erjwXzQDU=d
zHUSVFbv|U1ux*>twCVh+Qjn`4REGsX@lxl?72wE}pM0{ZhvA6m$vbc$R774=M0u2{
zXJ{dkL8XH53ALLz=S~aRY?s>n=9cMia*S^6?d<IRaB}$LO9a*Z=@1NlD;YUBzXJg!
zEH)DX9d7PyCG?}Zug;w_V`ws+SljFi^eU5^kY5;5MtEAhOn9M#R3RY)fM}|O*Px!8
z;><c~outr`CW0$Sfk139vYucu4Dy4P@(8!V^R6pM>zenNOSMlxQenRX>TM_4HK2e*
zkn{4h3`0bZyQyx`>7t7I<{GJE)Ntmo3Ldw*i!UtL5yE}CjtXc&sSg<5!6w9!Y|t-4
z;+QB1KpSrFFwB3-Y(OsjU2&T^VaiKwZ@hMjtn!Er=h|!}h~?6ZXGq%dmAQy*^%ITz
z(fv6(z3@5PrTvUr(HSTF1nE33+S(qpCAqfPPknOwmI_?8^GLP0)@o}G^=kWi%u;jG
zo5(?$RY@F`HsujihwCp81bu%_L}$=2%0p(xEUX6-8~~!x6R=bfn1dWzCbp2}9Loic
z60}EW7swn+Wl2a7KD<NuzJ@!J{kw)4i29t%Yi&xl@Cc}Dk<x&?#r<`rS;{-5^~<X*
zvJC>U0;B<y2yf_4SX;HXiB@GZ0T3e0#>B~KvCHHg%i3I9DO}u(6e~y-rHu(Ly+r!1
zo@QZp5^CJbJGxPCWftS5<|{4=RlzG68?dF@!)TE{9REH=p7AFIZ2&?>H94Sz(F&3n
z#-?nAd1-o`#mmVFPIh!C0nE{hdZ?-C2Z=`uXtSn3$G?0k%Bxxzia?2h(<<ya3mBP0
z6BhGiC8Fl*01BRd7o0Q_Zs%HYp<G3k2IL^2YJ&-WdCbh0Uw$bHFVKa-<G-T($g;#=
zz`RgLC3q!Y08lRzq%6*WGxf0YE74Rx;1CZG7X=9!rK^IzRk7{SgsPDq?2_5?);;4f
zBLOH62ChJvCER7Ar(q*f#M~kM$BF%x?015D6L3?hgh4T|OWTbooMk{U@KguHcB40z
zPlDpqD-m7_h2(t6DI*#MTI^Jv46ZNXb$Y{-%5bFeQ#m#8;q{fct)Md7fghfw5jf)c
zm@z#)upoh9IifVi`Iv4&bCXA+9<O}1^60Ugt9jSc^E0~nF2`qhn>uXDx(lb_Mm8ex
zn7M#M83{*d1>EHwvFvyVT<T6~4|w;J!l061dFaj)$(bnnNFQg4{ScD|MsWLP@090{
z=$;fiSqTLpV^Ed|(-`Ve_CZCHi##bJQ)hfx0k>gI@bL;SG9xIixnma9#CbzO3;Oet
zOPKK}SRTvDU_OdipKu$v@g-JRp=0laY1YHA%qq@G%NT!~TZe+FXRY)ym4BV6gM`Aq
z^1n{h{p&>CzfRPJ1Cj50q7K50Co&FiE^!~t&GEVm!Nju=kE7-1UxE@OKlBvK4(=?s
zlm1u`Tn>x)d=ej3Lqpg4SJwWsb8@Cb@iKICic(mL#7^EVgx|<Ih_Ez?1Ev-t7Boe+
zEKY7wdug@g&nfY_7;SN_k%NW|A6__*wsxoxG=yB)3g2hYxOk2>XjZCd*`|*xZ=!C6
zCR0r%yc9TWA?M<R5d@X)-uS`zd<7xO_&-+Q0_GK$^u`%@vgJC5)`Qo7ZZH4W!{vWF
zUO8F*>C*>26mBiEhG|I`?J=&V!71LyUwj!}%i84AS9K1vGla|YA8J0Oy;p-eD)P3D
z`;H+@1GrC4bjwdmjY5XA0&3T<OekmChU%MIiXG{>7=25;U^I3vh6Ipx-BA1&-R5)k
zNo-1O>EY*}f6kEy2hB&bv}fI4kNe&3)#9Uv5AQ#r=`$PPoaWzP{f)7$hk!?J5T7_o
z>XASKymJ_7B_cGXxrj`gc7QM3gbNTk&*?~(ik-dnKb>s;2M=OI3it;5F#j)mZ`xhg
zb({%*=dZZ3X#=zfilSsirX?yM2uMgcH9=WclQI@af+QjkU;t2}Vks-3O?T2`DD(t2
z!A_*4GxQJ)OgCn(^@14awT$Tk{(kRAd|`gV`_xla`|NY>xfdWQJ9f9}M7VW^J=9ZG
zyLRop>m1fwCy(kO37@USZ=?KFX+TJn9blV?aJoRzuaKEaDa49KCuSZz+E@X;u&H|c
z-#v-;vnOb7a++pWZ56<t_B5ib7@cptgb%jK2MsDF6k;*z4z#z+uGzNJjZMw%?FZ`4
zBPxUis*3~mo+bRKJ*#-V)|8^~E=B^iOx?Qj+@fn)YTIQyH#P}Wn|Zi7=&)v4d$O&J
zK2~52kB@mK-(<m;UF3U%5<V33Y{*CVeJ(BI3erm8aKI%@u)BJWCF_atjgFgsLqbul
z*ou6j@%yQFaozwK9mJBV^A6yn0>m6jLvru+4xFrUTPCzv6|C`gQ|%|@m{GO2J!g7e
zyI{8D$V8P^ONORg&R$2DvD!=6n;^q*wcGdrAPm*JJ-(gi7*75qbp#IptF*`)A>YI5
zw8=7$w}n24EM>iduKw^SRhk17$~KFEDV|`!#ND(`EUX=03G#iVd_lhd-F^A+;Z)Uh
z19GmP<11LL_DYZx=4dCh1d*)R3RQoVPvvw516)fY-<a<v1YH}mxye;&L-bH$650J4
z)trz$2n;+$1)J{h^6rbG0`>cMLJ+%<MBi}p41u)n`Nb-F7cJ;$K|FpW6JY%|gxKeP
zWLA-HY9N5-m$_(j3Q-i<=YrKjJarK^zEF3h?r=^!vTyoAL;)Xw`0gHl{Q0|Hz?)B1
zLgwu^qL-5%SFOIGqWU-D>US<7RF6Ha9r%)5Zs%2quI;8YE}nWkMc%6+Q8=G3cGBg*
z%w{AAi6NB_6Uei>moeH=FPb(tDwwor0)IS3p=O^$xkkrDCnj9dhq4*FufXlMSKt;u
zs=zNv^ZobkD>&IZkgWr)^h;4S(N!`kd1&{QwaIYfUE$ahRp4XMUjEp9p~n#A425#g
ziVF2XY%hQ8zR(sO;JmFi-bDo`BcRMdWiZ@*;cd)dD5GRlpbeThH{FM<g6@~@WQ4hJ
z#{o%`qr_&@M_VTLc#7shB;sNLJO}2et$xV_fwy<r=DcwnYd{>`Q5+X>63d_WG@h!L
zFCgy2k`Z64QgemdnD|^F+MF47pDxqKbNSZd0RO7^*VW=Q>>!LX#$zlUoNv5%5sUlB
zI(u+NJY21wt%PrTFn@=~+OR9N`#k@<!;{T7v9Q5BV2B?1&mG>;(h-5WhNZ$Co|QP2
zItu4c-ne$5voE=@`!~fT#qU*!qRxAK1{2EyGEZR7U4OyvnT$Xs#Mw*cBS^Hy`Kw&b
ze8Mn}az#qZH=sNck0M9R)yyfOzf;z}Tk%;TTc*=hZfOM1jiVI!;r7iA%<JH@;*L1S
zjyI!xXR#MR4XIx~d{{n4(e7tiCHWcGUd^F^M=9&nKJ)j{OqR>d>x)Jq`e^1^NT0k8
zR!x)SSJaAgmxwl2*0L_DIG`&)N(AHZVaxb|-oU0K&tLG)FWl(rz9cBwqjVhQ9YnLN
z`%xSq@}?Oxk+K?&;N$7_bq#vYR=`<{-+_}$f3<F(UU4_q=bidxF$Slid8V!XGnVB3
zSVg6{kz2I;Z(1~%@|7T1Q9V;GUt*s_fXkt&EhjCl$15cyzx>P2Qxzp5Mqf(>5Ya+e
z`y*od1U)^0de<+Q^A^@H<2Na9l~qAFAR=1Bb(!5}D%{0uKGx<|ONXkmTCGqWtaJ~>
zGWAz`fj*MxvYZBe8<E^Z186)Nj^JhZhE$jOx+{1>54l+n^B%e>E82pMlqrr6%0#dn
z8M|D|=?970*C<7v*|DmNH}DKvPgV3DJe@hg={)VVS%-!rh^}}~)0`9+8Ed)`FXi2U
zw%e`@;5)n(nYVzq>810J;vdPZ@RZVRO%5#}Upd=yTi0*|(Sc?O^}Yk!)s1hnxb)}w
zS+MADCAs8f2R>qna#pAC*&b{vw6nD^*2#@zM@X)(5FY(5ilCfPCu&QXZ1e&2Sg;S6
z(3F0PG{thK;YjCAY+fFaFBaKietY<6m5OTgh``ZhnITZ`4&#P;&bdb3=AT5PT7htq
zWR7~wM^Tj(H+}s4Hp2}8GB1><i5R~SH1R3rxLT9YgiaC7Ah|VK<02fI7fC1h=yIvD
zcv&YwHTltW1FH?!cI1K=6{;W?M9yO?0Wwut;}z)N6_~dytn#94T&mGiQ<YdU!tYQn
zBSfJPWq=1ZoZW14%25k?5=}MWNBN6psl}PgSXUXWY)nIe`DIr4C1#t0u9)aX6`B(5
zG0Ud4fTB(bRjAu1WFN*$ac%roLk|*3b;u}-MH)cQf4&SGI+-`@R<_CJs~myRLWmVG
zV;o-GRgrr=Tx0uRbfgxrm1|^j+8VG7963QN2jqMTS8^MUT)YOKtDb{Ga>{_kF(`<g
zhxky7(3B*ctCCG~$+RMPg!)WyauIyG-UkNuzm1QR!tBV{rI!#Z)@yPX1;Na6&zvp?
z%QICfsCa##_Y4m0-PC<-va3ya4%Vxi9a^9dY4H$$p598@Vs~Nlc(36ICQL}QhlFQF
zV-2<qd&>t}aC!tC0SEPX4;F7<`ruoX1PmF;BL)%&5?=kC-VZU1=&Zfr2uhd5UtYwp
zDTN60)siJuHuqez3dONcUDxuW{^*$1b4hP#{SeRV>;`K<Gle_gkIRI~EOwM0!SbMp
zMQ>k_3cslwm-ElU=6!qK!bs-Vnh^PILvnef*uw!q@}+^z<k_bN*mJxC?i}9X9T6>i
zg+FC8v+0&rg`IL3RN)0;Nb2Dkm8csgzc?N0W(6XUuW}hbvQp(lm+h>W1&Vf|cpROh
zY1VTpo(}eMxIz_A=$#ag&-@A~9(_s|4-*--ouUAxejcTRR$giU{To<oRGXXk*+Ce!
zx&sjPv_8C$)*jU=5f?wWVa;;AuVV?WJ*GLodU_+Mo8Iu}bL&mM)?{W+m0Dv1l}@BE
zG$n%wFGYih(3A%oUv)r@ED%P5V1gjeVFgNFth)B3Yd5JP>l);ZLPOmFc+>uFy!wRf
zjTgmY<g7vE74%|$L0Cu1FG!Z7s+3dNp1h{K!1buSBG9L*MLb`fvWS+fXFYA*y6>Wf
z!;_g*H(6xK+Z^n!&aYE)#7mJJ3H>y3B>NI%9lRbB>U{-d=)vG#&*Z@S3OLOy#ZkIk
zLgYgyuVJH*&XNyefMf|$B|>{?&J-H1q4Yw-K@q1ti1-cQ3lAIvg7k*+^HxrCrI$rg
z=-6_Rvt0?RHbsP$?@7q2;=LT%^w+eP8Lh|_MQ<SAI8axB@8FDpd`jindTUBosyMtE
z2~)MfTqZ0sqpy@BE)GQ@UG@Ka`|y<@f5|b#@NTmjbJoM|nIwpucH3K0<y5O2lE8mE
z7FQAJs4UUUYc$V{rvk#O=UR|qJE0eyeV5^s7IHQP@ri8syw(m1?|sfY(RPR$PgzL7
z*A<fqzj8IF++*Ag;V<Zx2FOv>=TFLhA%9%FPLbE@Pg_tkykK;#cLZ74b>#0jezo2?
znJgFkYrxt|s`X|Y2Dwm`jCeKsb4=o!UV~VtnonyeB0Qf29;)(7)C>c!1RrW}^)&n!
zZ-h87H5&oe#Lt5*LRNnzFIJabL9O1l+=ZQLkEeD<ReLDfAN8CHn?3H7zdP}0KbK;D
zEf!C?&bF&{O?}dA<PIy19)vnC(in%fc6hHjESRR&PR%;lr%W>J&8|nbAVid2lgz9>
z260G&PIVXh@gYu$XVH5=eK})LZhWk($Ior00%fQ}))r+&k7GzZxp(;R+4oO3w;yf8
z=a=FO#~RPP*Lo83Sj1_twu{gw0y~-gcKL0}8DA?$u`nnHuZZTsH;`p9I()bdpGImw
z`aU<^#@;lS8suGH0<8NKAJGyLs^{JpH>FiCQ#gP!F;Rg}0Kc&A8}FP2*W<}R^_Lr)
zn()b~=40(=TTf(4>DYB4@)@hQbIZ86>oQK9I@a2P&uHZ|<XTcl@aU6ehx4sHA9Zwh
zU25#R!q2i{xYf-I$!99w?fI}rQgsSv3zje;q+OSb7Ho4zDg;0=hwK?Ri0#{4(t=xz
zL+jvi17e~c)KRt{sQbpiH<13(-&7RwB33Ezc;QeI(8<w*r-<pgFju_dZMX0*C!>u6
zt%x<<#2R!nPFuX3$W(CqYbYmckkLA<8{|8vwLQAj%*VY3+n!06f-@h9vQ&wZ+B+P2
zL$eb$bMda>BP2lk!QS&oQIy;<$Xb8+8NLLF6~2bnaJNR#Pol^YcWLNKp2qOZh~jaL
zPjrVn;A<xpS0d;6(LHa^4?n|OJ{5Q_7k76f=G#3tBD4KZzMa43cB5az5i`TyuVlo$
z(sN_S5%d3s(dm=KobNEQ)fk)_e|y)ZR3Dw1c3()1;puEt$}eqvf;&`SNB=6AyatX|
zIYU$78&_ZY#^v+pkp<h$Q9M5BRhY24hG-=a`<4zhn9UMzF=yC{9IC|eOYi6iUK{s(
z#U9$bCtJV&tlk;iO7qF`0w!F-^o7IV{q1VN1Pd74=k_ILac)r6?u^!Db2#zqF;+$+
zVng5mi-Y3ntJi-mrjS0_^Xc~<o2NzX&yF}X=V#UTc53o9xL!&2zOZ%0TY=fj@3x+9
zKX&SD8ydt_G28!<oG79f$4r2QPj@IRlX#9EXxMpymZbLQobeom6|Lg38Jt~Kioe>V
zyg-S#j!~;U{)Wj~)oLwd7b@Lx|B4ry+na>9&Qkvy?+2>>6aa|R-^HClY?{GBu6ie1
zW_?>I(kn$9M+s6_m7P#5=1fP=mCW|txQWJOuQYJos{^MId5(B3N5g8?3dVL;ZW+$2
zQ%cGf&roN2r!YgIJ60(#?uaQ%4}HWWK+DcHB1~!$_TyQZI7z^C$EL{ZVTP3_7shn2
zCT>uuB-<v9nbwkU39;@DZI8dEk<~bBtr{dOySo}ZW}B|TBeS*Yc-Gxz##6>=)AFNn
z*jjbU6FbTdOOBz^aw~9>*h5@d*m;&UgxLsAj*2jlk>2YV3|uJpsdlvOBFuOF>UB8c
zy55xI80ET%o}1yBLdS6Y?68K9j(V^nxei~{*ZwU-q@2C(`UvtU=`!`WC-__U=9_4`
z{>v`Hh~J9+aB#9^=`Fo;51XfO!#fSnIKw6^CfU25pb|xx#Ob@JJhV(%S6K?GNETQt
z+Th(tT2Wa7+7U_!{r25<NTF|>p!_)z*kZc)LK2PjM?C(DQ?kg1@rL~8$eaAh|AMKw
zf@Rk1=PP#+r?W^NLUaaF^4$w>td0T3&F~!f>A!1nM7FYH3p)NQYzVPibsf9_;YV@|
zI6#k=WaV6+brG=VWP3yU!ElTu{iG}#ryhJGzeFyhrys0RBG<E1Ps(Q&(yauRsV5{u
zWcG=#R2RkcC;Euw^%&uBzA3o`?pIN~_#jh2N<4}exMK<tZhS$p?(07|-~X-pH{bm9
zg*QJv_r{SA-hAuRj>~XtK4s#&PcL-JuANWiTkPHY-jYWTy!h5W{N4L?NlC+U@rQ-k
zm;atBo_q`^<LU;m866w>5D>-&T4p%JHn)y~3}#4)X&<4ctf|ZnAVDVV0Pm}pdtUf-
zmS1(Yrkp$U0UkEnAXY-2ZaAVb=K8vK8AUE2$Df>JPf{siCz?p@Q}ux@wp2kK1xC=>
zcBTu8nqvy*+JQz{LSlbLQqN(henLw4l%`c+t0N+o^)wU$SE-JGWo}g1q9Z!Euq48n
zd4X9~74FQPilW}oqD=Dhjid0M=c+1gxY(7iNJ(7w{jUyS*`tTQ6tz+9ILDMtAG|0k
zQh)BlYgaz#dW&~{3HpKzm^g8nt1*!txO5%*z_u3SJQ3dP?CZJ+gU$h&n^vDn1$?-%
zCWOJp%M-{&@-hPdisp7lVkt{0fy_Ary+|*EUC}`e>>(W>K9pTbT5922ZzALEL42Pg
zUZ1P`uq*q>p;t4HT+2Qo-%ZSxb1kQw!naUkMR8Yhi>oa0O7#-61?z_F>I#(T)+rm#
zv#hOVOK60iN&CWTpn7_C2siIkbI`Z*@AhCVhdF@0*U^2WGZcYQ>$%&oyVV<u=1BJ0
z8+}Owo_`q!qhDU4*Q#mzWQ6lvQvSEqw^zCI?BS_ut@iQ%{24Lf*P{MJ`I`clL-Fb)
z1kb%AWVE8)%zE9i<6sR5B_TJEzAS@2PLg-v(?}SCvDAmpR)mU~8vLWK-W#&HlBY?~
z0+U&LUxwt+4qu9mYYxRDb6{N?k5uE73xXV%3NRK|3tz*2LMwnXF$pQuW7qB13U-yl
zFI*kAL&sipH``fE4jQ#bIrEdyz-Jt=0z(vEtbp5Yx^N>f)*#AEfa2DP4Sj~?&QKNF
zusfAelY*QDC$X8Tv+o+>TuH`)U!3Al7S*cMZh<R-;g(hHU!HB@A`Dz{Im@1H#V?Le
zH5Y)5CpiYD=4-lcY+Z3<q*&DK<jzot%6Y`#qO!`%&#Gdq{nn459>G}b_k;ss2g7a+
zkJT=QCp0A2@R-(P?#;gAOLIintj}Ru9=q0X<o9~JaP$afXV>PYNsG%dH=$Y2Y;Iug
znB8O7x}I!m`^5<?L_kF;sw^6==oti{p}mODB0|93XlI(=M$St3c}NGATz&iasiU}f
zNChc~7iM1aZA1NM!X=xGkkKye)wU!R++q^NqhzH-vc>7%ld6M7ETN$Yn~kco#x|gB
zJmi#|fRr6AQfioRR@-T1^G_Qqy)<baPey8`G7?yvFf(4Xw-5BT;R{{Z{~q<uO4H+V
z)$^A4eLxU`9Gt?j8FD^Re3+fyM|c*a9CV^~G2t=kjQq=5gRcIj^f|(6<yAabtwao_
zy6>@CN#TKNCHO(GfPIWpbUxJJPO-6CnRcA4N|C9%knI?lY9%J8G6RsB2{BfwBywfw
z?$~)hu$yo%Dpgf-=RzxSM!kCI7v;(nF<U6VxDqLs9?0ST)l8)3c<+_2i^1KgucHn^
z;}~JOfKU>?b<!^ZAh?i7R(}P@kNs;ZF<TpskR#Q&i}teMSiqw;o5iZm35oDT-9d;a
zCUK12jxImXMD3}E0YTswHvrUhJy<p3eX_@{vri~B;W}jSZmq#WEQk9SSSk|JW5Ys)
zY2*mfcLN`^kjxbFq}V9qzFiUTAQY2tnA{8|dXO%nn&9Yrr>mOB9jXobnbez%r?z=p
zw6I$B6=pIi$5)Z^%y2c67x7E19X@)*p(1o*SC%Se-72v-05GKRDt52ZmZY-=Bpu|z
zcP4v6fQjD~PPG=E-Vd=q3me>beh`~aox?eUmMbr^x*(o5a2$D=F%}<ViU#2_X3cu%
zJ*33C*wJ^Xts8mVA#|Dp0Mf`!Q@D_*!{NiB)J&@@YRNAi-HSHCb61F3Fs<yDYG}+R
zqt`oqXs2dM>}tUmK%#a->#1cp)SqlS+lb=?Eyo+*2G8q$fWWq1Z$62SY}(bm;p)*-
z@1DeYi1+2$KMYrY=hW{upOH&AVIJOZYic}=?_kJfd>++IQof<`tG1m!^DI1A%`lO<
z4rfdbtY@zBX=Z!K+`!|}<w(<a@{narh7q{Z&cHD~oBV_N13!Fa|7UOQt3TIut^ekY
z#$)e(@U4A&`8^)dR*4L03JNv`9PqE#UBkz9>>GGk9zcsK?lo{{y;QkJ(jB4an&D2K
zg?gq|u3kF#t+&n{?&wC$`S1sY^Z54x+rZcM=~womGN9s8gjS2;xlXo`WA!Ai2iZ>2
z&meom9`Q2FKM2!=bdT^oJ~UkVja4W7L2FH^cD0aH_crom^vH*)ZT2Rgppu>P9XK_N
zR2lB12xt{<a%xIq))!4MSrYM+WjcHtwoBqTax}o>al&Qp(Wi$Mo{;Y1FmqzxVX(p{
zV?~w^B&EeA<F~z|1~Lo?ef^!MZP1IJu-i#Hy~%D}v-5$kcxt0An=AAlAHTAL^nohR
z1zoM>BY3x?sZdDU=2{JRP^*Y65dxl&I~>+ldryI;@`i`i<k)jDK<Vbf@(W<uivb#Q
zf<jGhB=YO=_TVY8JmH~WE7_iE_|@<~lJ9l&pJKAq`Z}B$saDW#93gt))e88va%qrb
z$arE35(!f^Z^eW&n1R5`O7nrlRh7kyWvIl&>*sQ?N~Tp*5Jydaljrel%0Z~CT0fju
zS$l<Lh~?X=5tjztOSJ;MbUV!()oaF*berp+nST-JXf{BVqmZL#fv8eJ)>5U@LM2|P
zH8(0wjYlPG(7<NP@+5;vrg5%+NhW8zvitUV>5(v5>q1dA+D7JSN-1Rye`ndJ7R&iZ
zYWEF{s;F!uIcNpeXeK(zhY-}WUr`hN5(Y%ihwYbk-^x%!WgC;is<kkk?hC#+vwvN~
z4puhPXtXZ`FCnWp_(jHj@})g^F7Z1Uw*5BV@t^p-AzZDbEKLok?hC%?iB-f<s16qU
z3g2L0)pgik#V{570gQG>)k7IW+Yr&P-%vp?(<CeLDCkbw%k39Cy1OrQ;DbeVe|U~g
z+_%o3Kk$d5-S$M#^TfT#)pHL61a={~pW}K!AH}|9l@UCv7>6afva%{f49tC{4NN*z
z@sPN&z&%@$tAy5M^M)#w98w>;MkVse%XB3WIx|`pMhio8Zdc&6io2_=Y4#b5xPuiP
zYB)q6cgHpz<YPJQ@~&IMYV^GUA8*}lt6o8MVAM`<NUMxn6~gE~skHeS7g*hIhpg{z
zh4p)~QW2Is@WDuU=FGD49AAm5F(ppQPge|8$emMogbV$tasRYOvZ_5+;qg^_FnDy;
z9!Ps^)t*X4lLXGEUoA(`HG`GUxV-AGrO&$VkZ;0YN~e_A<x9!=rX60uYNTxf=N`(0
z-2HnepDxpXh|nkdMS?2qP^lm4IG=Y~VYlzfR?(c(6uW+1G73G%`{;g+w8pz_g`D<5
zq^k0~%hME4d3I;FZ&x|<vg_AzSR<Wz#Vn2e(HNeFKnqb2@4O2@ID~=4|0eNbq%@c_
z4c@iBee%?q=JxM3wY6Ig{A2;y?%@btxF_q#?mUD~!AA!jczOx`fhJ4Apc@|clT5ic
zI~C~Mdv+ox_ns!H%Du${M7ejOuO{F=Cx%fwnXRMpvL5ln^Vs_s`h+vTvTLidv$BCY
z|BeKf<ilV2cSTGqKeK0tNTTyyxs}>p4Btl<wG)m}_H%bhnHZa@+{`=Hm~bqx0aFZq
z3Gbte{Wi`eokKw7qaF;mb?x@I>3eM?FCjftzx+BNd7>J`u3l66S}!_^ubA`Zg+J)y
z-z!}M{Co8Z!}E~+>`XEa?nQH#fSb1R@Zk!9$Dwp~AlwS_PS-3R)vc$&Yxd80B=c5A
zJy~3`_HT-sZ7{+KHf+_~-(cFK9Xxv>Y`uM51D}ZKNx!OjO(&YqzH{oBd{wh>t#jaN
z?<I+hLO<|b6$s;{0(Zn0KElxf>{crrI`B+u2IdiV^>hs!)33H!)BB~nK6`ONJeRZS
zP`*T{`m(3ap2HS;9C0yX|DHPUv0iryH{g78e=KiMZ}&5~zt6uc1PFr}x60>~l_CT8
z<tUC!*uz|;{x#q(0^eaDW*a#IK*|2y@<9YLD~(B+^_qkHS6tK>ce5?>in|F=zkTtN
z6P>5pm~V2Rr)n<s!hp-dLg&pcJ5?iNo2P8jDyu?adXj0<xkQ3R4Rk3vH9tqr<y}j!
z#ssa=(7#_1z`V~v&&4prl9^c5N-4zb=2xym)r_6>iN>s*+h8ocBUb;eM<4s<1Hrz@
z^xL^%tU4DAA>vP;g%FN2(nKyEhr$2^2Lu)o;{#GrMuVPx8`V9)+Mhm(K9N-wsjRht
zPxc%&($TD1mkpOWHj3U}{od}*k2<@f-l`LLPE<;U-(*@=@m)24cNcZ%RgZBL`5*A#
zxN`CI^6HkUQ_>WBZx<D3SuIv4gw&#n>1ng#Bq|dmOUzTOXO)^}?s#s=mUCeDh3Cqq
z`?7ge_`T<>BJzmyYKd+?Wkr?z5*IqJ4!YLQT{>DSZg^e=c1vY!S5W7oA=4p*ukqhV
zhB!xJ<9AdXzFAXzvw~}LI%Z6XKNgyq>$T%J-4!%VM^wAHzo1EN@#RgKW7Qf$Dxp%B
z9~+5e;@^*y1^jlW)6#v$nlqPWO+edVQwC&|sWCAKpFK#u^UXI4e^}$C>H}f5USMAC
z9l*~_+vUnGnjPl$un4<ucKlNHW*g?&c@d^Ic;6<2DWR4V(j9V)IN%EflI?{|&UjtA
zqapUa(QMMxiw3Rot~(n3Ew0KZAU???fZ9>%{=)65Ki|~d+mDlPIP{=DVpYKdsLAg<
zl_rQ!cl48!Jah%=GcmJ3Ll~sn@KhY=NvCgph6=+5?x(I$Hb2*_uu@#LkNPpT&PzOA
zUdg>(^fi@yZ-+KiYemylNhtIjjfivi?a+p-)l)toMSIs}e7yGA`H;%|w?i8$q*+5|
zHRai&YN}c$?p3SFb~1{f#cpSG8LE;qy-Dj~wp6vd@Akk#@LN{MgB7f|l^2Or-qU!>
z7pV1i)gFlzsHndu8mln1>+Sxi^O7_I#fA`H@@1A_!DNJ@`N9t|U^2Skpg1U&rz>u8
zdCGL$cZ1CF=!)zsf0JA7`5wkV<aO-o>BN`CWV3Sn(KC%r&HO-NLSjq&!F@>h%PaE5
zTX^%86?zAJtc@Fk`G(XiHqr-&r<yfO>U3sxKz}$F`J|9DcNW&r48stcIl`sDS#OvQ
zrgDdC+6e?t=K)|E>86JADf~@+=q;&->!&&Qx;##sjU!}FPQ2fHqi_wM8^qxW<nTq@
zve0`OtCaYb7e41<%WV5BB-dZiq!9%z9owA?y@<$R^P<M@efOfY_hs(pv~34Z+g_hx
z>U%kf_>uy#>p{oIkhKZYBJSY%Mt=G%VdQrc*^eQV&uj_2;)yud!gter=*j=gvsN5@
z0(;mCQKjlir>aUy#tqsN@2kI-HZqlJ?n|HkeBx9Y<UkJ&P0z(n09m<;5_?=OBdX9|
zAWz-2+C7GWsaRBUe<rt%or`uQtzOpIz7C1;iX$OeN??c7bOmVCI&_IMqbT?Fb$k-H
zHi08FRv?xszhRtUAQw|zX}jLtf@`fkmwU12kuSkSv7+dfd>|R65x*?x&P(qq#A-M8
zs7^KL@ZpTg6+Z6kg_<E(2h+3{B*&_&nM;l+Ln1ZVVL??;q{N<gw)mbP)uoipQOT4s
z#x|^nh>iFMsSwwV4!Pv>9}oKMk9IP704aRk>4M-u5$H&T6eShlh4_j*=bcPH`WAL!
z0~nkYg+Y!U@o0P92a<LQg+&*lchJA^R)%{O9nM#Mn}ua2OJUYQ<5yqxZuo?=4)ZG0
zio@=vm(?Z-!qrTzVL2$bP08IFZA#7I>%Lz9gHK=DcmBYE7wcd8bnm{e<HQ3%AfDO<
z7XpDV(Z=*WIvn<3A`Y6JAqptU*FoEET)z$xgCC(|88bpT=29xsGc%R2$z{E-#YE*u
z3g?_*MC`sOs^x}i_KzS$D_JxevR#PhaOJaB$7ELk%nw#lk+|ljlTpf0NLRMs8sxJS
z{=N-EI_~Lk<e<9k&Y%Gf$zIs$B(+-Hg6unIE%{ynAN|VY!9lV)WYdXSU+vw@dy;!C
z?q#*t@C~(WmHJh)h(p!h4ONCb;Hif0ssu&nGPd@LQ<h|fJ?37W4-vn~%lipNIE?Pj
zker;DW`lGT2qUkLxFNf+m*9Wnk`e^`uK&%=%Nqn!zC$bHKmh$K&c*8$!9VvZw40Ir
z`c3{_F-%$99Y3zdNC3REM<SH9CCLNjVO1!KX;<)_;oO>4xBVvnE>aJ0!;2`14yism
z*W{dp=XXSlCC5<7TN1rakMH8}r78<sw9G^VQE9`IaZE6_5I(h%!Z4C{*@Oc{gx{gu
z$|*_Wzxj)7C8MJiPHW0LFfSn*;@u5NTx9nk=rs(qBpKV?&y7Mw>RMnw3A_TtdlVBL
z?(G($X<>nYwndADw;{U&{S`O5dQ+mUp){_QSz#w_b<1A^`h3=IH`xv>zwHW`yq$d_
z)<~^ev@F!EN~t$wO)%Wx9x*FcB`e&zOtXMR6MlnPgut6a(b3ByTYi~>mozcH6{H()
z<;Bdrh`f6$xp~u=Dmx#N=(fp36yeER6i4o6<m4N`@-_7#A>&yLB-%FVg*iv^Afz@Z
zoY$$%u8`u~g?yWb@NlPW{5SavMTsHLc-~Pokhj3tXfKz{b<=y7;dWO(lHRGD`$)1c
zSw274@NI1WqC7NO4z@bUg8e+7cY>Rp|1ZfW0rn*SY-X(Uqhx{_j@(Ne5n$l(%Tu4{
z$7|tKE%2`UIywf<Uq&PlODD)}>+$5ho&6U(u1n-BNX6c+o~9dp-4G&NlSu_n0!t~F
zS#imenK5KJudllkL0*e@Vv2#YE%-t_UXm}xBP+5ft33SO!iz8N?V?N~^8lXUJLmRx
zSt8J+Ve$-JpNd}NTt*!><MG5W3SnW|G8_S{tAFdwb3ZtL;DdcS`%4N$WhU)5Xw<!g
z{ca+eUPvWVe0YgiXfH5REn49PkjBaF6umD)f106R%+No`(0da)-OI0?KQFERnoJQp
zyZbwZQTQ-?s6G930_|IJ78lPQ=81EWrgRPn+&}k&mp<SV{8oLcg12^A0iN5T0)BZ-
zqeV&kaW>w$SF`<i4u^c}fv<G?h0r^BNu<k#R>s|kJ)b4T%h8_Xxf+$>9aJTvR9RI+
zLsCShs{PMU72ZKrpNiBbEWyF)6i=uSuBhpD16byG2<(XDodt@)^;dCESCJs#G7xw5
zNYb*fPGO2JCu$*O(z*D}M2kGu5!B-VvR53?w-40qjZf*5L(gy+Cd7#&o2phc@AySX
z`kMAAVex#>k^m_jx-filf2UjGY=ko2`WzaPc!42q7~G0DtmGA0j2D#(V3Z>Sk7ZFq
z98gAV<(}N!RXo<y93!m5r4c!|Xn|MYq(iFL4FK&Y*oPnvW=4k($Gr+3ZSY8P;`&Ih
zdtJYrAf<J7Nemh#YgHF-^hvSgV>Y(r&8TDnWr(6JjA+M(;l6eW<vKesY<X4Wj7eG$
zJ;<$HayfpMlU4y!aHr;xt7N}f#Q`Oi9TiD&Lvv2$>_RG&7+aoXKC4_trA7<9(N%#l
zB7Z4DLJbgb+qnZ6zPTgK$&|R*GOd8;CMuByUR)QX-vy(k?Ws3U=%#?m)wHOGQJj<U
zQ#I@Q05bDSet&I`)R5bNuU_&fK_);|&UNiVf;@f%w!l^Me>EkiYP>8qadE^GM_5B}
z3iU*R<w(4uhJ<(FF8XYkp7sfgPS;08=s@SOFD!C#AtlrY1NBu{Fdwqic*)vQPfrwb
z+(#MYlo<=nM}n2}PJLy=nBr|S?GOuYNrkPdpq@I6R)JwW);Z6k6HAhp5hlAI*?yv{
zHpQ>llo3Wve*oHv!OF7RLC9*#S1TTsBU!OH*_s>HH++9}Bs$(+QEmSX(NV0TmdE!~
ztE8cCUNzpv6keK~va88J6*a`Sn&@T=dRKl|4hi*P_H~ULNb~_5v3g0)OsNT}Y$M`e
zI?C}ZyQO>!1?t<Z#2d&3TBarV{Tsd95$#I^co_4x*g@0k9F=33&7G6I0@||V41_E`
zNc#d4SYi5QgA_f)0@_=LWeE*`>JG?}M~M%?CxzO3BavlTZb$2DUx8b}CRTC~UV7>v
zO1iuHJTw?e)2Sl{K3yW7y_W>G=*nEhAQDyRg&G?i5m`q|7bMH>T5@oCXe`cjcH@}d
zN1Z6Z1}-@AJ3cA?k}y^3NhP&_6HnPto?jKSXU&fjk6HPd#wYJoj#3q>iVQJLSCIh_
z-W&Jwq%C8tZ4pmRqZ^K--WID^Q9h7KmB06*)rKBREJs#~LyRnxn|jN?O(h!|!s`u3
zyw@W32~6?T+{X>!$~Wvr!s$@gS7Gj}oqRjlc%r%OG<KfJ(L%0Why?Y!<@zVegc@xo
zNxTRhtVGY!P4vv`=Nsd_E*jbjL&pVVPzF+3TK4Mk&Ymj+S2OpJ_(ST+zbcxZxnugM
zS1aT&p6=M_b6pq>WwEA3BamGY842w+T{Bu+j0ZgE<%OX7TvN!(#CX^6$uI}f(}7`1
z>HBK9!QWIfJyo*(N9>a8QLs$toG?oA0VX9SZO9JW;XPoNwgZ8;+WhQ`A9j9{V?6Ao
zj6a`iCGc$)omx8Y^a2!EZ7o+wPKRi=n9Uqp)7SYiGWW72UXEYdp`;ui1}3OYB`~LJ
zK{oUGk-V!TQUyQleU)PANp@RLMTb$u-i`~{v3X;F>Ic#DJ_gtzd(n^<gg8o~6PLjJ
zLC#;*PL*m-c<e^hP}JB>=Lgpy_1MM|H3t%ANTo`WF$_bQ7;g9_C;|T^ODc<(XS4{b
z6HiNkm%%`J8Bs|Ij!8j4KLoR~@1zTn%A_C*1M9=aFkF`rD6)0+6)X+?vPNZTDWAgl
zynzRrg~Y;Eem=L#Gyrme&>i;2%I@TlM<RTrfLEg<i0JbnF;SlARo45=VpfjbvCRxY
zlEfzmFLb#OqN$7(5;gF0HACngy{a&~oi!&NyA3sowCh&0%6>xXtKAFEA)8NUF$p1b
z^~(sC>oJCiY*eRZ^~ltwS9hrASXIZ=jPF+!)Y2g*5iW`KR_u-l@^#`y_dwTm4Dv|7
z09L>!b7usywi8~0i05l$mfkaxHuieIFxjv7^q!D!9oijnwKM(I=uY(jBaK{)7&$3x
zFE2!~An{$X&M48<+z<|YN#%n`Y*L!Io_d{07qmn6WYpFWgY?Fm3Uu|$A)V{89ni~<
zBwaKQle&wNF(T|)4GVm*s^WyChhFIhSSZOM5jw`~1Ak?W;HA0)@hlNnbuxJXcVu*Y
zUV~IK63Rt96$Z<4ElMNCFfB1(jZkG7l5jX>&+9d8<1GQ|^b)L@`14iqUV0y;LKk4Z
z3XXe2F|OkK5EpUefw{@6a8%|BIl5Wl___xK!4g@PP}}#K&$P9kI@#WO47-rOeekt|
zza_epckA1SUJ=dByz~ZNI!LYZ5!0)I*;ly8`w+ho9TTv=yXRxf|0H3gj{~X5%_(>{
z!Q5~X-W3Ptq!fRTZe=EHhstRi(l15l4MJoH7DJ;XZrh}?J%!4*Wryz>`jv?#!HQoL
zEo?XvoYl<rgeUP5Wcuw?_eFI3Kj|OnycRbr{lf05OCqi0S=?R%QwiJiTMWu?MtG|a
zCc!jR$Z?{k2>T`j1v;X$94K4#NP0AhA98)lx)(dTH8wwF<(Ps(BzrJVw5wL1Ul)U&
z#1T`kFp`LQpS(|5HYX|Slb*)V>tPu1R|r&6cBbMgrKj1kbNMa^F;jw$WesS=_b__J
z10L)awvGr(_7swaYxmL(W9zwHLpUR9$H9#JlVq|O@!(!sdGd&j0b$sciI=KtC0FJ7
zZ1pKgl9kA3s66q<Dw>1FSPT83AbphseriZJyBu{t`mRCrSp)s50XQOVw@%TpQ;4Ug
zeM0s$A2k4P<4+nyERF+=b+x~2kd{!@PW>-FaR7G&-!<^y9d>Hq0zP68E}WAKA4C`Q
ztMmB8yoTWYI|1CgM#CI6IqWs)#{oNf@JKjobH(r?zVQ(C%YhL?mxG_;70g#+VS=Ow
z1qBGyjHgfP8D>!!sq~+Y@vtVD?XST+R=1#{X7Vp<-?Gyun**DJ8lnYN!q#Y+Fk7<q
zlb=3$e80V|`E=u%#<Qo+2vg}XTgl)@>az@VL?Nb^%_medQr(M7Kc}7Y|L%kXn2;^_
zSI}Q>_LDp3#dq!RKi1iU<Llw-4#_H`2!+dyicNKQCWudR(4oTRQTYj>fV>^5T3vWV
z9U5_j<VGR6-m3{#qP0Yemqb@Cxy#koi9v@ef(XpWWF`5R73I{ZgmLkzNR=F-!QZ_0
zV)*9>Awv*NhMu_jriTKW-aT`s`Q+L5#<up;jc4DH;5$xt>vTT4*VcTzg(f!Iij)1!
zaeduc__KD@7^|DTJ>8!mSvAjs*;pXkC_UDMAxFeqCbhOrr+S>&I|;q<StWC<q|fft
zvCFw0Ug|Sc@E{J|ab)HS-c?<#1naqbGDpG1puT_|5InBrS-Iam+4_BQtxZQC9d~n4
z%_&lb3JsQ48mwTFD3O9&vR~NbC~d>+BJEJE2U0!P_9mfO4I<Ol**4OssEVRx7hJU-
zOW!2VyjXEIG<g`+!N0YINl3ccs%+fRigv{O6EKy`JrLIeIngFr-JqjZM9y1Qtt7?~
zFZ$VcL!+fRcMDTbqWx5=@LV*r5*-qz;d?!PJAXsGh^@9dPl3f4#3dva2k(8;*%anu
zkoBZC=U!5SU!kAk!-T(pi+-@O>c`>j8>sAc^+s#IBWM$&+ZZp2X-W%?OL+V`u`szK
z13nnRar%UQ;9Pd9+m$9+l_yD{UAwYS`yA8^qV7dtOx^lZF}u}%kY(y&=gj^kE`X3h
z<_<Q>NHpP~n2*dY4H^8FaLPr?XU($wVS5t2v0Awd3)$|ZC*#(s<kL~wr$lI^KF2+K
zFX|x{79y5F=0%fJQNs$!gOG)}2(>~v%kAheww)c<@WgN9_(czt*caxKlBHpo*I(|H
zrE7h@9e0taW1yoQyUsco??Wj7SJKCHvJ`y~msAh=z8vM~{7}Z^y4c&zYjENLS7l1v
ztzT#au_+uUu1F&&HbDNVk5lDLU2t$<UZa>1BqDev@rJ)BcA$k<NLK7>1UEeiH--6s
zRH}Fx#^>lNSoq7{O)2&%9r~6*Cndm+y*lo45JTGOE(vXna9Re%?9@5#o6MY*uDx+m
zDxjNS+81;snR|JQo>$CbS;?H|OYM>KkXksk2lNov$I@BbDSUYG==<$-2cs{1`*q3J
z1#V&CEd6{cnHC;+CHp{}=nD5=&E8LCp~3^-hUf?J^+0FjLd)?q*A8?$YaCM|9A_#C
z0UhxW<UnI0xLylt4^NqWQMcHcpv;kTFDyaqms(UI@phOdV__K8Fo}kN53_ag+I3vF
zDQ8EwxboOp`4*OQ^`g5@9MLZ1J*O9w=Va8w|7F(XytNA1)3uhPJet$he@s^9WE9(r
zT5}cz8ziwus9Cz;4<#?LjA;CkSAzfdPr#9kus$7Yd<<2v8g+{`aw-UaFl^a{nuw%z
z;d!cW%2sa|YF3|7wdbkdc9<vK4_nSt0+~AAL7m~CpHdp?lFJfQD)9!-+X0V8%d8V(
z1v2mKEVN}U-9X2OIH1=deizS!h(o{>Sq0=y?z&Wv4J88&T|EV=Ye)AL$$EeFnuOXq
zkR8%$v8k!JWSMVP17$a$+z*2dMM!J}y>|oBKG5sAkhF__V?PfWL0vr+58s4|NH`%L
zjft4djM5v)U^k>H(Rci>srULPeO*_s4ixGy?kgNRc<|MRLkC}et<chOp$lIwJ&o1X
zo{L?beS7wt?(DmUjiZu|AM^C9oqe4bJ}F$mhu5+Aiu^x)olGTm@hbX`k0IkL>>WLy
z6mVR&A20M?kncpu{Jo=a5jE~Xt}1CRxOaKrV|c50_!4;6-+QqOM+7b*C)-6PK*AnS
zye`9Pf1!Rra<{eF8()L3j!G_dc69Fn406wJ6+Q+Oy*CCBHh^KkaXgf<9}7cr{#x4L
zxB1~ztwCNV8~XR4P^{JNXFK;7u%)5v@+bH!X<zpj2xC9OCYO+Ay|3#6*4OsSC$+k}
zE_U|FTj<A2h+e?fcVl$BE~h-%*Q9lPMw$-Z;KQxPqJFvZ@zvgI>7IZ^mv8j-boF2D
zyd*8V)C-oe#^1wwo4g|}*W>97kMzPnmj3o1-m~W{ZgpHh$~N{zQO5adG?nZo0BnS!
zYPb5YN@lnVk|>J@*70Qq2z1mP^iFmx0r;+t?t<j%k@~XTJb?P&X@)P*a`wH(GtGt8
zw!-N%r@q&Etoc~sYmIH7zqY^dUMs!=@$T6Io;cHZ^6dMCQ!RzYlkXRPxAo+){e|Z5
zpFY#v)>b%mW>4#h)5lwzajo@a)A4tY$>*1k;=Pln&K8cho@hObLeHKmNQG8ZYct+!
zDV%6NBZl&9<I&dRt!Llgzo(`3>`5uJ<<yx%V*vr!v#m|<9&bETIQ{OK)2G^+QT;I#
zd$RRp%Nf+td;;OE1E>|(3eDdGQD}Ro@%VA6YER?4X#E*!U!m#L>G#jHzWvVG!aJvq
zAH#ZZ;b=2j)_C-IvsQ&pH63qkJ+Z%VtnozS+tRu-g;OZy%pQ46TUU7R9efr8Uz|pL
zjrhN&vshM^ZZw@bc^2su&+bPr&zudfzSr8;yuZ+RrnOBN(sJh1iT!(oop=MKqY%7z
zvRO+JR+B*i5j-Lcdl#MdwiJ#vHy%e>fahciz5~y4$gu4yv;d}PcYAf4daqp*E1=U9
z2tJZ3U{RUo;4AqWNaWaBU2xmP?eooBI&DUS<BJ!%22zI{Qa5qf5C?@}wP{kt*-zc>
z?Cb08dk3BB?zEk{l><TSWI7A5g<Venp<J1YN8V+)Zcm5}*MHgZrRF!4f3H_EJ|-`R
zRza)f+c?4Zv@ga~Z&x(Rh<htXBBbTJDVwKeaZ<CZ=lTtL&{mdShr#RazIYA0e)Rq>
z8f)h}YLthesYZFB*GP|0zZbYij29`;#b5ja%I?644b5p<C~BY=e!3M!8A00T{;OS=
zLxv0~SloN2hpd<?ntaJDkw5wA8SKgJyczjE`p`Lj;)BAAoN*<-G5vyCZ6sIlpS<qs
zoW(dR&pbuni5}3bUapIPJXE9LAfAD(2$bUE2%>yaxJ)!sM_SSx*>UBe9~Piz;BtN4
zaVb&L!*+B_Uf)j;7QsN@&u^k?eJa}KYzcZ43+0ok3vc2%dnnx+D0M1JsuZzL)2XGT
z3Q6vy@Z2)w!yBgev^SN$WrjO1eAVp9$(b9oCsvPLcO+}V5bC3AEbt}^80zJknMu3&
zl}k!Z@LR+!lErNJkrKZK_jR0+F7%wQ8z@|Zs}F|_?whY9rJJvHB`nXj6Uz}eZ4vKk
zuLQXI+wwu^XqbqyV8^3Jq(kv#2-Dt}uA)TQ7Gv^}s#8i&_c>i?B5c``I&b3KBTm)1
z8YVc!B}g+BxA>}<FnQmjxjHJr7{YazAptz|s6IX`)+_70C`f*_OB^`+nhwFpjeNvL
z<GSlz*ZEntKG_t31jQ9MK9(DY@^8pefD|D4u&hO_z4J<Y|Ml*!0esZB{(Rs0o=?yB
zeR{rUpM2IBGXg}I8F#h~uCP)pLv~Kc=X3YN2X215y$#On+4hsC+M7?DKKni@(D*Z%
zgGy?YbLazIER6YxH4PCU;g>iwU=JhQs9Kq(HymGOu7rw$&*JcA!Bs#K*$8JB3+Swl
zW|)~FT3o+Jsq~kGIH&<*uNn3%-ks-81-9VgPQwwc4O?+JU`hk&{GeB1ZcAIae@_?&
zbZ9O7(Dw>a`nY6(C%;?xY)@v(YPN%t2MI@&F`Tj_)T3i7x~v@%q%Ve)T~fV2qo+M^
zCGL`p0@zZljMeTVB8P_rVrPT949^s&eL9z#PybRJYSsj{X+T$^h3&Gw4Pg#PCwTE#
z<C*s`FNuP`yxjEu+j=>Ke0dqyj-7h14UdOVFRveMJBDi#-NnD)zDurTy;SKiz~Y7P
zOXQ6w@^0i(35goaXkEDh)1$9b$6_Rr-GJ4)4pmky*$+(CWeYn8A2*jTxF~zphB`5b
zvSll%O^Ec%LT*$RnpcOtzxX%_c6>;e)CEcNO~P_&c#{q$&Cwzd$dMw@0dF0yXPt0w
zH~4ENJ1ORXzX#n=>w+H_v78!7y{4?^RSD_Z0BMFEyQYy`!8z!7I%^6M`ypX_eTi?N
zx4ZXa*dAFKJ=G?42^W%OLs-%4y)eMN(Am0+pIpINwiII{eyRAe2{p>VH$ZEYp_08*
z85jC5Rci!<I=rpyO7);d)utwV+hjB9oamw4dKP=*>SRlYQMUAONHy9_1#}!dYzIxw
zZ$~RG@nzZC;T2S>O*}8=x*CKkemg@-VWL&)tYy;iNc<22VyY0AB1#_9xexhfnfO6%
zyvuB5^I>q#L@bEC!<=RiI$E<GOOlFrC~X&QYYInlM5XwNI?;OZBV_BL#PoWu)E^Y4
znJa_a!p4kuqs9fDO)Vs4%dv0Zshv^{)tqG73L0L%6WV|h+i_Zh36P>7euamxuFttT
zWY6sHxFSpTu@-pSRs6G|?w5ZP>i$ir`!}KP--NoKKGc1>xBC-V!|wAZ<ZKyCWipPL
z72n+-d@l)0C_1`&Q838rMV~m@-qhIE+}?r}1sF#0b6T<2_CzCw^Dxj_#C9)M>9A7O
zbLBX~Rd`sZAa#vxO|7kRF5um>Ee)^Bxp5z${k?8{d+Y`*C*G;+>AZo1a@}wSvri#B
zhC??U-OW~uJgr4mJ}ciYKni7?9<WyET&lkA|NDo!{dEfu>-K-`{LQ9g=Wo9L?XT^x
zTfL=p%lVt%Iruf+lAkXpKk<~jAXi@tKaVEQ9!;J-s?RpReE#OqMm&F%&y#G%rDm)d
zdSg~*&=`Edy>4Yv__O*W`8g&(S0;GjC;aap@yZl`|1p1~69u$~g?{J!%|_HNWuRnm
z*Oz&-arA2!@b#4*eCcMO-GRb}G`KJWz0Oh)m-~9JISKA@`U`0gmo)h>PW%u)rInWs
zDSiYU+OHmGws>_B(XvaOa)LI~u(%8vQ$r*szJW$+DHyK3;n?OJbwh{asGIq(By12S
zN{8!?LiAsI)ix5Qip1P#uS9|*A;W90qWLn{06Xz&s%|_IRZPF2vgD}dWX@$J6Dlhi
zo*rL*=j@5&4Oj+jJ&P4a-{u%V6g=wWk%~Lw<5Vpm?q0Tt_#Pdl_yW3E_a0&vy&v~C
zynN`}hd7U8HL9JO&NqI8MF{u0L?F{(xt)0FrPjAko;uSENhep)cbiu5R2BTg?v9I{
zOtvX$GV5`w3EF(H;rz{(mJeQh39qU7X>Z4pXIoo)I|{x!aIHItSa?<9Z(hZ&xD%Rh
z<qdQVe3D22mW*K_n@*j;VFd3ClV!vl5T3d1N_v*dj!1lP47`#n7Eh|7?H_gYb;;Hd
z%tI4N%c%N(Y{c$Bt3==Bw|rL5e_xD#i$(S8Vz2%8W&17DpS`G1&{Y6XE(?_ZT$KM{
zqsBM)Vwb`<_g)r7+uQpgp4E+me53v2j_waJ0l*0Y$z@m1;^L1xeRJ<e`(%4Wq5hkA
z86BzGD;N04-Tw%c`6zezaIQ`KsAiEwpKT6Nk<X+b9+1H;`HEL$e-KU#kJd2GYD3zu
zbzSScie(=}6(E&hrxi@$;N4i(4k6-;fZyT6h#IDcN3;o&Rb4vNu%xR)$7SyfmIDk<
z;RJTq^k44mbK9ES!Ky~*<xXsraWP=zGPPt3zvRjlB!K0_s=f~FuB>*SY~{Id1J7ML
zP&nP)35!Z}qYj%9#FN<C3En`v5!-8PKGTk!7VsV4-l?-b@QGibgPx=bLGM_ytj=g3
zJSOIw$NLMCcpM*)3@~`sndrrw;m7ZaV8o`7GtO{R1`#K$0r4moSxA4%4N8t3kZRmp
z!^nU=9};-eY$!HZRkNJuFPy*r!{gnby>uhjY{I@_&MmBK$5E(Gu4L#)CoQa6=+Sdh
z>W$CjzXPAaChu4}V`+&w_TfJ6*}JDtV{1y>Gj}4{w7^AqM{JXx0pL9<u0ItpLt@{K
z`Z`{2LJ!4K<KPb#5m3}#UXCD<QuR&ewED$yd*~LxArGe~O+Vcw#o5-Nv*FsyxaMP!
zcZl4m(jneshy1Bbhj_WNL%dAG;I-Vq%HfHQezC!7uf@+j_%&INf^7kDEGC+jc!hUb
zlz3Hy+`b7QVVoWmh5qA*`a?-}2>Dz9E-MiF;9jgQcYS~xsOfc-_ZHU3uw3D-I%oI4
zDfdwy7M=IrxOi2vPI3R2*hp}$3;!>#pe1MsHXo2O-Lqw9Kzfev+Uk`RBT(CQyxgO$
z^5)AkuEKWD%4Q$z>AhXw-Ge=GqGacgI$o|yo=#RK7o=sfT8^!C2wq`n$(ow@v!X}Z
zpCC2tm((xalWnQ;Qq)apsC1NVu-l>c>5CF}mH*KaB}(BoMY`gHm+m8pN1S@|Q&6?P
z@RV+d4@|-H!dhW6g_HKU)veh!QzI{}*@~br#{~vO@F$Y$97^kw+J?}U)HLR|%@^~#
zQTJhkap+yRR;5XTy}y?@27WZcdA%E#z$eDClcHjbO1LFABNVM!{<5k&_M6H#<h>wB
zxUXIC68j_Os8DyKrwbq8)89TDKkR+=LZy%L`#JlEKtT`-q-ILI9koRU6mZ+wmBAq1
zj$P5{0{mQOzm(`-J7kw>(WCpqHtj0h@z+DY)i*!vSXpas3|waXBJtRF1LDQ2ofkh0
zy`(z}?g@I4P}!B*nvfpU;AKakhn9rCqsS@(N_Lu(>mzwHaFp+ps8A-vWcO+!j{8o~
zRnGgF48!rWNRlxN2x22O`o+#x=asxAtO?DwVvHSe;b2AVW?+w!Z+MTNqIP6U)5(f#
zH>g9quEBLeQZ&u0ydOmBar<s>Tqe2^m5>N$#j8XsIlq-ELpJ5%f^^o!+-a3+4Za<w
z>CYqqNv6v(TV^V<NS>q^srdr7lY9s*uc0I&mh=KO(+ei4;^}1zjOM#uX887|ln|C1
zdw8x0)x*%Xww-Et{q=8urvU|ihbZ_+WzFNbX4irW)R^^{y!G~ReuYBffTf7p&Qg_9
zoD|ALO8WAvDMr!G&Q!;12vG_l$mp)H2{@!hu!aN?#T@N~j$|0@7CDTj)hD=UE}SHF
z5WT-B&_Ry%g+hRP+Jsaf(cg>)NY^RtP}AwxzkLt`hl_>Rs+P~Lsyor1*QA{+B;Mf`
zf~y@!B8Es8G^+=b?rxdA_FlfMb9MQAgSRy5V|=Yv-^^3?S+AC#OnuX{`1bY_I3RH9
zm?XD8Txh;_ec%&pljk`W%nfs9=FXZR7k)u!_&awn>fA24^-E)yv;PI;;(%Ma&+*{U
zlR)D`j2H%gCFhO!_8WzREW|MnqKR?Z=RJN;H-Tw&X+|ZL$<7uZ=JxuFwF|YEu{H7K
zSAXl(*I#+<)!#~&PMzHg(L^}`$m+az=>&idKjlHA!%_^QX2!G^vFi~Rb{95-_ykW@
z=7L^Pqz0~<43tG7KT%-X2cOL!KlNVonW&6iNZUZKxs=aaHzONQQ@5Od>Keet*LI27
zA-AMN3_Vus!&n$qrLP=VWmxU$?Lnj?8tsEQHCBm{w9oJX7X=FWN%<Z}xf*8IGD!<R
zV6gfI@E3L%yORx(sE-jwF>bpsRhH_^6@?YWf>>es_u&KO`jYMdsZs#<fK^PDwf$7d
z(3M0lT)C*hcHbKo9uF#lE9o(bEwkq^k!M$|jp@P+8;w9Fa}X=JD%&G|;}aiTovGm<
zL7lpueZ0xDE$P^+KjASjS3L=&#VwU!42L<Os`5~wxIPT(?dk=&y$Qv9)gun38&JKZ
zXUPDKonT?U67?>Yd$pi(+p6+E9_1rDs_(*UX}fVjD#WI1*>5PvphWqicDZfbg<b1I
z=;@jJ<n;vX@?EN2jD37_bkS*SSr6s`O@FRI=NH2)Y=M1tZdJ<pu6YoPO%1(uJbvLq
zQ&n_*XvZhaA~0h;a4`)sQ=#wv#6Pj=sP5zBIbk&eSFjsjf$69=K>*q^B<bkQl_%JR
zXwVP-I0r7V7ST-?!G!4ooD{(E-(4yjb&{u&N2<FBum<gt=LqZ3y2rf^N2K)Y#c~-E
zpaHi%bsp^v;WO-MI8rUFwSNy*_y?}H>yB<dS$7@h`Rmd*xC@5so>F_iKKFzA0qiDz
z>C=wua@vCD@n5<r>-*<lym_sApA;c`!P{|=y|23$E0EHTgW0R?-53hAz@FYhcovUc
z>*(sHpN9FGKA<5+zojd_>2CV=6MyDH@1;+wK7oy3omV=nj?g++{C2M5%5Lc19<Bb=
z<;&73kLU8DS`)@~Wjhvb$KjIr5ieH)Yj}*plJP~g8hVWK1VeYW%_q|cu}UI>lmZRP
z`$+*4Gb}7$?iIO{$80@EF5zS%Ak+n&_))taOKK4I8om#uhevl4$)s3X*}E=7B`^sy
zM|RGNN*ruwCH!oa*37a5+-?dnJ5yb!z~GO&F3IQZt4u_~2pSA*wLw|IQ2EqJfVYtz
zaG3BV{j0Go&E?Q|jaYZ;{P5+M4<7ssyLIs6+by5r@1a)?eaHSb@^AC&`oJp(Uul*L
zuN-_;FEqZ|#5Wq7tcYX3)%c1$(efR;(9+n#Ct8|c)>rU@g|{@n;ul`iCtCcKme&rl
zjFx75qNVvYw&NiFv89JvT3U{2OZ?I+yz~JBo^=d*>Iu8Oz#=|+QqK#8H{K|`Ebe~A
zWMG#I9cq+ez;58Pys=~F&Qc7r2f9q+c6e9KdUg3yk#4`Li3GxTFl?4La3kKT;#=Nz
zHL-e`bed%2T)s;n%6<^5!Z@0vhjzXJ@8KrW!{A?##88HGfH4k#Exea{_8?ZbksekH
zJaVM)+P(u)1ncq3-$XU9v}6i|ydn#tH6>GIbE?Rf59O38RTZq@9cz@e3sEy|SxdU~
z*Yit%`B3%ZcUyUSvkDI}y6W#>H3CB-ugzagn<Tdty3@kvW_x#%6-4aQ<wQ1>gGZjd
zdfrBxT1MijYNFDEARbY=(Q}<U5SzXQ3*dee6ax;^tMaL1`MwJd-y)uc=z>ZoAbkU|
zZION9oT+HNs1?4p)ByQwfryK_cwOuTZuFl!^Z{?YZZ{5v8?U5pyvBG!|G8KFs*jro
zZ$eCWsI|GG@N}(Sc&IWd9BQ>g;Z0h&mk~<l$Yv}c19%~`{X02{_cZfdEdL6_pV0;O
zc%fsOc)8{aR|tmN+UjrW++1vP#mm)toWE_g9$ySV_qg!}=UZX<5e<}7PqJYRwB#yH
zNv0Iz58-9*L(1E#3K91$JHB;B_UJHN=(bht;5A;=Wmc;io8@Sa<<^TY$~h<me{hQH
z2F}>2^Cp5#IZ57@kR=4%-pLjI_+D22bMC2xrEj1C(db~BAdfJ?D&wY^Mr3s%T^_Zw
zRz<>?oUWymAYCNIKEY9T9)9!dbvW62pzgaGwsR$M7;7p@v~7>uzF<Q{V6q~b4$sM_
z7xouCmS*wrw0B~eiV&wvrRj;P^N9G<R%apbT~!C?m0S0CEYqXiJRBH*dmf_I-j2-W
zvR3udna=*+8+}OO@DjGTNIqVCZi}xUz-PCvUQY$Y`#OLB26EeDBQLgWk<G@pXbw<?
z5&3SlSt9!V6}v!Fm{jcKb-B>BYIF&{epq(!K~}Vu<KXK@ThC(dC0i2x9=G!!Z9Um|
z=Kai_Xz;1ig-of8<zPI0y~fiz+{T;<)yc(h(AUIdukag9=fu8Hzo)KI=Z76K2|cR6
zFUapE{e4k>AJgC1?z^Y1S$`v!#-6$s{mlvN+xq*8Tz^M@UzOjj`nyYh|Bn9tJ^B5+
z`ujuqEhi+76zaO=_X+)tZ5(^*PU`O-`F%=%_sZ|n`un>4{%898_vQB){f))XJ#}&(
z17$MkAUNRnfLwo9e`9CFp1SYp?~mm7d;0rh`Tc$UeN%o5Qh4_h`TfuJ_aDgb`9D%(
z{?AGMu~C0P>XuPokou8PUy}N<QGZG5wo!jY>L*71H&Xw?sJ|xlCr14ZsXsO9Z%O@2
zqyCQ6zcT8-lls?2{SQ+A#;E^E>fajmzexQ%qyC=MpBXhm>W)#Pq&_zaXQaT1L8Hb=
z{d=Pxkopfs{R650Xw?5E^`DKx87q`BWE75Ifx2rHPFsQcFGfw1x@QzlZQ<UqQM07(
z8#O1?Bg5zsPHTZOR371Y7bwHw(QQ)3`A7ePl=1q}pOP}}KKfUrjGvGG4JqT`qkl)r
zc=qTHDdW<kK~lz-NB@D8apKW`B4wO-G(^h!{^-Auvfe%#CT0D5^j}H+x%T!ErS8!e
zMp4cl{iRWqtVjRND9X~KzcGr^^XTu4qP#r%A4X9^9{n$)C=-uHl%g^}8nau3{m}!X
z2=t@>ZPfo^GbfEAn2)B7B7Bc#jZ(lKDz$*4c6;g;OsN)bkuo(}_%SI{oQ0o|GF4gl
z6H*4)0s&hvz!nJDf&sQbz!nU!1p>BUfGrTP1p{n>fGrqc3j}Pz09znn3kKK%0b4M@
z76{ma0k%NE77VZj0=8g)EfBB;18jkSEf`=61Z=?oTOeQy2G{}tTQI;D2-t!Fwm`rZ
z46p^N?Sdh;K*Sadu>~TwV2CXcv4si6VIBuGfrH_<K>QXAzXjsAVE8Q%zePjrF*&$s
z3b#mwTQr4Rq{1zl!ae@?Qp2L*w@CaJ&A2`OkGyASJ|^cE&A2W8Yu+<7m&ofyGhmO2
z%A#TWn5Zn80bBf>@0rPZOyj+1NIxbHi-z=L@_O+<DJLG&>@FJai&}#r{g}L7G(+^5
zMse}K+I!^gqT#+s+!qb^MXk?pUnK5}hWk&c0gHdFr7UXgf2-6_m7j+DB5_|d+<&SZ
z{Cj(k#(Qzp-usvAyGg-MsU3^s_9V^q;y)Ng1HJfv>^<d~$;9F$OPR4JX?B+k+a+q_
zlIf;VhV9ay%6m&@T8%Pnmx$()VQZ8PRZIVt?-{mBe@4pCTw=?X49z9V(vl51OO&1^
z!*A(7^F146mhO_WVP)waDbv&?YU+|9wnW6146!95wq%Gc5wRshY>Aq>WSY7}=~*&O
zU81Hg8DdLBY{?K?qVy~oMoZMqCBw)l)6^x((voTFBJI+WVYEn1T{4W8h|$uNVzk6z
zbIA}}QpD!$9^YI3Bc+!ASU9n4db><HTQ+Q$skh6fx62$TmkrHjV!Lc;E>p6W&F(JK
z?k=0AF4OKVn{F=C?k=0%U8dDoHf>yH|CUV~m#O{Brhm)Szh$$#%hd2?)5c}mf@MQ<
znP@J%{@rK)49#Vtxol`I6U}8qbD3x^8=A{RbJ@^bCYsBJ<}%S-HZ+%s=Caw{W!l|k
zLvxwfE*mQ4QLWI-=`zjfvYFFG3h1(dR34>6_}Hy0$HT`q9xl>SJT{bmN+cee2L6<K
z_1G|bOw1k|W{>qr!|X9Jdu*6JCT5Qfv&Y2jXV#~mu}?p<KK+b+`kBe(&p3Mg%trB5
z-dnNSSLlGQSnVt1+KO5F6*{0R)|M6WaK+lP%08@ETUOb!6|)Abv<)j}4OZB)6_X}8
zz`v(%#cachw#8&?h1PDxEX4}-YQ-dNg|=_S?8XXvv|@56=>SlsSrN$=0Lr9rg`?w&
z*_0LP+=|JhBpASxHbSgW16NE!S2%L6*m$u*eOxhlT_v_Ff1%pBO6^?v(kN=@ib?M(
zy`q)BvU}9k6_e#vTICfROI9fbD<;{i94}UEWLc$5teBjya=cji?^Y%yWYsxOeOxuZ
ztWqCWjT5WX$5re5D)n*Idb>(}T=o9Zx~`f&u2LUYy+>4!Rqq4!an)N*eO&c+QXf}M
zA6F?^tEP{ul%-X(po-tB+00d<xoY0cD&=L>v~!jCt(rRiOes^xpApSflk%St%{8;2
z>l~lfOjFls-`32Yu5+|nGre6O6q#5vy<O+nwPt#|PTRL;nzPQ4Y|X~eb#i{q^k|*q
z+M4OnI&oMtJzA#`S+h}fov5stR;_c)TeC5Cofxf|j;+&5t{KwnL~PA0_8L8kHPg5?
zYTTM>+!{H!W*WCfjaxH~TcgIUnZ~VA<JL^$)~Io7rg3Z3xHXfkHDa`88n;G`TQfOZ
zqsFb7#;p;fHIuqEVzg!&x271G46ae*)=cBpsBvp1k?YjBHPg6tYTTO1=Q{1~n(5y<
z^>59jb)EXRX8N~I{aZ8HU8nx7nf|R)|JF>B*QtMNrhn_yzcrKVb?V=m>EAl_Z_T89
zomO$}f2wY-(<-i+%&*fbuKm5;qxD|*{!!!By+_ozb?*Z;Zrxi=ja&D2QsdUWWz@KJ
zZwobU-D{`Dt=nj^P6=7J(PEu4v2NP9PAOQov16UMuTNO3iS)W@<2tcjpRs#HbKSIY
zo%n5-{%sJk4by-P@@&IoeuG@vFe%?4Up7pxH>jx_CdnJr%?*>?4Qk_tN$UpnZ^Pts
zgBrJC61hPg+b|j2pjK^|)NN3YHVlUi;;>;jY!HVH!(oFsY#0t3#9_m5*dPuYhQkJN
z*f1P6h{J~Aut6L)42KQkuwgiC5Qh!JVS_kq7!Di6VZ(6PAPyUb!v=BKFdR0B!-nCo
zK^%&PLy_DqnjI`sbBbp9iqw^&*}5XNqiEKxNPQ@pJu6ZJie|xzl=-6BtRkhnX!sS0
zU(xU@62GG1S0sK#!>>sEiiTg2_!SMmBJnF4ensL}H2jLhuW0xciC@w1D-yq=;a4Pn
zMZ>R1{ECKOk@yu2zasG~8h%CMS2X;J#II=h6^UQb@GBC(qTyF0enrEtNc@V1Uy=9~
z4ZkAsD;j=9;#V~Mio|cz@Y|#wZQfFi+oY}A{IOBAZks<biuP>tPmH1k+cf>#r2cK1
z{%unKHckIFDIuGN-zM?fH2gM+-=^WWN&GeqzfIz|Y4~jtzfHq$llW~Kew)N^)9~9Q
zew&8hCh^-e{5FZ-rs20q{5B20P2#s{_-zuuO~Y@K_-z_~o5XL^@Y^JQn>Mf8()tX)
ztuJKk-?X{+mey`)Zv7?i*-U&(+hW*maUQT~{>zrO%#hyt8@^}W%$By(aNqhn-m|I3
zmbTiYV2ghFrg=JB+6R+~E&An~=ILy)M_VQ#Tl56BOvAS*pIas`Tl56BOy9RCty?BN
zTl56BO#8PeyIUqpTl56B%oc1>lDAB<w&)3LnPu3bTyL42ZP63hGCQ$FDc>@w+oC71
zWmaR0GQVXqxJ6H3%RHSeYQUCB<j?5|Y?-I?bLzvE$>+~$&$i6d`8l;?%cS+^v~FAG
z>HM6!vSqUSbK1JnEfd1>1J+TxW$s4#0c$DUGO;W_U_GT<=5mxDu%^;26V&np)>XP?
zZb$h6Yb)I{(Jeo)`poqxKcKFbZkaHbAJE#CZkhX0en9Ol-7@hmKcM|B-7*)X{DAsg
zx@7`ien5*{x@B%i`2jV&bjwtr{D3yPbjw_k@&oF8>6RI_@&kIyrCX*X<p;C^rCX*Z
z<p=bbOSeo>$`5E4O1Df^$`9x@mu{J|lpoMilx~^2lpoM@F5NPPDL<g?DBUuZDL<h1
zT)JgSQ+_~eQo3boQ+_}Xx^&AFr~H8SrF6?wr~H6kbm^8UPx%2YPU)7ZPx%2o>C!Dz
zpz;ISpwf>_g~|{9T=cT^BU7UC16qmFk4%lq4`?S!KQcusKcJ;3{m4|Q{D8Ki^dnQI
z@&j6n(vM7m$`5EEN<T8;FF&AmmnN+K(uB&}gw<D?Q2Cm$@=FscPZOq|r3sav2~)|^
zgv!f=sbOhC<zvEBuQZ|ZFk$Leno#^FOvOqQiuZ)6RcS);oiJ4@O(>odraq+!#c#rl
ze`!MTnlSTUnoxWuYyc=tC>|3w1e7L}{}ZNvr3vNtgy~ypLisyk`c;}xeomM^l_r#*
z6Q)0<30m3Ggy~CZf|j*3Vfs;;Fn*gplqP6FOA{vlr3qTi(uB!-X@ZurG-2{xnxJ(o
zO_)5FCTI~$6DGf<30lF@gvo1Zf|jo|Ve(m;ptUPam^_vyXyHnej{l_MKk4{SD*ls>
z|D@tS>G)47{*#XXq~bs6_)jYSlaBwS;y>y5Pb&VCj{l_MKk4{SD*ls>|D@tS>G)47
z{*#XXq~bs6_)jYSlaBwS;y>y5Pb&VCj{l_MKk4{SD*ls>|D@tS>G)47{*#XXq~bs6
z_)jYSlaBwS;y>y5Pb&VCj{l_MKk4{SD*ls>|D@tS>G)47{*#XXq~bs6_)jYSlaBwS
z;y>y5Pb&UXj{lV6KjrvODgINA|CHiC<@ir2{!@<sl;S_-_)jVRQ;z?X;y>m1PbvOW
zj{lV6KjrvODgINA|CHiCW%$i2d{YkJl)^XV@J%UvQx4yh!Z+pcO(}d+4&Ri*H|6k6
zDST57-;}~P<?u}@d{YkJl)^XV@J%UvQx4yh!Z+pcO(}d+4&Ri*H|6k6DST57-;}~P
z<?u}@eA5o!w8A&-@J%ax(+=OX!Z+>kO)GrU4&SuGH|_9ED}2)q-?YLv?eI-2eA5o!
zw8A&-@J%ax(+=OX!Z+>eJFV(F?dm(N>O1Y~JFV(F?dm(N>O1Y~JFWOnJO0y(|Fq*j
zt@uwn{?m&8wBtXm_)k0j(~AGJ<3FwVPdon8ivP6ZKdtyrJO0y(|Fq*jt@uwn{?m&8
zwBtXm_)k0j(~AGJ<3FwVPdon8ivP6ZKdt!BIDRwgKF*j=T$)jNm~lL3)P0;WzqmA`
z@-gH1&Zzr1W9~s|M&)J3@t#rlamM`P(u~T_jN><>?%Rw-=Snjw4>OMEjJj_#7NIN6
zsC>*gzBB5+%~+JKG^6q|<M_;|do*M2X=z6Bo^iZp)IFLp_p~&l_|G_gGwL4An0s29
zQF)kgJZCtaD$SUCTAES$m~nh&IK?T=n0r{7QM_jyuNh8rN;BpjmSz<H8OLviQ=QU`
zxre10m4_L}bB5EM(u}!>r5TlvS;u!)@tt*iXBFRB$9GoooppR?72jFMcUJM8b$n+P
z-&x0NR`Hs3d}bA&S;u2m@tCzpX=zsVch>QlReWY0pIOCc*72EDd}bY=S;c48@tIY8
zW*whd#b?&>nN@sd9iLgnXU_REr~H|7{>&+V=A1us%AYys&z$mS&iON^{F!t9%qf57
zT>j@&{^wl&=T!dZT>j@&{^y*ZbIQ*-m;X7H|2gOHobq?h<$q4)f6n<mr~ICCdpD=k
zmpSMEobrFp`9G)ppL71tDgWo3|8vU!Ip_bJ@_)|xKd1bkbN<gM|L2_lbISiY=l`7X
z-}Pfo^<&QUV@~yB&h=x~?z?`>+I`oLIn|ekj_0iE(?iE|*6ur=v#L)I9nV?Sr-zQ`
zL&f)@<MUAQdFc2&RD2#fKC`NS4;`Ob)xU?%--pWIhtA)J%HN01-&uwoN)MgCvueK|
zI)5K3KOZ{3W*NdLJ#>D}s{MZ$_{C5~>0#iP+W&{nuZQGE`NuY%>1tj1$JU$=WiDC%
zu{CC%b(eo^V^jGyr{d+?Hg=Y86PxmF4?Wx_Hs#wkmX>c5i}Gy`J>1s1Z`;^fzD+F3
zw>|W5n^=@@TieRF$=mX64?Wx_Z_Brhx8>X9MftYzwtSnsDBrfuly8%_<=fVs@@?|A
zeA{?izD-`0Z=39vZ<AN$+a|l^+YIfLZ=39vZ*%M{|HP-#KcViIKew^DJV;q7f9|1z
zLCQ+`bHlDYNL?;}ZrGIv2g&Zw4ZHFnb-DbxVOJicE|)(y?8<}G<?`n?7L*66%jM5)
zEGQ2$6jA=%Lj{BMjmm@O>y-x?iYO0ys9=!3QhCsPzw#hM5#>Se-ynUb@}T*G<w1rb
z%7Y#%7^E*%9yH&uJV;qD4|?cekg{GLG~Fx@(w3A5O*hMfv?b+1)6McAZAp31bhA9D
z@;7L@SstV<DG!>hEf3O`ln2e$mIoQWEDxHkEf3N+E01{pM^s)%y#FIAuOr_75tY{w
z@BfI(>xlP%MCEnF`#+-cI^z8wQF$Hl{*S1<j(Gn^R9;8C|061|Bi{cJmDdsP|A@-#
zi1&X)<#ojSKcezF;{6{{c^&cokEpzkn6F(P<yco9v3O*8lw)0a#C-4aD95_;h{Yt!
zqqK$P5%a~%qblzs7MCoK-lqJHm~UPl<yco9u~pdeD95_;h}qKeC_{zi5woS`QO+I8
zBW6p>qYM?6N6eO%M>&@$kJwmO9%ZPoJYr*Ad6aXD@`#Of<x#ayBR1BRM>*FhkGQ=W
zRs9@sdo`;1IpX$eRP}Sj?bWF2=ZM>@QPs~8w^yU8pCfLsMpZvY++K~UevY`k8dd!q
zaeFnY`Z?nEYE<=e#O>9n>gR~tt5Ma@5w};Ps-GinuSQirM?-&A|3<BUk2y6dk6QOi
z|DfVCYTbKmWN5u=(`aZd7XZqmp>+)Umq)ENkCmCD)|khtTcc(>A1mueZQAo#IX&tl
z%cyGFs0U$3Rg*@|vMsA1kD6h7ta3c&dOfCcJmz{lrgA&xdON0aI_7#hrgAyvdO4<Y
zIOcjdrgAstdN-zWHs*RZrgAmrdNrn9A9Fn#Q@I&)y%|$E8FM`uQ@I#(y%<wD7;`-s
zQ{2Z~?#C48F_-f(wYy`E^O)j1=5joyc6H3<c1-Q)n9J#y+RZVS%Q3Z+V=jkdDkx(v
zcVnExmd9Mq#yE#9j~STdF~xe!(Hv7W$7~K;9#cCx=Huj;+Ql&+7su2Nj`=t^rgm@4
z$GtI?%rPJ5##9`}9shB~f86mOSNz8v|8d2C-0>e*{Kp;tam9b!@gG<G#~uH1#edxK
zA6NXx9shB~f86mOSNz8v|8d2C-0>e*{Kp;tam9b!@gG<G#~uH1#edxKA6NXx9shB~
zf86mOSNz9aQ^!?P$K7s>b4ji|?s$z8uk9h@=k^f!xjkh3+#Vu7w}*_M+e75%_K@*&
zdx-qp9x{Gz50RhSL&neTA@Xy3$oRQEM1F1$89%p&$j|K|<LCAe`MEu0{M;TQKevaB
zpW8#^=k}2Cb9;#V+#WK1ZV!>4+e60B?IH4Wd&u~?Jw$$P4;eqVhse+EA>-%v5c#=1
zWc=J7B0sl>OrN%gs88ENrcc{L)Tiws)2Hnr>eKd+>C^TQ^=W&^^l5vD`m}x5`F&UU
zeb@PYSNVO{`F&UUeb@PYSNVO{`F&UUeb@PYSNVO{`F&UUeb@PYSNVO{`F&UUeb@PY
zSNVO{`F&UUeb@PYSNVO{`F&UUeb@PYSNVO<`FBtGchC8EPx*Jx`FBtGchC8EPx*Jx
z`FBtGchC8EPx*Jx?9ui;+N14zCePdVRG#md{n@@p`?Gz|?9cW++Mn%vW`DNt(f(}T
zGyAiBkM?K#p4p%6d$d2>_ssrm-=qE69(MeO6~AG}Z&>jgcKn7FzhTF3Sn(Tn{Du|3
zVaIP+@f&vhh84eI$8T8i8+QDL6~AG}Z&>jgcKn7FzhTF3Sn(Tn{Du|3VaIP+@f&u1
z8CHE6c6}LEeHnIr8CHE6cD#ob?_tM#Sn(cqyoVL<VaIz|@g8=(hZXN($9q`u9(KHk
z74Ko2M{N(&Z{8lZdDHeV{p9Uon<th2LGc}S{Du|3`;Omz#qYl3cVF?l@A%zU{O&t`
z_Z7eU=BIDpr(e2#-~95&+THsO&wYjGzQc20;koZ{+*dg6I~?~Fj{6SBeTCz`!*O5X
zxbJY>S2*rF9QPHD`wqu_h2y@%ao^hKaNJio?mHa!6^{D`$79aAw(r|$@|aHO_I(5A
zu~p_?Acvam`y;xQ=cn{Yx9{8H{7=caCy2rFK;#n*Mm|9d)@V*Zo*)KmG=X@67_8Bp
zfILAA)@TCr1Tk2nIRSZs7_8A&pY`ksqN0dzJn>-g6GTNxTTh=LDoWbA`UFu?($?1}
zh>DW7&OSj@l(Y%vlUYtco_H|$$t)H131Yz39(ut~5Cb-vlJNvFV58|4KS2!GXiCZx
z#DI<F1mp=~z(!NCpCAToG$$ZW5Cb-vlJ&%c(NAVM0eRxV=qHHcvi)ZGp3HIr@{fP^
zXQmnW#Yq7E+MF4`C>{7~+Jax47~rqXoAHYZjK8Kq_(i$EUz<DQ7nK@+O{?&WZVCR{
z{29Nf==f`v62CZw$6uR6<JYM40e{VU;uoj!_-p!!UmOtd*DNZ2aVn3$rnC6PK>~lx
z%HkI%A^2<aY5d~Afxo8O_{Awb{@R=xzi4dm*VG-qXrA!b=GFMc0S14~LgN=r7XI4Y
z8oxN`;IEkp{NnT;e{Ft^Uo>R+Yo_BL|LhK@Mz~^gY+Rvn!xb|pxI!(%6`N<{3iS+E
z%*5adH4Rs6u8k|yHC!>XgDccFT(S8!u2A1_#Y_>dP~&jLR32BTbGTwkk1NzVTrsuB
z73v+XnBwCKH4j%z^>Kx|hbyN1xI*p26;pp)q5l2joiDsScfM5q-1)-WbLUIt&z&#4
zJ$JrT{@nS(+jHkj<<FfjyghfmRQ}xg!rOD_OXbg<FT6c>zEu9)`NG?C=S$_!oiEHF
z{o|c4l|OgBuvs>)D1YvJVMYm8ls|XAFtda!%AY%5m|?;d<<Ffj%rxPO^5@PMW}I+E
z`E%zBo0H>;^T!MnuFyh(V`idog*FOT%t+x1trV`9nZgy?DO@o_g)6jFxMHRXS7@tn
z#f%lM&|2Y&nJZkOy}}jq%y5Mk3s-Ec#}(QvTrs1CD{8gwd|~51u5iS}Gv>kJ3P&wm
zF`oceIEvzmc?Gz_!3tL_%z!I0vPw1veAB!CMmN&GezU;r2jAUuqX*gl>Pg50p3Os_
z`|h4M-a;0NJ$tHUOz2Pp-6*E(g8n`#zc1?VCi#6ye;<?Io%*|3eqYw#E%N({{(f72
zU)A65$nP%w-73F-Pk;Z8{QgjX|E~P**5AkF_ci@}LVow?@00SoSAU<9-`Dl`Y5D#8
z`um^B?>_x~Mt=9}?>6~8puf+`?;HC2UHScy{{Ei){#bv%C%<p%@9)d+PxSZu^7{|;
z_dl24_~a``HXMoqg}e!*=Bd>4rqc6N=y_A;c`Ea~sq#D(dEV4`o=S{W4lRWWJa6hd
zPvxCA)t#r}&YRlKQ)%Z-W#_4|^QNxzRMvS@)p;uFys7Ctm2}=zbe;-2Z|XTu<(xOw
zoTp;Wn_A9ODd$Zk=c$nMrjGMe#(7gk<WUg5m>SMg3FnPx^HjijQ@?pC-@GZ>Jk@O8
zlni+kq!d#y<WUgHlnZ$jgfhjNR~$^KkVir8nL^D|opAh{+{JNl+(Q-xQpl<x)Q^lJ
zM#!R|_lOFzC@4i7kVQc$avoU}lp?Q@ML{WY7g-dPB0rHuK`C+&Srn8a&yYnyDRK!}
z6qF)gkVQc$@&#EGlp-gPML{X{9a$8VVsDW}L8(7y|4izTRY7`da)zu5LYZVCtAbD_
zOUSAqlt~Y=DhOrrf~*QcnS>y#f>0(CRP;wC1jwo&w+wb<RS?Q-EV3#HWzasNuky&i
zL{<g4XYe7b0w`or&<@elKo$j3$f_Xseryz#3Rx8N9@Pk06qKUkAd7-h1Poaelp<ir
zqM#H3Lly<42pF;`C`G`KML{V7hAaw75in#?P>O&di-J-F3|SPEB4Eg(pcDZ^76qjU
z7_ulRMZl0nK`8=;EDA~yFl13sihv=Df>HzwSrn8aV927N6ahmP1*JyFK|>5#6@;IL
z7_uq|Wr!iGf>4GS@+b&ph#`-HP=*-tC<tW=M;nZTy{rxgeQ^s(6i6Ydf>5S!NU9)|
z0fsaRLYbqCR0%?vD~kjOLKzG=L@tyu9%&7Px~nzd46;!7jUscAxIk-H%2<s=2J)n_
zdXcO~sserTuX)edxkz>{noV5%d)_lkhZF=-irF@#AP~x|*&;2^qS>g$8QwD{;pjV_
z49Nz_38Wlg?MO%<_lz$~<O@;`=sl%Otd}U(NI9VQl(K2W60JN^4(L7Fb)+0niu^>%
z0i|fWk#ayOT5F^nP>PNbQVu9Z`;3$WO3@-CeFdpSsu2<p=sgNK5)ddw&LaVVQsg`m
z5GX~?BLRU@6mldWP>Mp11O!S^$dQ0RDWZY|1WHlJL+XK1a?pev=?R3NCge*LawHkh
z_lOaa3@A0lCk-(q9gz1Nu{oiTWPnd1>3~os*GM`bl*u)c4hUuA2$Bv6W%wcKfKZ0t
zG7&?P0WFhe2}uT&B0rI2Kq;CfBpFbOW(i3Kl%iQek^!Y?mXKsXDQX9j3@Amjgd_t>
z(JUd!fKo&ZX$6!bMo2556fr_t0i}o$(h4X=jF479DPn}Q0!k4hq!mz#7$L2IQZz(J
zE1(n&5%LHqMe~BiH&SJ+#(`;Ad^3_m!ebi}kXS%&8n=-yKq%uj(gg@*+<r`MBRv3%
zLAn6BXWT}*0HI7<%2+)V(wHwxjY69Cl(Cp7q=A9FF+!O#AvcUrrcB5NBa}h2s10z#
zMU_N`7d8hOU*t(sEeU!FX@2G+H53sqJ}MzFx#@<C@FVtTH>vlCof%0$kO4+(p*kZw
zj8YU8tj>}uW4Ts}HJ@?&zINKcTcogHan+tBmPi+)1=0U{Y#N62F;b3cQ5nm;AdwSB
zAEodh3ye~K%X<c%gzMx{9}!pwl0xuNB#DrF7HmOU7@-W=MM@76M(C4t1b*sc2Q4k)
zfxIUHLcI7hQ&nU`5XyRmED=JPNFqywP&OJOON3BnC~&r3C==IJ$|ug!vkIK8$30|>
zAcYJPLYYD!gM?7#dLx5`P{v_okPyn+g1ity8H14*LMUq)@<Iq@L(mHGM;-{RkLDYB
zAe17)$OEAi(MKKurDz(E2SO=w0(l^mqM<+@2&HJ~kq1I4N&@mgC`IUz2SO<t5#)hT
zih{6e$ggk^KpqIa_ZQNKRa0W*g%HZ5!YK3dkR?LyIbwfB%7#j0iI96HKgbdxl!*_r
zL<nW#gDeq3nfM?}git0v$Pyuxi4Ssj2xVphxjTfiK^3_>gfcUssH~cmK%Nb`XGYmt
zW_G~(Zgv29DC9}A1IR-ml-U8~p%BVUH1bdgWhNSVD1<UIfIJjJnH4}D3ZcvjAP<F5
zX0VZmLMYRE<e?DCv>tgVgfjU=9txpMB9VteD06&}D?})lPsPvV6L~1)o@)*@U^OL?
z$X7w!k-tKobUm693V9iLZ;dt(c^Q<V1|TnkQq%zCWl)M5fV>P!Q3H^dK`Ckg@-iqz
z4M1K7rKkbO%b*lFkGu>@Q3H^dK`HhYc^Q<V1|TnkQbYxL8I+<1ATNVb#0YsAl%fW#
zn+B{Av31jcHF9v>G+>SRt(yj{Q6JV#1J;P<x@o`~HDKK|V2#+Wn+B{=1J+Fg)`;}F
zX}}saVBIuejkvFy2CPv7)=dM}C<W`L0c#uw)=dM}C==^8Y_4&fTsM7Kqf@kQ`hYwl
zz;@lt@Y*d>gT@z53)XGST%&1UH_bsF5qZxvXN~rA-82VzMC6`1r^q8Bl<5)jhzMn_
zD)NX3Wm<(iB0|~diaa7hnU1Y-z+AThVvYK~ZW^~nO<gyQTUY!{<JJ{1)3|lT$TV(U
zQ8A4}9tx??G!A(vgffjo9txpM<JOhCrUC05kC1nQvV=So@}x=m`U6rX*T_R5_e_$J
zhe9ZmUF4w<%A^%}D1<WkT<6HOVG_AP{E&Bo^&t-ho<!aWrN}enoluHgLf#3bsJF;F
zp%gV0c_);jZX)l5Qq)G|oluJUhrAO?QR9$zLMiGP@=hp4t=cdgHi*N9;jlp*HVlUi
z;;>;jY!HVH!(oFsY#0t3#9_m5*dPuYhQkJN*f1P6h{J~Aut6L)42KQkuwgiC5Qh!J
zVS_kq7!Di6VZ(6PAPyUb!v=9Enk^_&bBd<@Me0h?^u0*!D4K>BsSicd=^{0tXj)vP
z%oj~>i<I)B;a4QmMZ>R1{ECKOk@yu2zasG~8h%CMS2X;J#II=h6^UQb@GBC(qTyF0
zenrEtNc@m@g4iMt1>_8QCzK+7$UC7F@k8DTrHCK$PAEnEkat2U;)lEwN)bQgoluJS
zA@77z#1DBVlp=n}JE0WuL*5Cch#&G!C`J5`cS0%RhrAO?5kKUeAcZ^>=n?WxC`HS+
zY1SKgDCC~mYviF2$}BYUPzYuEhddNQnf@UUg;0hc@=yq6_#qF4P=+7!PzYuCArFO6
zh9B}!2xa&o4~0;MAM#KLW%wZvg;0hc@=yq6_#qF4P=+7!PzYuCArFO6h9B}!2xa&o
z4~0;MAM#KLW%wZvg;0hc@=$<6-U;O(oqgn;P>T2=?}Sq0eDW`iBAUoMq4yY$LEZ_a
zh%NF?C`CU6c_)-2(#SiZ6ulYboluInBkzP#oSGu<gi@3O<egB8vjXIuP>M2vyc0^%
zCqmu{QpiIg{WJMQ9txq%(?K2zp-ftlhe9axbdZNaD3e{}p%BVE9ps@9$|M<iD1<Uk
z2YD!jGPy<`3Zcx?K^_XBOv;gmLMZcekcUDjlX>K!5Xw9q<e?DCGyr)hKq2ph_Ky;Y
zyc0^%lSAGKr6`}sJE0UkIpm#CiqeX_6H3vOL*5CcD7&(!hNSGWL5;GRn6zv!78+>*
zNK2vs(pFIki2_Jl;Uy#rAZ@BxLZSfDri>*d3LtIjSVE!z(x#9lBnlvHDp^9J0Me$E
zB_s+UZE9IUq5#sSm?b0%AZ@ByLZSfDrko`t3LtIjSwf-!(x#v#BnlvHDq2FK0Me#3
zB_s+UZHiJtq5#sS93><QAZ-d!LZSfDrUWG<3LtHQFNp$3o3Klw0MaJlr3vb+qy;ek
z)9OlE0Hdk5k`};d>Z+s#Fq--)X#tG>|Ji$+?lzAkU2xA|@tRE=V2Pwe$#&UPSt^Qz
zEOt>MhotOwlPX-007bMwfDM3>*e=axR<oSj+qrXQKWEO!)|ti3>Av$@RLi^Rzc9}e
zADQ`n0Hi)_-$kjrOk}?K5g8E~85tQFDRYux0W2-^kYNEVEpv}y0W2-^?JdFrXc~b6
z2p_@%SX$|iumF};`XelWrIr2&3t(xbKf(f7TIr9l0G3wzBP@WWmHr3|U}>d4!U9-Y
z>5s4gmR9;BEP$nz{s;?TX{A5H0$5t<kFWriR{A3>fTflG2n%3or9Z*~SX$|iumF};
z`XelWrIr2&3t(xbKf(f7TIr9l0G3wzBP@WWmHr3|U}>d4!U9-Y>5s4gmR9;BEP$nz
z{s;?TX{A5H0$5t<kFWriR{A3>fTflG2n(QT1PUPi5f;GGN`Hg}u(Z-2VF4_y^ha0#
zODp{m7QoU<e}o0Fw9+480W7WbM_2$$EBz4`z|u;8gaxp)(*OSr^i_HxE`WWvw9^lf
z0%SZ+KST<Uw9^lf0wnG9L!<ynJN*zTK+;Y>L<*3!(+`mXB<=J=qyR}f{SYZY(oR1_
z3XrtZ50L^S?es&W07*Oj5Gg>?PCrBnkhIeekph@TTmT(EA_Yj=>4!)Gl6LwbQh=nL
zeuxwxX{R3|1xVWIhe!dEcKRVwfTW#%h!h}cryn8(NZRR#NCA>|`XN$)q<#J(Qh=m=
z{vuL<q@BKq6d-AzzlandX`jD{6d-A*KOzN4+Ubu-0g`t5BT|5*o&JavAZe#RA_Yj=
z>5oVOl6LwdQh=nL{)iMHX{SFT1xVWIk4OQMcKRbyfTW%Nh!h}cr#~VENZRR-NCA>|
z`Xf?+q@DhV6u>m%0x10vDL~TB4~P^XX;UNKB2s{)oj(vMK+>j4yhWq{NjtwFQh=mQ
zop_5#0g`t9L8Jgln=#`pA_Yj=`2mpvBy9$bw}=!VY3C0_3XrrJHQpjpfTW#Y5Gg>?
z|GSS*>Ek}BZ{Zv&-%ZVZ`)^7w_ep*GZ~upWH&yrTzbXCPC-v>W{h#{X)ZMp;6d>(6
zf516Z(x!sGg>$B)o!)THl(ea#ZxJa#(oTOw3Xrs^qHhr?K+?_+h!h}cQ%Bz-Qh=nL
zKM*N^X~YFk`l?UtTe#v{TIq|p0G3wz{=3sxePZ7tE`WVk`u@AqSAAmNA})Y^S9&2X
zfTfi_hznq8r3c~ySXyj#ZxI*3(n=r11+cWz2XO%`t@J@$081-<5EsDGN*}}pu(Z+#
zaRDr?^g&zzODlbF?_ATki7xnZ`utbLKkk|9cicRe?@k}wJeRc7=l?VOIel>RT)sPf
zaFblpP9NMPm$Ype-s0xBq<uc)=C`D6+wc}Q$tCUc8#l=%ZCi)8xEU^KpYOOC&NSj4
z*!UEF#67UI!jHHImR9%?_rTH$KjI!(TIr9t2bR|PjkpJv*7=RN2bR|PjkpJv*7=RN
z2bNZTLEHmND}E67z|x8z#67UI&Tqs$u(a|M;vQI9<vHRWSX%iDaStr5{Drs&mR9~k
z+yhH1ei8S;(u!ZiJ+QR$7vdgRTKNlc4=k<xg}4WnR{lcV14}D?5%<8-N?*h^u(Zza
z|LR9OiXMb1(9Z~5Kyo5Xfu$9l2vcBbEAPWaU;-)c!$n{MNjvxmJRoWB9)Sap=GA{`
zkHG*WZQwH)fTRt41_O|^O%DbGkhHbOU;vV~>G4+t1CX>$4+aB}v`r5N1CX>$4+aB}
zv`r5N1CX>$4+aB}v`r5N1CX>$4+aB}v`r5N1CX>$4+aB}w4s~903>b8BZC1jjjR8P
z4+H~X`d<zI3<e-+!#{%oNZQz$!2l#}?95;Qk~Vh!>wlK%%B%m{{(qL~%B%mDR(kxa
zu``1KNPV-#GZ=uRP5ogo07;uYp1}YlZR!w%0Z7_x@(czbX;Y6F3_#LmmuD~lNt?RF
zU;vUf{$elyNgIDL7=Wa0Ix!f4q-{De7=Wa0Ix!f4q>UdL3_#L0o&Nf-!tcELulV>^
z5jVX0Z)p)Xy!vlx5jTH*?fm*$$N$>-6~O@5{%hyg*E)V&{nz>k1|Z*^UtjC^Upv1d
z7=V0tetoUuf9?E=U;y&n`4zzcB<=i)U;vVKenl_<Njty3*73h~enl_<`R@GsTF3v|
z`SrDqA6NekejWc$Chq_GTGVM?{kQL86ynu?ON)BVtN)f3!w|3jTUyj@Uj4VU7>9WE
z-_oLf^Xk8)<*1KW|1B-@iC6zEE%J$1|1B-zo>%`ZE#jV6|1B-zo>%`ZE#jV6|1B-z
zo>%`ZE#m&KKl%K8t@7$8pP#Q)Uj5|r^R>z=T>ZEDDz9+$-_k0te)9SGTICh4{@Ztz
zS3mjue68{dSO4w1$}3#`x3tPDT>ZDS$}3#`x3tQupL~A4R(XZ1|Mp$w)lWV@U#q<O
z$>-lsHeQ=gcsYdZ^OH>{ybdBspFKYrv}SYtD=z=bceAtd^1q~gcHjcPq@DY5fuCty
z$XDoaMPJg+O}L^jY3CkX(U-Jy3$Ex(+PMQ)^d;@wfGhfvcIxAbzNDSnxS}s<r!KDO
zOWLW4EBcal>fwsMq@7y0qAzKu4zB1++Npsn`jU3+<BGndT^izwzN8)dKP&csc4_ys
za@Wr;-F{YX`q`z~&&oYNyY%{5x#efuJpT1(<&K|i^Z3`Fl^bv&-|(%}$Ax@L%jWT~
zxR7sYrPt3cy>LZe;B{&Bv#wr0yL9?lSF4|08vU&7^Rr7IT+x^ME^U5Re!_)(?eC4!
z|BciCjne;()Blaq|BciCjne;()Blaq|BciCjne;()Blaq9~bhiKczn|<Xc+lj|=&h
zR{Fnj`s0c|;eX@wf1~t&<Me-{^nc^@f1~t&<MhWBeW~yCf1~urg?#Hz>5mKfmR9=X
zLcXPSKH);XrFAyrLcXPyUWoB5X-0i!e~9sHX_a1x@oZ^rA2FUSt<noIo-M8Xh8WM5
zR_TQp&z4sHLyTujt2{uAXG<$TBF3|&RX!lbv!#_k5#!m?IzGgBwzTpqVmw<~$BP)x
zmR9~njAu)$wuKnamR5d7jAu*BkrZP*TiWnt^k&p&!Iw#IMtzpF(VtPDC2i81QJ*Dk
z{J^Nsk~ZnhsLzr%{$SK+Nt^U$)MrT>zcA{vq)mD=>a(Pce;D;y(k8tb^;y!!PmKC3
zY1iKWC!J4-@vP%T)Mxqbn-xTTmbCF3qdrU8v@1q^mbA$qMtzpF$sb02mbA$iMtw$_
z0U@M6hKP`~@fSlxNZR;`AtEGg{KF6tk~V%}hzLm=e=tOZq>UdKB0|zee};&Vw9%U(
zA|!3}Wrzq#8$B5!LefS*hKP`~(TgD>ByIFzhzLm=Js2WF(uRMAh>*15n;{}3ZTMw~
z2uT}086rZ`hEIlwkhIBjhKP`~$#aH?khI~OAtIPYKnTS*LPSW~@r@7>l6HI}M1-Ur
z-v|*QX~#E0L`d54jSvx%c6=j5grptc2oWJ^$2USmNZRp@5D}7gd?Q4Jq#fS~5g}>E
zH$p^6+VPDL5lsKDjz5HmkhJ3uAtEI0_(O;YNjv@!B0|!RKZJ;owBrvUA|(BP8U7F=
zLeeHr86rZ`|F?e^dBA`Wif@F7knjJGr)9n~AcTFF`OSb3mX`U<fDo3J`TO7gi_;Gw
zBG}%4arz-dgruE*2oWJ^ryoK@NZRR#5D}7g`XNMwq@8{U5g}=(A3{V(+UbW75t4TL
zAw-0voqh-rA!(-{LPSW~>4y*zl6LwbM1-V$z92+|q<y|1M1-V$z92+|q@CUf5g}=(
zH$p^6+UboD5t4R#BSeIxo!$r$A!(;KLPSW~>5UK(lD6H{e?y1}N!#v;AtEI0^g>7o
zrV#)_>4%UIl6HC_B!r}$UI+;xX{Q%LLP*-yxBrHa5R!KMBL;+|9sh^{A!)}qVn9gR
z@r@V|l6HI}285&?--rPrX~#EWKuFs0jTjJ;c6=iSgrptchyfvK$M^rH_(sGB8^7S2
zaTz4dNDV0eA8rh0qy|YFei*4i(xzW9QiG%oUyRfsY11zlsX@|)KSpYhwCNZB@o(0@
z@joLq$am8({zHxNjLTr~iGIPj43-vtU|a@Ei+;hl43-xDU|a@Ei+=GR|8DvPBQ;2W
zMn6VskhJL+jMN}$qbDOZNZRxZMrx3>%`Zl3khJL+jMN}$qdy}xNZRxZMrx3>@dG0@
zNZRy^|Iot#TyeJz8-5AL;;)eqzeFd%U)#jtmvAxu+BCv1(F^d`6jA&V&c<JxTKFZp
z0sfkDieJL*_-oS*zeHcaUo)@Zmq;x9{Yg=bU;i`5fWKy9!7q_w_-kSrzhvIvubEl!
zOC%fqnyAJv(M9mrHk<e*(hh%3eB+l`JMh=cE%+rf5Pwai<Cj=G@YjR{e#xxFUla5A
zB~}moHNk;jA}R6LL_dCs<pY1s^nzb9oAKAg3x0|91Az!kfk%$aaOBt&LylNLkYh?b
za%9#c$EF!_#0r8On`p=pNq`)iYRC~w2y$$)AxESGa%{RGN30>pu?dG9kr>FaDTf@H
z^T@GDha8#p$gydM9GUmXv5AKqnfb`EsfQey`^d4$ha8#x$g$~%9GU;PXz%!8s0MNi
z!!pPe#dUmzhhZ5!N8w>u2G3D=7?#0v6ds0U@EnDQVHrF};rU+>mcerz9>)(uH7Goe
zABJks9LEnsHE53GhoKrY$MM5Z4VvTlVW<Ypar`h;gXTDX7^*>YOh5q}BB6jE6H&+!
z8HF4ZQpgc0g&Y%8$Pqb(91~Q?5lMv{6II9&S%n-ER>%=)g&Y%C$Psyk91~c`5s8Hy
z6IsX+nS~q^TF4Qt2|2c`AxG9><k-dtIig7+$Cf(e$leG!wmCwMtk%e}<qtXBAALxG
zh8h2NKI&w<7pKRA{>8!epm%ZH?ez{G9*)MtZvSc48f9mQ^3Xlb>Yp5+_m9TiLBH8~
z*&U5X_1fuS6UBO+er<j}`*Eh-Iqvp5_1eRG&9&9tM)UFZc4Mcuklg^B#c!Ki-^;h<
zif?-xdz%f^yUFo+rx`js8uXuchT~?Ze>7+Vy87H`A)6Zy@So%1;57Rt>!0^}041}i
z6lY-|D=U}RpTC{G%JtXopr20jw0qh)>9pFNAv&DgdS&TilLC67Nq~QJ(i$SHhK+!O
z8n*hQ<Ib?rKZ=>X<M_MVZ+H4$VKPwD`^$HPRPPx7(WXn-aO@woMjgfcDX@d`=;O+$
zuc}<?soNiShW%DAA@o)HsIFFOz1QkLJ#Rh582IZ_$n(X^tzjwzI7ju#l;{--R&oZI
zy4mV=f#Fe|cojCE+@_bPs8ekI=%jP>EC5=cd)^vGX7Y9}MMgy{$^Z&zvD0>E+&PNq
z0oRO&N%>8nBmzN+bHwy<t2gRQ9`qF&MIYK)iCk8BFqGNs<4yp%wBL%VrC!cBn<*P*
zo_rR9=wz+ouys+Vw9N8ltVpRPv;q05(+$vGN}xjds`kzX;4KQyey7v!aRLvHk5%~G
zP~u#rmr1Q?3swD&2fe`yk)c=brC6uB-Se}vcWyUdr`+y2R>1D~eB5bLHn)x-k?JC*
zFGtSAs@1(HlYn7c!!a-eaz7oPpqHD=pa1^zuWo($`R}iaFLqqn+GL`1C(#iAyaP8(
zY-UQ&0oK5Wm=a<BUv<n78r2O=?s<50Vp#gXo+mUkrRVNP>v}><)B656q#nKpt%o0}
z>q*T_9(6TufVpLn(R!E9=<%?+oeBLu?Tnk<{_&tbH@-OQp!r&>*Q-rlS@PveQsQJh
zK5Gun$7kn=yLOdDpje^QB1ZF~+a901)4GzcQEpSD4|MeNKEOO_T|l56Jp+cqB)bae
zqpP%1>RXZ1JZbf@pixc_ph5P>T;v#t?tF^YetFfRQX|;y_JtZw-zmf`7B2zsdNnk%
zZtJi|&2KztnC^!WaBCBjpUt(5Ul+2^<_?qJt{gDGo#c}Nxi4bAZy+Y~;=4gJ*8+ft
z_}>x!*T(;-P8`b+UrpXLLI~=PG2Prjuev%rriIH0rFiGgjqqGSXA+ql%a__49SzS9
zt6)o7&%XKQ8}02bcdF&B$b7aU-*5ZgbmD##AS>X8MvJKE`}LWA>$EdEYaPXH``Y02
zY|sb8A8Zc?N1f5=pz-pkbH=?ocF^)q06J!+@M(j)EN0F8IP=QmHhx$yN4?f)lx_5%
zV-M3_9X>rj1yG^Gtn(7HvOUTUCYI=)p7lDcH2M|b+#EdZ9#ueB$p7VGN=-0jS{hBA
zl8R%?yx;2fI_>fpb|8ddwFn<T;<I7*d28Is=FXHy(B9ekVXu3Xg<067g#~NF`O&y;
z9facZXo6NCufBMvJK8-zBDWo%_hd^mGd>v(US$2wi-{~yuQgttVN2L)XDuwj1U@r(
z7P7<haW+2bWKX-#JAKd2Ucdx33LP!gU=(_VxTKxoaKIhx2;=IHkL$HP)I1uTo?;Ei
zW<MRxTJuY_rP+i2LBH{`JI>Z16jriNNA*uf^UVByFnmTUSGzj|P=nzGKh7@9#6e<d
z21VHDCT&~ug>P(?jZXZq2c9o|v~JhE*ZRDe^2Ep%yY7y*WX)gij<_te6~{L4r8sgK
zo)%p>=tB<W{vGr`9m)6wUsH<}2d;%n&^OqUe<R=^2Q+s~+Le`*gA`KQ@aD!R-O=J5
zBQ%KSgVeeOE(FSr`GA)nLx9;_8?5-@GVA^4-5w@a8#o9PEXcNsnJ1Gf+xua=(fqdY
zsIjxU2^1{fo=NgI);Al-yva#fnx`R$Q+jYd-h)#4j<XjFN>Z6<okA1)lvE(ek|aB?
zAh=J(8JUYtZJCQ5tVf+TCOhZ1wUiBxG25~cjI3A`mSE3zwphGVG3}K8_&}Vqtv$sC
zCyz1Vdoof<4t7W1cKV%Rs~5W{@?nAvS^7tv-to#xO5;l2?v6H~p}tqMWRc1?D^c>}
z%;-gTd~^bVt%yqihVnN<J1kIN1bU@<x(Pu7)F6McI)R&3H#ch3$8kUlVWn4+CF3B!
z?AD!|YLENR`hyq!z&4_D&wkpT1<QafMYN=&6K41r;z4vFU2coBI!yP2(wARET>#4T
zpMukP**Q8Nb5*!cWBo8)-bd%h$K98m5$3E~8*59QmmQ$0wshDU%ip8Zw*DQQayiw1
zMP&dM!Ch;rE)8KXW$8qIcorpu{}wKlMvHB>+HGwU+O)J)(bE2t*j8+b%@te2X%n^#
zRENO_#!;xNdmRHA%kQv-g*KvNo_%tu?e@D(%$<5|aMtNJ4>2L22I#(IZq#YPxU=07
zLtN+OS&wV{_Uhh)=5Awqb!T;NYiA*w)WGeTIkc4Yczfn}FzmEsHYP1Yv6(wV7o3kX
z8A1AEchp2xC|S&AzO<w$VIB~YbzmLG7nUc`9KTh7B*JzZ>Fl%5GPVW68=oXidc<~X
zPgc^<Y7X+|X=jLYl9@>e0QNfl=k*#9gB|eQl`P}{R-h6hJrKY&g}^0|Kgoj}P3_<S
z$I%$W`o>OU4JiI0Vq0)Qh#N0yA`JbY=f*$mv4yl@!5>TB#YC+po)?jQnF^7Ve-ncF
zd?LAwYk~e)Kxoa6_(3q8kT^&MVfbG%G(Ij>jB82gwG<i#;ilpt`^#Ul0%6(1yfBMX
zmN}d{{J$X3ZGve2+TiSB*nN63&gw_=+07d_K3}|fWBJSMe(SIc-Ftfg!X9-YSZ206
z!_)3)gsmj{%}zSQ&f!J&6sMEW;TN*wVW-1&oHp{On1<s)*6LqmFvpKjVQ@HZVUq9|
z8LIi&#SBUk)!ArpJbnQU0)?>=7>x!;U2IL-S$l9KTgDc*t{I0h%IZ`ZW_PW|?7Vc+
z?zDO{U91WC=ASYwcqdSgu))S221iL~gOl{Ach1A&@VVDL?ZPO;9#~z_Fq#46PzM(T
z&V>voo899J{N)5#$VBptP8P6HVAI&BAQPr4?4CM!4q<`gxQ^`z7StZpy0b3ema#_Y
zqzv%`+@OC$(TF)OPH=Xe8wpT!d_L@VM<<;&A!`qSGU@T3uuowf7C#>JaH7JYQG@Do
zM5C3NnLT`J9S)vzJLFXE55@ph&@BWIAe5^1X><Y`Wp>yxgn<}9FS5dLFgn`iK=^K}
zmz@oU(qD0yOX&YWBir4&zxVy>P9xjc&9-;8zS~%DtY@>UyGYM2WZ!S>J=l7@m!ZVY
z>Z83MvaS2s>Z2dBKW#i(U&tE&u)WjR-OaXkW;PyfZ*DY@xAACg^YQw|qi?f&sP|}V
zFWcOBxUq*u_qH;2XiaT2Q15>Bu(7lD04=ZH+t}RL`(a_`{>I)Twt0VRCtJ-R!S^=S
z9&fJhWZRE-wzqa0=zbl|KH7M6e+RuZ9yT8BEumNBWsUEU$aWvBZf>%xnbpUDeuwa7
zYg^kt>}-7dU@v>HwYlCv=Dh|WTfGNuUb@1d);3o+9xi0-s}EPdC3HL47FyYvVKId+
z`~E?LS?q5W|G&1kvGs_<SlfEEw}a#YM!B;Ws(!z*+gQj}cQ$s3ko!AZ4;N;LPSil_
zXaw~hL4dLqqFN9XD1stH*kcUZVae7TtD9&G<av}M-_rY-g<PSSF-yhC*)(+}E>Wt=
zpmuScPdAl`!ceTqr5~n;VrO;H2UW)@x2-ya3W*LZR$R4V0O?3SkZDYtj>)2K)51YN
zRviDHR)+xB{JU8VDi^gu>JIn~C508Z;M8hFk}Ub^7}KuT9z6K&Ve>oi5Xh!lh>7!h
zSnM8bH}7pcTHX0U7u)TSo?nutRb9$A8w;TlfQxCSy8WgqiOk+z-QIwhk|XNcQLjrK
z5q}m(ogv(An8MR8?Zc;pj&(;(O6YoQF<pdWZUP&Z#yK<sq^5&X{lVAHPYUZB%BKy>
z!Hta6mpc@Y^rY4st2<2tM-j_{(>o!Mx>6Q&Ico<i`Bjw`st2OL)z(%HUbarp&JNHx
zRGr)yR0Aop^x_@iPvRZu_aiH|m@sv5x)k`7dwuYotgAVlR1}pd>=$b(_=}4K15k?w
zN*p>rh|)&Pkb^{13msb3e;y9=;|&EA%kN?1#tfQnzX`UP(jF~HC``g2q|D|Nc6OEq
z`og~=rc=Q*hf{*?84vJ}Uc}QE--Yy_Q##2Zj2lvr@g+{!t;9gA^$C4M02TRxjX=b0
zmqjY&5lJFEBJ{9%1_j*r0QY6DXgnoT%E}Nh(qTGHIoScgxN~}jJw|<WG2-^A$+T=*
zpqbS~I$*nlqD>r-(Vi>^ajFLcIf^9!JM27#F4Y+}M<?gwHum>TH$&HJ=lx!{|BM2Y
z4WY2W36m!!s?p`~xZ4}8taL_4t+URm@waxv3fhk$J&cVTj8CdJ7?fzlgM4Ry<l~u)
z|7TdL5t9IbNw!#N1LH^K0sI%6qcCLrFe3$i8MOf@n!&#!V`1+N8sF>oTf++xn8~t-
zZ>dFEK%LkQeM)c^20iW&hp~nZHbzp*aJk>I1rF3qTNVz9CjFwJqYlFiEfCa3DRM+5
zg%H35a)xtby&DU~#U+9x!+$?3?soUAGxBNNqT35RbTXaGtCo~9Sz?ZnjP%sY@jiFP
z=71et`v4b6;$~Pm>Ai<dV28&IvU2CYp>)VN><x~dL3dUqf?o<Y$(+cLrkxsCkg#zX
zoDYw%Yt%0yN_AzB3^c1>K<^-gQIC5Pdc#)N3b5zvg1wFIQZ1CKYiskds|iec)EmGN
zl;Fp`umY>-6!QiU%TXL_VL`Nz-MTIzUgrZbiY=&|6S-K(PFkZ~I2CkG*TByAI-S0d
za(I4*ouhyOf6mcZPA7mUIUf=M#axO<%x2RVu%G1<cA@xw4JWE2xP;c8h-RZgl!Ah|
z5T}!>vXFN;dO}A6LI!~d;4)4&JERgeDvjg3FI*xZa~ghPxG>%!x)e^C)tBQIFm_s}
zO$@0iO$aLxvFdrPVm*qsgrC?|W$mPlH{#3^IXo(2)#H3I?2eT`WR!;VghTdkJV_Kg
zj#fF8Y@R>>Y0jB9$Hga1(76{Om9i3>6iTzpWG(EY^dG}ed*H1496eEA;D5vuM>!Gk
zFX#miN0N58&oXQ}H<)Dp_>kr}c&vQr_E_<t^|=a&!8)k}fXkVvsID+>zfUR<Zfzb_
zuL6y+y&|uSE-@3a9TOm&uP|oRZ{Ox{<)ufC16&4ktH1$Mgd82LJ$U@+PtDzpe{2{e
z(A{NfwOiws<Khq%p9ys*BZcLO2mVcl&g7)grd?FxPZ|R=Tw(OWEuF!!V}Rq9l|LaC
zz;ZoF5T@I=;u<RGp)TQ=NLo-4>u7P2s0Bn2D)-t0o2C3H@;{k-5t-HIL^WAI!s?34
zAMPlUZF8MWS5Q5|jBEfTEq{6ItIs(H`LDSALB&{y%0^9jFqBR7#B%qc`XwbQ8Y4rl
z_S@f!oSAdk2Y!*ohAb<N<da~5{c#JHe_a~NpxW|A#(j_7$$_8>T~FVmm@?}i`gY`T
zzxz@e2)mJ-Qo}t-P&vlqi-)i<pox%mR?dRqlFFJM+q3L6y~LUXCaop>vUv|XcG6Iy
zmRcQPy?!_t;1bSU0L)oap-M^wIITpk?W9~TpTJifKAY^4HM2apHci5GYe^tZ2z(V*
zHKn-<R1uSu1iHeI^LVCIuU&6<o?qt&F{Q{^TTU9ys7S7H_Ol3%pKxRnTPh?VA)z0M
zq*m!*C@u*DZ<pIAwkvR;aOlJOt@{3gFkJS%1c{N7lFxX0Wu-gX9$)Oj(H^#9cVDR<
zJ|K&oqN#F9JFKpqBCmi@4hB28>RfHJLa;zene+kCbA3*%az$B$_Kbmo7Kl}kx{PJ3
zvVd;*@^XtCiGtoEs!*~vga?_2x_X#d5$X+(invgKB=4jN1xbq2vp~zejfV}`n42)o
zufgsN%kuBC<;It=Wvb%D)sl?q%ItJyT_#L22n<t6lq0iXKv+I1!9&r-IA7R?tkJ~N
zzI$3rGth;3e2bkk%0YG|J>m3-)?J=4%T^92D<u_2YgK4C0+4p~B!fA^G>lg`!NEie
zTA^zu#5{Jp+$QO<T~v*Vjc1G5GL&J`U9f0sR2<q9R~t+_NWLh}L@H=sB)BLuf+ph0
zbq&b~9Fw5M1|>~F#2C*<0#r<ruzbcj5+uK<xa<sjbl=g%Cg_r#x*Z;2FxV*=uLb8Y
zUvwz44Imo7v^Hp~ZjAIp+#BOg6<4d61<4~>zHB9*bPZgOUUh`=S6Sqz>jGD^SDL}{
zdstisFGkzM@fDN=3i`Z(BPa{}^Z@6V6AO^l?p=(N`ZGqVm6hGSb)GzZa|c_fTyA4$
z$8yQjp|jFTK1~hD6;8RKY7NGOcJ|BF^BGlD$3=)5`2>|Nh;oCmA%;;ST{+rN0Hc@N
z*>mI5vo&741a!7tX4ZO_QoL+H#ZnOL3=~l3=J^B);5Z!ybnfxhu;Eu7f^|gingC0u
zqo?yc%FgPa#nv!V+$>AXBP@oC4Iqwe_2#<h@5b%h*|lqPU6B~UVMV5Kkh&zQxGb=c
zJ2wPZEx4K<jE<X}Y|T$^-28{7Pe)KWY{;Y_{zv~5bW3?f*7M8cYU-7plbV0cmdk!<
zMeg>`JM@o)k${)uRKb!;&~IKIx(~@GbK*`D+=JtwwaZ8mu8s+LtW&pc3$@TPMx%FL
zadMP`LO|$e05laSg0W38ng&#@iXnY4goRfe9dKBG-fL$#jprde)kk_u!a0Du8?mht
zt1we#aB+ZXou10Xt+0bSK<fH?!nMYi;bd~}%WBAlFORz+aF>ouk<7MoMh=l6)`paK
zD(Nbc`z%x{=c8*O7(WN2z(gcs-T26QzKMWg=M>Hn_bCy%_tW*MxMe7ARIvuh3ag{a
zNf?<4QD+Ybu049ZNrn69BkBE=Qo)3&>XPf}oTHO;dxN4%txrEq!-U}~uET~&A^Ic?
za@n}T?i`I;YnN%nK_M<%1C=&z0U$wySYjoaZuSczdxC;Xv;<RoGHhEh+12@|*Xf+O
z`T{55YE@YfjQA2ku*M1E%P-B0dZP?NmdMo)!eDM0Q)4gKvv9ViMHjNgWeQwTz{0{E
zwL8H}g^g}PjqIA3z~u%y0l*k?I&f5h72%?P5?og}lR*^J&iG{%!OJ<U5&TKaRiS*^
z5OleM3S1`Eq{hq;08Vs)V>z-on%`zMwN6BFN=Ixt?5hj`WJ9{oq1l1|O!$0sy|}39
zas}qS#W>gPkSk~rIoSw=RB`J_lmXIDmU5eHOy|~e`AE$C(N~I4g~-XFC#Raer!o#j
z@2dLsr0Zv3QRqz`3<qbSlWx@&eATaB&%{mP+SWVQrz?K$`pMN>C?4ZpR#SEuYRwFW
z->cOLM$D@fPj@t+rTsJa3{`!=kZN$2-9=V|?gOP*vqVSFYEX`lWqG@Mt9y@kn>&rw
z^&c?xYLqA!p@x0MpB*|9;=7zrciP1gdySokaJ#ZXIPg2g!8|NlX-Du)L<=_@8s8Vi
zYd*+PsCma{h8~-UeNtX3(8f~k7LcFa-;SH8BfXFY9|>4}!`2FJCrJvf2T4k1eUg@e
z(j6hm<L*EO2DvNwU~UUZPB-3RNpj7pA<03Us4?%=CRxy{#T8*!u4Sk4Fbu@>+zniy
zAD)VzZc<;B-*wx(qrPpcLTwz=p-fT(8Q_y7qtl0sKI$01ah9W#!)~uTzCe3450wk=
zcAp->#?#J~PKp`KNe+eiIpvy~dQw(QO2KmwA{STZT0O=6RtUvXU8)f;-&GcckvFLb
zQ!2PE;0Ds|e$Xlin;|cjuyM$wPiywT@1rrWfT&DKJ_iX)&fLh*L)O-R`6UfbNP15)
z3i_EF5_xXxPq_EBEI;kem7d%q_on1N?2agu;Y|%^XnH80cYD|jV(&9EvwZVQDFb;v
zpx1>>3>?AtxsOsu6ez|h_<Jb`^2z;BemzE@0$hW^x39br-daX2cxStN8he?^mT!J7
zMWACnJ%=aW%<?UTUL+!q|0Pd{@jsliKKx<Y=ci`?X(#~P($Q@6dYz|mpT+>>JOVD;
zQuUU!zXvat&UM)uoegm624-w&cc065TRrIAO^dqJ{6fCm1FOkSa3z~<$)#NG9Wg)&
z$%Z?&>aI;Ro?`ohJ?G$%uH>`QhGOx4Zvbk-n_T<ml|6AvLz8T}kE@4R`mr57Iqi_@
z*rN_*dTJwz;XfgI99;HYM#B9DD;1sldl&Q4mm>Qyd`cG`xMlgEDBkTnmFXbv`Omup
zFrMM)b47Ws=1tc8T=6c)fE+b7KymiRW;z4d@cL-AF)k^qWKc%VhOI6x{bab%(CW2@
zr&7QuVFTbLf%>rX&*$KNwjBn>!GJM7SM;w@HiR3?p%5-ez(RoFMD47O19dlJw0|e4
z_*~Jx=Ee<dS*)}GqWIiJePd*jhUN*{47xp><zVQF%{_&-C-pN-M$A(AtT?>~*RQ8T
zvU>or1?PlQP?t>(1VK38=Pd-t!1`>rKPwP^p(ww!zS$^%9;pubYxR{fk8y_r%q%;Y
z)^l9?7Pm2A7WY1$b!6^*p>$c>ek|AgGP$FHaqCS_j^zs_4CNO__%EHRB05V-#N`4;
z?DI}M2K>A`9G_zjG8?>|CWq+q47ak|!qQ(TaWU(lxeBVJAu9r)zyMz;nGiTpN7cdA
zb&iOi9Q$pW`ot|rLpHjM<lmy*A?F`v@{9{TRQJpLN@bCUP^T$^dq4mSN+9DhH&gDF
z%fTJF6f|jvQuELho0eSnb=jl}+pyD<J1}~UUz%CqDG9E?;cg4XZ;lZz;Pyt(bdihD
z5M<K3_n>W~5@cXZi0>hyT$La%BA&NI6c-{fmsERc5S2Wf3Et2D*m$(w+<QPx5l+na
z8}OBf6>aK;G})Ewu23n1Jvn};Nr@;W9n_l?%2-kwM%)H&&CxqlH(~}o@;5Z9jBnE7
zhdm`GjJ*1T8*%Nsw5Cr3I@E$<5SM>z=_-0rz9?o|St~-j0>TtOTWh5|@&+2>Nh-Lr
zZv}s7xJI7!usG^X)Ou}%Q)O5Val?`N&`6@#Qgeg3?NGdCn_>qD`8*xUIo+1$U6qRf
z<#C<iOKFTOA64|Qx}+WjfGET+SckbJnitb4Up4KA=ZNv=sy;AC-bq{)L|=~u0Fa-E
z8M<Dyy5i%Q9JWc9MW9d5M`P>&V^c|AA_lHP;AmW_pdcB2x~zZz;8e#6F)ZD1k;6<r
zgH&(pn!UIPvjRL-7O{*8b1h{*49+tGE~gqYdqjbE=2=TMcm&|2;scG5(*w#5t^~GS
zl5m7KbV};v$)R1W+7WaUXoao1y5lKF_$evda@*2ngf39Bfc`KsKqKCpEQVD<HU&U9
zDa6%VI~2fJTRqv|W*8Lfst+oM>{BWiO@#6<N+@T&JoO4U@@u+9iIg)lIMf`aSq9A)
z*oOc|7#QAABpsk~;`XO)`<xO6Qq_d9Pl6~87AUf;{0<fQr`;pC38PG14)ud{^WI`r
z?i$)_21Mk3ic!pa_eYE6i{^h!xOByd1#`=g7(u7WUn1QGGez3UT9Gu2-fFnApI!fM
z1%4OT8}i@v?D|?ZTgP>Tq$JP_qeRhY$CVtiQZ{uG$$?4~&OXGB=dnM{;_N29v{kw~
z@aBcZ30<n)xymgJJLiL)F9A@Rnx)}Hq)qOG$>j5X?eO9ZcY<*-1p7UBWG74Kgd#wY
zZrfox490pt0xK+#B??rEk%UkaWE`%AB%HH<iwA3J+YR1aIC2p^B!>W?@<-g9S~50B
za0B9HfB5WHcIywHEoQgw1mCgjv+NZ!Z`=mna&3_=XWjNUbD!MKI!7l1ctBtSd*`!T
zw-Gf7bCJp0kgx5|H*<Hr2$NbycDiV=-628r5PUphaH5v(7W}BN*}T88*=RPgy23+E
zM4f?etc9`jm*{zzT~Rf|vYTLuTB)2&<T+B7>hKLZRQL@QDg2JhxARVNUn19RL6Db)
zX1uu4Yayoj2?P{9OAIQLQ3fha9A!I>84Zn({Xiv367WOHw|T4AK8aQ%!ypF5cs@IZ
zv|-S{NvL6Cg7id~W2eP=p-WIok@GX`vpTy1wR*r)$Z6YhlD(%IWVLQm)E9U<JM9XQ
zHQXJ2v;`{!0Rnn(41?{v^no!Lc^2;!K;k6xrP#_;t;WTniOMuZ=i;s(TfuD=o*N^e
zRNNRb86H29%hk;$yEYV5;++JmUSypjwj^P&mNR*!f^MhYYSqM?vvnU$O<jWSdQS<_
z7@M$){lYO@U?Hj)vXH>RQ1uK5g=yF7#2LYgl8fC}&Z`rJ=00a2Byf?%?Ra(h#tl=%
zlUZQoPE5u7HRz2nf51BSBoX3O1!<;1VU*)m#X3(7o$Nj8gW&MI?iE})7;wQlh`N!3
zp<1n`BraVNwP-3-7Szf)TgekG>aOV**&4EByQ}x1@>LVG!m%???i>PDIIV=+4yj68
z>`ljoXg-4UWXIAT>`In@GQZ1Ys&YWNd0K2c=VTFXOI0(;6f)KHuz>a(m_u;-X&gxX
z)TsvT*=G-a(ma7nU!yLzA+{LgC$iEaIR(jc$^^9;R%v(damCp{!3Q=LB3uCmY;&w@
z7cdIE4hEYgS9yZyE8??=PF2xWh5rC-%2xfukU4QgejJf9Ze)?-fXeX^aXy$<UA&XW
zS!@a{U2F~WCc%wyS&(-3L&U?P_$Oxw+<NI6{ev`i(+q7UUK+|ZdOoFo8tI)gaYx3a
z_5t8;4&G{zH;U&+d2@ux>dDZ<J&p%ER|narU$I|3jDj$qUH}~(<1C=(+Kys_Yd`^y
z3RhHdis;er5kp0C9`)b?o)GjH8-)H52Al6$Z#UxggU++650?%G2fl$~#R*u&V#)E{
zJ5s8EpR4d0Y-8`mXH|eFjQ0LxxGJ{=z{~hgI-kOFrBW4qhfaPezDv3nDpxTdy02!w
z@Zk!B$FQbxU~y7F9yLP1S_n(>ly8&G9)VT6!lWBF%@cOR)7IZTEcCyX#m{h9shsB~
z?&6)YYNd+L6eMbJUqhtm`4UaRTLN{g{~Y3@P!sWG0$=x#>MRhC%lG9cx9GT;e%E{n
zDnB!-kVnb{3k@vZij1K+1Q0O<R2Bv`7;xMdHQ^l0QP5U42wN^M<g3c(xAAqoD{iCg
ztdbYPs+Q9KF4>VV+p#Kr3tWo>)`ndi_yzm2q0xw5rQNHZ%+1Lq7zOs(t0~cmV*QTz
zbRkkcYT@qaYO^0(xJ8`gM%JhrS=dcoYJGVRdGPTF5Wvjizc@pjq4o4Wzf|YV0t8Ln
z!Aksm1$NN(i`-Cz2FyS|fDdAZs{sT4Oo1r^lI$T(>gBwf+?6ibPs3KD&u_T2H60(C
z?mNilDu%NS+jMZ<Xp`I_vu+hIZ2c=8v?lGYan~%;JDHVNSzxS3Mir6Z3FZ<Oy$e&8
zEMi_*fT&D)@#?*%O_$U=Ipy?|Wh`3DeE3Nv6;kmiQVoX>dFX=t!r@DH4U3(Y@S8Sb
z&rDc`3m?qpx-J9*!SWXt*$mp4*tZG>pI|W3dP#efU_~hY5TFn^UBe4v5LTRcuqGz&
z0ysW!1hf4*0w#{>RF|q7;+Nc6My0`5Bx5eRLv^vjo`MEX9zzGBU{Jw(I(VulTSpy{
zYauD=i(G|x%(h3Gqa#Ja(3hXJ%~bE27P8$&<4;ZaQvghleWDQg)>r$&*Q#xoBY5Hl
z)2C*63y;yHn)gQ-lIr5qha_bt4M{&5=f)u=V1`|ckE=){%=|SyP}!6hC`v|-<hW%T
ztzl70+euGp61pdiQGHF41+I);RRA2$7)rnf5iw=BxnS|7780{tgrQnh$B-Qomrcn~
z4HOPj2_&!9z(i20GjZtJZCRkglP9`JLuRO4WZ#q&@kGEui2JL&#cH};k$nW4VWO8}
z*olo(Tcq2STOA9A#RQ=rii;vEkN~)d%}^jd_^D?Tf(Jd>n!2Mz$Y8!b>%L_41`>~0
zxO-;i*PX#%XiN|C#DYsWRPBXBU%0IjqhpBE7sHRz?nfsSi{OE3cK!28P??H8%C0l~
z2m6Fm1TJ-Ys|EbCK)k$+)I@eGqStP|m!^oc;S;PZCqWfnkDrq`EvRhluRdXZiJBkZ
z0T4Dw5*O4=4#oi~LJz@K@eocjMUou{4uw^F2p1qxTfFTUb)FMS`pNf>@3HqvG?e<M
z?_eS#kP;~<4uxssvE7PM(YvpV3!G)gS|&KEBnNAYBLVLA!b1=^h0cUu07SAGbQilA
z@gIbOVsCOoGlBR*WQ%Yp%m`E#+}A##d+I|*?xabc1xKe5e&9{SXC3e*+(JUTvEn{I
z$^kYk-8c+Dv>z!Y1FXgqM_hrcT}D+CZ2b`}J$;IRpja~pLw(Pww$yAsL~QJ>^=5Oa
zX5L&P$~tiWqd6D8wYWpk$q=AeV$|PU`a+!VU(gR-<{U5kN-*<VOE;FX$Kn}=H>TX#
zaz3;QmwZfn9?#+F^4-QxbG^Z6C%pGVw57umFB)~$#HT3?Ze>anLp(fV%_;1l2(svN
zXy?~O6GVj#;r-AQ>e~RWG-U13t}0^4llAT>j*Dk*(ne~m0VK(ySU&hG2xTD<rrfq3
z|68<K*ANuK$eU2jr9zQU;&3Zie@jFDR)`97b&0jhQwPKG`X(&gKL!Z1iE25NV4=kv
z8&B&16QnWuqErRYQyki+^@D*->j#ogoIMkvJf$NH)z>d8RhYODk_v(t95Az#P2iIU
zTv<H!qf@)sTO4&|bqe<h!>%m<4&$y6!+E!${Dcc;l7iL3Og(YAkr5ZPTQLBXbzIN+
zLuvQE=&Crn9gbO!9T+<YXY9l&!=~pRO6Kmv>OOtJr*G+d{;4jN;Z?w=x^Yy})7&Y8
z$nT+l*^%%8E*VB)F(UoaSWCs2Y@oIk!NvX&p5(%W$s<CuBcZcU5M8n8G>yJo4XheX
zfWD?`ACZgyNcpVZgf($&#1%2~u5=$u2@)4PF3Ty22gQ&+=>lE6?hYZh9-a1Nj$dU&
z#*ns%h&vQuPcJa9tXNPhDH?njeUVTv)yUd~mcq^H%bF=_C22dNqvS?c5K1Uq46!`P
z`ApZ87X1WdKse^dwh}3|Wa92MauToWL@Q-H3yfc)M`&0gw9t|``R4nEuV{>%LfmV^
z><1NX&a}phcM?;Tw1vTltFI!_N+fobx^`J1mK;tt&fiUV2oELtd~^&a#S*~T%By54
z@g^chE0CG9&W50RT*x@ioDKL2hI8pE&v=Q~7Dz$bOT23p*>0&cmcT4<k2@z~c#tz`
zXTO%1)m`ugFPPx-C!Td4Zspw^0F}sUaLJoLrBVH0t6%k6m&oDrsg`j}^4w@Lx;XMe
z&$M03VrAG$>I4S*#POGOkyu}Y+%3<liZgR#f?x?!ZV30N*@el*!j}k+&oK|W!N)kd
zxHGR;*TT)ViO<Ru8U}8nTpk=Erab+A!8X=x9Zw@|`TM2Fo&cXWBR-Yb95Q%1T7#f%
z(j^C5J+MpZj*mH&+cwDWtNRGWa3WGRIx>m0=?r}~8TsDDdbXv`fZ-a_1nXGwymi;y
zfE@r}!aC2*cfn<{^l)gUart(F>Ip#484BHK2!z&;c{7_4b+w!=(KnX-yk4`FFZDNS
znNcL%hw$3j9}uz={1x|3lft3s9^5+3WP`8wu&}RgF^$Uxpf4UC!&_tUWba`2p5dOe
zfQ$Z54?9EOR{2UFhT!w1jFm*(L=mH5G-VeMl!g496$V+n!`hR>PLXlbh02<U@;sVB
ztead+CSAo;fZsB=VPB@a4NwtTa$N2*UScG=_XY!%tk}ZIIDxh{VqynU<zt8EWmW3n
zJnf2sAO9rx3_M&Z7qdxI%_N#G_=yxX>;{3yB`o)1MNWPSV6q@-g*>HXT3D8Y35LxV
zN<{EJUS82aB0TK1V!Fzqnj=|4y`+b!veM)o7rY(vzbG3ysB>M8a&F`|Fl(m37rYAC
za5&=h-T)+nOiWp1_G_Vqln)$7x)=P|t6|DX08wvHU1g9cyol$km5bz-wbk+|fC4Ig
ze39^h7#U>|s}hKqbA=1Ekr8D<qWB1{PwKVMTd{n8KH;lwDZrIhv8)6Xo6wU?<XW6>
ztdT3%uQQ~=_0!f%jj*v&TdFVLxcNB)CF2j*y%&eqxYAiSXauIt4{XF>Edr?SvEUe7
zg(!Zjo;eaEDZn;0E{(yTlu7h%J<mj+a7=V5GfCJ49{@Xar`CNk9J~nkTYY&KJxPuO
za5({O8wAiQ<gP<9opZHwt6b4SxMq%ActMv&WWf}Uk8l+X|1~YxF9;M?v8X6fh+MY+
zWf+?Hn)SFfdiIRQ>Ff8i>wDSt?Tz*9({^_A&S%Tm<MPIy+*KFl1Zm2UH5_$;ZA4&s
zI#HGrkBBQF0xl4oMghqYws<d2AWUya)IS}~(=?uc&rtO(iCYa?+ZrFR6E<js{7Q~}
zN#xQBrJ19f8!(ox+6O2z#r3MN*8$p;o?K|cqUzdY)R&>Uylir$8sWm54%v(!&bx>e
zjcHj~;Uv<3xC<nRTLeKr<rCYHua&MUX(ZN(+c7M1C+><yi$xr$;8Ju6yCir<lg9o{
zy%4B)9X$I$vdio><2YvnZ+F;)_)njX{*zo-G8_=26@(5ta6#_q{YJw=5H5KEJZ%{&
zmh=@>lp3x<CJ2h<!fG02MG6^|K;u=vHf$2AN*T?&`vvvi+369(8QXTwc^tfJyM%%>
zWI+Z6uG!X8hM--`T8C4s!w&UPxjj#!d^!?Z<hmA$L_u5HV-)u<dp>FKmQSvs(ioDG
z?>USxLEmL0?-hw16lJV|tAuWxwA;BXN@0+frQnasJh@1P_ZV*g&u9M_I4ZuTnZ#$4
z6KW?}ddMpq@7xP?NiE-zF|tpR9glm0LKB+q_VIFemvLt>Bb4dGH=Lv%cV8mETrIz!
zMiG%mK(_oV9%`hqi=-`i@V5BysO;fX@AfEbj0%_HCt+s|x<ZvDwPKrbb$TZ@#nU$f
zLWe5HCX557#(Ds!3S4>AG?a+aM5Iw(?J43&^jqSa4;b)o(taRBJh-<qys&9sSBy+v
zeQNKe6L)mK>TkiimqA2IRJKE2fif;POCzlq=SOg^D4%;qM(!dc9Ye)^hiGW@)xiz=
z_z_@}%M;}0z9l86<F5*Vm_&7nVvqG*j3P-d@SWI~Z*-1upNNS_BmWf<M<>QUVv-dH
zXT1L0N1N{yy-9%(C&?B@z#~u^ioo{o1Ps_kkGk!6J2HrzWT01Dxjy)OT&H3Y8R&mc
z4O%(d81<gGB`|R)N|VZ!C?G4gp)x;%F>IoBEtL38%w<NtI1mSC8UZV?w<SIX6WP_4
zj4QKju?Mr6swq{D&kXR=!l7%<1m-I?Te_+|sa0Q&f^`u_{8auH8|H>lRjqK=r#4gx
z4j8u)BAB_ws`wg2<8-B{7NS;hIhkjaD+xUssuz{J6gJ;)y59zciL!Qd<Ou5zS5KY%
z2y^%}TyWL(0<*uc)<Oov*O7DEIcubKW>MCf%}1*b8@t=9Yl!`_gl%p**|EE+;pjAZ
ze3?&eMWDi4M%I6nznrKAmo|^?HV41)@6bvU?pNP6Xd6}FA;!33hXX&PBTA6aIelhY
zG^|g2B-;obALiD!pmU*J6^xoAvMlsva}H_4HHD_^y|r|nwMFV!$BTCg^FhFJ-n2r@
z`|VU*(|G*Vb48H|Z-{oTM#U5IRHx&TbE#^U=ZU-?m4;-z%xXf6*ATytx6XXd5ZQ)K
z?M6aG;cL>3vVwkYubQp{X5njJbN_sfBX1;3ZUkdL=E~FN=nNsC;qhCaoxje#yZ5Sh
zsNo9uZF6^fb7QahXsg+HxD6lVzhp`7dSi3rVRLPDd++g1Lt|^<Ie~DlU7NMRc;;LY
zMn-cwIACzBXtYq{Ht%b~DU#uQ@e&g>N3@IS_o><~IG1E1#^JK5KYuz~Tk<z#wFaJs
zSX_n~XZFvZhF)Yc3bGl+h3L-LK?UV;^eNH1P>eVZxKznr1%klIvibzE){w@E@B&-_
zTWkY>N>@(v;&&nSmH5HgURb<(OW|j6rti>{-N6m*sQ3h>LvFW;%-~p<ojVB8FA1w^
z69m@Es?Tit)~w42shgGr7%jN^mfEa%q+27M0u+tx?3K8RX%s{^d}(XopJsDB*cJa>
zWu&5O={BOU(d+<2skTYCV8$w8l+q9J!o}>G=3-klxGS7i25ZWvcq6IbLr`Y~!af@#
zmWqrKnL`WAfdgfIXkG0oO6x+roj{5p6N{VB;;e+c0KXiawB%IUNO}_oN9bQ(P(cBC
zjZw-3$%cph#F)#9F*8?(j|1vFS{s~Qz)kvOoYjx!vzs@Tm$Sbg4BOr8-r(ueZmU0-
znc1d+2ZVza8+Sgi?e8GI_0wSs=d)P&@QSm9WMpqo;h4j4js1%Zp}g_J7|1+sb$LII
zAsf-v41zau)kCD_<M9hbgA(6ru*KkrF})Gvd~hWDdkNSp?^ulx{Tq>WW_PW|EVKf2
z((bf+ct}{HLi#5OjfCZt<N%3|80i}p=KfLdyiFKhAk;WNOE75dflbT!M>A*?CZ7d?
zbAi`Ey2lszOBHA#6UA?IvcNkpX!`Jcj7$i&UiYZeXDtM1zfOKYED}5xaVEivIab(^
z3~(t!(1Dv%1k+O#jhOS|grTyNvc&Z9`LN%`g48BtxX3sfNRN!N&N?i9j408A7aSTw
zfwysPjz=I?W@h&A2_fBah)$d<c@cvMpbEN$00M+k)jo|*Afo8vt_TA$fL>&U;b3&<
zhhSP*`*8?l5#TWk^21z0{|_43?$-Ui?^kyk*~TswwXN?q*6}39?CLJkvkTex8}Px$
zxiw1ctUkiZw{<^Tee^^2r;SJJn9KiwWpH;l+uE7gc!(9UfxL}JYnzYpD8{$hJ=A-&
zwU=Ra+~5-$ds`Vhw5B#1sCPem*w|TnfR<P9ZD57{VPWR}#@-{gd4FpsTg|ptclI{c
z9&fJhWZRE-wzqa0=zbl|K7v=u4ti-kY&_aqLa)fn8s8z2?LJuD++<fXtB(Qw4&lqz
zwzhxR+4%OsUiM&X6AyCiX7?I^Z1vt|L%YJD);3o+9xi0-s}EPdC3HL47FyYvVKId+
z`~CqgB{X)}-zxrJqQP+(Yg>=@c92}aD0lWk)$ccU8w=U$4jxVgLhkQuJzSU}I#C0y
zqY>15)X-LlYC%w-2#OG4xRMYW%+?#Ln`jH<d6XmH(zLg#JH7V7-(#BN9wFU4)nK`P
z4z-6jDgM;C2qsR1LC53uZG`P;;aWxiY4`vQu(+p6HN?47#2IaI0`h{0xok27B*aQD
z>{AfcmWBk&-LIjj81z{O_DqPG{&;0&Lc3Or+kvw21bZk9qrz7^-<QhLDX)6xF-6NV
z4r|xo&x_q<{+}X|SS6~&;;T=)tT*;R(dM5TKQuwz`<okUd%Ms)ab}_4!JJ#)H`gCy
zzVM0(3-UYOe$nq>`Q1iyZ>zbwySuf9FQ21pSI_5qI8?-&CNbl<*Ln)zLIKSCqi0F_
zxn|(BJ`}-g)hv@Gt&A>C4+jaMX86PoLW%Gm(9MOJ+TZVPJ=#@`fmyeZwY$gpCnY|&
z5|705N3y;^)}zM{?=^NL`DG~4>uus)!y(=&U_-x`Vw;<i_{s|1!wocejHTti#;3iV
z#|<fmYw9z#`>UI~mRz=ajLM<~9OUnKGfJ}va%1bhCLG``c?(1H2r-AbIvlnxG>cIG
zs6PqlEB)zUKJ3uXX%p*=<`YIXp!t$_laQ|HPeQq%KMCHx{v=pW^rr*W=*hz{q2Zf^
zY)*9=Dal5$k8Mi=&2pa_h-*hLy-^daBOCj=KByvZc?@S4W<Dq4XO(<N%3~C>N-HGg
zVF@HRH<4I|83bjCd8Xu5hgrcqliCoF%7i?JF#98aY7?*U^QP8V;ZHgIoL%70Thi)2
ze`*^~@H00G?3vqmwVmr^G7bz}M|5i;TMqArBaA>y{60<K9)J9jODx}9$Zm!fmUHcW
zq!hqPdO+${=nrg>q?YrLAr(t5f0=6qsrU_ikqnQiTl0A=j%gkbY+$G(?on0{uUX3r
z%YZbKM=#KWFhux*{xYF0a3qBFFJI$>Jaq9nzz{z1FUvRZNtq>Nfe!G6V&B+{E(-&I
z1k2JVX(pY1g&)#as7CJ@WtWik1<%Vhl?i+4?^|hhesEt-_;hpO_k=1LmF6g`YO|J;
zG{~F;g4~y{O6@Tx-=5?YYE5d7IL^bzoP2wdlWXtmiSQ|9as!sEe2bcuYZ6(LU|bf`
z=G$aep-sukw+RD7q-6dR`7u9mVbMaHETs~}iG@;=+9V_An`BPDMajuQGKsBjE|4|z
z?Fq9cJ|0>o58&q$;rpW09&_^WNlp&F+-%5wsUQp6OQy`tg$g>cOrafQ`FzWLDHBNu
zlEXlmL@9+-4nP_v?V2=7QHv~@sUmKcAti2#yjlJ_i=atUzmZ(iVQ>+7FL{>7`0~O{
zA&WEx^2Z_ONH<xiEtff6vhu8?Sqa|34LN2{W;8%5@;0(^te2EVLtm1Ma(sZSIFOVH
z(NOwHm5MN(%K~+|EKxNpKa{W1?h;7QT~PodE0ySZxNmYT%TEYo#i117EenO^BIbY|
z$;!i<%7rPjLKz>g(nSb<bH=}g`TKs19|uuDV|m?ME-!qVl}BT`^!@gTp{ny<5+OlM
zeqvD!<{1={aboH60<xS=+I&MQb#mxL7IfB}Vvg%YEnF5|Roy)3mNTlr`Mk}(R82H7
zySvL7jCy@scu~VPtN~*@NeG*O9<jj$Fv2;ee#B-S4sc_<(oip2BXIDM+S&1LXL7Tl
z@LW^cCd?DEUk2Pa5<iM#7%8^;_qHC!t{oyTuJy3dGDu|$9fr>U_4Yyc+u;lS5aLJO
zd54e0?A*o082eV_$wAe{1><qy{X+4h#zz>CpIQQeFJLQV*bU3TrH36DiNts?hkY1+
z!OY;0h^YWsJQLld*?_)f2BmyMe6fO?hr-hv9HCmiBmhBntS{qNcPM<~7aT5&5rI8Q
zSQa`eiH9ki^I%+v`KT+0aaOD)yc9pT;?H6Hd2xY*8YVHCXp@;cizR@mm_V`Jd^dQ#
z4;EQiO9Ol`alob#(*7EO_z%hH7-Cl_q$$z&b1nmdhxtXC>2bEL=RClwdlclS%f2>f
zFNB0~TkF+jISV>H7^3XriX&M^A%|+7m5_|2hM5d!nU36ClsefMIyCqz1d>3r21|7T
zTwk&SRGtRQXg~2D#DiOV!@-6hkm2rRU|Qz!$THGF#RWhO=do4YOH3Nj84+pVtPi(X
zZ#+LFb#H~;V}I@p{qwTuSqAHspvXHFj-9Q|W=oE&XfI(s<m0P}6E|uG4iWP9Ew;|A
zTfkr@4}+3^yI#Z(R8_dyoWsk;M>wD0q8S7?&N~;_F&yfFq@>?jb6Vw566QT1BMcls
zdSE-T{akA}8wRL2&}iXdQV7}%k1z5hMIz}ug`q$W$wxR%ka##eFTjh+v>@mQq9r)h
z%fYz}rC$4ScIn#P`T_p!|9STCxIY|UXX5<H{3{-(=b(e3_VF29VlbEudGioBn3Hz$
zL$3t(q=lmroI{@X?HoBfgduYT^DJKS8RJRMC2eY7PR%C3TEDvjFrQre3t^m{|J~dx
zy#IPBlvjWbO#BoNNbtzVU9EANri5GI8Kk|v3WW1030}Dv3_QQ_sxf80m83hqK9<?-
z%h7b;#vrb0VsX}Ie;N2pYS+NJb*>k(d|X?eCo8WU^ih4zY67QuVvJ4(FSc>(fp#`{
zDxrduCyO}`>g-WL$@*sJcUH*2t+DmGiAJAo^O%1C<Zd?G@JY{ieFn!M6=X@s#m^j+
z!8r+M)#>CCnMUZukQzM}FgH$J;@T{uQx)DdHy#6%xjRFIb~angHuAH<84L^Zumwj_
z=mlJD#f3}g;OnO-6%C13W#jif@(gbDnHf5dY@PS`j2El>V!Hi=FYz)+9u8U;@1%mL
zLLevN&AtMICb?k`CfTLYe5cM2EHOCJ0<vBs!>OdXo`SxoYlq7BM1Di$z`_WM3xD{c
z1D=59!R&ZaUsTr+)kZl8A=`v`iBz0oVl{S$H!KW*OLd*n5ReW22#QkiI}tA5mG>d8
zizW1fmYJs9_@I&oY$ctsYcl%aAV#a`krk|o<E!9fp*mR*D7l4HgowQIO-@!cs)j**
z^aD|7V?IN;Xdds>m#OW&lLCW9y{xkti{Z4$xsCYMQEKM9=t!<De1n7YL`}+JY2R9(
zqK|0zPoh4OW77=J;LT(~MTt)JiFyL{z=V}%ny8w}#U^)U+OMeKKEvv$4UEFeL6Gy1
z<xm{K?2xHOgC&YGnfsnCN|(;H5j-Ba&oT6v0(I1a+PAzy%&UkHWmA=I<y25uzv!+*
zKjqe2oiw4GZa^x_<@EccCqo*R)Eo<_@=hN9q`)*_WS0Exr?X73sDENHl1Y~E%Hkp?
zqAS?B@f8LI3{Ke2yqd8YjhHf^*G-|9?nQh`CWl%3PU7V!Pee_jN-TRDtN>gYqVzJF
z=C&J}H;<JIC&7IoNvw}JS#Z?xQ$LKRcHA9~#?8{&pw!JZtc<~i>uj09HJ9g~?58a}
z;Ua=H^&Q(=Mk}!z8&Rl^2EFrBT?h?N#+>f}m$eZfd)GDtY_TCqP3{SucAX*gy3<wJ
zMp3_lP6pww(@K7D(b6BHdEsZ6G5DQM&yX}^W6P-XXDB*%3*juGb3cZ&=P{a+xTo$P
zh$DZ<Ght}-$ECvIb8%0qsW7oG$rM%~(lgZ@8{W{Vor0<oBIV0U5n=^|@lYVybQ=&R
z*tO{jwv(*+#haK6Ik0m)V1+Pa0uagIywltHd3X!+(EUf%Q&mhvhEUJJxp^rTR;t@U
zdk57JD&~adqG=jVQc`H8F3)}}AFN++=Suj=4H+f^(A8ANy12l&>CpDsoXCn1fx`Uf
zgvFzo@Drxtfl_3|nG9P{SXitoImecjM4^(+oVnsRJ(dIZb=WF#leC{~>Yn_CIp_PW
z#pBh*`!^Q9e)8kzuVkO6n|Yxk)OOwWY{2&rk$`Xo7ds5OCZOZvBukl#^UYT(lcpw2
zh3C;2!wCiC)H;ZDL~mTNMr1HOlu9(x5;9*3{Yk7EOlEJDbH@!WMc7KMLu-`{xb}Y+
z(|4RUMZJWPd(u2IMLM)wD<u@@lA0F!yj(~*H!bu<xsX|D(o$cp_-rVS&NybDY$S#(
zWe@-;@<!*b+Ce|FpeuPFr7H9W)6PVs<_&py<SbVVyaZ-Cy5$NaI77;O(9zv|LKz%0
zc$+@g46wvO<C*#<uA%I1Y&~isOv4|t-`~J9q%uwj49sDP+?{KQ-n_KQ;OfF%cv#EU
zu36(AE$t7bQa+o~82RO6;6F!gn79mlxu%$dEZR)Cx)f$ho>;lBi+4h!vTP>URXa=`
z!@vI!w&Ek|v3)EXkR#Z9mHH)?i9Fw1=ef7aLr)ENwkOGTp&_r;{$tx%Q27@&!T5%m
zX0YI>nPI46<VtYxL^vugOrbE`A%pX#@4pMokE(99NbCo0GJpvw$i&IGtny~O8ZO)J
zf4atX3navHrOzCPFBk(!OA2HAggD^XO@)tA?b<a?CwoTNx0D6)g&JrMruWD)=?6}?
z0&j&_=V`tw?GuGsT2$Ec5^2JrQP+$hlc4V|NUGa$8S*iv!Q!1CNr+c}n+Vm|e7A{^
z$7W!}Uv?@$Qiq-}7v5nipk@_^$&!&5?_ZUbk)R6XR@$%21=DFDibz`F=8bZN@(N#E
z&~GlIRx-OVC(Zbg%(x0=G+!ks3D|?#27(J>*r}O~u%zr)D8VqS^ZbX3d(h1#2h%<%
z=eP}s%XUFw<Bdsphts-H!!^N}&s%bqeHgQkPKNcbAm;4LasCTD<ie+Z{G^tDt1?XU
zhqIylIL#l&{Od{nb!6K~W&reLeA;Wmx4%2SXtr@4*v9)RpZ(p-+EPy}_kK*~cz}xo
z4ED7fHQZRd4XYWB%4(1I?k|3&(LzCl5~hRIbFQD9>&q??@>TexIpQ54pS<!(nuf*W
zCPKBLBA14A6I@+WO)yp!wL&G`)mK;2Z)v5fQQ+)Ct}rOa%EBN=19qk%pT#mmQ=FZ@
z!n%oooA3%a!qCXnDxl>jTJSYd+EJt72wQCyOzo?MFF|eP3gRrUwIrv6|B?rVZcGnx
zT?!h^!OO<|gO~U2;eYqR9-^8#(Z$z85QDP?x%7)lqG>BBlHK?$JT!2hOk-E$##ieZ
z4$aPC^1?wTO(?va*dBEB9e2ANg%dUe)h5ej_uy!;xHW8dhKmnd!)G`PHOT}L(l#*l
za(4b=h+YNJD<GS6t-x>{C(#ij<P6UdrNkMdJ|oVHwd;RA7+u4C;~8=5LvrzMeV-fZ
zzw9sLzw-0ieEnd_e&K7U@nrwn;*-0cb8yJ|H+eK(`|;1SSKY4vW&h6yHx>_G-uN2-
zYvIT8F$E6JMCX4uizC|^@th)nKeBys31aFMQ{vCW)IL+<w$0QNro?v}sRbs)aS4eP
zCd6|S3GQAMYIPY$mh4q6&r!jhux07!uK#(C&8XuzDlIVux<#zVi**svvm1{#_BK{G
zaaw-Ih&?ka#7J^(EWiP9;nu?E3tudJx$ygiuNJ;uzzvgS`k~S#a$#mxcxi-Sw^EH`
zK<AVC-z{9g5xz_=<eb7tn<U#f5lxaJ13P%W$gf-JdVvz^)`YE>A7&*vazGZ-dLS0l
zYTEZEf-NeDo5bGP$5|SeGwQ8y-ngP+(#@1QWu;w}69r$aqQ4QQjY(Xcy-<Y>U6xH`
zfu3g1uuP(?l<9cE2P+m1AP_4nWSOs(H_x?LT!PXJG$FoNI-0;(LRfrLsZlfDO5%E=
zLv@|95(Vo!yaCImZ?0|8I&SwPG$%1;)4wc=VaW1o?B<U1iCY-<05H^$5FUu@Ji6qu
zx(=^!P!3el`OB(wT8N9Vh{6bN%6BM{4I`06uo43!mdt2nmEM-TYOjUGwyL*+`-i(7
zaBp{n|IGdwD+CwE0_)`Al)>RFf}rxQE3V4xaBTj_g5uETz7Xz;4_y2m!@P#B!$c1}
z7=+g?Y&uIz*n#5k4z@px(t15Z5jYpzB!#M>E{vCubCLn_6l9%o-Z2Zz!|a4+GcONQ
zfOo0`l9?g{1$=Q(Qdgm|Y3bGY#maLCVVK%md6+t)_cM?mB*q7f%CJ)mGQyrk<Ds}Q
zbZ~Za)Onc#9|MA;VksGaqA8eI2&Vm@MX6m?1u4*ONldSppiJcot>1FpPy{8tsF@l?
z6A>XyXIWqZyceMBeVk%z2;%_-c!wV-UbV<(YPXk4_(W$oAfx-ujoXlh3_mSjP?qUy
z*9Z&Ika%EOKT>XAi|@~MpNO$j6hC7X+k3EuSi6a<;|P9o&|mr;7FBb`64Q{pZOUg<
z^alp0Kk!l6)e4YBl_Wt3Ukk?(f>|YhER>2#KMyRPQmQm9o*m|8Wo%K--mdYE<)qWa
zTzm1nuOw@2u!!1jQb0!FUY|M~9<D%r#(zaq793Z$kJ#VdjPGt#aR`vhf+U_*T#E#?
z_D#8}@)P<}R~c{I(a+XMFilskp1#t$R7fUM6z~Dqq%m77ivV#eLDtZA^5<KT7sGm!
zUrrsZS~>rXUbX06MT6{YQmMcU5m1vi<d}+yT(NP(L_a&r8Bi%-9COL8W2OcT9~b3&
zl@50bIOt&vD9agM8|>5X;Ay}6FZA{UQI@doa*5OOdB;1y)tb1D&Ei6A;weV4#M5_l
zk~(>Lw@^i?40HV{N?0tLdD(;YCqMFlbnXj<lS7VZJeQFO$NM69>^=4p1O<+0Js)%#
z<fGX~{M;tyVXw=OB_lA%4?i^JJ)E3zWKwyOP|Ovu3#E1<7<_DeUk0;{l)Fhwkj+%n
z0`pVAk6xjtg@T-Bi?o<{qqvMOt>JXcx>^3cW#P=qv%az{oJv{N&0Cfg&*5%*TR0N&
zQj7vXI}glgp2ffc%!0wpn}<RxS<4u^=XtT^hyaIT8AJDC-sf^005sX}5L9AaLN2}A
zHduGSJPQFu!-^k*NJJ$7To+H;Jpcr-TGaBU$9{mA@hv;mD5a)Fc%TTfpV%x-2_<0@
zrL11AYp@E@M+{y0E;z&9#Ymvo306XY{A#8Y$LEW`#qmMxo4>{JL4-GKw*MB#=WlU*
z?&47qt~U~YK5I7DH+C3>^ZL@#_2-ECF&JKNosS3bFKI8KB&7C?yp&kmZCxPh7fcXY
z{W;cDT-=z)ZWW*BTK&-rywLPbwo5bSTCcl39KiKF{#sdqbVd%hCe|9Gm}<RF8_g_v
zBTE^bBN|(qp=8qUNw?jG1^VjUY6S)7-8#P{h))<Kz*IRLav+E0pRGMO3CM*QcT8}F
zfqxd}B<xS})&n0}!)q^+iJekyD7K$~Tn*XC6OC2DcimoYkn*_Ve&cIgbjDpLdl^Zx
z6V1X9ENYwri2qK2F^9G}Y~Thj?&H6_4MS*KxG$!zVbp)kov>ab;l4-UuC4EPgLlil
z#~YjLF`VQ}+KU3q9ULqz9ULG?+rfeMC_3~K4BlibJ%9jGT0@4ofnsg?RI_=1V-t~M
z#nyler?NjhJ!?qgql=L|3dc8dn(ZzwC4#jpTZdzt2I~t1z%?s4zZu@~&;pd(_-!mb
z;PVT4YX>`{PU{p3=Gg_J9pY6xChOKBH9V=F(gTMGM|^I@2^314BXArX>lB!@D(=3>
znC7%^A)U_?V58GJURg=_L4uazq5xinVMQ@HDH8pZMus2=JMTl;e+HX~W+_)>GSTDw
zb1g+#P|A?tcqhK&_YrE>&qu@Shu!{l1ilS)9wY7^FgKfi#3K{g;t^u|9-R!btv}W9
zGEH{9-FbeUn~&M#j-FXJ58)KaS@19V=b8RF(m(jGMnZ+-B*;&wYHX8SGSovqiYi5h
z$8Ve?Simpp*hM(O7yw09Q0A$FNrQ$oDWgorC*amjdBYRbV*BbyL6~?8sRlA;rx!5U
zjb0#jBHR6gSoBygTO20ez~6HoxEYkhrlvnWu2<q~G=?+SBC&QGD7=mRKGK>LR36O5
zE^E9iD3IZ=K?{!4c=3Q|D$5u=#x~M+nHAu2cpk1p`wh)hK9C&IUFe|nCQuDN2)#Dm
zLPFp}SrZ`q{BVu>31B?pncFeQk9`{s(GCa0G1PxOob`={j>KiDu%_<}ivl;9ArQD}
z49(Op74E19!@CFlxmWAtI$Ek&m@U7WL4d(*QzJailvV66{}vnbH;Rop6JlfX%H!W+
zWBx5R=C2eRQ^J3GAWe^rX)1Ny^c$GtCMwmCD~MCs8Z#EbKOx4WyRFF53bg}z1zg+v
z<59iFQcEm@)s<Qig?Sg5keWD8fSV9wMDF64@w9dDoqtl9uf-Pc;3F>VNq-9&Ud?6b
z*!le$R)}LU_ra#!>p=6Q10+_B-dSsWQm>7qP3xMLQnhx5?icTLN75m}ZbInGf-%dG
zv!nAfg#N^SlhF$GIak~&K!oR~hye`6fDvo0ak&p1ydAX(sTKSzth>%q_FdN;{YQH3
z80$u-CAZoVkbpb6KZjEgxz`Vm+QBe@7Z%$6Pe)I%n#y4%ePIW~{Tdb%@w~3_T*LD0
zI&zJ}FB@lQVjP{}ATp)F8jOv+*$d(UWeo9|5U3$eN)jw}dw?PmYg4sx+eN@&3F=_w
zi|qt^SIp8qM$x)ubOt)$L4fs32uLoWc~`)&P7%5QBo1G245WpkQt!%-G_iOG_vtsF
zQRCW2yx1t+@D$NHN=spHqt-LNI-}Q^*&~7wfI+ajpPqvc<h5V4+&w)%Wi$XWU@OqE
zEwJP)f(oE-nlH#Yq95{**@41ZZ8~Gfk!TLZrWQDXpxotj;!vpda?%J97@}P29Y)xC
zSv^AQaoPy2NHRjt$&b)-r;hNv-~H$FP8?(yRPF=DrD?E)zcWPjI^J?Mx)&LNlPp#h
zf;3O@<cGV1U{8V~kS}AzvxZa{LOEybXyoJKD&JJPoxK9rYV#x%yFHRiN7|agRkuzc
zbUq2EfLcjFn96sIG>+gm!A%uHiOQ?m#&XP+ASJj3Rx*z8rtS~4YrsOmtC+6qc*1PN
z51j~eT@xuf`|)4MA9X!L#P7a_RhimUn7))RSD4sFMUUd|EFIyWp(-9l>N%#`2;T9+
zY0m|>n5!mU$sIb}MB;LJ@@$bs!8Q#v1QcN1;B=0dc*eCwajWEplaBX8`C^%+?nD@Y
zF?IKBk_rMx;!U)A6`Wr981Hx~{jl{w_;it1Jgl!8pyf>)U7?fwI7{Lu=`t0rDJWBM
zknp^$Sps-k8#5_HDsfz0sZ;KyC!)Tx+ekKDEyKcoDQA&4b&|sm+-(bU&x$;gPP1tw
z5C+5j3a;cf-N<AKJnYaH{Lr6H39Kvy5~^K;Rf5bclt?&<Vo+Ii8ia-V6X3LlP*@jH
zcwEUUTEi+Ti=svV=u*{0z-OHQKI6|Pd^UhPY%|SSwEvE#WqL$MRuO<%ghv-!?tx>T
zOr^L86+h(GqQXa8k`mNU#tyaF1K2Is;DrGNu7DI#!18iJ#lBy>SH~7wD2;9PGu&F9
zMJbXOd;F#@FB70rN*?3N^<a1bO&anMk{uczZ&TkN4C}RXzfUdqPjk&zaG(t0V`!4Q
zBJXwq7Q*7|6QhZq`FPkp!2fnSFn)9p)2*sN__#qm6hkaGGeuo((JKyU$y{%Btyf-h
zmI&jcEGn#_4`{A&p;I-RB1;uG%+cgymp00d4s<(Si>%osq>UKz8j9rFgT~sQnh&?u
z8+bx#=i%xm8ocrHhW_Khb+Odm58DmAsI<G67UbD@vEbTPV`r_Amcjuy|0Q4=b69@i
z+&4XNzOP(_7=iX9D&59?@k1zc7al;I7gm{v7+gbw=|1QRn0G29rl<q)6$Kfo>{7m=
ztxM1k-z=>E-*C4emmFk03ACAs8zh<UCgTQuu-7>06O<k}wYIsD)V&4*KKN$UbQQe9
z#qf`{Gu;TuPSh@JjE9)ImsHGZe}@@2Q}ZpF0ObRmsl@?G!?3nGgQ{b-X+z3)GVpZy
zRi``#^x!d#A_6{VQceh91d}f#z49Z!FQd*#mpR*D$_0B_SG1w3RaKY8+zWhK7~;`t
zvD0d%NR%@S#O?V?I;}hj;IOxk_52m;q*1T!2T6e&aj1hc4Z&X8Ti~X^#>&fO)fYbT
zLy;oFLkWMC_O8+i*F`iwjn{7L>glBx@9-I*lv77_4gcR%WNY!oI~Z=h6u3@_0<#Gu
zr&1IimAq60p$MOz8vLc`G35v+bOBW|dCLP9ZDEy<LlpN6fcTwu@Hq=nUb0lKV04mF
zn}ZxU+NNVL42C0FN5O8oxW#3eTNqb%T>K!6LefCODtU@00yOZ9U_hVD_z)B*d>~BU
z$i_FA5>Xq#BPU$GzUzMJT`^My#QS6GYy!LNU7X>@S)M$<G`Eb==euI#<BvSuS$>eY
zD{|}A*dbB($ozqsN|U9)S(phqm90L%c!#gW^`3WjI>-9Sw-<@#`vFsI64UHLC+y>8
z%U_pmAQ<fsZwk&nqJWt+dVNFa*B(8tP~t70-+tOO0LZUB_73{|?MIHi8l295?XhD^
zie7)E(O0lMqz>1JH8}vHqU0J$Svn=<f~r!{BJ}9X6#l%jrrNzIX^?k_MhiJ>IxVcj
zIKfq0r)v9DfDJ@Ht4S{)Q;!l)i~UdAc-4x2W1o(&eG#>f3dRDIHtn!VZ%b4%ItA22
z99TA?z)^=o^d(}T=!*L>PkZF(2fMw_ge|ga#f80OqD<*#bMhvzxXtkLzxck<J1QPc
zh3$kFs2rhafFCqRLAbo<-sMz#*CRU0(6z9eeAmOdj=sJIhSSCFUDY+Ulh0W*mG@^8
z;BZJ&|HQ4n!0u^_T8DUkBL1JmR97&P>^KKG3+AZr0Z1?(f8n%72SAk$(#?(wyb;QJ
zJ$J<q+08PCq9xY^KkCgJzm;c}51F+htsXLp1GePmvy_;4g7HQ+>;@m4%|xioN?`Wb
zDe_f;mpeJ!YWw-O4+K73*hm|=DtmPH>v6XZA7vr}Se9dxjk|p>b#qzs!6u&+oB#;6
z_O?&)O_mBvoTjdOI0B6<&BJya+t@)V2Zfgqd6+C`tN%ntgHB)KV;J`>-jR_*=<%-`
zg6f*enRWTyZMlWFKrf$SrDO1yd_Wm+9a#MNReMi8-0>(GEO&)_KeoM-Qb-M$W?#b+
ziGOk3GdzWU3t!K3Hdg)EeGU^b-)n}`1-V!yo_nVWzY2%@Iw*e0`A(y;>S2p7PzY<+
zXKM0ST{gJLDozgKd>`Sr@8x4ILK(mu=K1ec-8RAGh}wjnb>p7zm6Ky?f0PXQDsg``
zt+TNG%U8Bd#<ZHc`^i@WYg0R)R_&1#$ybvpGp$ks4d<(vJ~*wCDysQvDqg1osJxJ`
z7B;7)n%}vjkn%<cp2=0?XMRoMU<iL2XVZy?#~*9(zM))K?&N|X|B6e0Y673hO=DjJ
zZFZq4v2)F&Cl}%1Ae&1kTPPdn8M5N=S5}f0Lk|UJWIa;Fx6o9+X){%)P{4Mm6_pYZ
z;Jkcb@ElsXbEEVnF6|S@vIc{g0Qhk%T!Hwnh)XO&DOeSE$<kh=m5L_v;FY%pxj{~9
zM5b=oS8y$KMME*gxi4}qT>Er^Xz$ZpS_~L*+(j&XHi^hlEtt70L<j&LYz_vaPDsuu
zjY9f>BH<|RlVrg8ebSV%IHk(8f&o18qaJC_ajAsrCrb|TuG-uuk`Bkv`jy-o9jPa+
zll9u;{<HqzMIUFLe!!uJ!t4rc20l_Z?eJ>MBmo|Q4Ra0(fr7@JlAh_+4S3Ojt^%LG
zFe42WK=U~eK*w^oFG!O3p;9GU(aA6o-DQNA4`-88pGQOoFkWWVh&bszCRs&!^(y0>
zM2CEPN|7=UIhnd&fIBx~rkA8#gaE2WY#x*+I=K`g<9FO0LhKu>J-sKVM1_2jX|D==
z$>b1bGV6c%7JKLT=+>=UUzgOyDJR#;h{a|0*WSmaETHms@s4T-GPB}FB*yV86oY(_
zTiWyx?jS;*v`8mkuE|D1(*Qc_$ldFn#_bz4{vQ@k7u)LEVx^1Q3pc^kd=5+FpTc#^
zUvr_y7z|w55BRtE@ZsV*FHhw58NOYlVMPnaU52Hp4Tb%P=GjieF5cTX#~45ur>-8;
z#`j>QyE1yW@v#TE1((Nna<X!Ix-z1z>6aQEal4odk*bi~FESQiSyB0RAGc;yiH&>l
zM6<aP4W*yTfkr~^`I!HaZ%b9=Uy}mAAV$u|M~kOj25{&cy#w7Je<;mqoO_D<tZ+&8
z?%@_V$`FJA2D`@D;D{H~eknc}qVJazhP%UWS0a0uveKMSQ*jXs7N`%MkJwWV7|K%x
zPkusT9}4Fvw{4_;dqoBfPtZwo^y~gWF)<n=dd&o|cKPkyz>?-Z0$67!SXw85wastu
z29`93{StSs9|1aS8w|D}chtL~2~X_+*+Pb!w|E?>DFHoJ_ZrQ|JDVZ%(c{g{CL-G3
zU)|VzywkW6(NW)+^CJub=FM*Jy@t_v2$(-;Akz0o7(@4L5i<mb5hDyl*Z1z(^UvRF
zR2$oqhLruqDS3Z$_1osg_IG$78Vbb(MIg-4jct0061tBv)aPIR;^V@d_+NWmCxg*=
zk=w68LFQOSm9TtZpTj)663pJ{2T>w~Qr7om^uMNdiBnlLh5AWdxX~4rO6o3DU9h!P
z;Vs8~Z65U<c&Su>xo?NDrtBfIjx~VIi0iQkP#DevEj}mqvP54fE$Y!+d+eHX-Fr`}
zhH!m8K(@fpgnaR(c8-PID~S&`o^j*}LK^)JvFnx&+SmS(5n>Zag+{h*3Y}Ow!aD_h
zwwTKTo!DTcuE6rg7(@>N^knVU)tna#0So*mWfL}~4_k(1Y4zQ2Fa!VzjPrre9oW=!
zy@rmiw!B*FpU+$D;SgaqpXrbD4*wpXADxW4_%>5}3Rg8|Jso!7@rEBg>gV`-+8PdU
zT@k<ft>;o?&_i%rX24sy|5SdRoeyDBU^~N3o1aEqxWWwiYm6(1{EKH7u(8393q~Dz
z0ky7*u7>g)uKxUYb-o}Z&j;Ol^_TGb%UAPr*Sl!a#(Nq1OkEHaKv){6<{d%xD8nYG
z$5Sc^BCb&faAzC}20*?IXkp{)>Km)rjlvI`UT6X)Fm?utB0%%+_7_`=|MKJ-ey!kt
zOV<p%*v~??z0>%%`Rz_)yNN(rjh#Kt5gW8U@R{%BQl>C%Lo|ql@$wiv;FG)cgZ7WN
zULDN;MgGob@`wNZ{FTlIZR0XDM)^jWC5Uw4C6fSn0XFKS*}tqV{-Xw|j=y8}!QzwO
z#nhAegQfZ3)%Q^ag|$}v9kZoY{2(k4tV+qK=W;-SNd(%DW2d;(9#K4C__gg3Y9%Zj
z6?a}9_0AcX`k~&^lnF=yS}13iIK$_#n=dw?r?YBG%>Do~h1)c|g(++GL*5y+b51L6
z=baUIM|tBC)3De-roKryYe7y%icqNap?pG{zv{{^rs)>%1Fu8%4q(PxO_P+j$GC$G
z6curStC->lfR{y~(BWKIN*M%HDo#OV`0PDY+M82Mz{&&`$W!<iC-D0q%q-%C0YvoZ
zeciy!g>nd0EtDteMH_JzE;f+E#GRDat%a-dy4@d!tqW)SC0^IBIbPSSgx4)2XK1r=
zOyc##B<Cak$naXT?*is~@VX(dn%DJX%Ik(HSwqYBCSkm$2Tec-G(|UY;1XICBihQ9
zg@eI(8G+viVb#3e?pg%5dl%+JgQFE_(WirU*n?pGbI#5c$E9nH8m<PPvN^}D9(otq
z;YI#tf6^MbHc3TJNHcRxn|v@6arFS@)qCidTgO)!Ut0RI;fYk&$OZ6a?DB!;(v-9h
zTgMclArXg{<mE;e^L=n8(&f;^H%-BQXY>J~=og7UdZirRD|%2LHf%~)wG}b?tp2(w
zikRg>1;dTLx_y$jy}=CUSpD&3Qhg}6SUQd*z^*lLW67@F-MAr*w7bX0_*9qoD(CS!
zku{13h8+sq@+~H9VWE|ZoL}+-8KI|S6M)`C_e5mYqwXof(&2%R?DUY)*yKUaFyz@K
z_<!BCPq?5H2Z|uu6Dw%^Z9ebd?c=qOa`BsR#G26lO$PBUweJ-|`>xL$nuZ8^fQe|h
z($^6J2qa(thntL#ag27UdRp}Ed4Kg(?j-9S<SX23f_Sm({M*=i0sDFriX1=jhL5bu
zHcKvXkc_;fd?aznWuq3HTBNm7cOu(kX`82h9JO79E|sb_S<6id2-01|5&H=-6}<U%
zKHw~9N!yK;Od{w;uvELD>UUbz1|pp0O<N=Qlc#}gFAKiO4%)Pl({XG<oQg6saBw^~
z9`f}<Q`lkgz>Uv^+BN*|f5iVj6|ee8f<Q1BmCP%()T+3qa>v(snFwB?WxE6O1?&xs
zss!r_gi;a>gMcs=MkvH8q?8L9ZBy`P%_I-_qA*_5N{wo|*-Gs@v1wf5Hwf`7bcmHC
zeG^zfu3ms8Hh_&hn-!~3Xs?D)I*g+u)WUj%E4^~s6t)|d7Ubw}S&85{fDn8d_%EW@
zrx-DV2trVR&@SS*7;ab}5n%A@04#O^yOV~8#Y+Zd1l23GU5H*CrL2I#>#yV@?zf~X
zqX+1+NHJ+<6Md(mlXE61yOTIePn|F77#SyIkWf>YJwKnUhAhVUGc;_x9=2Yt1?QM_
z#uodv*^0@Tcx6u^5Gdjk9RL`T;ax^4%jbvda}_-hk*05b60C@GUG!~KQ)R0-J~&US
zQIpHL4zcoTKVRGiBcaqM33yXjU80Sdsmzp2hQbggfA8=Y(e%cBL4GV|a4mXNU5bK=
zA&H#Nw=U*viG&n)S`+v_u7W}Dla2YM6$`@L*am-ml5n-L&C&$Mh@1ob_`Zsmks*0q
z8L|>P$5zIZ=+R2*wAYfrN23@mEN7iyh^8unI~WKM25%GzB<6sSAD|fy7;X$A35Jd_
z3F|eCq>#DF_!9;l!6v&awBgP|+3_rm3SyYSJtmz}aNf7R(xJPwDNXuMV);vznY&y&
zY#ewQJOZ84EL@fn%~9M;D?Cjdo+#R;i60YUlY`8OcniMWQKm(7UlE%*0auh!vPc(}
zqKJ`dITw*c%2t2~`X%Bh0-R&(A|y7WFeluFHH;7S*4!0VZn#8j<h+)nLo0QS`UF0R
zm6e+U{ljRF26W?&Tm;Y}J5Zv30E`+R#aH}<#SN<plO^~larYGw=s-f_4Pwg*@nNY#
z5#dui`%Ah=20;LVo?8@|-dS$kNI*vl!ZVo)C|(T1HXhs?40@dwwvDRch##pXL~74c
zY|qYz1IutANZd*tEhg(=QZ618x@sbYWhbCBvl5&~;`Wh;CcP)AP!cP-X7O4S9{rlh
zfXiIIbXVCCJ%Qm0by-hbv+Ho1nh#@5bV2JX`92YbAdJd1a4wdi<w&WT6THo7sxV`1
z<c-BUbl#r7oAA75t#A$HqJ;^{${?Ppwsno@F|YX|Q=|=`GrXzn8RDTad#LO&izk4L
z*uF-TSSQ38*P7;QR#~%v{6bB&ZB$mwvqz{$4LIt+MMmQqP)2)B$VE|!lA<Pp5MW_{
z7=RchHK+)H9d$?MtB~l1esB~fyElZ_&Tt!1YBbXjaQ+<~SsjR!l(EY#xll1Ee9`sl
z<W+AqxZK-EN`-W7_E}<|@l3gYmMP<(9R<m>t->WRE`cFSOS}&a-Iv#qNgD~3c{)gm
zfLN?5F;us8I5>P3{Nsv7&_XUN&kjVT3tHTsU+G}*3dh8nMCkM(NIclfT1LaBLX{Lg
zF|QtByNLk02nKJ*JHf<yNz0=A8(<i+8upVgFo_Aan`X=A^^&&DB<uGnOwd$RYQ20n
zr02@NC(^O^z7&!xA|2sT<y}8$(t2(5vX-lmo{lL&dM<ks($hi#=^w?&`R=z)FSU#A
zLfbfPMJ4S0?V@_vf+bA%c9W((WJDURSth7e<^EYjFy_}pZ751QtwoWSC_>?QH<|=0
zXMN*kcZBHVKJC$pN{OnCy()p}7b<wurOLN_5hcPprExCd>2j0yQ`z%B@ftKN=yctY
z-+m=JUsHs+gIUxmBV=i%DqMmVSGhfeS!Yx{rtzQ88NeH=dknjyMorg!>PMLdm&G98
zl~6su9Zy%fqTujzv(tad_#arnjeB5zmBSn(2j5Z-hn7i&OC%pp^YCyBT$Az#m7xWZ
za7QOC#=GncM;R_jJ(ab^VG1G+;Rr*X$Tuab92qP4h(3rt7gmw#QFi-R7-ilM#VcHC
zl-R|ThUtAy9HtjYhZ*wn!}RRw!!+}>sw9b>7zTa-0jh@g#A))+Y6l1W6G$7X!x8fD
z#D&utfXk=xMoF~@u@b~sDWT-`B*2=>6W=n?x&2B8?>V`ldj+SjfgzIP{f};)!|@lc
zIIdjd!Pico&zd&*Ok1a2ubna+s|4603QRH8gGZSm`V~ishY)?--)bNT)IiWNZJTWA
zyK!p(Uy|Cww}}LzNAtAj4-Uf00V9Lny360hO+&?ovfTAZG+JyV>~3N4MW60yOWv?Z
zwxtU2OrqY({)W4XnPe|wncS}CVgzG%p55RMMBf$aj#fvb!BH2nGoE)=lbDQfL6yln
z5wKigj3B~gyP*KV%mfHNxfFzH5V%$@>!<X^MalOxac`#Dq9ZWj8Qs9rRO7W?Nq@!>
z`guQm#7o_fWfRchEx5AqKJa)OVXIbZBn5Bk22xRf6@YC|qZO`zH034im2GpyOEVe>
z6USl7F^0b!*9j<Q^Khq)GOw3e5u>CeuqM@7!Cf=D<R@_+7f0**4P4;X{h54#!-0Ym
zF5bS!9NvYvNUqC&^hpdGPS~SfPBKi@bbdy(L<7X-jX0A{oHaS_FHepS;D2k`A-IIi
zFX^brw>hSp#-8uT4>VqhsLnn_dSh{zFsuS?Cl_?=RnzbCdJ_fI<yCVi=Y)RO@M9k#
zHd56tl~K(e_x+yscRZMH`2LqB&0U0WX}5;$No6o{{ih4J5ICGfPu%62^qS-QhtMss
z&x71SMB|y789Fr*4J-?u#sv_yWZD<I(#VS;W>xhMp2Qbjy+T`-#0Tju<ZBAqs^NG;
zmO?&v%bqL0*+l<BvUr<u0mkwh)bcz5S<ZjPuff6Y|7Y*b``SpBbm9N?DXM*ZN6;9M
z#A3tgwnsYxEvU4COdo#QKmu)+#M<uj+3ypPOJywr;qIAv-?^=B+fr3#WMpJyWMpJy
z1S@%RG1WVGsjlkZEAHso>WoSrx8ns_vEjm|m0)^6vEkW@{a(h`Lp^z<;2Q!$-Mv$0
zkcP$#Imz}(8f%wSrQJQA;U77!_$C`9?Sg_bg7dZMz>weGCcOJ-7R?TL2}+191(H;H
zr+ZSZHgT1Nmte$l%XSf6OL-trXXy=@K?P%S$Ap<c*ceQ}MJTup!O7x~0I(Ga*g-|8
zYWfiaV~uFZ9^|e=1e~uy6cc3}^@|RvnHF$5!uJ1z)xP7+sM^fPNR<slEMHU2VQV7(
z$0h7_-n+rKM;>c;X?sddD6@f>%i?cHOYaGn*=x_xGo2Iz{0HfOgKm{LL{jF1^&e_@
zDEcNb1E}nH2@w|J3}s3$=o+46L#2Z*raS;VFbWj2rNoLe-ARh-Q`CSnch%qb$P-)v
zr;RAP9z{bjyn--*`mHzzLYeoADW2=re4(6m4aw~vgqaO>IEFq!Fe+nk*@*!eOi)7e
zp!L%lNw9TMN3~~CGKi&=sAI0u+p~a4VDc(;$fCx5gB8>S3In>I`VQs+&8^7;KALpg
z7VQh^KjRa?XC@y@#*$li$_LfeUSjaUW8Qv`;17#5>|rYFw4A`sb2YU&n5C4N3gp+Q
zh@2zUAQkS_X$NhpvMe^<Bk{~4Ew3lZwd-;|MCMm*j0B#zBMRDH4Q6v$lnx$m5O&t^
zbCVu*{>sqEQJ*;jF_mD0>>5CrBr^Sgluu?exSwfa2q3a&-1s}Zl+?oPe=mEh!E*=o
z=md<7WZ@pMq=%dE#1Nfy+7u}Hcp5v_&ev3owU7!h6M9#iSaNM4w@ZX3v0C@I%-=$m
zc_SK&EP|J_-l0WF9N&EinW^MR1juZ1A3|!p1uF^K%9IUm^s%d_Aa}2`uSp`$s)LP(
zbPE$p3K&%mj^PV{EY(WyVSfN`Q7lV%ua*VEi07rFA%WWi>Og^ltr%Iqfgkx_!n{E%
zoX~UBQwN$-q?VyT@BOv3&J}=z6cCVn#5_q66^`<QmRxok2L&Dss9A4;6*u>J4k)eL
z1;BcN2pq2=7=ssTNrf@EDSGLro6si%YG}T2W{ixp<co&7MkXw^T*^U4zHv1qaP^W8
zVGtz`U)HagSuL6oJf+wZn0`r4|FioP#32+ZrZ|UzKf_eerbLP#v8AY~#92Rn(!jlj
ztdmD4KX7&ix{6^FAVf@LgV^#hNk6TXA{;NNIjlJJ>E4t(SB-cXobQO_GU)Fd&+#B3
z_(5_q8o@hWpVmu0(tX1hsT}+3macxR9?ex{rSdi;5jECx@193t9tZ`rA+iSe0}^2Q
zmTXu;W-gT<t~t*-q0rKKA7Q?H0e@CEm$u-RP~IlwTj|z=CB>a=604{T2f$l1ftyxQ
zAr6HaD#L}ol=4ISQ?RE7CbVEYsZskHb~_G>hy<?PVNe0NTl4_~W5JGNFrN#O@2vzQ
zR!lj7K9T7qWQETdmT<MfY-(@dS<<}VTpmuZjlE~kC~U8}<6~Z;P`$5VXHJIYTd3D&
z@~czh-8*Mucu}WwpYB|5mkhECenC8MqlmT@$cK)^Qj-@#TqT{+RCHj_A}jZ5ykM)M
zP6-+pXRGJIfp>ih9~6{xkm;?3ze99c5_piL(>&qb^en6HY1{{fgf4aF4U&NOL=wSQ
zeXpT(c}7d^#aSW4P<d#mUlS3{+eWG<ssPhb{h-2PuO_68SLc!<>B(!-)b)+2q>o6|
z(7b6O(-1#|0xn{xw|>I}MSKw%2spGfZyOEU5&N=vNuYJwmubl3>#fWF0<0;#I)<7@
zHDx3(zRS(wi^ky)IGO5TL28!j2N_$`cV%!PVZr<czwMP8H^zlooYqxma)?ezrj}bd
zDJxwB>kj6e&{YUHCn2vh^<Cx(3_O~tpur`^66dvJEl~tkZDBKPtjcb@o%#;$k}&dM
zO0%&tnyk}dtbFNoz?9g&(Mic(BagSB^#d&2;!YA<!W5=FIO>dBBF!co+QA!Hr%CY1
zy_aWB<66;*q%~<0Ang<p0nja~{(9;Z=nfu8(^4)*2&~lUBhRlj?@#WAT}l;9ER28N
zNFc%C9ujqXQ!G+&s!~p;Z;-gMs9K&<1}N`b*!*lZ<Zzd37Y}K;%+&d^*6cIZqXE)e
z&Bb}Oc@24O>>Nl!dFX}tiZBH{{T-8=*c|$oRxZp7f~ZEzsxwKJ%<@-Qks-O6@V$eH
z{%SZ)tUPCx2Jw@EVn{jPEY85-12%>0oAA&9UXo+klb6v{;#KM;F4>2jwx5vdI0PqC
zvzK$%2N0V83Su+pzl7KX$xUotwH$&gzmS!QOf)ra7Ypq`!g~|F7nX1jqZr^qK9N{p
zvc4=B4B{>qoq(!_2SX{OJs4K3=vdO^CDN`~jYuL><OKut67Keo8@I&cCfcNN6V%kg
z-f305Sf*7n0RM2kxOR!~D02Ccz1i3x!o2r&TpD|w7s_zV>lk+DelFnEpLxdJ1iO78
ze4*~FnqV>+SU^)Ib=f8@T2mEunj!DZ>^;I0a`<F{ItHUKt=`BetL>?R*8#Wue_3KX
zOk&nz9)aTwJM!rlh>0U~Y=`)y)FhQzbD4?_KrF_EjMwH@11<rUTHzaEdw|H=zJYpA
zoe^c{7#%uNk&7`LiE)f#{)Pf8Ra%<mx*%3<DP#?$T<u+SDbZ7N*W$clsv^DEk8M=C
zFKT-J4$6L!KK^odf={R&w7K1-e)BvZAqbrEZmGcA7*ZNvP&tBj+js`KM#eJQ_2W{s
zFsITpoLu6qEuZjdA~XpEl7wftBMVJ)6Bm~yIIYxsa&Q+SNVXzQ`gA*Fq_5MFuG-rJ
zgcty^0OaWCO)<Q7P7?3HH3K9NFr5+xUB@NC)ifHEcu}S~@l`dM!+EzfRZ#0BDAn1C
zfxlQ-R<K$N!!sFK@e!tGLpG4jQTf{83=i(DIOM0o`l0IWDXb8*3hAYS0ea1#W_BZK
zg+~*-1{>BH<{qZ2W9(m=`)n<YjKFduGoqk}9BxXUa|rUhpHGK_1r_@hzBn=rDf>uE
zK8?|WDx|sUB#P>4UnBC!mpf<PxTEkw3b;8Fy;PCF#XX8s>0rtv!>~%^==@dc14KJT
zVeK{(SlMhySPgk%kWI)n8d{?3{!>g`a2D)%ei-1@ZORKku}TOOFO&3zs3PbjvB5L3
z?g4d=Bo`H-2ZzJF*t#LjMa&W^Bs(qie$k)~HsYg#xPIa$A$3{S79%vGT&aQ+8C~>8
z)<bh)WTcm!!0z4)NpQ?d1K@PoX%_6Zz`bAUjgAiU@2GH=;^4#oj))(K#bbs4Pi#;8
zLR6j?+H+8rB!0O)#5|^ib7&S#{9FEraC@>fH0!kIotmIUMo#d6L|JNDu(%=a^h7iY
zprjrXNo<qN^!WM!J-}(lJtjPkU<-5BFq4zWhwZ~j!h=bSXZmp9=@i4kwhC9>e=1&~
z!YlSCK9?qWgVjU~<M2N?*0er7r_q)j4rEn$CD7bj*6!uyV&ZSgVOM8}U@CMTn-p_!
z6%$J?0wlS8_Ybs_BV;<AwG01L{WPZ@f%CA=tg@UauSr@Y)Pl0?pjJCZE*0*kyL2E{
ztK_fM6$5Y*t9_x@9kfVm)kQUG_8l&Ojhq)DM5p@GjZW?mV*;zUzqw_k!T|g@{b@l@
z(G@;8nfjab_s};fUCTCI1;o5Nt{{0u+O|vfWwBIDf3Pkb(3q^f`@w|6Au44}2Lqt`
zw4m^negjG>eOf(2OaQRyymd`YaBwbM5u*sQxaWl6>ZGk>73yPQY9o$1Myb>5!7B(6
z$mE{h3$;?E#2bZNr4X+sJ>Vs=DPs@9A!3Of!JtAIEeDsZ+WE`mQQUSTibMIuY%;X;
zwKt3R&L1oUR-6yo*!<zJU@}AaN7@Jq!SDjowQA;*yMcnn5AQZVISs+X>c`rsoENL{
zGj}W0cWeHH`{a+jLr{%6^gyVH7GiX|t5rk|?lOx05t_o|XuLQ)bn%TGSBOXA0Q{qv
zgoWV`&neRQX*v}riwt>#I$WsE@d6n9$6ND=11%>mysi9Kh%Tw1*T!PwL;hZdvhMHs
ze~C2VZIjq{`CIf-f#lW-%SH}MPirxOUyQMA2zLVY%B1IodI+h*R|<#{v_>g)QQpEQ
zaQYFZi3+3%Q<?Ds=`_gqRp^UgM&lI+j>LxV8G=DOixY}o{l~@W!qbR#6rn7O2Y{gQ
z5|$h<l+GPQ1YFvTstyLcMO8|&RKkI|N*JoT$G)#YDB^m<3y_tPCW$00a6w>TGg%ow
zg%inyAXwObVK4n^Z{-od3w>UyPtAG{14M^i+JvXu7$WdSFtWmmxfGJpDqY4-Vl4+Z
zZ=_*!Y{9*}4VDNwuh=vspGr=ZzuI?$VK;>qD9N4ly^*gZ2kKp|0S;YMzI39WdWPto
zW*}gLsUdx^1^|#VWym&)<yqKaE-KSug?_UYhTlLbznf#Ls1Xa{SPjvj*M1T;$+D6j
zO%Y0wx6z^Zaw-6#DlnZ)37*_#3P4E3BBl)2VstJ=Ljp%yY-9l*R=o|Dbi%ep5fCB;
zdtp^tY<=aHgZz8h=&AM*umL%dJ7Ks`$avo4hOFbr;C{zDY8PeP?+IHzn^O~B^hfcg
zf6RP68K6JJtjf5eaOCI8=~WEidwxbJnY(BrHaJ8+GI}W>e6i4^`v#|Y0e39=B6{$`
z<^zIO*qda^M&R0^J5v@ixTMc8@ZoLBsIvRVCjO4ezpMQH(D-|OjS~*7e_3mOU;DbT
z`5V7mKSV+ns<yt);b~d?D5(v)F9sb;^SeJ)B=+;rl{NW{OZUPTno;~fckdUIyD7Zd
zV8#(uz;WWBLHHd>>X8!6)Z}bT@o~3Gxii7rvMg*rn5&~fXNKI%paSwI8ZTqNQL5~)
z3>FRz>pg8rnslB3DmA5fMS<EuX$SD`Z`de?580npmDu|YcE|8WByt#u02%`J!KYGU
z^il>Ykp%SuS~_xtZrz>IR+p-({$NMy>H_UhC8D?lLRU*L*|;IX9m$xhNj69*9&_*@
ze^5y>z3mCwzyf)?x<^Wo*vVZ4InsKIs6Fb%Yl@UFQEpD^DHkPn)mA@!<)9<I=<ciE
z3)?Q~Jz_41UPu>6%m_#Yc_i=3y8ch!3*}dAR1kJ&ml;sa0L}WQ-twkQ@`e_utpqaZ
zyknEc5dI~3<nJ8AKFtgBr}P{)&`~ndt6V``aVbf1)(Mn!6{-cp)guzD1t=(@lgogp
z(NHU_$6AJyC^67$?+1=KA6-+MWx2@FPi6NP=4COaL<B#|=2^j@%hFq}w~9}x1x!OX
z^XdlvX*ef})k4=JL-1l^i^COxIxQE3rdqf3afhcI>|Hp{xWdT?`~2yHK2I@R6Bp$m
zu>qNnGF13;fwi&IQ3t&SBT(m=>Uify)PW3f1-gJR%Zq?Q6sDYnzlnla%p}d_yh7~F
zE(hNfzZKj0rWiOzP>cp!FDXJ-fikV>#+)EVRkk1~As)F{#<emBV8yt2`Crlp#~^Do
zAgsvVKybc5Whw@0;UCchhwqF|889|K<GEv^L|ucWyyZ?!gDyh5nuiS@B~K&3OyuJ@
z{Tc(zrw+_P`l*ApLjpA-x)6scfW7z+qHVxgt}e(Z_(xBe{vA?@vD39y!U2IofB~NZ
zDYrZ40PMB3YE54c3)2W3(ycI@@)|c}u`?0wYjDNnsbXEtTGu18)Sxi#keeLu8Onf{
z=J+`_XBp7PQ5+qTES(`Q42zmChWe>#`YJO73m#qH5e#CmAR4`F?qJ4qM!Yuac_zo2
z5S!~3-C1_)#=6yY#Fh2E)lHLhx*M~jmagGyZy8~=ngmTZNK+zp<ib|R$~b=!AEdEa
zP@paxqqOF8D=vTPf~M8QTs`%g4;_U*n3GCRf%*!+d%oKyTk%RFbe{~SvQWSp>H<OA
zpcPC=zW#Z%KJ3}(zP`3GNrfwUHydgrp0E3E4NO`v6O--}dgkqwsY)rf{YfcWO3H_^
z!5re+;sDO0B&D4B?K2>*z3WiEXZV>4PwXMw{oavuqFou&+@n=MnPGQ_HByjRyu)u5
z2V|;6HX!jY@YiIDo0A=!RwmI%>9i_xQ*+IKhL#pOO2<=aTizLcp`4wr*6d={ju$JL
zRmvY#tEURK*pr6ioepgM*r#CsqQ~y=u#!AN`svF;I+crGQqXUjG3*I4#paPVM2vjz
zv4<O;ojzXDTEc#f;sPb6G?(;HpR@}xy`6Im%$Fs$<+ZRdl?zvJ8X}gn@?1}@$X7jh
z5*Ng22JsAu5ZN}VAbc5)%b1hFEF;DN2b^>GwsT+fPD#{uF{I!MW$sjrXT5+0bGJ@A
z@@{+W%i+cBV!V!R^V-(W$ox4XS5l1;zVZ$dG?myJ?b8CsZ+R=LI8;joH#)%`20AoW
z5$hL9J)p9oyxW$gH|c=xaP6Z0^ZNE+31+<T3)&bm1h`>lz`COlB$kn?zp{m?-_ipX
zlk6{$IodwHoRg%8l+uV+?yQ!M^B)sTtH}mFt&!Xb0*UQ51j8l2xI{eL?htkjJ#e_w
zKbwswU;_M$MPuC}0lW4eB0_@%#Ps(-mgOP1h;h^uK$BPK>?JLQa;V_HNrH>c=jV{q
zJr5={Xv81B1|1o?e7?FXu_m&TTv!e<SOQRN%<z`%2W9~y?DQGtns5mFAjHT33Aw)?
zECf0D$k#Z+M3KX7I23q`AP558^ML07@26Mk@oM-5Ph^F#5S~BD#L4}|Vv10cLWEuR
zdx)htGq)&ySr%9W#8=v_<3H%A1DY2>Jvm4HNb&$OpL=cUJ|uWKL3UzBcTUTqV+58D
zdFvtjBX_m3Nw*LYwm-M_QlRT1x$zN$Uz5!+RXhU5>5djLVingi(rcDnrJ=XWOwms*
z%W1}HOg%EvL3t%(oL|L!O$B2DDAU&|QkzqnrCv8qu?t?5edf?hXGpEmzNnZ<ZkY4J
zO;J?bBmAPG0a{D8kk!GePUC?bV}f^h_BFkYgvL*dsp)G-LIntm)lFjH)gj-g#W-8W
zDpCcRD$Tq9Gk<5%y)doYY|#c0D5!|7?zt)P`AQ}D1oFppc4i325vfLYKWUKj?l}Ak
zI50BwU%9Y)wRHES;xmrb7J9(*u(PweGoHfcrymB{glRPy!XH`p#l89)V0`;7Pet&e
z_;~vV*9+1XF$^T?3_U9RopQ$i>u~$aKYrU-Kg4;-_BMXg&!5-v>+8Sx+u8+vmCyb`
zYT){p?_}r)L@hLcHaL3(JsNvBmw}mBcMl4jJOvf{B5MT!<q>L*=3$TB6Qyz%`;DT?
z-7Wf)*<c}J%gHyp^o9&1X7Rn>_-4Ju3+=~O8vYvocGsIdh1qH+bit3ztc3N!tCf|t
z#H@cHj@#8_a5|}HyPr?5dt>r;=!sCIB!UzrdD8|$lVEY4epO@=M>(V)gmAQC7H|mm
zoA(t{3|KOSd_qN1gax6-g2#){*z({Z$wz%7lKG)&5-kTPqE@@G*R|Zl1O|$VUM`0g
zLK8_UQ19EVd9UWW8bIg~%d*x4+p}KeR9S{>U_m4oqScr|&{v6k-r;bO^e0QIN9K4)
zzIFxFB6O%##vrN3+}E(|4AfmNc`6a0$4;%UVkXeIdf=%yfGhMB;MK;eYWId5pqb!N
zWV{66(RbywtQ$kaL_O(rVfM-yC!zLvQF7jUDdxOkx~u6T2(IUYrzb!J8i@Kx7Vs)b
zul8<c^c!4BI8(QIs+sRJy~gkZWS1X79kYxi4K^_jErU6X>Js|Qhi38LBgpB+n!L%B
zW(n@K#<S963oQv*m+QVR0GvExej7*MTNhiro!{DC|9Smq7T{(Zh9sxV(s^8x8P+Ia
zFm@#z!(2$m2oeo#@4#Rg6+B^<{j9H@E0QWyJxU2Crds#(;;4<)?`mW6-J)x(a!y+{
zI|mo}c^KB|$;cw${8LyW9J-rhDL6%wL?MFy2oWWN(s>c(t=Fc}DZo(^o>$IUZ#rxt
zoE_FqO;iw8ku*%}qDfX73(4DVhj4Gp6oC?(Ir8D2N=}h{uIx8nKKF1X)nXQ=FX*QS
z-L<tZzx;Huwa&=W=rgKZX=xX?x49rYu*;IFGj1fk!dTylf`DV39GL;kf~2Z~t9F)7
z1s=&&bw^O!qmqHw3?ui2$D9(Rl9|PsQ@!eDP$k=^l+qS50ajfNgx+TOfa@8eT;eOd
zQ}H{fzfhAl0X}!{7s}J@4D_G{efcBK2F$LR>gm#M3mw%s8Mrgld2oi<RB{W7>*aPT
z*+Q??>}9jH#Kg;b*u5d-<4BW-*L*!Xn4M@+vZOw5Af7Za;ghbUV}RiH#0n^*of&&X
zmTe)(s>R)G%#xZ*j@6Y<qo}$Fqz5~|c#VnKn&SRf2TSv0yiG@>I2UkM2Hk?!U=-UC
z1|hPrz+yxdH^m-oYcjjW1rumI3Qg_rXPh=j&){X;rR8&&Nh{*w;XqME$c99Dsas@|
z_b=Vrl!!f;$xM6z<RZAx3(P?h3WwbLda9F4gJ-6T_N6G6637JSLQim@Z&S`1raT;9
z<D*eQErYA%l12AR2?_3$*r*W`L26<zMvlW|_2i8o+#sF(nOIY$_9)7J4Lh2W!3LJm
zJfp5wDYr@x0OOy^&ys5pSMV2K)Ykss3g<ym?9*6iwU(aA;$jet2k06U1}FWDxO=8o
z5LHbsIZmVQN)cM0=wh6S8tCUjgE(B!pYjTNadO@gwVhw?vLKzZJg#Cv4TUrEd|YLQ
z=kYYhh?enpj*TRY%@;ySj``4vgp3UPCVFC=VLKQyP2Y$O#u@Rhwf3Os%DCXjAK9_Q
zIkdF|AD3gPgNm9IULlKV<k(d(N~Fwy)rslhpClD&!Qb>OU?NEN?}}RQ*+whBJ2OLm
z^X%!s2M}b))NzVVvsa{3gZZ6L)o5}NbjdYx?;<Oi1IFZ*2kb$w!K%U7Ihzz&4NXDB
z8N5cECj=Cw84|p@rlHAxIMjBE%MJyL8u?OoPV6TPAOkqew#J>_<jRTL&9mGo1J`O(
z1`#AiT}el$q+$TMB&!pBEw=;vg;ie2ZgyIb5}K73Ud6v^3b;4JFBKG&y>n{|lD59J
z10yCEUFHq}*Q{&0#i_ByXh?Xd`?3aMR&hv$(W4(KGoD6)hMCAgwp8PG9BfPC1WMJT
z(AnaAOE|nc65tqMecGEm^a6C^xtC~1Y~`Q{Y+f8@?^MSaf<0`d-maNHw9(S7KJonb
zQ=RZIN5Q~^3LC8E#X-($R3^~>uUIez+x)NvlbP=SPbN&SnexBig2^QNArq!yHZ<23
z=4mzEz`h2L`_80-7duQEy`(yOVwy@~f*Dgo>Qwa3DXDCKQV0bYXy?M<j#b>Lg*6f9
z&1+%WOTe~H!MB~4j`5k7PH*>E8A$$j7&BQE2`%IB5cs{-rl|#?Af>aOlRD${K!{jc
z0ljkRf5gmzk;KPyA||My<+0djQ#minFLN33=j&%*ff#s=Brl=xqVl*h4jw~KQfF71
z;|3^{pcztNFZ@Q1<=I=M$Bo{`>hop?9C`)y05_6+PG^{Qu8liStJVfBymX;Pe!Rnh
zxraM(BJK@lpr|D0d_eGT{l*mooT>{)Mw<JAnRQg7fIKgv74l4TYWj1wu)>BiO1V(6
zvQ#v*yYT>rKa9I*oUJD5*=&N;eTZpMDcRLa+={B@<F?rS{}D<~rcblMb$?;4CD*N3
zB)VtuJe&7sq0nh>#sR1hb_-jUes9)mKUvqv!?)<+M#R;u*P}G%bf5m3a#TzZxa!Hm
zMJrGkyBkV)P8qXWIB1+Avf33)8vrw(TrD1O_nf0#!Jn0rQep?*opg}&0oU5e%EEp}
zb8Bq@6Vtnjtnm(gKqK8=Yl!_Dp!F>E%A$}Y6J!{|YyvNL2*wAx&p*85tX=B$a4;Ik
z7N|P+d=8*wqol-w(<dYpMGS^3`nSj3Vg{MR2@@e#0UBwHP!gsd%Ao>hMp%%>-EBlC
zz(EdT3rwbnHWmU^OzgM}p$^)heMur}*Lvu~er}d!e{k5X26GB<O5L4==J4o3i#Lc&
zKu@Un)nquF;M(jpnjCle&h-2+6sqD?t3A2zabLv1!aD${GP!vG7zso?0F~vdd7lDP
zA|{g{7}}*&^|%)#x<j!S*tUg%AAt7L$&BcWaBS{wZlV37G}tUtZN%-gm8)2%cB!6A
zrBl|scm=<4KG4Wjk4m+wg%a>Qt)5$@j1@1QTPL|<>VuU&JGIl5idC{hxx#5am&UVP
zF`2KWa>cBbK)qtAYUOi<Topj8C5sxApmJ%{%UFf9jo1%}n~=!mbJg<?p-iq?B$%0!
zZN;tAxLwU9X$n}UHT$$wNu&7`z!r1GjEz>(g><pHg;w#*O4sqjsvO1hd1@*YuK|0T
z@UxPo({nqQJ*rwqr92#ND^?;6%;JfBS~P`DCG+uI;e(Zm7vfpM&bCT`Vuz@hz|LwM
zrRfp17svm}Y7XHY(2Zn?TyvAv59p;`)m0n0O8SEpw{sODNX9M|K7@#zr~&8zf_lZY
zfI_5Z0_hP!5h7R(omRM5sdPLKSQwrncTpAI@~k_i%eoQ~xJ2B2gg}z_;_xFcX2qNq
z8zGnTxn#OnNmnj2IZy^Nd5&V6a`w!(v^W-}(`QM97y%7D<$LNrpix7KaQasprxjo!
z;eC(OkJzj*^NRuw3|-V|$^*lB;ez5_LA*ZZQvTD%Tv3NDk78M%$NE5l6KokqS?-y~
z=!wm>1e18mAt>=;a5bR2tpoj?wq$J0a^1;SgrO@@s+PIkXH6LSmRPDe%T{1DtW)wi
zM+jrBbPf|Y4K3TYRV{IR<}^sQxUK$TH2jAwQANN5T8)*?Dj%C_IS-MS$e@Cz+S};h
zjN%(9v83W(M+@Dd6^If*4~H2NPjnl8?LFNwA!cx#DJ9RzNy<Kz#^4|{nHpDI0&bCw
ziulNUhA?|#_rb6YmkkFDjU9N{+o(gn|H@{a$>XIEQa8;9@hOhFV1|5&WfUvAY2{)T
z(|?|Xk7tTX?GGa$*;&m<C(?j=!FE7)#KM%54q57*u&!gDhj=y|olm?M+hMpN<DRo_
z0oHi{Ck17EEl;N$bCo)$^k-oT$Jvb&3%RK3oE~k7dlIS#BoG85pRFkFu7}siw5<=&
zX*r?80fjov^Yt)pu&axjq?jSenl8W-TnZ_D-Bi^kqO5|w;~FWuF+&nqqJTaF`#?AB
zu95P5Os<g-8Wg?X!6MjCQf?yiVwMPqP8}th<e+<f0MLdl&uZtX!OXlk)g8=e)~sD#
zrgHY>rSLBRT_FF!wrc%xS#l)~SQn&BC)0zNfNu8G_GAncd8oU6_O(;A8#NjwmeAt8
zYo(^Wh`R|uro~#DQ|X7kWyDiHunStME$@UQ8vcrIcvJ|dBTh4~cZFuP*KOTddFPG-
z@ao7oMY(g5$F)P|CB?xEvjz3r9jy77McUq_gJr8qv}?oeYUWI&IdxCHMz{2gWOpL<
z!l8kB;$VeJ8UQn5u>5l>=m`)&A7`+g@=Gb?jE7U0Y2>^Ba)KJGP|2xg(8+iRN;$uL
zv~pfK)N-CV7!tjl=K&ORUOJT&%~+xH((Pq=BPae6T~3atsmM951bSQvXEB+V@UWLO
zPU*B(yq(^#_XW<&uX~Sx{L3%DaHtg^y7px)mCodf>GhC>&p)g|SF9_CR`lmJE1o|+
zid)-ODwoYw@$V-7Y+0f#oS{_V&tEU5`HN|BYxB3?qaXfQ$JrtSlYQMBbPugX>lUE}
z+53!AgmMXhkVOXF&+mU%tnLS(M`#65x-7%gzvh0ZxC@L7o+)X@uzOdV=(Fn6`r&VR
zv7UHecu)tF&$#1ff?50W9}bxHF9sa*jV9BaWI%7TCw4lt&p)&zEzn<(-1)I#{6fPD
z6#)U)Vbxv<82yK=h42U$8(*yNmg|x0Hz2KRY+wbRQpi7r52FLZ$T77nH@ak%*nZ(1
z?`c&B^k{Iubgh6Ud5<6T+-WQ5Ewq||$Fd=~U0Q2F@2z!*r&V>niuIXm5;KE%J|g2r
zI?jhT@a~YzRo|(r>+*@BzV0!_$oQhc5y<b>1=atVVCz<$AtXpYTNh%o=_wab(1oOa
zPdAc)AR@U)6k+xH4#Ua!$?SW{)@bVTM#B<M%@NM%DfS}i5M2vkgdpl_+$9_jn+9PZ
zxYJK?<j94LLTVqfWQluCq0qn9C{!w{6ie|+B@Js43^FFW8x5|nzYqQK7-I2<O{};C
z0g=edIlJeQ=L5u~p&Fwc>Aex5#c?58^yI4R+&iN}#+7JY&`^xMfgCB!I?3cKopy$(
z0c~)5b_WF*>l*Y-*^s(Z`?m=g#_zvPz<-;7|26^tpb5A)!fiQ7H9GxE(rS;hql*eZ
zHIDs^uP7@F##Fk+F=e4Us?V~0^{j{MShFhbyyJumXH3u(D6@n78lw%e6m9xS4S>^e
zm@XXtOg0K#!&J(Z4ac{gcbYBD(%<e{!|KEY=|Eel{a@dVz!28zKiBDBez?uQpvS+w
zy~jwU`RAzo>)YGJLD-)j+@D@(b{Tuo`s`uS_mqH{V&Isp3g&|OO&y^_!ypgNq<(;y
ztpSxV&(vPtsk`}^H3_82_J9bx$QD%x#BS$<;6otqTt~w&!(CtZX3~9*gT#};@bVHz
zVwK9}C7GUZQi<K66W9oc!w4OZ`0JSHIA6vwCr*iludKY}L+8Vs`h9>DlQ>9y7>IMr
zS#LyET{^a;!%>Q&jA-#UIJcCO9o9c9eTj1~`w}7S?Q{|W=Tp#!=;Ogc#TRm@YhI*q
zp1NE+j0+U#0IZC{YA71;8;I2q1VDp>H7wMyrq7hgPd`BSv?D2R&`EF97YO8*ok3To
zy$+6r5M4(D6T(hjif|fn*g-&4GSO?&hVk~)=okTZx=EPI=FP`sbufP~gAJR63%2gK
zFj<%&3{5$pL2f{?yr^vE>X0M|AvwfKXPW_R4R%pWm!6Eg-6qbQ8#B_y9X$K2(&=To
z44NQ#wSJJV;sGx1IS=TFG-f==Qi&b&0hQP_9$>TbtF@<DO<I@aLiQdLY&<Vb!H+1#
zwzRcS_*cH3E!Ig!3%9m0{V9I3`{6I{XZb1$LZrJ2=F@oi+5Id_1wVTVW~tyWo|4E)
z5eh?ex<JxHIXfLYtfxnyu+t+0YXZcwETkV2mGKY@q45k<qh}D>&O;(L9zsnpo{56!
zjPa0Yj)Ele1)_!kP@<Sd#CR+vk>{z&+<FK-$4P&ov~W+3zhwnq5mUBNzpS9;VVORP
zh@pro4O@J#Tat^fIN^{qHmvx95sj%RcH#t`K3-XCIuHab>`FZTBuv6S;9qqeVHGwE
z2Fr3y=&NL*u+Fz4`($yH5k5luppVr!P38pVP7#t)ar}pnK7PP?Bz%^7!>hwXi1#2$
z!T2J`AO*Psf=T^Smtd(Dl3mUx6jwU9T3-ZJePopt#HIuLf)*64HvXuF@l~p03CK)1
z+tC>fTVVdzQM{F#i$0=<$dh@$3GoY;Vs(KLltH|!T@26r_RA$fKUq?E3OIn_t?X1b
zmu*FEZjo0VDvIPD<kQ@iYc!t>`-X!&px--9AauXcgwcj$M3Zw-4n2aHNp;&h_U452
zPDi(tbW}D+iF%oH-j64|w<yWG`_5H7+Az?iZ#TbVVN>}+6s;)StXAcUKLAdtft#OA
zkVv#WH&El*Dzy|E3^C7C1z<3k7boL@3bYz7sc<Tl*??N43{fpVZrqfr`Erv%m7x=h
zl(cSyiCqSBKUjltD5<x?6wF3HReAY1L{MA;NXruk70GU33s*@Dw%8+j#EfJ<g|wUO
z?D!!VL<|kh7K5Wq3Tn<Ux}@y9Vt%^tc?;&Rv#5U(@MQ!&7C!QrJR>Q_VfUB~)ndt$
z!Ktd4Q92>17rNt10Jur|n@Dx3iD*#ECpk3vq$F_)1TNe`i2+|H9dCyDs;FyxNEZI^
zPGtZmZILO^kys$6c!vENe*#4vOalQ`oOud%rH{9+Zx<I_%&TPmpnU-dh>aL0q;c^5
zAKa@;4$&MFK%@UXyJ-LTn*;0ELc?EwWA~;X=3_n31*9_?{OCY;w5g0k5JGOKNZ8&D
znDGjgZtljoOK{KWlmo`o)YMtuClyi|24v(y1p=xTv4&wQF^ZBFYy89lC?tV_f-IuI
zDP=cc8YUeSO!qSfBw5`cAM^!JFO0z)AXG!aAk^DsXbj-|HdlpA5I;F+sMiXHpQqlm
z^JUR&E-D7`NW^;`Qt?_5f<025-&C43P-938JDFN^%)54K(mUs?&O-6_c7IyEVcm?=
zvcd)Y>J!rc4+$WHhT7SV9GBiU93~X@b=MB9#v5i}DgMJ>O~+V2vy<;E{-c=bP$i-K
z8*t|A68dI=iG}v<XK&(w4uYE5Ufd0fSCWu2+8oLE#Zqa0xQVA=Cy#Mas%@tfb3^5Y
zG}z92m^R1Vt~qWiz^~C!k)3JBfq!Y<X(v~yI=Pym$r<z<nA3?yU7Yp_lxVy$F;r{f
zAqoc7H+EPOR8k9wHTU6&fV25o$NSR#A(MtVpn&9VTn3kMK!X-01wduVNao`p>ElRE
zXPz%3IQ-pJAWKP0xi~2S$g7fi1*C=XO8J<p7r67s>mhu!oF0o1<0iI-J}tQH@bJ}<
zM{h{4lzkV?!;SvV;1zOp#bC;hT(f*VCg)hW6f3l;a}ux^_&ec3tKO>(IP*K^z$)zs
zFWf)V5KSzzCC}W_lorT(T7uYsH#zPD;uKt2r~apAjh+P*!G!T-|4`caIkHyKnpaw6
zR3h`ani<3|*wV^~j$IbB0*)&2!h{1Q08S86j;jO8WLQbTVeblOhw=wskd#644T7rW
zpUUbBFI*PC4DJ?bNJ(Oetu&q59v(U;_Od=u)Rm3en@)?&*lOPEft7Otk#&e)alC1e
z*P*P_TXC<3Qc4m*MI9!lBI6vFSnwLRSKL8dMNvZH4Ql#ZRvo?Z6qmn=T1&h{JlnZf
z%X@YOua<WNFRf?v5!=YRxooKIu_)kQ_V|V{F5L-lJD5<vZe4Ev_66RN&<-e6=J$4Q
z(fZ#0ZYRIDhSPrQd+R&GVRic7r^D~lbp(Dq{0ishFBjd7^`AG^zffdn`VC9$`WL>N
z?hyu*UbQ_K;+iiTrN!h3i?+yyXea4!YLx!vCg}^{k<*|0ED@PgTD>c>0VF_?Q`KA&
zz<JmwT|1V`-~X|-@&1niHG)673fiLC5MWZ4K&82uRi8~lkM(tBch-eAcTMG*`@vI$
zk>E|iTKIR;v*beV9uT+|fWj9Wh>QJ2Ac<veF~NOlx-g}>2ecUy2t+umFqi!7>yFEz
z-a)`$GAc@Pw|M}>4n%;7{>D9Ri*Cvo*{Ff}DQCMFH0m5f5JMLLqV)m@QAxv?bzUe7
zP7(85BOtkm>C2q5giW-ZF<6D_5?ekov*=%77n9=pIq{;Ba{3{6HcUuEGX{E4jWxR?
zg_E=513B+TRpX?G6{C8ED^hhaoq>ps)gs3okKmhvJTr<9#PI9I)2z;N)D2Em6Np(f
zXfCGw7Xq_?xS0R8v4I?78wfbMvEe*<|A#SIH2-lc4tAPLmN)U`kj`u%HNG5<5TmL$
z`;I-|`1U(om|L)ttP7@BsJf=of%x#;Ga4{wtTR`9^ODbcT<{1Ho`^D8itL9tJ)*H&
zN}>&-(Z$Bse=)WGBGTM*M&am(y+35iVMtezd)WN)lH_pHP7P8|WC~U!1rv#3>dmp3
zhPFK715aa^<VJ>HWC!Ntz~S|v!jLv*VI&8Bq42J|FmeaW!q|BrjXGb+A#V`N$&l#g
zP(h>JKX?^7<YkC*FypGtzUx`^!_FUAoBj%7c@~e06itLMIb|7SgmlmyfCr3b9(nVL
z`^LQI6T+aksA<Szk}pHbT0jRRERBs)j{H?4O6{rd6jDFTTrIcnxXgg-4izQrWR5Bm
zUH}Kkz923)57<yh9XBQvwfi@s#3q8lVi={1{_+l@0yxVhF&J4|N||IzyvliKkkQ2i
z$QkivUFjUbEYGbhW={sKgiTTlU2%l$%GmXX1?N3vv|Do$wBfcK;ma39ad8f&M<=0G
z7)bj>sKsKF*p5-=Tak{qrn3?uEDXUcDB=Z9?tmf!r%{yZBMb?-2_i-h1oY?*!GjR{
zq&0?n134o~(h)s%j!_r!{`C@W0(%zpMyr1YK@s?QDV}hIydqynI7+sxHhV5ubj{d$
zn%r4)1jRu<QJ{Y{coZI!2u039n_0u~4XMFv5^jUoXVhumomTh$*0f8ATBESG(;aI;
zTJ?~~N_24zF{ZS$etQ>h);o)c`Ga^paT<zK#&Yew=!DwVx=FwG<@6WX<Yy$0-`Kc7
z1ofZk%R1?NTaW8sXl;`$Eg(5khKE5-5j7RVYQQ;l$ChgzY4fT_ht|7u-(*lR7dZHW
zRyO#l)Q$%_5H&rOd(pM9SjkB)@?uV3<a+LkHm%N-b;B~u15?^Jm;S}Vo``d?Mf3rk
zzJGl@`x!EX*2NVbNlFS-l^_-DtcD)a5=6s}6jmXMI2-*Z4%!h-bMK$fe%=qf(n`v6
zMgI2#mXBa6y}gk8o1cjWXJJHr3O|}l1=pZO@6k0@CIqe!mdXy*mAJBj)lq(w60Wou
z9<*0^*4aaPjCa^KN{3gc2&oE&flu;)hpC0m2$B4s?yw4UiCb=Ql)j7|Aa_D3c?N97
z=VEk>7r;SAux6q|Wg&ed{4?v?>d_JgS|i#pe*gK~#+UE<8>Sl3ab0f&&0fBQiQpj>
zLF4^o=Fp1l^h}vK!mTf`#lw}wi|-fTzu;ezsA35<I>zSK$t*jY+)bAl`Q7j#?@c_J
zHL<k6fC43H_}BN={QK8mKmTHV0OfwBN1N+kKUlx5eZd!c@Y4q>@_wD=zKNh%-(+A9
zcT4LQLTNAtP6d(meMwi`)dS9~9HWp}sae80eKRf1DwQ_{KX*z5V<T)x49qA?{DS{f
zux^8_PA8=P<io?=6Edo4?u)X<CJE*vjF+|Dvr9G%cmNlG7)sL-fyZWKl0`xg=zuo`
zX9)s0$OI#+N(bSn-WjNotd5zTRb}X+Y>1+LI_)Fqj^X?7!Jm9yoT4x}yj(P}P2i<c
zEI^F%;V@zz6Xf9H(PhQqp9d;Chh(Z#G{)?lBln5I#gG#uko}BY#nG6&@cI*UcHA(S
z%;SIseDICkHOYe;Zixah=n26sLAseO2^YoJT1pr#p@yp^wGAexK}HHn_y<d*`*fTp
zQEOKeNFcDm9^Ea*a^Nrx8VF)OvDSsWmvPA5-1NtkpyLdB{e!POQAXhjObE{eZ59|$
zlxJVtH%tjNZm^)YhV!1B5KeD4KQB2-PU=H5)X~-+ry|cCsZA*>r`D!?g2Jh{L>-jY
zmOr<0E?cfJWX5Vm6ITAjcAj`tSzvjj)dKxgp!kun4m-7DY;K1p9HgJIItE=#sAWG(
zeaq-LL6?e(U>4#PkDVNol?ki#9BrA_`X4FBtT2=2xAezLn|-P!%{9_W4KXGD^Y`oD
zpTF>KK<ZrrAKto=(b$+4iuTgNLXZ?6J2YUkag4I5faw4J%$ZYG*}8<nN#6h9dY{F-
zr8aj6ETEh}a1hUl`t)yL3FXq-3i5^j_vd#Xcw+@8kTwhqPlytVsF)O$laAnNSNZPW
z?=0n}h)lqO-`NW0QGCge1DO{N3wK1N57T4PBPVl?>Ip8wWgllJ>H$KgqW~cE*bM|%
zh?W8&V^shUdh7;5W*IF+3=H8Mz~&mqa9g>%q8+E$8qlD+?CUU{>`60*o;I?^%C;8(
zz`*(7?&Gmxc-v?5*b8=*k1yc?1y^h_xCh?eF?GE7qsPwv;yunB>e>`7wV_dKJeb}g
zNm@_7s_=eVTa|?L=D`QwQzyL*{}Umf=rAKefmJ}`z{SI8n<Oz!)c1Wv`aN|RYt(vN
zqdU}kBsZau0yq`>5wuP4A4s>wRFcGakGnF<5vJA4Y?Tmpi9EBzl;Tzc<Lg(<d&_0f
z1siPKq7j&_L0x}Y7zyKbIO3V1@>IuTTlCKsniV`-h<kicu!fKWkOzu?;x3zWG@)r&
zz06_EIdjt|rW{Rk&(qZ=fWb{@KytEUgYb*&0JbHTAoluk-u)4#RLuQ9#E>-h?%n>1
zOR~@hU_kNgDFYVF{|~PTfqe4@3Fwj`KPs1&umG!LE(A%vh-~qHEVoF)yX25p()t7!
z_#3m*RmZ2Rvu<y+yF{IL2}Hk~3p`nab(CVF;f^g<jeD{3Xx$4ZlAG2BSx^^~B_&po
ziG+DYorU<;d5?6V<=8KCKFLKl7}6{2rz3TgIg%25Oh#e8CG{rZLR9kOK&#Jt?gvDj
z-kYqwe>B(uzgQcMkPw&OA>v8*;0+T_b0j4mi6sD$1i{)*0<K=`zedRx2LkzD@Hr2F
zdljcBDTqPOL2A}W0%u7O(H0XyGG{dn&?Q6w4+AE3N9%y}yuc?u$~TGqL6j~Vjo@@k
zj}rQ7{z4XU;~f^hH)|fcbpZ149n<v-Le(D99{A6xBEAVtJR_BvCCZ3j>aYFs`TI}5
ze*gUIm+!ye&#%A2>j!2{O5XmmfEPz_YBC2czM8;b8OdmhX|#5#Hl@sLrddwUWaEg9
zw^>vb2dK<aQj!=4qdvfjD_GZ|umO;z8ZoIck(I~uoWN6+pUgFKTeuX6i(o$Ap7)fy
z5LO}W(WAJHW{iTT6cXp1ay{aWI~)A{T~qo#A6B}AxJU4tJwr@91XKVEN-7Cc%gg7W
zBSiISUH34WKd?g;Hf6c!_r#=E0BQTgyI0Hzw&=;jR_O`7%BK0fghdlw9eas?efF9+
zEaadMvP&RP3x#k&5DfgDwrKN0QWFO!&hgPMA0K^2oD6GoQyd==hu}yFE~9EG=dn|R
zivuB=7l3DSA)Q4L6RIjCW*r=S05(umV~A<QTXsFp)Wk2e8~a^G`Qy`nr&(aC+Cg1e
zSs=hRe!)ycQSqNr5p)BXBG{8`@xrb8g)M^b8Jt6W&vu17ryMNbDEZtM^yFRLZEKE@
zlyt7|>^6uL9Zvk+3LmQONZ&(5C*e13;|)8<sTWhc5Jjd#qYOJqV#h6B!6(cC%#LgF
zKvX|<uu+svdh^-r>zHJJ=d<}cpADzy`8%J@-}!7vR+3M{eA+^8B<+x%Y$ae)8`6DG
zvge-8XBYX&HT^*_GmjU0<!N*^89yPgA-vhgSk^91XOj*V_KVa0RDPwk03R7MN&?4f
zr(`rFF_JUE9^N`dy5xTbJg21DaL5@T#}h71IQJ{O5P)uJiiA}2;VJoZ%Z5esFZ{j4
z*_UFrJi)Vclx*3P$wCl?P+aqng7!~D8=$5<ZE3NVP79)2Y~SX$&UV^WKqwkypzMwB
zlfBu3It_@x0uR>0l}i3HnMU$5N+CqHT-FE7vRK-Y{sOTYyQH)t6RQJ<HC{pI!$Xys
zBiW)RT;-sUh_0|gC6zxHnxtbGfPGja-9r%4NcP~xnMw+jaoIy-fv1dYP*3O`t+=n?
zg<^~r9k5&tuJ2}iYXo`NC?CXH80QkiAzPgmu69AUmy|RL11qT3!ybZcI#0<{AtDiW
zXl%acYQ)NZ5<A6Kqp(wcwVps{=r8FB2+8zyK>F!xynBhhPBDp<oXa#P9CwXSc)Pdg
zY)@|o+l%3Ry8~9n->uCy3ze<TY=M132uU}n6b8a6K}9KIA5I14LFF`zX>TN9+|6g(
z!%3$#+(z);S<itr!;Ia;I2&-NtgAgAMw98n4ef3Q^zhea5dOZlUo9<gd>IFmDqBGv
zU8#U%a;ArdLK|bpJ*MLtJ$4F`PlyLdy8NeW5~}hq$`Vy*MK+fW5<6j=wJL5{DTFYF
zqycZ!5{8+okQeira2I^9p3~gD9!}a$(XH_I&+sdJL82(262T>cM+sJl7lBtdDI4Y0
zV*Q_BcV`#5e6<OzjD=D*D^unG>#ePAq|Cvjq8Y=(%M(&9aZ6h8%6$@l!lm)`^-QhN
zNLo2&aVnicnNP$nTrU%}s;Jc(S=D7L!Z7Eh;F7XZzDy9Sg+fXOrSs8%iL2?{i#YZt
zH55p6Wherb)>TiWFXuw<WiP-GYPkas$gi89=`DPFjd*b9lRF(nmBQqo9-UHBC3qdf
z-}vf^9fmEW1x?n|n@AtJv$?+;i*T~ZPHRl(?sFi3b8?*ifyYOdA?PeQD!9Y?!FuS?
z!ZR6PqwM?+Dh0f<$%FZBiiD%x9;I(W)0cH5T0}s8_GeLj5rH)X-DEBmkWHDuAi0^6
zpp|2{bc(IDCzIhio!HaLc=@;&Ik;q30bT%gO<}(cB3ZZBzPnxvCn;|%*drO|{7cx%
zW2GF@;<V-r*B04kwqVMbGn0@LCewi90Sw0CyL9DPkzLSJ1#Xpz5X2mqX2ctO1&XX=
zlAApmrC(Nf3UKoT$QY-~T*~^4=%kVT2-{BFo$>%cSDe@t;A+L$AumL@7Dsj?tdRt@
zwG;caPI0<|s_iYWwj0}7S`Fxv3=L{+FY;0Ndg(~K!SpkGM^qB|Cc0~30@^_K(^8|N
z*!wb3%jHu-KEDpuaZ-oQ#sTgSdj)ezo|c{p=YQ8x2CS`c&%jEBzv1vN4vJ`mbylD7
z@Ba#Ku~AlDzz|i|x58^*Bbx`mZNLE;mKx?9F2+VHo$kV*6j~XXE6v&(V$ZDOA^RbN
zJ6uRG7fCiz9b+j+G$VhyWgT1#?``dD?QDMBm?H~v@0l`bML(^Bi|B;r<$AkYyPMJ7
z4NT>acc}JpwOTt{(a2`>4yx4RZig;29Z<bj=ZVP(QgrkF5P3x~Y;$aPR@hhut=<0V
zqzEWh3F#rhE5>`0!@wk*uS6F4N;GNLy2sT)B4yW)Yn58_xE#sUY<?G6ByXd~m0JD5
zP8U<<s=e2YWOgdG<3uwuJSo?v``v2)sW`61nziGbYAyX>M~c->ty`@$a?f@qlPTXG
z+x1$eSgG~X)q3W+8!6h2TKS|t%-D9MzEioaSL*3H>K@<Qk@RD|R!^1By0v;Xve&Cm
zbCuCSs~&rdRx{1pcDC?b8%3M-o&NlMG|kj%(L^!bO%&?&OnKbhsl=z%Qu-j>-8pV`
zV#nX?dc9b_9UhgdnYvvoj=Pb^dSTdnYNosM?&x?{A14Ow+vslPXqsrIkH^(vGg&?>
z7VXjev6VeOYG<eY`t$8$a#Y+a-`cH48mFJNW~rUIeXgNy;l|#r)a^v`R@SXWi*}^r
zs(Ze38|@9JXWdi+u!hmb5N$V#d&S25q%pKJomw>2jC51Aoy<`!G8_~lwr$s@iP~^D
zIDaly+L`IX9@h))dUOB$`MA-mBd77`v0sQDKb+s>qU~(vtavuX`{Q}F)~`3p)0UmC
zMXTd!q8o|U%4hweozCo3Yj&a=?T*W}OyWoWByWGL-pg&J!9x;my^n;;wTEim&h%;r
zX&ZR=MvHxn$$dB0y{Vq{Q@z{FNqN*Q)o+gz4bZ(6J<e9^$2;9fcigCft?HR$JAGWL
zK1@N!>FhifoiKfl>;}etALHJlajlQ*#oBFV$If<}cBX09z?<cUUFbgB<?@cb=pEtx
z5bqo6{UTjHvujCX9BX16_w7crSiVUp9@_zpYxPL;`MlmOb+WnV`gnS5*Y@V;u|2yP
ztv%I8(_1_C7$g2|rK8DGx^p7<GlROrfv0YveT%x)Vg<0Cij7C`P<_9ejaJ&}Xtll5
z&6n%LlS=HMRcZ8N=aIcct+7bjqkg1!jdr5VMWfz{bnDY|@B{dFnjjwPjO?TKQF8#E
zdN?2L74iP4e%2i~BlT(U9L~@j##|dFlehK`<|mcSmm3Gi<y*|-H0Hofaneq8Pa4(4
z`FW&y*E$<!D&s`87O@}OH;L0u?f9uU%#@2a^=KzDE!W27`%)&+x3dS2<#Fw#8>u(z
z$%E2)WgcseAyKl^RN<(eFV*Z;6MS13+qrtQdw<^8tDcV@6V=ASezBg&6^D~q?b)tY
zGRLiAtcJO5AGd~yZ_RA~TPr?!XxARcg<-u|eXj3yQi(?K`F6*Sb)QR($5J=dJgJSE
zxl$@qtlk{opAY+IooBn$J(@<U>1fnWHHXce%tPm<e`{xtp9;@*zq(V{O^y?@=23lG
ztk?V9sNHJqU_PHE3bl0qvDhdUTj}OeyBY`Ao=3}($NNI&{Hc~L@0~;ziFRyJ?hF$*
z)th?0nqCag$CDl4_|SSjKdD8cL%ZHTtwiIIRs*u(tiM;Qc2A43X0luB7n=?Hxz^}M
zPj(VdjcW0^RV^m#XTyARn2FR<^;BcDcU~ABKNOC-`N~oMyih$(H{z4W_NbYyoR#;`
zhokOozjnSeJnhzYXYJt>I3F(xJI%9Rb}!Pa4UfCSCS-dwiFtX{8fLy>e2eAAg893d
zuJ6=p#UkGABya1OaMNa`K0O8x4vLw~*sdKX%k>QAQg>K}YzF6%Y`(RRx{ztKZZz*9
zi)vyH_p6ZgY5dOP_qYQYYS+$B%EQbJo_#CV_L9xm0<v^5wV#JOh4gWvR<CE#_9@mU
zf_GHNHd~GQFjdbU-?nZt=Z!{TR;U;6+ObTzHdw?e)qbj!F81yCWcTDbQExQ*x9u^Q
zA=@3FR};x{qnPdtZ}*zj+eoukPjsuzR_APZW8XBht=q@PMy7vTZXDltjuL0>SP}eN
zj@gm&V=q;BF2@%0BxuzcPml7E#cr)}Fu}X0%BcR>iuI4HjmLC5hP5UAX{QxSKbI<t
z)AMS7S{OIau$Jw$hPB;teHyDrK&Np%cV4wKkQ00MbLXU;={%Jineln8nYV9?Po=tj
z*T~o-%-dbe>CE}~e5Vp^+Qkw2o$W+Bsd}|FPMlRz=~?wG{ZQDcR|?VTImWixJQ_Ai
zslsEaJiRYu9~+<(VBS7qJul>AkJEZ=_*@wt-?wAuvqGa-DLmIx-E8>*vM71JlSmea
zHrCR|;(6qt+)j03P4HGXG95Q-fH%yHArqgQqv#F#l`2#Y9;-3TtJ>aU{y9<Wji<Bv
zZ8ug*)2d!fHDj6jNwj|6tlH<*XfxJ|6>pl2<CFHy@on=ta*!O5J`!1g4(VLFn5hlz
zWBYcPDq~F$x=Q2(dguFjXEvCU27~mI;dH4Q4uyKN*@Q0<B3ce@VsS$Q1qb7+35o@5
z@Y*<|+}~CCY{Q6UPFt`0*6gx-H=3He2dJ<F<|}e6&+mG(@NZwky%CbE#k*Y?Bj#c9
zWDx~;3zva6(&b_>>Rm3_!zKj$Oq6lEWb56RX@7b#ez&EP7|R|35~*givzMM9a5bk>
z1@wWO5S+6GoW(K?sFSh`;)LNgGt$EGasSf0*rW))wi9zy$M=)F#q^Fy&3eARn>YJ%
zF(%g)aVt+_083DbE5PbNsPU+XCjj9uoTgFJHRO&0QR$#nBdrZX5{~~n;-V8j!F3`&
z+*icIO$@q|y!L;Pq>uvE11&ruGq$l2!9`N@mkAE1@6zdo!Njlwc!%SUtJZi6r0>oC
zbKbk9SX({Zt=j7L-i6J|IHu>y*wzrw@jGKn6#m`MdlzdL>xUMy*YvJiLrSX{uN0$J
z7fIJG%1}*#v%UlSYh2bIjMpvVC75?cQ@YZ=iRk;t;W_SExX#PILk5Z8ycJ!Y53q??
zVqL==!07@x69gK<Fzn*u3fMsxAoC!xG=jD0r~VC$cuf{k<6ytDZo$$NqrtQ?Ggv?$
zCbL_MPv=1CnG8u|@!h?@_|QKD3mDz4wC0nm1&n1BHlT}}nDa?zKv|2du1x=mvk;6D
z835<j8ea>k$Qtj!hPW=$g>@(-^A)LA$Y->`h5ZF3O9wx5bT=5<#+~6^mu}pv*TcbR
zKnI6tfk1O`vAGozmry?7Zp-xPgT;PR^l$I&$NrdpQwlZ3z3o{FKipMEZ_x~>3=HTW
z<N&ASiPl4ZGGb&@S+Wm?uI^w?ocBQ&>hT0r<6CMsuo}`6LjGzpgf|`a4u(1$a-vi7
z!%(P7@fq4k>cqX_9$K&_B89HNfyRKbm{TG+^(runVW@>9QK~~TaZ3(&b_SSul)i%+
z^mcR$2ppvm6f;w8#O<_|t5~OYsh&%vQ`WmU^5j);4T$oymTFZCCG2>ydTy06NHcqG
zoxqlZY*S}QGgYZrB|DTWoaS?BJj)f6`C2Mh%vuT5E0(HOK3B*g6O~mhS=68el}n>u
z#ww)k<Pjjp6S;h@dj27VU0;!4W{{UQZk@)Fn>1O=$8GDhW}lWSY0h$s?4`M4#zrgY
zLb_PpLaS)X!p5ptwJQIQocz0nZ%bu+at{3E!9E$5#AitoFmXCy1{L?57eMBzj}(sD
zVHmD%n0ez4LaIrRIvsbAZPLh5rUfHAVu?%JIE1D{_I&TYH#|JV7~0u%F?mi_G%Q7M
z8YB&YY}${cBsga)goYP|))1=1pJ_gx$Fch=O?z90W`fDu8vdEalXeaa9?!4PzPEW^
z5&ulvY1=N@D>Uq<xjI5|hKm<JJ!mToI>Z3xaQ5})z1<cP^3U`eLMfL_zuvOH-=zQk
znU*26O0T!;?Xft0`4hBGr4zN~l-BVugG7h$>RFak3DJabdeog;7t2aOcDx856IwK(
z!~rT`=Hp;YO9e0!b|O;FkmDr3=2=Uxc>e3hYCh2$H9z%BW_Sii`nY~9S&?2y#@4R{
ztI!+4cKQX4Bt0gM&<~l;^pr;ok6-ILv}u{ZLWtZrsave%UE%E4k-J`wfGR2~EOS+2
zsVUP=1Pv?5P9YE$meFpN9;i04)lu04hV1;LmXWuk0+7-fUqAwZ<QXFo=rt0N$X6dP
zxdJHVTOKs$tTJy^Cq;wauCBb&xWIU!D6V`fK3+I0i+>k3KKi^tlpBD&$gD1u4In=*
zUkxB$OjZ__1`Hojz^H5QhoqgD(t7C{_xeO<hB9ZIYqH_3w;OL_Th|QxTd`uLQY{4L
z04~ufeYB_5VU_8*l<t2kR<CIBUXuT>RIJppWu#yB_WQp^&<KsH2(wdOI4&zSP33(S
zkJVaFc>YPm2o!wamM`WPRhWAqa$){@+2h=BN9XOK7So>HtLTE(00_LJjlybH8xIzj
z+H!VD?~|~Fi#}^u-fqLND*%h|{%QnYDUY0}K+_8gUapJWXgY)_hm#wd4y7IhTN?JP
zIcT}a!PzTa!9OP02_Gh2S;at1?}&pM1ordzB94)#)e9k|JgXRQ!iUP8-47hbZ1HyU
z^Y5bWA6`kg7RDgXaLW<NffJK9A1LTbsO)P~hkx|fgXvrb<b+lz45&bWP{%DXX$v63
z6+ePonSMuRLb9fjiQv#8G;mAxM=OG&u94w9h@hRq%`*xZOECuC0f5p>S<UlI`uMn8
z5XKCD%G|XO#)*p=)(;}6AVI)OgzNH=L+c1^G<B%)JJI<M#FR8zk=rGI3)1@GhtAsW
z3Xs)88-%`k7lggigI8!m?%VvxOACe{c~*{BfULKEe%RVT;fF4D<`v+p{i7eeGD7&F
zOB1l0!-K|D>_PpYg_Xe%m=@U;aF)ARKU`@>@B?O@dnx1><yEymn>@@*V`)F&IUz))
z5JR+C6gERc0vbY!<PA3Ad?YEP1Y~A7GnvqJ2^bE&Q<Q{qQiu(^HyvV^P15--EwZW-
z{$0*(>vs%SAEy03EHOmZ?^lz_R=YL(!)npNm6*{UkqH!KDJe!<GZv!(;*c#2RzP!*
zc~+t|qgMVZ`9;JL(IK%^E~=A?IB=vjgFF*QJ)GQhl?xva3p#@P&Nvmqg)BO4C`>xH
z!fo$8&X-4Yn@bL#ub8vw(b{~GqT=+;p_!9yk%{AYHhVyWGpOCS0}KfyH$MwO4~Z{W
zZGu%9GCqW(X72J@*hB@7A8@Lv3X8+VI>nP-3#V#@!nqtuAKN;v=b~K<ThM?j=^|nM
zMQfA6_{r9orT`*f#>Y?0yn_Gba_X$aVe%Vi4ke%E51}v}Rq}88g5TmqTJSK<IXr@m
zhJQ!!{c1RA$zNN0R2{?H>+)+kEm+7++H_bZ;NtXOa{?Byu;<-H3cOauNikmG01<xs
z9g2Vd8AohzZeI(--{5e2d(<0EW=|vtFs(SMj9T9fA>xciI5}%QKdJSTPpDDX*et#@
zN&$8M%IPR&m5x+?(tE=2C7jUZrCA9_j5FpxsbsDuIhEljY~#Qa2b)s8aHd~RMMq}>
zhbNP%a30aRo?5}z!nJXyb$8u|(N@*@X&D^y@1VP;gN`n{7S6!W1hXc1LWGo6*79N+
z(k5Oj2m=aCC}*_LxL4HcuxyAOl;%VTK<`50NCcid#@>fQBYnF==CVtYBbL0@LlF(O
z$wQlxLT$J1D6&VZOV>GvTi{w4i`6}#q%0JdW$04w0E?EWdWw5+(cMq6y`B9}vE9Q%
z=4f0Vnb96ZCmKea4h-rHS0yo<)g3DJ?<L<cy}V@C6><=A7jx>;cu3y~shj_Iw4r^&
zFHR@JC%6|3FD?pgA!;rz4C&z@%EX-um-6kYthG6kl}H9SK$(zym090qUInOdP+AHq
zL#aDvtDtwqC0_spcM86iyc#f7dMUsZ-Q8IQmN>LnMk@%S`DEBz1rXK|Dk?DQ4Z9b(
zEZ&;c7K7pZ#p)=!jYMLz%{wZ(Jx6Q|#)d&0!wVyi#pP89Hz)|tp#|07t`JOB)qs{9
z4o=~~HhmpjXE@lzz79n-LbO@uEAub#&6F>$UT>r9Hl!UlO?Xr~eR$Qq)QKB-w++S=
zpSnGH*rwiW&bbnI$brEil~lO6s7$+$FM;hy`LR7m!ObC@OvZEEF;ql00QE))^0OK!
zE*bzs-H%)eioSar0LYx}N>Y&^rLk2g6CwUP<kEjf8{tTAzx%O$wcFbJ6pgm8T6>+V
z-AHGz)9LPCwR-!MB8`(uD(TL|SSI+kp}#F4I9pR*V%%gBOPnI&DvJ%Rf}Z5Kw1#n3
zT)8~Td!*Vwgx)8ISbg(c5cdt~huTmeE|g4?sG7ep3ZttO+DO#kOT4_Qx)xNUxLieP
zAy$jE7aL=(lv^6oV>xt#^w|^r@n3IJ{RC|IhF%)W_ciV_Veysvtm=^J&r54J>8Bq0
zfR*`rp!gdM=R%7%QvwluV?9bWccI!d=|>8I3)#Cd@*DoO=%NWks^!#*aD~b9AOktW
z1=YMAN!JU1(su%nj_L(<84w}Z@NVrV#W-q{8GeQyb16D~NuGw6a5kEBv564z0hhgI
z3mxrQy~v57U80`pVHJ#{^fI{k3@^#1c1h^1iC`%BX;5AAafL&jyrQnrYiyMV&y=ZA
zbdBwaI^f`_H3bc&66s=hji5f$pGzd2yY*C?Zu{#ZEHUVkF5i6=4RUCEkyVGI!(h9^
zCD>+ycwMg|fc0(nPc)5Xad&C+JmU+m<G|&BtHgtcTh(>-wbd0u2r#7t!Z)xk%L+LJ
zo1gXMpv^8Z_A>r~8$5Zd_ULjuxvU4w!j<`f?@iBviXMFr@E6&DeQ#y=|1jb!xL^=7
zP{CXmmM$5!A8QCbA}VzxuC~~;sLr_e8RCAo0|ETP^_gGf7C{nHl2y%QY^#Ng{~`EN
zOv}KR*?cJx&sV;}9HXE*qL^ucTMM4Ps<;GV?ckRrvwjE^#+rCT^ic1$oQbArrv(FA
zVusW#@|u8fHkxwnVuW?@72FP9O+zp@om(7m4?5FZ)CA8GN-3XkjXKZn`g5ASparja
zgi?jIl92OTUhZAKfs5_0v?W_w%(dN$k!?v`C^kY>U1Vs2ntQ3(!IF~ZCMmYNq>y_H
zBybDd-b5LpplwT(Ve=dB7KAF%cIVVXOc2}LZF~}?jy<KEeSy@uqft>1$!4K3I%A`G
z)Z8&JTE-i53lOEDa^87V`zY`+-g09gvQWu`nF!VfKL@?aF}ep4>d%|_8)iw~N2)fs
zGeH)BH?$$m0m~t~ICOQu(mL=GTn|sb1-7dC0!7T3usUdAJ6Mjjd+MwXSD6wd<iOds
z0yNa~S{6)*G<gi$!ioTCYre$LU;_8f^A(|c><$8=M;5#eYU(jV2rxcj@j9TDtrDUI
zM_w3qiPu}{NL?YoNa2HP%BZ(Mc5D`2kRH+Y2B2;eh8M0|0td)11bm{HxE`|xgCRY1
z$=({fS#v9qbhg=Qn@Ia?shnYBIGH0<-r->z_Qy4?ovp2NQXjn$5rL9uQ@{B+Xw6px
z`<fzic<5Id_}JD2ZL9=uhO6p=K7V}RD|R0&g9L|%tawQ!U+Zehq330tO_d7zSL$;P
zu|eYyZjBD^aWaKn)G|GZcH^E>stZb$auvbd)M2~TxeY4hZ?H!zDmY6XH=AvF0J#b|
z7-4&x6T2FDX;I#d)hKhn#imTyI?OiIOPC@06Wofyy*=D&Lo^=Bt7i0jguBH!D^C*j
zc<bZ`)tYwhPML@f=WA{S%ZqnCanwaSG!+9&9JdPk3v!qNdE9&WOVMta|GDJ51$5#U
z0K(khXk;A=QxIktiCsPuQu`tktQt;+>pfFK_cF$?2LSO7<|<&8;Nt;B3AnVVF}uZs
zMLTChny52<8b`e1s+cU^{+Ro2CZ4qB(e7LB>^zWo+b-y(cuFy~7$$Y<uHMo)t;B-n
zy52?mVCZ6mEPARw0XQQqqY52GMiIn;qX#%&rcmz8(hB>ZWJyAstp&22(G3ro_4Je7
z2?@&~LPd$D)Pbe98Yfxv^v+wxL#bt;K6u)c`P&4y@=#*aIfFYT%#DH&oY9l#nheNq
za<rIhwCzS6Y`;YnnWtlgf@sD~1I&8|mnBpcjVx)}O9wZ2_zA9nRS>=3)aWXW1doKl
zoPkq{=d&a;vF7S5Dns_8AH>YdnN)A?-^Qf7K}7Q~sSmWIlegFs((nve#)(R{x_J;0
zmt~Xz&a+DK64a~}p+v*9W2TsZirZCrWQs!?n%_f4LWyP9+}7U3aX)*!DKhGuR_x7P
zm5`pIRahiD*w>}SdFw?)?s*SE7+Z_UZ0jA#E;nBqrWVy7OhadW(d-oc(oPP7KYdgP
zm|#Lqdgm4>WzLKsK*oZ(<`E;FdF89^Vve^q^)g~hm_><r-?~JR96S>!NTLU`CP{F&
zMDedXZb>12c}LF&!(5euP!5{+tD{X5&)1}0Q2w4F%-L!gHgyEAf&k{Jxs+u-V=|Oy
z8WYJnKCPOLql}`w+6p%HHUnMI(+t6NtCTAP1I|i~bE)XgX^88w(F3Jix!#lP34G!7
z$J?C&f=CHv(FN%RrP5VyIbd`xuD1-xpsl|fop2rkbEVtrBSw0Ad4;gs2(eD5kOLJ-
z9xt{KK^ejJ>9`RIR}DCaGOCMOtiAok@WnMM0ljoUE$^B>kRn;-RX<%PySk*<qA>TF
zr_0W{vI4~m4so6rapRDfDCla&c(=464LATcojo!ec>DnPkb{^?EI(ueCDXks#oD!V
z<R;Fd(2kl7&Y6suO8!oHA);eK7jo&dilx5_=2j<13j}yksCkRf;PEmlFjuc#yi2iJ
zUG2Dc!9lG-8=`%)fUSKLBzYXxVGWMXy+Av9D!5o}k(_c{qC-#?`vVH7!5Gt5C^t%?
z5EyC^483GH8n|+kru+UBjYf_Qn2e7fX*2{y8Fk^g!I)t}EjjTwvG<JpvZnf+1pz2h
z4_>AV(*qmOm#9Gq3VtPc9?S_nt)>JI8OW}9&qqgXT2b-_fszDU2X_vajTSWFNL?lG
z*cNh}Ec&$OqZRn@EZ{7_>;;lTVf|ns<wiPLE!pRnl{CWS#POB$C<6`#!`Z@dCB1`l
zF``gRT$-lX#j)kYaVaph!hq83ZXjT>uYph|h<-#?=t~rDdohFOtL}o-zFug@r$2Hl
zT}mG!;L_kJ8Ll}N?_o#{APt_Xba@`~Vpd3mO?weT6!U@R1c`YHSk`NCf-otk3|kEk
z6rHsH84SH6zXT!IOcNWI2^|rG#x5cgpD+j?PmvD0lpLJ44gfc-=z0ibW8Pa@no`1X
zQSjyE61D^#m8KbtikrQcfx#HkETK_h&I;gVw?*Iu_Hq~WaxeyQuK4T?{7cpTG@E*u
zAhLuBR^>|MDfnx-av4IBE2SiNtEQw@sW}RM>a|92e20(leeY7bsvsW$Wf5@C&ETe#
zsMU)kO-}k~Nh@JyAR(l(0KqCB3OtC1{_zj|Z^=v}xMPZ}29yWNO5$#=N>d{X8IME&
zL(0W#6(-McPHz!`Cs&7710mpwlY7gGS&`ktXzVbuM=|{Nt!$wx;Z!N48~$%u2&mr0
zx{YSfwvdj&Oc}wE+N4@$HmsBi1g#^)*OKw|EekQ^aHATTS`jl2_XQB3o|X>em%v1s
zwK^28&%`x6w<vRtt2qM3OZ6HB<)rwGTMMj3TNZ!=SAK;&7gn1B*#T{e&qz`2MzZ_j
z77kXzUhj}#^%slj9Jg@#*ITm4)?{{#j3&b#?i$Ru<-<++V0+ej*rZ0bW#+@}c@H^x
zkP2)Y*($b&gZ6gk_Ud7~eK#1wF<N9CQdmm5=FlT?qX~iffEsl~4I!a>nl6y^26bd)
zJ<yA!IiV}NhPjJNNFW?ISK4YQGmy27FqCci4fixI>FWo|u1Dq^st?M<kzlRg+dwqJ
zOXSzUXtd!S2q$W4(?Vh|3x~Pzy`mB7lMfa$AVG!NqT0t~9Rltmkrj|v6bDmLPf$A<
zH@JX!nkrIV9IB%cjt$sF*iK<!8r>SnjHW|Nm56S|K5~l)j~Cn9KeVDBd$HC*r1Ob<
zjJx}-{f~R?o!yUn-JS04{>NUl)s9~EBHg{6gO9y~$j5f%Q!KI{{S@2zbkJ$FBA*C}
z5N^~Au3C>)g@z0{=HekN_U=z;lRk-b7BTuL!C5YYcxTl#KMfUfxl}HdOvWdp>-a-1
z5x+(}-0sK7^K-nIxW4_?za3-`9wLc2!Xd{~iBh3Fe@K?kQ}uE=n|?U1*PhemLVPzH
zkJi$mWWR7^pJk%0Mzh~br5A<doL?j#3%B}tGEr?diqX!fc3qCupSs!MsMRR;yIK4`
zt4CVdgQwP+y%#DZcF$7P+ucIy`l)b}julhsXrVsAv$XfwL-i(|FT`)Tu|&U+tk(;V
z=~TRwxJKme_@o+2MCy;-#&C3gcKw*I#%JQi9AT^V^Xpoye$zcW9)bYF&e)zdN5h-*
zGaJ7%(QdYX-x&=fy=pQM8vAII&cxB!O*#H>eSVv~K2OIVkV>GD?c~ny_OC~S$W<~M
zvoqPjd9wd;BX(`yr+4i^u5)UKt_u5CiC%Iyew8Q{OV5MBw3cq3JU_Og#j8a7G~e7!
zZkMl5O52M_BH4TH#}oGH_UZjYsu)e(4WXUd8}ZKZZN0u%XoNnMH_lqKt7I!#=sc&R
z&BDo}eROkoe{#^iJCD>qK6Zy!wPOCJ7XP$ADD}28r#sh=A8(WKhjcvNDkbCP+lTAx
zbf$0=&xSCh-J=J4TF7=DnytjmB%8fGdrov;Vu58G5c&}oD3l`)$p_}O6!JZ!YPS!M
zg~UW)mX4RJg+w-ydd@u*s@D&tRPL!zJ&zQswa3OyJWnHhRQ{As#0%)w`NI(bvLiQ%
z#PvgF60cQC&B;dPzJd+**_nNCeUnQhP78a3^7x?qa6}w#M@osrc{-E68MRJtXOY=S
zJsP<>&pn~#=hFG@@pi2ce;}UgrXR|gLcEX&#jid+WUtSUPbSS=|Gs&Uh@|6*IJ)sT
z&(v=UcZF;r6~AuB3-L%cS^1W&<k~x_aypSH*W#E(iFoFwbkduBIyl&mhMuoa_A(zE
znMiwlvwM72IIgCP@sn1iJ>H1jj!M1S@tHjuZQrK5dpG^Y{`tpmH`|T*O|{-W9(C$h
zy@%w{?6mT2oQvEhLWx{6+1q};yV}n+YDh@YJSfIv&+}0upPO`_V|SUOZa;cI`PLlt
zu2Wav#+~T%x6Wy{mTT10XWO$+AD@PkZ>^J~UhiNp^f9tAPwaGdg=j&-4!>XtWswT3
zHLMJ-$8uE<5KX_nDx11h$fjXT$)<cF&Ywi4wrD-<l^;@NMeI&7UQQ#$1E$`yTN;#a
zYmcF9HGZbVK^kc}+Ogw)J5!uCW4qVw@%f??OB*6%UQI+1s(hz>T}G;q>~p*+>K3Z$
z+wMGMR<4Z>?z_nZa+bt<nTN<;DHY!<q|%RtoA^T^m1yDF6Q1GswSHE}bRLt>@nZrS
zO2p5r@!R7}!M=US2zEb7KPHeA<X~{#c&xS>-Md_NFP1+m_6v6WA$84DBA<TfR@$+H
z2;^0wV9y`2VA>GmGTNq$B1Ywm?h4KBftUFp+rfOfMCt+Z>pFcBpI{v*r&^)ZU8{N1
z*}vJ1Y@A%(-#-p^3&YmNG<UR}{x;8#>gjUj;8XV0-W_#E&yUaD=f_3vWMjWLINxv0
zo<{Lz>wfz96sbJz&pQZ~8b6v;BPWGIq<&M|-rea|3bjP7^AU3FG5K+4yELqPnr04m
zvpYzlkSv|u7N0-e9!;Z-863TC_v<&eyX}YbPlN2R9Xh$)$xNe_-F9wVzzq2Gba!?z
zxT;l7kB@hUi*~1*-LKl&Bl`y5=Hptae|Oy+Eb5DPvA&y`HU|40h}6>k({ZAB8tOzV
zSW{~I*^dvsV!n}2oImY#ckdo!kDsn@Tl3>uY`^?Gp6_(?!v}jWJ*YJbjb3a2adw*?
zB@5ThlY>n5+wD!GnVg-4uAi&h`Q$9oP7G6(>t1H>p!n%vkePhkm_5~|2iv!Msnhs%
z0eqajK{^je0$x13sfTo8`=N}L1A^!#e#)$IWTW5}gjhA6ikFYJ3yBEJo78o=kx1Cr
zz1VrnF2~NECc|=mCz5`yRwmoq&j%IX3RTO5utY%;rsL7^>G&#r5R2qu+aL4a2KP7B
z%t`7bdUg6ZA15DX_vKNq_VHkL_HANU^6g>rr1boB6YW>K#cZ-&*m&4a-Gt7gwf3jk
z?0!}q_sbt|cPAU?_k+Rt$B+Aoa{YdsP2NX#pF8DzCzGuW_pZ}BpNdz*Vy*sU+lg_(
z?sgmHdg|(+UMM|fLpR%>iXWez_Rb^i*th%VtNZNuAm2aEHIHZ4!>i=z<J0Ig-cHvS
z$QCfY+MDLHx3|5-{`2(l@$vlb<H^%;Z=Q+9P9E<2N1^^`_uFIU{BhV#j0fkn*+%1O
zbXIG1Qzze&d&fh2=PbDsi=H2x_v~-G_2<UH{qDie?NK4Izjsrc9aL-E<=KAc>R@r-
zONP#N&njj6rhVNmq>l=<t7$HApQ)sdcdnmu$^K5GIXoL|r*Do=D$VI`{8uT@Ruzbd
zvTn!x3Y$>hUKNdZHxP~APJE&<wTo0M*O5TcSWTtDdqQ2ql6-~)G54YKQD$C?4ae<4
zLJO2!GICwJ$vvi@<F*o;Nr*l}@)_?6RrM|;W$MFGzf&xv&LQ%kwxzBi3gZtAdUivi
z(Rg-~C=@F3ha)A`9=lq2g;?}C((5{3NmNi73x8TDmi@vBbwT&Qcn68=dKYRFy(&2`
zU9k4XQxJQnaZ-cJlO(<;;wSCX<M{qYe)MTFs@0Q^C-aS5z0yt|6u0|Fsr^uTcACH6
z?Or{OH?onBxrei2`(SpJ`M6U$+dV%%Kfdj!qLIvLZ)0b%(b*l$Ym@%R%D47OrPb_x
zJLwe~{p_dSBtN^KJWoUO{inupxp!8X-hN9wZA6a7z3svEMl?QtXb)#MnY-#_eqTJV
z-1QdCt6sjGm^81>Zin9%)nRXts?P6^t_#WILArf%Qw&u;R<5tgk9*at<Ziw*?>{Z-
z^Ty6jf07)fCs#;4l^*AhvnO^g|J+R;oHd_M$KQ?-+`hASlpWi5^QXc+l9;3-`MqR5
zRK2=;DE9AT*|TpM2<E4AyOcXS-oCzlnA*eLY3uph#tgfZZ^_p5`o8<zJUcje`qqe7
z&#vQ%r`l})biev7@lfc;C;Q3J)Aea`wEIbEn~8BMnHW`tmJU@-NII#mrG&axC8?K)
z*RV1{JU+*}#EVDe-E=5(U9OEba!2)8r*kw<d`m3$nv3JfV&~wY*&pma6p5eg$W<cw
z>47wUbpQHz_k7qoLgK8{<UWryEA8aom=?+qmf2_Gwftyr2%<dEvYJ23J>^Fzl!9tb
ztyT)@Y%1Qk#w6Lj?w&OtA9@ejNp@#Aii`$3S38MOHh&W@hJ4#5DR)QbnQH&>aWdR(
z*&j=Zle=^||7|xrKaOlS9@^KP^3LEnbv%P+{_v?iO1G=E`80IU{FFJWv=euwQ7WE2
z&Mdy&=59ZZriBO+lg;zH{oQ^kb`X6mCT59xahg5dNL5e1Ef)1;K3W>(i@EH^{>l0H
zY^Qm27m7rm2B+UL?ef*x)n3C+!IWGtweBmE#m3Iy0E%hiwqtkVi<{ETZoY-2X}9A{
zqFq1DbocJ;qwBr?et+X`ySSG<&VCFXWH-7GSII*0{IMIqx=*#+<9j<B-Lr2p8_E0q
z>Gk|*`{VxNE?2P2PxTF?yqTZf#Gi-p&Iqg5^trbEZCK0OcgN-JkD+SfCbC`6H*ZFx
z$Nfb5bZ2sNzVR46KG=tmqVwsxS2(^-FXqE??(y69(H@dnC2rIAz20>%Tg_j;EO+a@
z*<em;Xp#;<tsU$1u8r&0*6D08f|_EOn<&V?GV(k@IA?d331{~Xd&#fYBzQXee~J5!
zEydMt$+@25&b7K_5n?Eh010`+Ob#OhvdBDrZ_-p{)hToSd%IT;>YTMyBqd0F+r*BD
z{fbB<SYew&e7>A^k7acm(qJfHz0*G11V0yQuP#ATxBH(vFa)~rKm&_u=tU%G!H!<#
z%@I5fP~3K^R1V=xPuq{v!*iV@m!}fi+4s^?w&pa@p^SuQf15s$1eVerbB=xw`<<iN
z*L-ht)6pqbs{H*24+6i7D{XkC3ai@8s`9N*W=_5dM?^0XRoM*jwwNr37vJAdbBr%<
z0>x_=2Ku6k+Qr>^w^%82Rp#f&C$7o8_yHn704sPxlH$|Adc@SOPPfG?C2Y0jqx5!Q
zB@wiWuUzv{dM|p;QP3B{N-cM#J7x6w{UJZ082_X15Bd59{!hO@_yv^TzQF(K_lNw5
zvHb4)`*)cUc}tDjz})2noAGS4t=L$njvG`f+mcsD2jfT!3(Uf737Nf>VIujaM6GWX
zEF}xV3oLh3&Tt6hCMFQcRgUP5?Dv&?s+FX2bO_7xE=~ka@^zV+BddO@T5#Iq@@>7C
zzQ*Xe5wxtd+X*W&*pbN`56VzXZh|C-@#5HQNY#6lQD)hjq=*UD`R*#akHYu8(AaA#
ztU%LmIV57vpf;<Mm!PuESAu|g#Tsp_8GTswc|@e)fQ}bf_m-W=Rhue3s_Jz&KSq2y
zdEvccWIfp)o0L^lUtrF=U|dG3+mTCyQPK0jm{D}oCTF(_>KwsWtDC;@<<p5iu1c*)
zd06{r!P+i1=~lG9J5K%pt5#O9HSRq#I=?{PK2it*i;4X)8f?KxumfAo*Z`J1q?3`e
z3_F_&Jk3u5m2%#J_51VWo^^7!7`<~3SUTV~9f97Saa(b_U|OA#oMhI*LN)$X%S45L
z_K85KvpVqO^Uws&!CH6em}X_fj!(QeZ0wZ0-;#|3FgNGlZrG*A=;5lRt{Mpl7Sdxr
zmrBRe8$Er{;WZ5HhN}8f?x<#SV*pgy-W)W0u(opDJqDw%D$8^&Zyr|6s+MV~^>g?x
zZ!L~Y!~Pn`c3hVxw_xHlCRqr`&sx96@q=79SVYca-#)n_ac-SBb&JwF({goI^HUt>
zbVFTq(l80{0bOiQ;z15O=Ptf;#X3(O8loM{nsEwfGLEYQ3*VSz{yB$8cbGGG{r|pa
zJ>-q7x(yYK7HIYpceUT)=3l|Dzex^%g+8SHD?3cz>U-&^KshP?;^BZYsuv|!<JHdu
zaIpLp4$I!4!dnf<DKIBf<j{|R?&=gk%=4;@@9&ZSqv{c1dFmNZ3;ZbXbzJGs<jw)`
zuv@9yU3t3Em%TF8NbHw*GV@*J%eTX=n05}Ae`boU@Fi17-jHs0L_(QETzQaQruy}>
zkQpY3*VF4~0Q%6Jeev@Pe}G@>7yz2Dz_2O+9xR&l{u6I;B*2Hw3DFmK)E)U9aB(2&
z!^Jny*x5nUS8%99M(MrF*h^zC|M2lenbXA2zR08y_=YGzIDYsD2OJg>671W*t}2jK
z$<3Sy^d*Eml~i7JxN)!rwCn-CUR3z5LSG&Bj#MI=puPWQCN0@3S`90HR*+3EUw=#R
zOS(AkKU|(fX@NZU_A}7NG+$r$`fdJ@6@Qf!_o6!L;-v<3pI4zK{jnr{L#v_HVYwIl
z&kL9R8!phg#G$eFN;fpQ$|as$nav)5TF<fHJK?&&z?#O>I;BmGU(D|M81>%SZ-{Kz
zMxlN>9uTD_SZR42x5PkgVQ{qL&8!QRIuA%14(NfyBtg{YwsyIK+iqa5uaXKvwcNNT
zaiw5GbV3y)wGN`nZjq~HE2TFCTFF0+<i?yXOO72#+D>Hh<`yR`5gyR@<);}_?(1n%
zd}w#FqMJ`Vto0mRy4lb~<vD?~_{jZmF!rXgIxv<AtLxO|0{c<JcGFdsCJuJUQ$c**
zpesT4E?0-E(V$mrrBZ09Qn!y2==LApGh?~@>fR8KPq;#eoF}*1MLPvUrL%K~m7=4$
z(qSujwXj;K`n$@cM4w2wEp4g*-f;W_S4TL8Z$pXCcIR~h;L8ukzx6w>cE`(X9=8<k
z+23}9dCRqPb=`4=KLh{@0uGA2qt4%e&{tQ>bK(%^FiX;osZ#|LlSQdsZW7oxDz5n}
zw%4dh32D?n-5kc`yX%f^tK@FXh=b8nb*A81ogOY5ahK~cM1d_$miUHER>S7%!fZs<
z);a36lhTTt<VjswF(;C4+g0o8T1y~z4IL&tBTGO8l7PlMS4P)^H;_Ee&}o$6{<I42
zMIviKj!Uz1Bwco$vYTfUm!;K;u#s4NdaHQA*nCZ1PV=;PkR36K2cLzh6hkV_lKIcG
z)ijrT{amdsiM4`CA{egwvB`KL-WhwZU!!qo9t5>V^Q;XjVXLrc!e;g+L+-i^TT`j)
zT2QHFSBt{xoLqK--jP`x;}=!-V&jQP(AzUnI+UkE^VrMcS|?BXBih49Ff0WZM3~9V
zan?Go^zjCS|GuU8TfZQt3uJ^6{X~2u?!s}yugV30oqy^Pc9tXA04Bc0l-HncX9DU2
zv7?7Vy3{Ofl-{R$04vvG=I_d+e9sOVKqe~p206o{0J}Q<33kR3Fb;!(9aY{SAnn}u
zT=Nbey$|kQ6}bCCdT3n?M@@bW?o>d6Uf0w`cZ%1w(f?_1huE$=NjYb}r<8{yxB$~{
zEl3@k`Q&^-egNHh{p;WUaS?EB;r-)5$UMD1-K$@C`*_e~tU_psy!U)`d4ud<XwSb!
zgER>l=6zd<M8F8AxoW)?A>40tcJnR=Wu%GWxSy-&q@&j$BVj`7_2hgRmTG?spA3G}
zA2ze(t$Lz3)9!f0y$@=-(UP3am?69o(dB+m(2Hg6&(@;Vx7?=6d>krQ-#@_7^uav+
zb{A&*st%LJ&uC*@ybV^`(u@g+vhl{m5R8#O?D=kp==2W9x~7O{-VJB@0h_8-X>;Jl
z#vX!7MEfR`v4fNHL?IAz5mt&p)V^^kxM_m~v8rUSFqs*iF2}60tGGI2Q_x580y*n6
zr^$wL3u>89TsJsWIorBpOl(mJa>B2Q`k92x`j`&ayl#<_vT|)rJQQ&|W-38Gu{q02
zyty{1Ug+ap$v<b0a0(^u$GO|lPkosFSg`W7H6AMBiII{m4Z%y)_)+V?;Oz}HS2Zqt
z!AB5EwfFgmg&XhC5k%aJr2O1TBku#Z*PKI;MVUOVl#LZhs{GK&wKFH>l{;=Wq_a7s
z$w)J%DWlJ5Z&zgIY{<>EaMbn16rN>R>(&*T4_pfUhkd)AJ)x!Kr^vI!P-wf5R7?r$
zmc5Q1wUClrtf&0TloyNLjWWJ&S4410Re-xkY&s=Zjb9<x_#$Hgik>Vb7g<i|jjIj2
zq$L6t1!vbA_7GZJs_)bThm5yO+2%w7b^4Z_g|%AmR#3cIhxkm4vEM~VxF(9#6{VJ^
zINxVxt*p<pt6rC)JKX3g(yLq*8k!ENtT3f)h3e5@whKL$+t$<;@_sTg0#lhR>^57(
z5=K~0+@-y9u&s?sA9@kJq~jc28DgC;9(K(5Ed=bHbZEk*s^8<;ufcFwYWp4%50CMG
zy?LiXDcwH76|ehCl>Dft{}wC(tAhhR3ZKD}=YAb1%bx<J1A+2~QW}AU?Zs#0FkN-p
zI$Zo&@ye21oIU{*SgY&`{P7f66a>{)v!(Z`^V>-7%AXX=MLq(h2;1V9Wag&V@J@Y4
zxIZ9s4vJU}M%vu++NS&tng4Vx=il!7D}?cn+x}1Q8p8PRZ~ISo4MoO3CW^?fM~Z@9
z_U6_0#ao14d4627=_n@BZ0MsCY|+i>D4F#=c)E&D4y7RW@!kp<WJPDQ&c*VR^GlAJ
zg?R>iVNlddO!@cTKo4Hb%L(-QHLQxiLRs@*M_y;{y>i5{wUV^?{7ly;e~SGUc?@NJ
z%<-c}6|+IXmifB8#vx4T)y-ruC)Jzo^e&xX)%jeeDjK$3z8bmkwqmA*V%z~G093cU
zQZ*g~a;okbH7`@#jw5uWh15S=-SHlU)^Xh~xf+3aRY@1$uY1~$44)3dRn=I#{4wDz
z$rM#%oGLg;!EU%AiZ&4{ZKDjj(=|#tDo;^BwIqEsxLD2$z085KX)lEq%Xx6xVIjtq
zv!@N)HH`8<CW;VD#!JGE$PWyLoj>)4;Px%GC41QSX)EX$ZFZQ&uQl$Cq66c{A4iX0
zCy(&kgwdm_yv-zEF!&`*faqd#@*m##;a9L~`~5-v0<}5F&*$queNewXr^p+t<IDN_
z{ex=1dYqqUKQFq#dzCCqJ#yl^^I}rzU~ZYR(r~S~ZmSoHbS7rGIV~2^;>?DsH)*`K
zlCi$IR{F%lm5~9(Mk3tnl}_WM7VL22irk>TXrO3uiejOu9lO5)(k`yCz0md#j7`*a
zEqY^lcfgjFzfm+s&S}W6!>!bBdPjPaDbEJyJd0>8L^3fA4(Mg!$@6*JrR^TJske_!
z@W4deZSC3}bAy&JCo7K!EbbC7k7umildB+F6ZQ@^r(-DYN4>C~TBHi&r$JGRA|Cb`
z8|MvaDdqW6MQQ0>DB#as&X{jc(u0jl2KJ*o{3F0Ua;5_G;~4@I)AzX9H{L=&b0kT|
zOqqE}fca@(h#Oe{PS6(aZ*_s)(`!FTi$9jh-|~P2!iM$^>vCTd(BIU^e~tP28)U-#
zE#^P}<Cwq8x#BbCZ{063U-IZw@#YxHP4yq~|GjYj;AE6<3E<-al>q_-1b%gmU(6Ah
zYQ82;Ukc|}`~T|Tm*ack{3@p3t|1>6$M`3bi;KIHEOPd@!WsX1m$IPQ;kbJ*;}6nA
z{E&^l-i_<Y+{FRFCIxgOT7IgkzobGdxZ+TK51~K4NHQ1XKd!0|XG+P9ioI~s!MR_Y
z^4V|KQzx<{fuOe|x%5mdIme(DO6`0)o}X^R7AP(%!_in!uXp9vFXzhjAHrw{ZG5PM
z&7d`yzFhik4y}Gfj9v$LuCrhdtzO}CcQd`?T?xhweLvNl!7*JJ)tRiK*_DwSS6_mv
zV4i$@tdf(RB(ZLWT)UN4IL*_c&>X7=64_x`?0c<~*t}V4<ErQ9rN*w2kFE)r8b-KN
zObj)|u6$0)dJ@D8T2#OABf-_NggcCsP}~m@E?#GZoa~2{Sn2*wQoSBa>t@j<+@@yx
znVKW?dX=(@{5n)8g{qgM;oZH{Cc$e0k0{~Rmwm4b@#N3L_PIvZXqmbj7x!<;w;3x1
z;2`OIpPRmLgf=F_EVuF9^UO@b$z6}f4J|kakAJ-L?VxxDVB)iKW?wzyXQll{b9=Ge
zE~huTCFC2MFMt)91E=nHZw!!0b4?g)(iHX^@9B;9R6^bVBP{pzl<;OQRhnRaf@kR~
zS0JCY*w+WplP&Xefc2Wc0O)mjKM9(n{O-I$>4m)4UVljr2O*|aw;cO0r!T0_*Zb5w
z-4ZgiR+qHM{K?#^%c(^2wHT)Bp)>ty8>+h18Bk+nXW;*Y&O~#59d9B`Jsa7!0&+`x
zbGT%G&3%^bSJtHO{eCl{&afIWYg{4BI0~+1_IZo@w6;B4_u#(UQ{6_|<JXfuN=nAu
z&5(fUoU^A|&2{(SY;9PpcDPRVQ_x%55wmC$QmYk`G09GJVO2s;8n@0MH;*wgFP=q6
z0dZ&2o9bUYRnpihOWCLGgw06`68eI9SC|FzH#{-XXh|-|`-2`31y!_4H4fWlJ6Y1A
z@LPSCR-7?mw~b8m?P0h>`MQd#dts>DV<1i|n8U2RqISZ|6;Pjzz)tk$c)ghaE;oD(
zZ2yA>;jfuOe(?&Q-*|<;>kf+lI8*%l?%*5e1z?r*hj?tJ{tf_k?Jcgpht;<M=WDP@
zyufq7zbD;!=NA5aO;gObGX^|_7x@{ug?~R}sm6a}YP(+xvd?pcye7kM2TSC#&x0j@
zO))Z0gcV@k^Ot<~8}#{uiN?Y#`#bu~SdTHMk1F5*7gC(#eCkn-ve#sFcdRC$^G?;I
z6K~b_28f5e$MoIVF5>~Y&-$<t?}tZRKH2*k#*9f7p6CQU9=iImrPahO7GBZ}KzoyW
zyX8*FR>tmuB8=nNTrt~9&8FhXVF7VN(v2={Z)eiLd#PI|H0c!QQ1`6ju8BRz*EFWN
z&ak#9?rx-urML%?md*$CxV=;Sa!1|MJ`jjb4Hg?{$`PrSXkRVf+QK%|hJ85E023ri
z<x~J-w?yfmOQ_>!UM_|qIqe<R>ZqG}6?hvW$TQOganpNhp>t5sYbb~MBAR8k%Jw)w
zGnUMdz8kp$GwMD*9O@0vD-ULo<g&JozMvfM8~%cQqR%vx^y>Ebq7c4OyWcs4HgDwT
zuxPn+y4ocDzF(T*OzHqKqrV_g{|co>umz(3JKH9Q5tIU*f(Rs;-@kDqKWZf$Vao9X
zY5!HhAb%`jEPPbn39?YaSYI>6&sEDC$?f$qB{(d>*C+ZL$>3Y#{g>bmqhi@@X=JBA
za_<^BX^)fpLp3OCfM<N0QTZ!f+Qr=)efA94%;3o$`A7un!us;ErJH*(2-b1kuGNcT
z9m)bL4+t6>4@>uF)Aq63)_Y=;$l9f22AIgKm(y~#I?_r)7<Zc9AgztNJeP*|;POkH
z8%j54*#68O;thpy4@8mjacURyx!7FU?e?8d9;X#A1E=TSO0M*v+Qd5@b`}|j%46BC
zPt&m)ID1%ojiWkO+E(@J$QRfH@V_uG%X-)dFyk$yO5z*AbP|rUb|uE>o+#Ves95Ob
zPHqJ~Af4*rnpJQ+bVWyDW?f4v%si${gAWCWku;o-0YNe3zL~U{I;GFu@tAAcey#Nd
z7yi5GZw}zqeZGMx=!*>JU!p`Wu;{P3VEabHiT>6#-2Wfu0&P^^;djUd^$)qgN_`n}
z!Kr>xYF>5rZ}I{Di+u2Q4dfdW@<}%TzmX4+9}2C7yME>arTEfv$T9(wf|rP<JbmqP
ze#!~RySV84Nb+k=Sn*7R_2bTeWW^oNP{~7|SJ8>Xd|G9PbFCeMPUOdC;$w_HP7}h=
zEg|{q-Wf>?zpCNEEj*~P3iWDiCEvl+>pmL76_7=8Bk`N%G|DH|BNTSc)Acz#n)fz8
zPfrau1Pa+3c*E4p(j5w^c~a!Gwx>a8fR~w1O2j`C!9$Iz+@DAGMjp{#;FV~9x?h{Z
zl3_5H88y+n&*lCQi3f7P_(QTCo#UQlWjmONj%gnqHxW1VOdG80pJ%DqRNJ#Oxe_{@
zaT~W_V9y+Tx`d|RF^^=yUk}m|RV-xmLW1R;<m80AxETtbO$a;z3e|FoQ~T7LPsxsO
z{b^H*i{@s!w)A0s@K_~my1@@Z=6g<XVFdibdcN|?%~cupw;QqS2vuNw4h@XSRZVsd
z>*b%qny)zlx`t1xrob2PBAQoD*g;NEF8m+fn33CBDaO;6t`>PIx&GSL{>z+zKr5ST
z!nc`v^g~VvyekauL5JoQ-vQa*=Y@A36koOA(n;S)?^;?%CNiDpwR<=ZEo&UojEyNL
z8v*aUkd^yIY)numj)G?$q6`xcM!z1fSS(Hoq70AFgki&ClrS?LyT}M?8M{o~1T^X5
zx>51zF$T<<v#ciwxCW~2B(rMiO*w5=X$zdpPS?HN9p5nlk^;-cSOr@yJJlZWT_i<0
z&6$3#y9auybhQ~&)fSnCZTE86M0mZaHu`FM^fuw2V}UXlVd6J&?QE6|u>#=I-P!nX
zxv!qY{<`k=oN+lNhtlIl>q!8+Youg~+1z(gy4mGRvI%b{FNDixbxe7wQ(;bR0qH6e
zdE_Ry*7mMA+ZkB=M>lJ9n%ylac3~K3I<QeYY%zOve@xAv2df`coHv~4OGfyCyZIaZ
z=tnBfU-JUAwE{<d*F{=>52A&i*%U8bq`-GS6*DgdqMUQ?D>lNZcsQc_jZE^5l6eG{
z%6sqxmgV+$q#W%M1@OmT<uFwk5%>>?CE-E+WahwN^8|kh)6w(CIhKIh@4pjGKpZEA
zUx+5-#fH}&@^^9OXN6RJXPA7G0DRRz{)2?`0acNphnj)BIcm9vBU<@3<o$LL<`_6!
z8$s{(Q`he~niY=nI~)aM9O63}=Yv#=XfPqxKa0JE2RcqNz{mW?Y62R}(<FEYh&VvD
z`5@zfcJ!s#goJYgt^|B7FSe>TAR+1g$bY@P5tV?+!;$KDqSCit%y)^X_L7KxvVh)l
zFOot!1K<mW;HNqyW%&PQ_(>!BuP4h0cZj^c&;_(R0TFVeQY#DSyNSH+^p)hoIC7vm
zSg6M5j70B^g_5LGbtv#mX{E^NVhfU^+G%%83%dp3HUk9+E~=_lwvKoHet)Lv12C?g
zvpnO;v7Bo_c=ld+4Pt9^Cvf2jRQ~6#b=H}q59USYFd=Lt?l_K!kmDgpaW;F1i|Wb^
z4w9d3$rW$|aDcdT)ZU+E6;G4kNerur!8g@a#=}VTnf*SGxA{I4(OsaZZOpJp&P;Y$
z42>I7g|tW~Mq26j&CqpZ>pjKNO_G|)?x;xxczV$*ejti!yl&A|ZQM4rSe(usu8(wm
z+44HlE;r>+uP~26GnrYpTdjMj9!hTY{QkI@H#9XOQtQ59U>%&#_BT2M<ZJr+p1s}`
z!dr`zb|QDXa4*sq_wB)cJ?&cAylYQ$(6~4%u+2CE(H3(C@##-`0qap5t$V#cITtNc
zPK`5da2TD^_9p%bnTvMtQ5m~(Y4IgwiVF6e9!zYIo<9B>j{m4D<$?Bd?Nl#M2F-ck
zAU=Ruw+(Xae`-sSm%a~X-3}oNBQa{1eJ`#;^k#{iducrOG~fH#>FtT!2tG<!kz$XU
zU}W=C^&j_O(;tcbvhSaFLbWv{#6n}9L%#6%ZX>fUUF-PBA2Us?O?quzCTkhh!qyWP
zwm<OgjyVkxP%|8i;?ZD|b=w*i+1ssW&xw%diu77?h{)Z#W;097Gu5)h25Z-(Xx?@E
z#x(1Cz8f3gOoW26p8KY|&#j438;rbDO_$;0E>R>hcWlI`ZlX8LiQZ`yech5Lyhw{?
z&`|V=WdkGIO=^M)<sL|ldnj4Cawgl%M7>uGgL99}T{V)N<kl?F;7Gq62J$kP$8GSu
zSVV<C(;yL}<%xk@hLXyw*E!gAh54<?z9*Ak&$=y}d*_LQrNTRF^!aOH!QV2H7``%6
z@hu~n!i(T#f+F@lzGSicH*M@c_<pYp#W7#yr3mOtZTHnoyldXynDnp2p#4}*{o^$s
z^icnUasL`q&j=8D;CN^&E!W@3mq57(M&|cxn)sN0kl)HX2WB20HUJ>1x8H;WZ-ey@
z)CwdD|9BC+DbPMuIb8lmYy9jXU~nmMJ-zXgbk9*iUuV&IRAg>8l-+#;&pp4PDg|f@
zWV+)$W$5qJj4BDWwwj)g0P$zaKs~Hv`ex^AWH*-=Y=auP_}tm2fv9)Rk;@)v@9V&-
zh&M_~acWEgd`}owrynW(LLHs%(XvAdHD24!N;%rvZ8%(F#J?W59&K2xAl@06B$zG7
z)d9I<vhtd&>@cw-_Taad4Nid?C~S^E&=&3cmD6NP#3vi&tf;vYGrqoUk$v0e5~npC
zBO9VuXI7hRfHsG(chwnn@Zu8QD$3_w2BfS;0^+n)lS>bd^G?OdT2p2nEaLF&&Kt}%
z5OhDU`m;$@JTO)sUfhqfaHP{`>6D?S*8?b^0<~K>Vf>&qlG+F$<NI(JI*G4(RpdA7
z)}I2Txg80ho$cG&ifCeIwAfA<7yO{4x%ho_@h17=S>4k1cEh;-pxX|E&1R=<gy|lR
zzaQ?(EPB14hCy^$Cl8bDy$X>hMtRazUb&a`poJAn1sD_;p{HiIJ?EMF@!7jWddi7|
z(v$^xGYbl+Y3Ko{WY{SNMeCeSx4pH{v;`dh$nt1Ua8}pun1Sx}DaGn@VFRcW02AeF
zYRK@zhVjlRXB!KSo^c$Lm7I_~m;s+{vj(+{=8IUUjyd9S%@b?nVM=9YUtNLY!wAAn
zz_$zqRE1p=ZM|b_yHu(AKuX%WOq*!-GN%l$a;%JYVy1M7qAweeIDl7$TAFuLz6dw$
znMg%hUcAmqu1r`SNZ@3hFw<)T4n5$cd}Q|&i7{a$9CvhJ8r0K~uRvE;*yX^p1&7r8
zdobR1#%YsMYk`x)l>Q@xqhybFlFz#rz%SiN{tC~C-iyeS&!6y&$mN#*s~o3E>L1HY
z1nSJ~$=`gE72m{5pThZ1-Taro|B#`+O*cQdIl!uR)ED|bB2W)}=vlvCe7lzO!cRca
z%Ou6u_>VExRnslfdwx!_UQri;dOX(ncmEOqEbXr^_|&()8q<Fh_>fONH`$G0xQLq#
zpX5q;Lj4WP#$sM1aY`4d%|6nrhVHLoe=%yumbQYLN4~JBj&`--0n)oW`2*qOv^n3E
zV%1rfx)q<7qrKa6Xtqn^{(>F*y54RptrQf|b?*^O#A<bL*mk<&1~1oMaPY+E%_ceU
z9_kTxSe>Jd%H?t4qM#Gk(>rH1`a0=u){`y0JxiJ80;8(R(I}P?m0-JfV30}#0|<l&
zkLvOcFKOCEJ@6Bt#~v%WTz^j0svC|qluq0s^SHbMqS_ig3Tta?`q8Gq(RgIH+lcM9
zkIm(n9T>gToMUQbJCMT|(Zxlc&Y@K}?crXmX%%&gkxBG0CgfGwCt3FJHN!fO!t;^%
z&jS5KPY7*ve-ZfJ=`uecDj-#;uOMl>uba^;L|)!N>@Ekb)trA+i$A^=`RT<!QTZXY
ze@KJ$Osk{8eFmO=+<St1Uloyjw2e%tZ{=RSpH^!VStdvh=$-Ex$Dmv(f|L;3=xQGm
zp54v%RUg~-v<;iXZCNm<CyE2cBx$}ioU`4VAfjrkih2|q{oc_?ir<2EA71jAsb+H#
zz^N2O6z8MlNjG&6N~R4x0Hh11Tp6j6$VQj#-OCcoF%-V1XF0X98}09TbP`CUYFuMD
z6qvdt2fOv@`_WAM<_Yjxz)$qCCo%Va9cd}n1JGFU={ij9e(oSv(yO|vaVBl0)q2u(
zo~e;Q!k1vB?m!+}1wj@}6G2MM-CLa{U$7*Ch?aP_mB+*G3J-h{^!he;ZLsRfau3%h
z19G>(`w{m@o)w~vpAhn|tGj;&x&QjLrvI<i-RXrx7k_21NPi12$k*`k(e8hn@IiQS
zzyAH#@B&s)AA!ysp~Ay~bn+$8y(rrs7e4~sU+?7WKlB;s{`H;widguSSNKBVKD|nw
zhD30NC7zE2lpOvUEfvl)9=^(kg<fcvhq@|cM(8g0Bwj)(4#)W5<>K}Pzv$IYRWL+%
z4s1x?<Ib+z^-T`hOI~MM|D3VWcOgDa2H0+Gn8|3)!r71cO|Y_ee0wAO%)(XGB&h4b
znS0?F69sZ`g-gea&QZE5x4ond%VXJB`w$(4R3b_&Ui!Him`bGM74L7G6UA!!DmtvH
z!ndZaup<hw1If#HlR3nDHkXOQ83{EykIfdMfq{J}Qe2jFn>uVEt*s9Fom;7=wM~y&
z^4Rau>8x|g$VkgPUe)v<-yd~XU@@{y%>zq@)HBUVZ{26RZ9^OCumV??NO{;9xwVbv
z+GD}fK`mRoei94A1<+XTM=A(ZcW>q1kF@PAF#eyapn3XsPcqEOjcORzhAQXoH2(b!
zKnDYIzKD~*ycYS#i+^E^a{X0r^AP|!EF<<zpGwBGM!I&FwY%;Jo+GsHtOVXz5f6s!
zF77(u9(^_ml<MB&?SaSGi?a`L_o)=M2D$|4iRqWr(y7ZOKvO1sUXjyIz6ndA=KBf1
zgL6_=mUcNQEBW3(t@P-Vj=bUW(722J0N=OH;U}44MEs0{UGL?}Ejk@12e)qH=71a*
zt4ytaQZ2lbxn?YDAJhZw3<yU(6j)Q7bUccIzK2Q?&Mj8@2oATKzbOg21E$YuH)JHt
z{%&in5Ri!oxy6O_2g3<>G$Xpncg+<H-q^!&weBPi>U4`a6q8T@g5U)=Z#x*r9^|9!
zR=1e5?$=wpjKCVKk&!LE5NeCI%fndcn_9PbXu$V3CjOhR$!C@4-eKub+X08~cRI+w
zg+Mua)q2EVN!#*&Cj@31|K`^98VP$}Ba`Q^>J(7agai`btIB$Fo_mEq4LEo&Sm&iD
ze<$LBwbT2>k82~)S$~*fe+7Quao@CziFYZ@ONx)YY4LL|`J9@`?w1$H-Vg-<THZnF
z`kI=(`u*GQEK5`adGjo}ygW+)-r0wq7YX?^aOZr7UtKVed*1rEkL>J`ma~7rIR&4w
zFe%)v>kADV0@fBnOcRo0v+0qYabpux$@r(o2Gg1s%V2+5)!-sWpXP;E%M64pG=<xB
zJ)d%>FB9=1Uqrwl*j6*wK`A?qnIA7YkH2OIaq6&-P)BE34jw1LSiEVqjBHa3(!R1+
zIT0EWu;Rq=nc7l#p05-z@vibZtC;g8tI~dnpXUB?10m14Z<_0ASGCxQiNKm-xT&gc
zVg?Ml2()U4^Q|iv@nNkL@|mo6<Tj$r?tZ;1jk(&&(&0ozBIaCEF<GYwd+6Hoie=fd
zvoVWY_hxkMfZtlDkr!>Tu{y02%e<hgjk*&Z-=#wS=&sH|Z<dm(n7$0im?LX?q^~`s
zx$pFXB!R0wB+mrJSz=1PZCwke*e9ngp9lV<*Vxsp(d<*O*j07d(=oBB*F<nKp9OXc
z`o1bmO1MOb)#Cg%6@orf!XwL3C;07@NBTgYT`=(M{U@ay)VTz%T`b$gZcT2Lc1u6J
zIv+&8PeabzDvq>0oi|q=IZ^xDUf*0UOqI>sC+kxB?4NxnyQZYrt*Mo_oCLzfA59(i
ze(kO%P|Be*2qCSE2O`&v$U*E7*(?4uAbUkWY5K<GPE&rfnlm|N?&Z!=<7{VDarXVL
zJ~m1Ply}sM^&Y)nhklwi3uRA+;s=}aO6MeDZL>%4tpwfF!>Ec%W4-mC^2;yNi{5}|
zRRaRRo4PAy90?R$#91-SsTO*ypYqF3FaF`Rzeq30Z}ZCq;FTb+rgd;0b{01WTd5I)
z&bm^<^{hdi5sd;iAG;g~c(UJ~R4YNE=x$c|g+&L@;b)#i)BLJbZpFabKmmjj)!~&J
zVYX!yJU0t^x*9mK>ey-Or{>9=1Hb~VmAn@H%auf$D0-$dN`(hPva0r3Vw<H*lq0h<
z)|jJEYdNtb*N-(rif-bacb56i>A@D;_x(lKjkUR40@JLy=ktI(c3bPL$KtsZflzuw
zO($=Mz-8sMif6YGeLy)~3}1#N3@h3N{}*aFL=#!8?<X!KRFI~xH_lCA({7JUEh#-_
zou~Q}71M!xwI&c}udG|!=mvO74PdUymgsp+u1=yEhx;ij@cYyb1nVlr%}@&PR>PAd
zFD4n;&mhOuu{&i}M`fQ!rb^mi;oyKpNXo3K`y}DDPHx*)Ec(+-gACE-F0_ALwi`HY
zRMF#owH|djR4Nh?<-K1#tgLR1L@m-SnJl4#XFwu$A_pr(1|Z*g!IsNrZH}Ik_vmG>
z<K#h=3sK{%>p{D1BnvyYT?rratldvlq_oNn$XxfzokrtE0Kvh0R5MTGBSBc%CadpU
zbk%#xyPa&7%*<Yt>0Qj-g7aS6oOfW|$w)3zvnUZ{&zPJRDFt_lAtq%0<i_n2m<7fB
zw2#IS9DD8a@T}t9WL*53Iqm1Q(FgaL3Wq5I-mN#ABKuSk18c#BMbQ-!Emn_ph}xxk
z>sAYP8~+_T^EGe(rg!-lG#WX6%iZ=TPV+lT<@UdmyT6&1LGA|AvhVQ8&$;{O76x*+
z`=U*LU^V|}Vg7nIU;i=WqnY}T?gmoxKlD+rHs&>w5#B8q)mIw>`V?XD<O4rD>zPdF
z$%Th&pJmgCyN^|Rw$=$)VlYaRZu_9c9~GwOM9`#z0*o|7lvu)F)B~3pF$3C9J!={t
zvI4W0gVGSp2W9!C1eEe=1(q3uyt~^%cA3bBvU!OM`v<rO$ubIQlUO5Q_IK6wu@Q`7
z7)Pa23o`*&LRgj0BWMp<Ue`{^d>)(niXEEGk)ac4B+@1TcMc!-8A>}|d4J>Q57tC>
zjkg=;bzp|cjh0Ylt^|T<<b33mK-yLtP8{vn?z}FV4qZlvgUk4oK9S&6iY^2zj>foV
z4l6~$um~}=oxQ=<`SF=H=@OVEaNC0Gwb7C!8}c#U>!;S!A`sqWw*%jumc1%$>i+a}
z()QYV{iSU~rYrsDz913bz)|IF^dh*I6-!P#5CB0`W|k<tHaU5;Dz$r6W6-CyRy~P(
zmMXn;M3qH3e(~ZPu4+3AM%S*=8VHHQnBG}aGj=M0%Q1nXMmR1z|B{!Czs8?8Yrr0{
z^S~%n&vh24gZ$CwWsPAputgZe>LM@X$?C5#%t21gGA0tYGwXfttiWQ+EUle>vk7x>
z!+5{#1`4b40K0n5e(hRf%Xcu#>Q(#lNKEx|E;@J;xBz3ji-6dT-89L3D(ms-ANdU(
zGi!b`o%!u9(Huuy%WO}JL{SKb3eRScbrPqxude!f<Lc?OXa%bAJbd!5R1h}3Cg83E
zW6s;AIXp<=6Ab&q6yQg8kwvmZa77!FL{LmP2g+_*I}2_t7jwV-V|4oM#lLv%FPs!Y
za8md-CjDxJ0Ohd~^V7{($czL3FYK`}%-9K@zQ-<Wf7f$6F^#~e)VI&*0)9@#@~}63
z!Y6ofUfs_RtgBEjyELV@ow0<FrLFX2Iyt91Jvrr=KAOx(<}f#67kz6&mL4YFiYV|M
zMQCTOEPJzF&5YDs!ZF*P+++rCS0@JnKXJqH-9|m3gW)t9ajq%Ry0r3Tb=HS1yKmib
zr(y6@)7s}-6sh7`kXpssY^Mj^<&Sn=nmJ$wAb4#=D@gR6Pbypc5nZrtZ<x+>wBof1
z8nIof?FalB^ZT7~mhQ>EZh-!wUhlZ@I1%E^5|1;rVe?Ht(asQs4&gE`YCJ&?aGt0o
z^!IG7))^2ANbAt@<g4OXP`0)VqrL|91qo3r)#w69p-MycG3d)YP%S3yfz<M=OiH;3
zf;BZ~tCRHDRn>`5(ZD(|N5Z{LwQ#9f%g5==&#3dwK<bLEh_PUgS1^8VPMwZr;M<XO
zd+i*i>VCnM&E8vg#NN&>H}rC?>9o*GN+bEGo7U>x-0i%haMN8MDt=i>op>uJJK)@=
zUPGQM-fvv8M^{}|Vwz5-`@q|tBg5KLG{1fdT-2eVzh$CB(o%FNJ&cWaP5DFb)(B8v
zPaAPdf^EHV;G}zP`m3hhsm`r^-dDQqo?O<wACF!!1~7}??T_{$Dy|NRF1qId7WvHU
z$8FN;UaF9Q0%d=(L|Z1Fx59+B7uCnh7@nWhvCy0Sr?Ikk>!Sb4U~e<5&rP!5>gN8|
zbBm7`?D%Lt^R_9`A>mj!#y<}{>bo*=;*#)Jz>CO3+xOiL9_A}=e(?H-ru5}RCwT?1
zngLYNgg1uai<tyYIpq?*X1@1#4cwc9&6|Mk<^1y^)wmbL1XH+})NA1Lyx2$6R^+13
zmKmJQYw&BnoLv1G`vh=&^zug$HOypkmDEHP1jZUf`Zwdzw~Nr#bCaG+z|;8g_W$kO
zBmd_1|Lxs_ozjoD|2NR{JLV}oz0JzsR?L;G1n}iJ9}s!dZ4FwG9rl*=y@<y+W>jf@
zCE1-Ch)s8NjO~_e4!az0^|PmTr!CJFv{cr>HxiApv#g^8^bd2-#)uVp`!!IvTVsFO
z5*JJ=P8UNZuk@BOK>N9DL9NF|53d|k@6`Cj!Cr?Kxti;dq^x1Qh8U127*|+kF!TaW
z;(Bwk(yK&1*h(<SkuHthW=Rs%rSP^P!8i1%E=$=`le}K+A@61IFSu{O9Pl<ODaKci
z{X>I_rdAQGpwY*^Q<SW9SekQq6z^f$x;JDO##3X*DJ5QlR!}|c>nwA_+evv){G_kf
zJ*tm=ise;#llK~@EFGe3C;oYk4r!1v;f|D67~Ow1rSJaCKZ2g0P3fBp_fIYpu(RL?
z9<6s&HN%)kED^PAabuLH#|y2W#r_xm%%6jt0F(K1FeQj^C|X+gZ3_cU{N8m&@KLSi
zUc0%m0jR_dYdo&644n$3N9jW_7VHkktuBdZpxk!i<H3ggTrlU%^^s84X+sm9a%fOF
zp)R`vB>dHRxMAm5*<br4HmC!CHdDa^lj>D#>}(XI$$T!Y!SW5JhQe+!S!!9M^@1#g
zzwh=SLC2>J2x-6)AxsN!lUYA0ZG)~i;ob+UiqnK!U^U9CY^+oP$*j)DQ#~F=msQ+Q
zs%(zL6&~;~Gg`~VUQ3K*L{Tw5H5$jo6nK|5R=_}uj1$dW#5a`bDcC*29^8XnabmSw
zpRubzWq<|U%PT)I^Pxu|fh~6p%i9R@in^{GLV3#d;Sw0(R{YB#=cO(HTetTWG(Vfw
z|22w=ukPpkt^0X7K3#q}r55R*rqtkn{NdgXL+CwFb9P?*a)1o1rz-2?;l?wn!VWQb
zcz$=r0mo(c%_ohB?Qh=mK*M|^sl6uDq_&h1@-m;-mme>J&)FaE;CDBG{P_;x<NsSb
z9gUfg(Ox?awdlKq4vhjlv=x+tj0*CREnRI6$uROs7l~|Y8pGu*sOy_<gAa+P(PVcp
z?{fMq+G8l}R=f0o6nl^e7*`PQ4@DS@A@Gxc(0xMhhPDezFTd*kV~riBL(V@s)?~^3
zG~}l;DkK`G#3;&XMRgBpUl!^PVK<p~842EA^G$^1tVpeHck(Q)?bBHjLPL<Sv6hp^
z@<8de8<6opmqAny$9P>7Q2tvCKQmZU`3$xPXIAVcf_}rh%EJaNtHzGQ^U8D9W&{3t
zwXu&pAqOEBrn(nyUq1v!imoO|Hl$O2dE7{~-40k&1BSLX5Zzup*zJ>U@XI30Hpy(t
z=hBfX(SR>^9j>?Dr=5=WvdI3Oso2{L8Tnze%f}35P^se@P28cwhEB5W_+YM#W-rAJ
zNvpxxCbpSyMZ=33ZU-7TaiAu(<RwE9B?yZxXGYGk+cZ`BJfx0}0Y;K&Zd_l~{ZR4i
z*vo~gxT7JosS&C%<9=?BYw>}Pr3Iu3amt+w2v;wN(w;7|kOHkW)1HUjTxlTb7SYu#
zSc0sd0_+%8CR3!VPykd*clY9&GByV7WJDw6jU%nPH3;6F5R|kC9wzrJ^E4OWe;7!o
zN<K_PpqaI9%ui)iY`oep_W4#UCS06X(-AwX76vuP6wH#xh}(K%-DbVv`iG4)a(P$u
zflYKW?!bHGqA@#Dx;A1S?Dsy9hfZe4C|py~9@x$4?d|4t{59;7zlEL$hyxz~1@yh|
zE;z3>QNQJ}|1BIz+M5%?x1#QMV9fe!0S6V(78KXtm?%J?wGj*K!Csu|<i_hEa8)7&
zTqVKr;L9coD9us_#6A^J(+drBoXQ8W8@LN%OcKwN!oFJsd@(o#ucNob3JVX$5S9z^
zFnRe(IfsJovl{f4?-LmiBsec?F;GN)Ok`+a-!88|;3-BNy~mh+(U80(Bru-+vccn`
z!3GathRa}Hpk5qb1{=p~i6j(!#Qgg%BYqsWF3f|#o@EB1Vx;kxcNIE#*qY>~k4cEm
zFA3onTT9h>Z(m=gewy;yMFOGuD^vXRB0Yj{E$llglb4s<zX+J`T9d<yyPvIr&T0}Y
z#B~eS*F1|@y^K*h+AXT^=yAmz7#8NAt_pM&(9D4+sGJA~-L?rgWM?w2t_mU$Ml<7M
zF&?Jb8QMvX*0{HO*g~f>U9Ml~zS((a?dBUsMrqjdLPxiJXBl@gpQ|0OT!r&CZLxe2
z-43}C^!5b4i=%WKo85L5o;IW^3y&g6t3K=L@S^*y`iLa-F+C>OgLKB}%`?1f<@L#9
zUOO0_yJt|K@Pb8SNNko|f5=sIp&P9WGVx@!$12-wFHD}W2I(Gd6*sqs6?^2M(zvh6
z9cxqPc5CU%>PA>>Q0~+1s^If264o<%VUsFbnEfs}!z6{ObpK+iJJKNawLOB3r0A_5
z2lg{y9{CYi;?JM>*biRXBkIEVta<L}hOO81zC3TZu|B}+nRm_WPd;P6t`@U*JNq-j
zZ_P0Qb8RrDz36H<|71s>zaN)P9>RbuCh_O%Wc+C1?0tKLkYrxvNi~jo!{HAR1Ru&3
zPq!R%m^ZKMFN4yz2ld0?6+F-6AD>6>`w>GvyzYNXFcRdL<(@*!7-)g^r}b3W+yaSv
zI)|LE%~juyuI*o>bjT}YT~Zx)*<6SR0eiyO%DHUwRV$S17G+b};(>k1N^QP%z*6Vt
zbMl!KDBf+C&2uX}S$mbL5$wAm+C@Jsh}GlIY2cfY)o^Z}Hf2vl<Au2k^&J*84Xr3(
zOCYvJET*0iLb;3coqxpGzUj;O;*hYwjHTs=H`VJC!h&^^CdlP$`4MBw%X$Zua>h|@
z#CeK}F~j9>tmhdx3tC4UU3J%Vw~`d;LSEs-ejm92@iji==4ve<1UZ8?3(cNr|5y?6
z+&&Naez`D0s1Oky1kW4Q8&gAd704JssB!77hLdEIbt>@=?$CN2h4EkP>!HDa<%W+1
zZ{Iu+{+J&A1zax~uK+Y>pKbr|G6eenEkhKV@NQ@f86y7ea<)HYiLZ0}e`JaCPtKEn
z%Mzd=x|aVpvcyYJ^1*V4EJ5gHNP&Rse~>BONdTXVOEiAc==~nwifuw?f~mXdN_&E?
zWN)G=jSrE+NN2OWo((Ojs5t`#VK<(qbhfQv_6=lnH#aNT2#AWK&O+J4JkP4z%X5rk
zYJ)3ue4_)fpQ=TTvbVLhKAg<7xd%Z9s<QSl6)rBYIWu>9*C$9DKT~>V25VhT@B{X&
zA^R1YKek2}gXEDBWRe<S(TXy6_j@TGgUT0f>XhL)t!2YQE#i=DZoPn_t$;k1dr8hA
z-G<$MeJ%588VhRFX>y30(M=*@>=hx!1x#Nbt^^inivD;+>qKi5=J{Mn4S6GLEAg<H
z@(_?4MJvf%nDBPQj)cXm^vP*(eudk|W~L5}L4!gfhuBJOBS7A<9RH>#n!xa2`k9z+
z=r`x49k|Qou5sXvQ_Nz0zjM0~XaA#2fxz~_H^UuRy~k)<SZv2|s|Rqpb^q7FJtU>K
zK!5mrl-~QoxAld{t8#;*GpDf2@Dyl6NbNceW(_*S5m!jJ>DC&LH}?6u4haCzv`rDu
ztKp@t`#f@==?Av)yw)^+&p)5(=huikC{xsgk8LeImtAm9<g3=9#tJ*`OS!BNdPhGR
zdF}L=iLaW+s;d)!3U^Pfaq9e-fm{a@m_rG(egj|ijefGX>)|1`yW34xp7bi$SKuaW
z<DOCF$Z56W(Mt(yuROKUfJJJW*-~6vObs66eR!VNXAPBE26F;M%+P?6=2psNz|OZ;
zZ1n174`%0LMuYO`fQW&#S*zL-_wt}!<F3*Y|3*}mzK0!tGkAhB4`<$}4&5fhktJG}
zPj->W5WDg`^n*ZBaREBDa+5R0h}6QZkP}<p2xm81V`Xr(O74DgdoUmb%VkILx2k&;
z%iEGgLv&wlW9S@DyDNB^@Jl40#TWo%u&ELz0t`4UieGP+4zvGks`imG-njQ*oA9r5
z#$UPkaA00ZWB6Or!2Z7_jq=lL9g;@!&ar>fgZ^0e{7{F#Tk8PP3t$uJo$va6Zy$Ke
zFRwoz%R{z5`BxRG@CCe4MSiyh!RhOi%&J6O5wOvJoM5>M^cr8b3nm&*0KrQ+fu)W)
zDLI(ezIpXLH0ce|ex1QL8yyXdmg|?;SF?eMtG>%&MUjV%7cX=FmkRh#UDHoq3tz1k
zfTaGT{|*JVwR(AV{HIEKa>h+tQ&o>R@UkATUeb@YH@~|degAyl*$>_mNaCKJwvPxu
zT-ONY6y8(z0By$^?)aFbxSf;0j85y^(AfE*MQBcMQtY{Cy`wzXwfY1C$IeM-3%_&e
zkQnnuAbK$U#AXEBB$SQUm&`biXjhN@u|7c*INAbQk8AaCzNWK$nJZ|~z-<xRg>8a8
zcE|FHZ|9!kxpZ?w<?WI;U`dGNLzzdtWK<on&Q%adah6XUUw-s!0!7u%KM<HhcI}1)
zBXz+!?gkbwoYtf1t=T;}kdJ}TVPWi6>1^coycc}TtqZqIx}kTN=k09K@$g)M13MR;
z@v_e(>xoaJN$k^|7aX0Q`IJe~TW~?xDxVNB^6tOm$Qg0ltkXvsr3cElqDQyD5Pwc%
z*vIt4fb&mI@oyeHmyaHU-;QIJbL>NGE=M&a|JGuBJ0%~F61bYaFTVX};(#~YbL2I*
z{+Kwvz9AUie&3|~k2m~<QuSS^DylG+Sl?p%IjE)$@>-kK`w<z+OpzVa9DoF{ZH33p
z7hqrWjuK^lUyZ_nRG+vB3(_uhz2Ac1rrYZeK7fP02lM-JUE2*GMt*W^Yl=i_8<78H
zR3g+mEedY3r_*!0!C=?%9^6J;bk=i5FHVvxqC&Zn&U@HYB!y$MQhd>yR6FMrtQ2E3
zfDK{L*Cozim>vNWiDhRpUMfJ*?=u`YIb5$5Z1`m#;}!<plDgkSEAydC=+wy58Z4Am
zb*he!3eS5d>|DG=ER=rUG-$6)4@3&Io1`lS3s>v2i>qbh+6l(>bDrE&UcEcoK$htJ
z)LyQ=pMyWH>1|WLj@0;=E86CWZ01$audZXz<x+OwZ+{von#043QrLj~YY_ZX2=V(}
z(lQP(IDY`e^v5pgV|y}hznXT4{Le{3a2)BR@yQ#`^0KHm-2jL_m0Z01557Q0-(oxr
z6O%VPoWCgYk<5_>P58$CEx*3;zj=e~eb>mFEAz{l;5AiTfGz$ecLhKS@TcV4HSHfs
z0{nb`-0cdcJ{v#utr}W*{)Zs{5#zz%YI^eo{VBfx(})2W|A%+DO+=*G`}WT8#{cdO
zev>4AY0?IZ0EyB&G9I=tY?U02*Y+%d4!n#V>!}gUf`xT?qYEI*6h|Y8uIUuQ16m+E
zd>`ODR0QwAmyc@9?L9C9;iU9<;0dv52?k&D(?!moq}U$oinK0qq_%`NHR13K`ZR~d
z=PBbxc@JYwcI)M9qCAR?hl_-#7wdE8B@-Bwgs|yp=L>)WH2wKpV>MAw-G#F2iCn@p
z2%)a_Ly?WrUG+wEf>42kOJWlnv4aIpV2uen-(<MRs0?oGG^jtfRC<w4^A;C<Flqo@
zUM1b)#{|SF`lwekTBr3b9#+Y8e)`!R;q2u2=xTS0Wu~-<X<Iz%+XcReBH(Mj%;F{1
zM`jPgwN-LF?(C;=mB3li9KIz9RdC-_K=g<hW~lDR@U<Hnpmj-t=j?5aF1)s?zi^?2
z<FBw;_Ep0t%WI<{pgt5gByeidw^O4(Pt|F8ZB$22?|9O0MgmaOS&z4_L*>Bs#*$up
zfc5{w+na2;scqZZb)I6^RRAS`R3sINK!6ZgjT}TSGNGschZJ@=>}4<SbMAj{VYlt3
z6^T&5oMVpB``6z#>X(}4^(Ls}%BCoAAPVFQ83bUH6jIb*P0kS9nQ$O_bo7IlO5o^Z
zQ1z6r;4&Z!HE^>1*L(7PM0vAJj{FEt@Yw~l2K;{(Zq;>QTLAaqZ|=sLBlzzxWdOiS
zQ;1_P-#^@j^c{zu;g{ZAU?#$4(oh!hqj%&NpnyQ#B@OT-z&p}Y=!U`T<(t<5uT5S~
zufxr?7mq(Rp5KpHC`cLNZVk{PSWgJLd;fs<Zw0PZu=X<eLC;^FP3TzD)9anY@Jx#7
zmp@+0*$5z^zW3(dER!F9B{-E;Z@R;e$Z>Hv1@YFDg93)!Gk*5#x!f7uvO*IeDUI*;
z!5hTwyL0NFDgIx+p>dB~D<0`TElV&Xe206yGN-@(rT$mT67ss)ApZzgzN}0Cn`J3Y
zbpjbm2D}yTRHk3^k^k7H%m@8CBqUaWS@{@Uili3Z+o5a_pp_I`JDSeUo$JcoVOQor
zooC$aC^W|vL}yaud4)xK!;$Eo(UU1Soe!5}PcjFyt5uSEuCV}U$Mrat$f>`eRvzi{
zMkc1B4>jHjbptD=Mg&sC78W<0;Hz7bT-Qb@WR74eF{%R{49q}GYJo)w46A|zsE^G9
zNyVLHLF&#CvN@bkaN=P$VwJU>@VH(3*iz{Vk@t_Q1RGE4^rl!I38PxF5r?$p<3j>1
zr-s$q{mC*pf{(2yHp#9;uV*yAD@6x1T3&RCTsZfMP-UV}j$L(>HHn2du1oFD@y5TE
zOx4v1!3c7QF8OtPB^3vLSbf*k{48H;eY5&YMB1;l!*7+uf16tYDGSyj{4?T)q4nZ-
zioSo?ZuA2Ci7|n2KRu{Fvd4bPaxcF;%wJ-LVT}HYat963E2|iy#m+1MUOuwiVXOam
z5ng+j0B~med7NJSdAon@_RvQ8cBcC=^#|&o^!3^MzH`$rIrE<c#4r6Y;A>u)?Vkc-
zAME-7{J{V#v}a}aKC?SPyaQJ)gZ^es^sNDC($nrZC>fy}_th=SU0;~+l>n?c#F1bN
zUK%2T?uR0*Gm92Zq(7zjAm_<s-GfICK9znz<6=e5a3p%L?N*VFJeozQZ~7>XOOWhZ
zDJ!=?$|*$0vcC&Lvz`a9M!TW9io+lusXKXy(s>`7lZb7KL`7F>Z*cod7;OXkeh(SH
z;fd~Qv%A!(Mo4!+N;cLm0)G?78EiALl^0WT+j~AU+7P~Ti2YXpl~(ZtX10LFc<#q~
z<w&`5aL$uY8mnhygj7B#XMoisCA44kCUMrmS=Y&^=THGehXJ<|3lwvlb<v-VwcCl9
z#P-|cHqJvxL-PC+t80h)bO`uvxolSVRqQMSLJ;Vd@2B8-x!!RDg=k(hzlJk$ogX5e
z5|AZqZ0Pb-amJYsE4Ycy6sP8Wdel!%B!>xv+S=YkuOV}yMj%2gbHyAY8;=btYrYV$
zMt@O3B4uoXcujAV<09TYQ&9J7AzI)d+&pk*yT)TDSk5epyR5@(&MEplX9!C+J#RQ&
zNotqoZm;&uVhdTP3HDOBW{hXG@`UfSYlx>zJ3~ohIl}_b8sS%=Swe%%Ts<Zq{H0hH
zgjQwvt2qaj7tYsXBSQ9=u}K%r<DsN@b|;ZNj5j+xoQGhQSfjXDB>6lFVg6LN7^W|(
zUgoqU+HmbLRD%h5UWD7~+(-6j>2sW6ZySC!nSR$>F(0Q}?UPIXO=|V+91R|>qrAUx
zGaSCmD><+mYOjNw`grx9^!RG8y!OjiAVo_5`D+oM;H8!e|GK>VH>(T6{b_jtvHu^|
zH3*{sr82vd`C^mMB~W8P=>n3e3Gr60X}B5(J=*8YxC$Fsg2!)ng8*RH?#e>z+sd8n
z5RIL!ij*U&&pJhO6{X_craY)8aDipo)-$xn7P*Q-m7Gk8)raMZW*d5rXajcvlf*Dm
zt6GpIkqdC!CICv7?%Gyq8TYE{+t6}0Mb7UIkXE{yQpnqmD#W_c+x8ail(nea@(y)f
za`ErNbQEyewl#3^NcP$fC(xcH&u2~Vre5yoYY%%P;qJY@grhfBq3k6KWn?9>7mE<V
zHGCaPOz>%J)n3(3zd;-l=&0vX!yvNbwmz>iY8M!zl{M+}cyMnqHX1H!6P|+NB-d}B
zcN-C#XPBM(9C|4`Tyb$izlfT0CWy5~1TN!An$<&e^{ynDE3i}!@Gc(FCRJ5izuixn
z)##dGULmwg%hrf(`L~r1T+bw**n43#!m{;848u-t<=#?1o1Z&E!$u6JQ5vQVLgcL+
z(PpcF+A%EJI#aQxSEl4vxPv4_q?budEYeinI+1=vUNr&UWe>XlT~IL&ze(p+?Oslp
z*JO~CYZFVB4UJZ$+hdb&9nXtvC7Gm+C?dO-gC67QAx0a5$7H759_j1z*w(0bfRMvw
z=bud>Tx6+)Hkrzv*Ykz~7~`%rJHZq;avNF45*2$6((Vxm4hJw0-Rd}r_2Xe!ouB?b
zK0nCA%DAodj#(U9%4+VKQM?z~40paWN{EgBleWTteuuyMTWN5t!3;|SsQz1PfuY0S
ziv9ok?EasX1|Lz=UrK{_Iq>cT{3s1fYNN7-|9`nOQ2r;S0jod;D;$oWFv!qMGbUGL
z7$^F(l)XS}_$I|R^V%yMQ!DN%r^^^~c^>c33^sf<O%XP0=v~&5V4{s+aI@6_FY=6s
ztlLAn3<Qh~<Ys$ApN9~v=glz&EuZ8)w?f4(<pw2rJHB4;0*({bIlmz?L?pEou&M}s
zR9E|8TAejc&X#VSIr#c=T$K80ADVtsZqC7NxoN|lOh+f$m&-z%>XhGtGe)atlsw1?
z%Nk1CE7@dGoE49c5eI&k0l0?<w<s5Y3wP(#m<PrW=4rAsmM?utzC~G)Y^r-ff}p@C
zOEx+cX)wVy$BMOh0}$#+-%0Ku_eOcHb1(@nR3Dwp%9{zY8hoETs{<4lwIaj%)KC2V
z82fC03<&&Cqqmq;<`!vj(9T?yiA4ECg1jm-$>Ly0JpQ=z6yVyQcNeMH8>VHPv#XN<
zgy{yk$T&Hc5|iS)E2&!-yDaS%Prx*5?qxZ;g(l9OIr0gq%x03~)e0yF&-}bvr~V$w
zGs#@EKp@-j+AF|{;7{US!N+LBZ~zzRqMRLUigOS}(%NG)rQt^N{8qkRt^{_!LOW7c
zaz&dZ(oXY`@I{utX13yF=XT$1!+TboSuX_getNn!%;sV?(xM<oFwB4;u`ez#B)Fep
zKFokI;$A-Ob#sj;Uz;s7e?kSce%#UWK<)Aq99NCPJorR%O;_EqP}byl48!t?jO?DQ
zFV<-@!RYz3G!W_X<%IYCBxz*m-3M$qrA5Jm_Jj@pGvSzAym}NI72qyK!Q&EMA-6<1
zes)l1;8Xya_y_@o;NpOwWAH>@-kkx=|K))Htxo~lZw-J3$5_Mk<<Gr|uZNGX2MZYM
zKuZFHGr$Ks6n>iyF3=XRnCa=;fBgR6-ds_hUR)b~At=T`5Qn|yxzH3DBL>=j-}*zy
zyFmm!2Jy#)_?H%gS_el6A+bLsL8pxh4WoaOoL`w?h-F}BP}Ok!%iIFxfc_*&5<v1w
z>#iMj+%52XmE*^mT%3(sIf8uW4_}MCzPPzoh4BQZ+ZU?{C=Gw=_Wnz=_fw6CKpX$^
zyT+{!IwUBY_{Z-&DlcIxn72)0lb*OIF>BIv*m&UEUlfcS!*<iF!Ek5TO$2PgTp;%w
zrarINkPjT^74X=^{9aEW=*7Lc@NdUQAvE=7ryDu=-lKK1Ta&9E$)@M%fMcHqAZxVg
zwcw%yd%K0%RSeNQXRanTV*c3N`DPt>87u(P<&>(YXmye{GYV8AqJcO%sg*b66Vc8i
zB^{=9Tj*-NZY^7%fD$Kx6`ug?$BKHy&2nY0mvRAUIYe`cyUaLCDERvztbRTiC1G7y
zZY_ZiX@h71n26?C7hK;n<vWx^M-xBaH`SqQ97qBl@Evmp!NLOJ2Gr_?6IqG5DRW&G
zF`e2=Z3i~v-w7{vx)#|NO>vzf3=P^d-lY*$5f9b{ov}}X{4z?$owb*7aU>u<Er(Be
z7wqd`S%W-e6wve@UBDrKxCBK|(+mPR2E77Ms$wq`i(U9y2#nTT8pv*N%`njQkh>Gb
z{*tZ9JNQQWgO@TG!Mc{)94n*B8%V@aB%*%`&WNO_J-fGB3-}~ART=nS_*$k%e?l=|
zM+G@Td-Aq9?)*y%6n!Ix?C4O8a%~r2H+xxQx^(i~N1=v?b&E_#Ntn(wg!Y~83GjK2
zfQ&JEI}`gMF?+5lK-oAB*G@k5M*fJ+N-EQmpKs4~Qv;>t1m_D=ee6iz4DLwh;hFG~
zUp7!5*ZL}+i(-@VTVt=2PxGQwi6ieWR;G}WcU(#1`WZPxx@p0Y`>*Ehw;b%n6Mgx=
z#ySuE;YGmy)7<8NM!sh4cMtkAdLB+BzoTB$`Y+G4d}TQR`?&834m{Hua>Xpx#K0~4
zB?(@+gtPRKb^v1z{3{0Zr4}9{8~?#oelMN|fKXsEFT7k_fL}Sk@-bQcg#*PuoL#@X
zcKd(*&i)6kEl_XZD7FD$&jX?dh^rDfzI?())YArqZ0<9%U&{R<K}Q+PPDIcK9wYlz
zweQcl*FrdtCk;>Fe6){8+a82=8zsl*@pg=Am|g6`fbt}wkXY~)OxUO(tkzl;JGle{
zqmM1Oi~YqZD#%=S$PQ}hV+Bz&*+I(G`(u#Ser&BsYUYCjXmyH>3!&JDSnV{;mEijj
zT%dM`lS>Q|MwTFpZiwUlVwH=@D04RR?x7%7!ZEWw$Y9g0tq69c&bq=`@X*A&f;xbi
z7h3Q#RkKNiu##8Q(F_a4<+R-hG4B}{-lq@O9z;KkAU*EFu<Xy4t8&_29_Ez_nZ*uR
zKBw-!3D&5k$tB^y(i(XiL}Ie%ch{Eoqz>Pw?B7Ce(h#dWUtM^60~C)Xg-G$$K_O@1
z!v1p`{1X!Tk1c1PW7_UC73LcF9kadWNj2qG0F)}XDS}06hfd?F+HBCrYz;On&H}_s
zY#PI*Dvh|%!DJHG(*f(;{cRN=iM8Nwv@^pc_v;OE>g0oZpKg}9U-?MIG`en<mxsFb
zV4gh{^KtBRQikN>La&y&2T#f~uKkOmTg$4(>3Maza+SO_kB+miTa{=?+qRvOJxNH9
z?^owaav#@fa8vxznsRsA`w|#wDej@xnlD`?gFWrd`?5*y!Nu~)c7n_?(_wdFmMt!f
z$DXb(wwZ?zcUoRS;zzh4dNT?0=DHdLrx}b(QYs1`?*|Y~Ex;x8EVzjty%;_)eO#N+
zsXE=}1FW#_^dy3DaX=l(pSFjog1P2jY!jEJj_>||n9}e2|7`xU`2FpC$%5z+YK70y
zH4RB#|NN7)6#mPicf6RK_HuerkTwL1-VwRJ22GBXeHr>fgO@gg4GvaSF4Z>=m~JVc
z5c*_Iy_urGO$Lm3fqgSJd~im;ya)^5&#Wn<C?4#4xc*A|YD~I>@)ST;?w3J2R89K#
zHw4KopkV-Gt&)FZOZ^uH1X$Z>N}`<0&5o9dQp-N#R++m+P%b}A&Wbugzinr)8jq+@
z3d{+&<9;y?7VXQrNh{o%l{!vGAEA(}q<)n;M`dF;=qT-09cq^A{c#CQFx=Mz-k3LQ
zWe;*_VsK2XTW91{SgTMA-%P$TRMA-njb)P<*==v>q|~lAHV#6IX*KRSvfn~%CB$zW
zkz^G~x^m;GK)TwM8)?ZCC(K!{o8vCjC>n@2H%oYmkT<OhDPxC&aeab8eLCOCT9+*>
z@PU#65GQu8A;Xzx9=qFG!}F#kn{X$9Y6A|csYMgcalnj}fScmR-nwku*X|PJF+10Q
z>IY2IAZcB6HSYA0N<%9akUBZ98i~4TZd2@{2AsJNT0e6X<wM6E)MtO7(HrjuroQ7C
z@Gx9Drn5!UL>!Q({-Z7WM@bFsKbr)Yb)scqeYEkWy2;&{l+=2Aa0djtU4TNAm(Q-Z
zN57vE);Zl(N*lNtWR&dW?-s*(dSKKYpT|et9Zq1AviXN`+Jj`Ni6&BI32B9HSx!X4
zfQfFA3s47He^l4vK+tjmZTZ`?ocP^i8G`j`5$-z(vS($E2pQn!Xru&k{)y!wxDdyR
zqKK%-pV|yL$A>$<@M}6vT<#8m$^y-VuHb7@vza<CVC}ZAfIAhVHT0}x!YqjR2^DRr
zK0U*G#3TpR20>y@czGb)@RZL)I@X5a(~i9*)#$dnH|qgD*do69wFeZA`_;XJi3`jG
zm~s$<IWo}WtvfSBvjoGmH-V$IF;4;lNm=t<Jt+Uy`?{Lyf36qK;PZPw{73Nb(d^$p
z-1>jS<b76ae_OM-=(T38m+w6>-8k()H$+j=38s-xw*wEJjQ(~VW9agWPwJhu^Fr7C
z2#QJ}IMBK(5OmCc6l(%E<1d>R2A-$Phe^D?0d~8|eRB0+@v|(L%tNTAWg3=ry_S|~
z{XP*LD!=SgL<Y$x+Y<;HUq|DQfZVtF<sI4aT`F(#@NUd}Z9z$SF{*M2xB75-jrj#Y
zl2q3>xb*xA4SB*64Nk&Wh4`)xzr2Q=O%aU$KVSRHBlf?!^p~pWAJ_iBy>y6Q|L3)T
zs6v&$Uiz=)C&KElT!1U}64QN52)~n38HgcNb{ctUlKmcsb4!gKHwiW!-2Ey}<khU^
z7QpSz91f{TDmtZz%=ZD<7qTswk1iapd+rz#=NP*Kai}72pn4KUUy?%e7@u_HR3V%|
z)$TxdXi#*P5$e`S@%GYvvk&+P89Lxytw1TV8S`B0sV%zm6(>{xe*=XBxMyZgI3G#2
zwRX3Ovtj7P0uTv+0jujhhJ;*zXi!4OTA--gBl#|2^GJ;vkytvY3U5^<(!E3E7vTvE
zRq%Zm&uS@DBVB=tx4`;iX-r9ReA@Zo70X`h9@(UKBHP~*M2qn~4ODl~`rV5t7#}OF
z0qLj;>3%>PAqtG%+4%$dpM}mQn{$sVYpGW33-kz==9&fH%g;0@W+Vl|?9WRxC_jI}
zr+%Mo{S?(M^NX_p;zt5axP=Mo!RC>&U7xTXl5)8>LOc>?{h)wi>FiV6hP7co{gjrx
zZC{b|7PKtz|JYm9e>qrypBpSo<&kRv><nmOogl6}tc)$lZD2a7E+Arje2mGifOn&s
zFZv$vvVRe?%-7HU8A8v>@#`0YB<Y_e(>HIi@hG%yZ5}HR#QieXIJJEymUT@QI4K`O
z1wqjuUe<QbjJ#qUZPPuD3D+)z(<nW$?#B774+jeDeZh?iw3BuT20$<Fk1hD5`=ZPl
z5N<3=dhMIkMw~dLaay+U>lN#FL;GSgX6|8q_3*>_mJ&RdULMm}pC6sdx)KHU0gi!F
zL>~4)6+vZzI_?Es_u>b=LCWVezRpw+YM1a8z9ygA4%-|8p$%>jtGSJYQEsZ;uyq3a
z42{aYJcidv>>k8Ha%xiWA&oN)vra;b8p)ZvW%G<BpcE<aF_{=T?&hNp9-xcdDLJdH
zK+&;I#AOiHk89M~U3}HC?&=ZrDgw-AO#zUHT56ggp`OU9?k{{+#s@%5lHV(+lrz5y
z0MYuNCV4jw{uLbaxeI;&$6t4&>D8nqJ|{GPY8(72{Dj?tt0C&`dj%aM`W3QoH~avL
zV&#K`3urN~#1Qs!1@F0;0@6T#Dz|j2KpN$@;6&I^+MhQRklwpCWV*r+dHQxVo?ceo
zUjxlISr*tZ-mjInu<UK^wV{HB?d_+a#n+|+!Mg}ZLY^x7G46TqH1OibYcTHlhwu}&
zl^@Lx5Ln@~4ozNoS8xvSK>S&)0oMC65v!7s)AUPbqOuB1^23Mt8LZgDGxPa?zZ<(?
zMf>)y)z;hWM(|zNhyt~|_)Q7QJxbO;@`L-L+I~pg-mMk*C~nYle#=YVSpxs)>wE+;
zf;#~FNav+w;L?E#{^?n`5AJ!EO*c(>Et04BSf6&MHS3Wi4>~&FCmBO9)((%#M^Vz6
zi3}SMNtajT!SmotZo$W!?$h9Of`N}$Tom*|@pI%Cyxd_)m}rv4Vi8?IEc;zivKR;T
zQ_Z;jWv4=$)zLWx+RXx@`d#31MC#qPCn@bhT^2CM%hAvuOX`?*q^lAv0hYB$&-)!S
zYt+NP+Qnegxkh?by!Sx{8iReCxY-Cxt!A6Q1v3z_>vl9OaJ6Br4tmX-v?Ox5>8^^1
zhBvmI7t6vQwW#w~uK@v!cb0F4zwKK}-Gqw8@D^kTH!}S3oO{EDCF{ut7KTKb5?@!B
z9-<VWD*z$C?Yzqm6n@?nvR1=6MTc=e0X^!g4#Paa=L5L~&XEzYlsFolbx{Ft2qdb&
zA7O?|0im9OCG-nv2=`sXtkRR~-hUPNfBxk+CGa=#3GTCS`f|@K!RGMV7+!0B^!;S<
z3Wnd|=DNRb+YQ0B_Bw`ki+hYl5OFrmr-u!04FUZj!r+b!Bn@Av<8j5orRZ?4z+DLY
zK>hx~2FF24Obn?C01o(eKW?Qd*c#O94FUe>_3Zt)ixuWA|Bm3043I^F7Qn;zdWc`o
za9ACth@*e}pl_E4S2x@6(Y!_%Da~`U`J`bN>OV$5?ol|U4NIBSN@3^Et&0YIIuA6E
z9|kxYMpQq38BA!ttr!caY6aytpC5>2?0*TJf$g;1i+~n}hDIWxWT7cOro$lpq2Z#n
z=W6vXwJuf%k}nj{W-cgxDONC(wa`r-dQpo}lIDXh?COLkfMob5NK>397CG$~f;sHb
zXgeGev$;qjVc(uJ&s6B8WYQ?mVYbtx`8v93f}sg}?&hp#yMFE?p}zHdRmmyypkGIH
zR*|cg5dG@GI7A%<;SKWK4&>~JRj+fQ!#odhL10vP<t?(VHM!w8?yUt`E&UuJTiVCn
z?d}W*o4XS|P>?^f;9J>|z}qh((@C+I)%T^;SLF5Zcw{9pU8wxR-&dC{#ty5<oqBgM
zYd!YWP2=<9QycL-fh2i?YS#$tS>tfDebHmnZeZMr&DY^`IRV1yUpwPpw!81KGv9W-
zJpE{p|B<TkeWUxk&w+cqx4JL+@&8y!U#VO4;g@F7|J$wZ7f$J8a$l4+?#EsJy2*bA
zr+w?Bn=Wj1EC*y=Xt_VbkCr>U_<rs7_sjZq14sT(cJN;|@Dwmx@A>09+xts?#pg=#
z^|Ih>^bZcx=g~+9j0k*z2>4Z4yO=zb0WE~dqy0#P4RFqQQQMpQDjObWAgRZ(28_z0
z8B&O3z?6I?nmz<8X_qZ?CDO~~AR+5POKL9yBPTpR=NYx5YWqC;r-~OFMn5GG+yzf#
z=R_ph>k?Lm-MH4C5DY=aTbNUExM>S3%ORS_{`}%CHCww%HxR!2m;iC1oOFGQ2GjWh
z*yPrVsdn2%x9(`l+Y>knG$?3$@79Bkg7B}-txONrVGM>SZcBt<CA+|Usbg%}(}U3`
z4>)~b9|M67CvI~Y_e|puQJv`K!t2?5mrgeP8@A~|8Pfi=dfcUlh_Ri6hseHUKtv7L
zWh~R%7`-fXW3KKr+Ph*HZ5quv%%{>tCw)s+hF+8~f@K>xi(X&tit9XcALtD8wU+=b
zC7oeP58E9aR#^)D*f{pw`IWfHUQ(r_<}Am*EY)wX@=t&5cWnnNHR#X3?J}>;p?>W&
z`iJTHCo-GC)~uiO(r$)P<JloFy<1(`HtAZ(pZ|2-A@3+N3OaPUJ9A)!a77O)tYD|Z
zyZzV#DM4XAj1!!$L%#5l7B)XWUf*}EpWhkm%CBup{`ed&5E(U#E|+}Sq@l3G-y605
zWefbiDIp+a3_eb4AbMfxi^5Y!W_kHSbUJmz@f8fHa2&96znt7aBJk5uQ2^(`?>6DL
zJ^((<-+h))YYb2}{PC2BXzVZ-Nx@l6Z0zFgfL0mmTH6xRgEVCVvyf+Wg1So}qHz^J
zMMgcr2uwc82^fTKD0<z<n6tiM9A+L|1nh>i?dGfbtb(E5Toos*f<g5_NZ_)4+EOIB
z?jdvX`=WbRDl+}O(fBhUx~qWB3d_n#;xe^MK`W7jXcF1TPA9EVVnR2!b!)`>1jnwY
zY#*(;tc_uUsCDVMk1y?2130iJTG)EOdZe^;u5h;>1Vfs!a$!$9M5F~ZfvNWe_GeSw
zhH<FR^=dqiyuDuUvhZOyg%+GlT33xnnoLg`A|TlH81_3?79TQ<uS@1K@CFz6$gqh}
z%58YNh3;@DpsRCu_6S3`I9?Qalz+JR+Ws%6Jjd{>_CMcuJnB6R`Rzvc*HfJGdPpCi
za}XVdKs`y{y0!OE*IqJ3?ka^i)qmas0MJRlP&Gb#io$2w1;jc!@0jCVSn)$Be84YW
zevx;d!sLqY%n#@#1~xBcfkVARDq%n%y;CZ`cu3z-#>xeo!Jx#x1KbwFy}o`hL;nbG
zd(jWx+1@XBK{*9*&VnCiG;nBo{lM>mM-bu)qY|+3fe{@0*!G~xe!a$CpX!stUv0y$
zm-oGGi2VO?dEc)NW=Ef}hX3{DeZM;7b$P$`2LGis=t?wkfOIZmyN56_$O1Z3L%S!v
zbH~ghA{y(^Q1_YR>~4{U$I0E02f`l&f=t05l~=&|^0vm^aT!haHpk<k3Ry23qKs;6
zpdR$XL9Re{yi=41k~GwPaG#XULWaV?o|?wpxF8EuL}}RAq|QD>P66n-x#9i&*#xoS
zR14;~2Us@dFBq1bCDApO^M1o~$n&^LFI{!*xvh^wqb{z+UB?NECh0~JPh#6F+pM2t
z<vDqp8J||$W|ig7Iy{83MLde?QQr^Lo@2zW2bEFid79x%7g<%AWjR!7#8q1}&vw^M
zf6>}eyk!x9;evIyJJ;}J#H5}Lbq7HnhO1<#B1nm9<=S*#kjG_fq7e`upc12~<1O{}
zAV7otZ~k=EG3rBSXv&Rgie@QTvVU$H9zWZL-;ad+Z~KOdr!_b1=<alAO5!<AUHY4~
ztU=}+KmhprObKnnmw#}Up}a;KCO%3(^kTOcE-D7oTj0=7&sNuA^~z=aU{KjGX3*i^
zFVqxAeSPOw*aj5_j)X9|`Ym?*H9F+1v#!w4WhO7#6K5(qXrfjvHz!=;m3qCC3bUog
z87l%AzlR7aXFKodJOmgRO+h5?9_)sNfOxF0HeG2KmlH@254Z>V?BI-Jc5+P|b=4+t
zSiPwCK3Nk-n&#ZJSsnw(Z+0{oMcQY|m|GO^boE)v?MsX=Qz1xPkKOBY65^BM)h2!?
zfoQ|d45&vP=6pA!lP?+vwQ{Ox7w&7MelUR`rfM&OHp@KjSctuo<V)ZPM{T?h;GvlU
zzo9fSc)}4HB51VBk`vVfKd=1bKplLq9y=z#?abu~DKsL8=MH=Q97H&#9-yiu2ES7G
zw+pPZy1vtc1LNxYc|~%jW`fE6!OXR>Pa6<j20AMR|9%WQHveJ$C<1}dx9RxrGnxLn
z8@v;S#?Rg0Pd&ZcF9)4h=H(~i4;FJ5c}<2yE&^(v0q#iyhkq|X;+er+`H{p45}(t@
zLVn_wOaKH12Nvl9yr67w<e%X}y(Z>wM&Nfu!9VTLpV3Y30@HFZZMn<GI9&no^G~nM
z?zj0DP|&a3_tpG6!|ydtKk{$;>eqJybdSS2WajtU45^(Dh54Wuf)%|1vVDpeN51fc
z;`@+4eT~}T0P~l$cu4a8`S}6YFU7vjChGeSpr;V=1P_x+MC(=S9Jwk3!-C^r`H{Gj
z5xO!b@A1SuJAYc+W9rl~NqU1sI1Hs(E>Q8a3E5011!E5=RqD9~UC6!>U}|<fT;Svp
z?Uhw5_dZ2lQ-3*hGPI^a4G9o?VxFUnmo}?TUEj9Z5;%ZrgFG4|-U?fBV|K2iZEUv~
z`YtFpIojM(=756}3A1+bbUF`C&UK=Bit_V)y}xKaCDMAnab&AJQwCo_9}aPkH1@b2
zp(!c`KtM=>_1@R-J1}L~%{^CKe|qdFqEwR|%dWb?M`fOBl?Fys!~&ASIGnd*L~eYD
zWr1Md6C@xJqQ!!f!&W>lo|~1MXUSM_!gYPqIA^zu6laH$R$GWGu%LI&k$833U@j1X
zn=bfa#zo#rSoOs9AS#52Vj~RHY<X@nbbyY<VBSDI40X_lxhQo(UF^yf|A((fUf<Bz
zap9TtUmushJuN^X`0pQ=Z$OWKJS}ibQ3^F5Ur#{9&HCro^2^N#CT|LyNpSzM(rHhg
zWUYm+QtxN<ZaGH;1=)Z`SU_I(Va_Y?tv6s+iYX3se7NK~iUvp(TV2j)Qh~9`O@PRp
z76AHjZ(P@CP%2i+f`s7g4&&8u&g%(tB<bW>DTLG`;<jvWLB26VQZtq2DX!8mmqE#n
z@gv_CDeeN~(VYBHi&oVI0OY)0b{qVe#eN`qm>dAarqgZ=t2H_zRErH7IrTvuvYgnC
z=`va?k+@`nh!YFoQdS%G=|v-RdV&_!--NJBU?@0>a__MN2Vcb9Q6Lfs?A8*&0(>xh
z1NeCv(3+s+(odppo=NUb(`?(>F*873YTKxgrN=x>QNfuJlC@m5USP%qqfr|Yan*S>
z2@b?EPTu}sn_8E_hY?O(Mt@n(|4wjXUk2Oh3-<4KhdGvRezTtc;BI6}h+MY@3=qB6
zvmb@X8zmeiN072l-li<W<#UAul_vEu*n!A{pT6gt76~k_K>PnqmHxeq{ZgdAsQxhA
z51+8MZ$652{pZ)9rHlOSt^f2c{_@V>n*%KFe|+nId*|@jeD<G^ix_NwD!`!u1z+&J
z|Fl*5+X_TQ71*2xfaez5J+6}bls$@oIPxbFaM8&g!gV%4ia~;WFbl0OS{4H2;cV-y
zo~4AD0I77*_&@^Fj0J**GSz*<Q=SsvHg0$Z&G!l(kZ{&ZZydHTf$$zh=OL0|T!|$R
z_S?O;HtG3BtY(PtLho_kl2M#==>@`*O+5Fnm~$A|!WB2L?C*ORFr?EaJl}Tu!cG$n
z=?+NT_TG`*qM^lsz6npJ+j<MJd@$U3vf|qJ<sEneG6*QUYlSv%CPt_r0SA%@+o!1p
z=e-}^*H3BNp!P*Y_#O2ON?dllS`wNVtBpIgws)p_QR6Kn)oU?oAGQHQ+z2wJuu!j}
zJFVV!YZ&_Gs02Cp$NNYBN#I}P3g#DWDC)on_BmilHW`8CKg0EdW!SC~alrf)9}Sut
zjAdk4>xGoQe11L1{(t!fg73jrGyCk%GQF3$L^7${YomD86~f!d0h3MJ>H$)5g?hUI
zQrMvnHW3gK6=>0x<0~c-eCW^_dwo-H%GS&M@%c^vD2cwV@erP2fq~_N#PEHQ|F740
z%8tPP8DU`od5Fgz@e}Q7M_oX<p>M9^ws8tc9;WSy6yv87#h^TAGverwSh3dX@Tgl^
z4uC#)q(IcQS|9_EuY`ShUDYQA(XJ0_BOWIp8nN+x%pd3#<9FnXvV@_`+<|TgZ8yMl
z&mEJZz$Tro8<2K^?09p6^Mkj&VxYo^+2tga$O8^It>5jhI}Z}qVhBzHY=k~Ed(9-H
z6K*?Biyx9A${@(M49Iop(d<*ik_u4kaQn5^MRF-d>vj)`MPzFb7<HrQc3aZjtrY#1
z9;9nl2IT=WFL&<X50m=LHM3SH6JLiil&88TiT8_Et5z%zaXAdS93=S`(P<urwg;T8
zWR%5$9n9<#9@qc6#t$Ip|IaIkAWyza`l)fqpb8q6-*$6;j4b{>v=Bf@!_dL;yISl|
z$k$(o7Lv*SoL~R3_SXi?l(Ezli~=g*yTW|r*TZ6&Ll@%)Ee)vm0saLlF>oG1LHxB5
z1LO5qhqi)1RG7)P-{yX=j(+-Hp1=L@pTFs&d-KnI!F?%Xx4>n>U_P&W+ZcYl_<rr5
zLyLc0*4JPBx+>)V<g$Lfs-Hfyzg^b9As=3PE;tx`M?SE0uJ+P&d>j)YPfBN8D>%D^
z)ybYs7oz8k_&Ju_eWA!r>OsWua#n%rjpy1N0DDAJ*9+1BVXTQ1Sk;v~wyY*}>}K66
z>!H_=iQdJL=zxE7q15d?IlnxKNrD|Qk8M?%&!Q6>ChFR~U5Jy|qKMF-BP2V(_#?KT
zmIzb;&E8(OG4mYc3yv~UjuOx)fxJg+yWAcT-rqB+>3y~KLJE@Xc*a?g0r(FoG(uvf
z8H71@6p9<dx+V<p2FYEHx82BIjBU9e(FmuGXF%?2?pm=kTW;B`Af6|yYJ2G(QAr`M
z_>Fcq1_Zp0+|E&R;W=?rN1Ik|^;f;#cHJaT+u#71^v^mPQOyQ+2Jd%wJQGW^J#C2y
z%<mxA^~-=L6or~rpm$$oz3I>aINZan-$qY)Pb=}P5C>T$&RrO<yzv0oRhDkW^#*5l
zPc#-yoZRW_d9^ur-EsD<lMzbJT|%0^I`N%Po&;b<r<Q6TcNakyMlPe&$4x?y8oEDd
zcQt5*MlGt<vB=tdFWaWkT>7eTX1&wbQ;Um5jv`|{v@3xXR&}%8sHuw8o@Svd9h}?-
zHf*Q4Vz%D%9-$Blw^uKxWDCHF-cL$0*xh#>WUAl1)#G*??iLhzw?!JVq6=qEX6N&h
z14D0|ujmb-j_BgfC2F6xZs`micIb)y{SibPfP|N85Hw|(-i1?eQ+AJtx%MWqB}XNq
z0Z9MD*ZE1iVn4JECJo~u5Jc1Hul_`T_fyO`!gIS5`REFrA4`5y35wESKBwuBCbKBV
z??O39B;)(}BvW0(<VWd;js|GB5_SR|9PFtW@&cuQ8zDWtS_S(0`rQ$G<8qlXm3^wt
za_i?2)VFW`^?PB|`Q@8|t_I16FWClIQ$^Z(_13Ld(Z@rgz^6oEt->W!uaU9|@Ak{g
z27$`OhR~(&=%}CG_Ltv0e>0Je;n4ud$PbC$Tj!9M?{|*8Nx!dUn89C^Ea;1aU*Pf{
zMNPhqziC+>Q0ZPS(^E|s9<;nM>gxmvXf&ZROAj~www~#*t^;KLEtvxF7X*oYSw&k=
z3S%BvRo|9<qoYgg6U6fZ`ekla_OI?g0!^S-CPVR!81`yGzByeY(<+`Fy~41+@u>K3
z{I2E)2<)Sr!d`FW)gbzcocza;_Xnek&H$3uLT4!tj#r?F>FF084Ro3G5oq7p>xaES
z{BMobY5N`Uh0oscUuJ@A`jae%)%81{m*8@EedsTQ$Avz^1nXt#HO;@k3xPfO&ru6#
zDqpt-1$kTE;pi%==IEJ)ADHp!bC5JF@MO~FSN?)?%N(cV-rPn8RR%Ox$#OVU)IEMw
zQ&?71vA<=IalbWvLB>YXR~6j5ZBnUEmb1r2SAS&UEonhASFuY~KvGp;?=E9ob|EVd
zvO$Y02&huoDU*%cLoWb`J;+28YsgB#x(g7kPU3a9*)<tvvITvQ&vLQ9bE@ib&Ut&u
zTSt?g@oet+73fG3Wps|l7(O|IyXX5J0D4H`uAoG2$eUYZrl-Pje`d+(><iu<L}~20
zQR#IJyVf4Y6gqy}`Q9SMp#hOq5D>i4H(UpOUt_n+J$j2Iy?XcWHM?MtVJPH!zDk4>
zji0!yDXP3!KlL`McMhGyg~yPG&G8E4?J}HKJKkWY%?e=KQ*&v~3gaPM<Cj7e6?feo
zPeeIqu;sDaMe-8+tgm&*RNs^-TU}jn2Q8il%aC$87zZ&Xu`sE|ZSAp`cpJx+Hh0J}
zbqqu=Pj_sviq%9936<x>jG4Kc(=v#d1WO)P)`2_3JN5|o#Bh>MJNI&96lpd_=DbzG
z579aqZkEWpzVX!+4nf#9jbQlUdFA56(rC<e#m>AfX?-%yINn()Xwr0W_Yl@tVIZz0
z7NThMb9B>X)J93cMv-zp@r==nTmHJ;H3F&b4A#wA8DD!RFTSq=NK`47OQxQC9>tsd
z-VO&1<;$zyRm!n@9>g?EsqH>&?hTTzcmG-hn|?m~^?#1W&+Y2JwW;~=<yc*R=H2K5
zygPdcCm6SItxJi1tj34m8N@+w@U>q6Qx@rJJt1a>`Oug!DRkX8;AVED@xyulUvf2}
z6OFJYwLmn@tePC?Rl(x3ZXDKBetm7$zF0BdooGIUxg+)F7Qu;8k+5h&A`o*1)w&F7
z{X?7j*M$Q<l3e^#XIa$Vo#nKF&N8T9q^2CtPo=3LNefNR_PUZ`e>ie!b|t0ff@8rZ
z*VH}XJBi7Ko0jfrLD<LEu2YeD?-*hXenIvi%IKJ!y5kjJ=CC*%(c^p+u6TWJ(W@*v
z-WDaQ?V~<lO)S`QT(MnsnRl_^FQ-TLzPaNF<koCMMuPi0qDNES^$GKsA;~G!*5=lQ
zXwjp(I}Di_bjRLr^dc-T$HyQqOE#ZlSFi_odKrovO$ZM~*2Zyy5U;Yjqv2DG25B@U
zi#U<xR4Y64V1lDxN!r0=r~z)?p<Wbu=)(;)NJ$9fPn(a+ZcUh3%q3?e8yr$eG(rh>
zSeFASl$)bV2u*HjJ$RkJx!J~QxX@i~j`YwRVBiPGSl0*Ew6o$Ztpx<5?L5}pdTu5}
z!rNowb86(TL5UCEBy)O{{Z6Y0wr{p=G=alBx}M7TK$OF)M3tAmkmucSup0H^5nC-J
z_vo6h4o4Evx6pe-K~#PO6DhSB%2j!B#@h1hQ|wP+;L;}6#+OZK$DOdc=Vr(=x`x-#
z!P}j%5DmR$#4myjrOSZOM9w7uK)GI=hl?{Y8uLzmoTfZihd%ZX8nHDr&pm+$Q1zFX
z9A-NpPovT4jljW9VCIaz*Y>I~N;GmlsML9kRVQjIoRVjHExQgVUKQxBw5>y0Mmd@y
zF<$1?VKy_>X;z$I2kOIS?17hj+O5u|)48+9?UC|wh8~lF4@$Y8uTA!Stm;G8@8_<L
zsXWi9BHK)HCXbil_9T=_*6z&JIzK^V3q2J?Fcgk9!PyQPB!w-Gup@W$Ont|!!aRVQ
z8QtaKGiRv`T4*c<kUXoFp`Ay)6lC}`i<yR&mvn<*hb5lb6RDV;e6_E?K6Bq1=<t|v
z5c)JhVtk0lin`V^SX*q(9{Zj0jAes}=+HyVvAhgHoieqNchyafWm>5q>{nb!%g@AO
zIv5SMyXza!C_~<`vb|+1>v`MXVo7t-CF!2i1>M@tLYNSZFyBBLM%lwPrIWbEAzmlI
z?Y5n^Da2;WwMbOQZtDT5dZ!#REs`r=_BeQQvs@^^X`q+j9^OwY1LHS{8lxyBxElk|
zq>sYqomeb-n!H^k-?gy#YB`Oj2RtWf8jWV{ecIg)vmf}yhK|!~=ZPR8^RjA92_Oq+
zNsyIc<G!=jfh(+lNa(=2rgYPglG&BA(2Ur4Cdyw9?FvjFYuQ)ExXn@#5+RZckb~n6
zi`XZ$KlLZVK44l*16O~%o-uN@=DJg6Tr>()^rIWZQn%A?5=~Z)p47rZt)80%dj3F=
zzrd9Dh@a$gW*F8CRzgr99^5T8yULYbk{`hQRt(4sZp`VV-_emiY_a>Qv}HQSRlo*0
zwwk*D9Im#fD~88+)L+V-?Tk@t(f6W2$RRnNX&hD>kt}20y=y?QSFNN1q`75H)miOF
z$lh!(qH`LT!n{3pJ1eLmZd;$Li+XkvNo)*J+0wF(Fdls^bY*X>9BZ+ibijrK$pMyP
z?pJq@)FVg{CCIgL583#nuq^b9CqM$%t0G8T73Ak-sYe&J4e|4Wtdf}lDMb~AJDnPL
z^-VO<J0uWr>xETDs%%}iPc;2sLx?YP-uEFitw1C`fT8g}>)*7g{<?c_UPqt(XJblt
zsF8Q5skW~|V(n)%!8Z*Uq5oM2vvl$6?tO#ytMSci_3~qj*KO1SO!!sBe`&+^rtq)L
zDQkfZ_Z^z@B|Hv>Zl${Q!9o2xdwo;RzN76xDfEF1HQ_mWxes4SoHzOn2v^}XnAfGt
z_ZQ)O`o(DJniL4n6@V_k-M^3l%FUS>ZOUi3fujQa%I}#L&{_+PV_73`W_lfoei|%;
zuj4Zr3<~H^ZVt56E?+)^&(Z;~IS=rFFo^h%5B9?ZLZ$D72!=_)0|I9aunf!{_4LN)
zK3ma~rQiS_uC5?B7Z25j9*jO$5XIg*Gurt99}Jpe8`fS^JKU2B2pPXTh+AMK(f%^z
zvS9H}ceib{IZ08T@lz2#7_uraZ4DMbFjg%*D3G)45^z*~aEZ`5s6mJ)5Ry{l&Y`<b
zcez4OVV6`<bq!H7FOxy`RY)Q_RR#xT9a)19|As!`gY>8vp^W{y7hre9>FM!u3dcAR
zcX7u;lxTb0^4xQrq4)<f>Vi2{e6rNXGg~<(zXb=gAVq7sb?2zoHevzv8X}Sg){8LN
zjOk_BfOftqA|C?3Zfl6T?Cd1~?J47ghi)t`!5zFo_}Or559{UNik_Y%Tj=#sxo+zk
zAx9wq(%~q$5a;)_ASq{ezdrU@e&Bml16TC+zP`tx1n$PHzfRecjS|#BRnsSS0{&oE
zm)5neDnChqii8xoUt+Zy3qn2bgcENY<mj1=0|S%HfwE{v<aEXxORSw58<Jp;x-t*q
z{$K{p`aqpy2GsjuQOFf!o!iT4D~1H?bmntgOr|d$j*+#njJ<39y)i>?FIq1YV0MCw
z{Jb%z?KRXK(r@E^?21;a!#%J%YAc~oa3u5njf-)5zg6N%)rh6WpkR{gb;R8Av=Pfn
zvV$8?5_B0?1m{Mpr~0-BL+F)~@2d5nD*2=YWc0{Y7g4SfdvZ*mRuEO5qAhw~v3Q|6
z(&;`Kw@bHvi1hjVBeN-csrQ~gw7s9wUeHD$-yE13O#5dL0AJ`dN;5~Np%q1c{r7iw
z&OgmL;VitQD+YE0$v`^u4RxiK>$I&_(9&tclRDoUyIOmPi-Ms)^aXEu0i~3u0}i7L
z4&)loP$u|x09DgX)egw-5`<seUnox2oHF%os;jJys<S%YJ53hL`I!{UK}|RsxNY})
zOC8dubAL_{ri$EBW!0+Kh1~Aqh;sMZQ&<FHG?dE`+pBb69?Z5R#lp&Pdj<esPJs;e
zLHXWU5YiiqE8QlBJnuTD934>t-mk=VZMHOU{`GFhOF~+)l`wR4pY&H|eSg#k*&1)8
z4t!1vOpq%mpaXVKOWvs;uK>lxa&xyeb^>AzkccQ*vncq=`hdmT77QbdHl+_WY!=-z
zFD0cNL2@4D+qSrrY+|Oh6UMqgRj7LR)X~84(-obhp+u%TK>K<WXm}}pPH({S9Zb8o
zJ+$1k+xVcY6~DaerwUZ57opm-n7v9HRB0#UBwS^APIh@)IkDe|Oo^N?c$ggP{B&YF
zb<-6)>z*AfwoF&6a$GtQ-#r5qyi(<(YLd!Z`&*%#NI?h)7K12SaP9KJjd9sf=C{aN
z=PuSvL295DLp*m=C7TC#Je<&HI=I{p@^yxK@YfPq^vPXDN!Fu%4{~lC|E|WWUXXo8
zH<CgZ?W05LX~8AMe)6ugnmWRgSL0(=?>EY(PEJ+z8SiCN%{l0+^MaBP4e7M#O-_#~
zFp?$13c(halXp1}7!WF=i?-n}>tZ8EPhf2U(a(M6_mgVajFTL$VmUmn)3{v{JgsJ$
z_B5uG$7Q1z+Bvs<dJ)IuY=J5*Qfx$Z2@c~?0s?m-HtUJ8rG}r&$q@b`VP^9*f6c!B
zL-qZ5hu|ua{a#l5rl|OweEqGufV6d}F4|9~xXzxF6gLvrm<SEG+5HivV1V@7>Z1I$
zx_If@EQMCSdsUU!e(b*^*Ypp;?vJDou%yy&zb1qQ)bGq!Xez!62j!wen$v5vtm{CA
z8931Ox9^z{L({L^1LzGxvQx=@sW86jlwHs$LvsT1Z(slWkU(&cw;Jg4#T)(8+lG$D
z|L9#K|Ms^3r|<eDVatDA#XkKv&=`jQE`PJ%3~c6{0dGk;r=zeZqjcqPmlhKIBq)x%
z$%HJM3x&hXDP^^sf``#6O)1k(_CppgGE7JlCS6#5+i8Wo$oRGdP|=Z-Cs(UjaTE1V
zh*VgbW;}GOMOct)B_#F2I>JhRY?7Vpsp7?6qqt3{>xoO$#td+Oi(Y#+1uewH-dN4s
zFnbq;M30_`*>w-gjzVXu=h_Ig7{Yw(4QuoWz&-4a&cm{4@e~sYT@AgR<%Vekc|n1G
z7cb8Is92DZjuixpNIvw%%rp~*vk^y?*}_0B4v1E*E&j=`?~Qd3yB$hhVW6L9_;M0_
zoDV@Cy9?PngMoO%?#azkt~OnZGTVorqNNe&C;`miIZKeE5NY2J$}T9ieix{a-6-uX
zV}t8??~LEi#Xq}dc%Yrj(?L3&KnOov-LY<$W3=CFMW7HR?;ik=?ITOsDQ)oo^0n_m
z{<jI_O-k&ytJAPQt3dc@i}HkS!zyr<(O@E59<$X1M8P_??&svG?5^^u04XsYB3uE@
z3RO9zSPA{jaP1piW2{>HQCD|v$m-NJlVmRT<=Nj!`J7JX<|&Ky)<O6Au8lciRhHx~
zNw<5xMlOLG5u_-|BbXN*XcC+I`<@ITLwKF!`f+oy#)Q)CpnGEF90DC{e|iMF;JB)c
zhS+gKc^;Sn?G#}E|2dqubq_G}Q=&bDO|PhFXeZAI&b`ABPaMEZ7<Rpc+*C~Gl7vNn
zX&WwDQTRMRz5Cu|4PyjZPU1eGid{8W066XbM%Y_dKR&PyS_20CsQqm|qcvMu$E`ev
z;xi|!>q)LxY0IC8)4cyn%IxRb`^`b}?Kt<Rb@#7p?)|;ye#*>rhMiKYr5WRX%p&_P
z*`6R={&DbldnO_Hseg4|UaRBr)sOujtzLg#ykKlDe6G3C*ERRkEb=`IO8<VEdntUs
z<&6Wn|8<)CfIP!%AL+92B1C=ve3}F5jbdD1BLCMVz-jLB^_RWG&R?P`dXp2PEfOu@
zae!R_@@8J+x%Xl9<(&SI{{R98dn+m4Cq7V4gH#B#`pjo58gh9=`ujod%Y*gpB=<|!
zAD~q-&Vl}kY?{qF?x5-cB3-O6zFd;o>0FjQ&0JW>F+JCad1JO$G-4meg0k>!e!HBa
zZ1yNh<fLc63sMqIp62i{#*^*!A>VJ^nr{wdoe>rg&bJp3PHhGsB!@FJcPk4_%2+bd
z$*v>GQC1@?*|s}Sjq%Wif%{fEO&CeNYT(_LL8Axr>Nss+RtI-E?J8ZlKC0l}x#I!?
zBBkyiHP01Sh^{zY9a1N<(<<vg=8%}67hNCvs03%`GdAjB23qp=%+uW&=vJ#|pj<sV
zfh>e7R_>5GNKSW?0V+yIMzk_C%KD7%(#R%nR*DV=%zQ3+9;(fQR=3+e3l@`V&oprx
zF1K|&IICNsoG0nqU>j*R$_&T)PbcBm)5aeVUq2b~sROR7R$!Q?p+feumLDGz3#Iu1
zvp}-_?WFN}(D?GjzkcnnLSE!k$*YW^O1uWNqP!Z`T+JOWy=Er&6eM0wAwt#dq8w;5
z6lxeFM?fc&7S^}Zc-(|>ZPpc2-4QNuZlMh5#p@zg-dQOeHiwk%aB9yjy3s70ic*X?
zqqWsb`^Y};)*5yL=AlL;!El6*R2A5mjnlV89FQSyA*X8vUnbt|5W-_u*dU7%uhoE?
z$bA>=qhiM<(_Zy(mj;T-IXypZqBrPQ54-kv*Xv5Y7n|#TPasz+*q+84XJKXUUO0EH
zUk1Rk_2-7Zx2tjD4YS>dH4GTCe%JDSJ{^kD2}|n<YSra(+eOcIi$(h(&<o^cStY@9
z!jK1kS~GZ;X!nZZF82_!irW-(c9^UwK?@TnD~SGmzV*4z3VE7+6uN#L#{c20|EARa
zR`dRK^?ke3yH9tzPIZq^s+yB7jqsq?%e`>R=y$9hmhS&At-k$?$RBGu%PrE^1yO``
zuMs(Lct8-gYvV6M32)_J*0v8*$L5{9^R4nlzPUPH<?w4%_E8e^%^dtje94=X`LU2T
zAZ7cykb=(g$$4NX|41E#2AlqdkyY<6`cPZEVv)@s`ZnmA<v$huuj2l*Zs$i3lU!57
za(^`1Ud_$_7PP1Y3hrpt>Tn_P3pOB17bLjLp-YIBm%w|-y=IlHQ8!Rc_WYr33+C}S
z#mf!MLakiVws|+Vl4)z|fSqMSesWluP7x$_D;ES5_)jQtABBV8hHk8!{kh1rOmoZQ
zYNa^+=9xomS5Pt%2PC+Rwvy=tK2{4_=8YO0)G#d$(+%We{IEq7V^sTnSQuVc-;Wr@
z^9?tUjY+IIxq@6cvA|0gY;C&5RI#;4+pLtW!*#?<EW9p{Ri$7)o%w|?XgZ=&Koynw
z2exkaRx|)!h1+IN+NRG}V_@K6n|S=}<~fz|duao;&MDgM`$q2A{LWM+Gi|)6bWOUs
zTw^3#WBwJiqg?->$Ib2p*tp$g-+u!wf*SA$t=11=ul9px^22HOt-bShL%clTKM9Zc
z6StZ-1?U<M0u^0cDQiS(Jls5HVio#r28zU42Mn@L{ra`YPhb2;QX>@}Q{UODaacbW
z(Ssy8YY_tRn(X8%?xsh&L<OZI)*cdXws()S?X#M6G0NHNdH~w$b=(zk#-2pt87rXb
zTB~>%1*Pce^rrm(XgjYZ)v;x3-|H#*o;U#l3DBjyLEeG5lSdW_f!x#o31wC-zx(X-
zcSJ|`qF7ak5_9Iv9OD~Dp~}pHxHi?SZbNksjdhu8kZRJ!x=yphhGH<(6_g`;-k~{n
zu?S7k{Zui%eA1HaY<lyw+e3;xNi38m&SqZr$P9Rc1ApQUn-R`lNTssw3Ex|N4Pc!)
zBg})j#A)wFch<=_AS^}$wmcVOli*o}dzvi1_ImnC9|wX(!F#-3TEK)nM+HyENY8dB
z+FGop5ode_Tzh-)Me9i`2IZA)hJ7tB;-yH1)mm3~Y}`s#R9~57CPj&G)wo$`#e?0d
zn=J)l5&!f2hgF$>Xx6kv4jO4T`+5vPo}GuLd~}}<3;mW={tXOt<nOBB=W_hlwYYdH
z6Y5W8f=P-;!dmLy`k^*NV*{60D^2kK|F~7dPCe*ffUhq*wOK-rwEcEidrHaaH$&VI
zKytyCAnIqH*f)hZWC%d!*d?sNm9}DEjRXK8zDS86nhimn@ZL#ydkp3lQ2p2rIGzEd
z{G}F$Qk-}4?}sbTt_ApvsV+J}RXdaCkAjsg-cYkodk3`78*p<^Pb;7Cm(>s;GVp#G
zPgTOx?=~9cowOFr$)bhJ_FhWB)L1^|U13fr3VqDENm5HiyRi>f>d89{GhNsniSEOc
zwoC-~Gt<2sXktO$2;eVy)>dLg^cKkNOx`xG+PR`S>DivUZk$|!{<v&<?0O-uSQC<r
zPLeUnMMwtK>2XHfblvanhrNj=dF+%sRV->rwBrt5#gW}D>^tH21qdg}oP#vMepTv^
zK&1g@#M*f@kzOCWtz(H+&1r}SnJDW?ImfOiFbt8?=rmwAe8DA?qo5Z`6Kodq7V>8D
zJ7IT}wOzy<kB>n*obQF$v*DwO!05Y~2s}pO#lqQ}-2x;6mmxfQWZPy$dEZn^7ml^g
zO5mer&^a1g#oZE6bm!7}Ukx=X2it!3eoH-P;t2gT?7&bQW<Cdn1f~R-!<gX(cA}{K
z2V&Glxo`Fzq(Vb;CD4kguCV;pUlp-<5i^jVkr^znFmI42FeE=^HouZ*zGKedEcAA3
zz1n7oyizTjCi0gZbQa&h9=}~i>DpBmqDj?2lQx$mpcIw$0VPnDDm1_-=>#vDT2Y*U
z8_UV76!61)QpfK#@h~=Ovhwb}9C8~VxtJoLPWDs6(fjqd98V48&+_8qE<#}*MH>LK
zAog9c&wO3T58cM`4s04hTAq82zemeHpEc*Em|V!}+<wJ8@S)^_HjA$Ho&qAqN>XK}
zp6L7bv?9vo0@*oqReO-b5!9h@b&l(-2rke>n}UawR9iBG2rW*AX6qrH0-!*e(=hZ9
zyxOcpI;8p*A{)2BofxRC1&ygCDk{V6vN;^7Lb|J)6I{|xR!_>MEt9Myt{WCa$0FE-
zc(kZ>-aZ@Rgo${AakdkBs2Ma}<gZ!vxI@a_BFk^4^QU!`vFwjh_Mg*_K~H~OEkVx)
zak8Ju1BMAm(p2?xY1c{$Q}J?tA^l%guAt+Bw(5I-@&HZDx3(%sE<2D}f(-$`M1l*H
zMap?&%z~{LJT}}6X;&|gjRFEi2;=#3s1~_&2G-C^#{~su0f=T13QmZ*cb_^g`2W{q
zELe*HnC@PmI<D^s7Oy(-ZB}bTmHPl;Es6lF_9G7g#IyUy7sM_Dh+Wn}zv;UGdgv+7
zR?2Vc$7g%>O~>`QVlMjaIWlNbJ>8WtJ14%8K4t;7PoU~G!A=GEei)I1nL62mf!vWZ
zbh(tJ=u|dxV=?VE<KnbFJ(j9+Vb&+>a4)7Dn@eI74{ZGaav?2Ux7Buc0yRcD@XA%o
zlIzt`l$Cka*jAHEoX!OTapX05juso|s4<gv5WTnoSe?*0)dRmiB$>~$#5!m(qh6&v
z$4>HP7l|qXW1IHSPFr9zYyiA{EepYf`2;A;+u4R#pBfspsU4$+JgNR;$F*T`!HBc0
zDdT;yoE&c0QRfM@f==Tc+#s0!5>!rx&Y!l#(XX&oggow`bqfP;vu9m02X1TyOtx%?
z0(w5wRiE)=!>GU&chAZbws43FUk+sXF-ufw1sHc9r}%dq-lu5&=OaI`gT5oYDpuY3
zCTRVI?Rvq<1^7Ql%1tj`0YkOlFJ?KPO#wGhVGHxz1Mh7#n0SseArC{9mKGIZM0qw<
zalzPIA}by#5>ZPH(&obmD0fMd`kSz`t(}sZnq_Xod7t>Fz<-<;I7or)`+7q%EgG^_
zzki%=WG|THwL4F^u{{+zi|`cg@sKD*mXIl`^WzOptw+z7gGXf&z^h%{%1YUS8FatP
zn(;iF`|dVN)u65*OAv&jM7Y-(L7Xo8U_eayZUg(0A*{1ll~N5|(%Z6%*5N4+7)T4`
zC<T`op1ZaHXU{1Ci%^5wfj|ekPTRGads27o#av<^frTVzUS8O%DL)=p+tW0<o1h&7
z@2UlR_iWB0QO8F=ZPjS#;VtgiJWYZ_;2n0&npSPc_}f}W(ya9qww$&<l!1R+%UJqZ
z>h#}KEsVQ8(|13*Z}M(UKKmh_yZ-guDfn7<ee^&-ac-b$IsIyB`LeQnsS-P=5?|{Y
zWvpBq&TZib(ff0`F+e@w0Gaf;+<?yJrDOqO@>9hET2&TT?aQ<GIRXdgyT{-Z-u>MB
zw8-JT(RyIVw?^x4SK%*O9RyC_qRFWa^a2<YJd?jznqQXyB7X{&=JnT}=cUMaVpsu{
z>yu#mQm;@i^$HX(4|{{kO66>X<DqoXd@BOCxAO7hnG4PP8o(|~oJqBne&nqla(V7Y
z{ZZTxF`CaJ0)yD(Eb?PZrmC&x_zKjA&ds2t@Cb1h<3hjCY!ogj{><g`BD)q|h8yT%
zm&(w{@Vaqvr#I0Jr*xb4k%ksaZlhm?hiu>ImN?Go1+^L(MO^}$8whWY?*1U1{KXj<
zD7l#4ZgGQ!#Pztw6fU5<;Xs|H4T?3_0paNk%#V=2bw2moYs0xwkk$!NAI6Yw3e5K}
zteo7U9ggQmaNhTW-3N}OE+95R6XqD}gBxG3p^9eo2~qN)GMCp{B4sJLph3n0;VyUg
zz*cGG$~VJCca<z)BOP)UfSh0To-}=$c?1#MwQy*DI|IW_`y!ZrV+#Cq;szx4`+*tY
zY0m>Q#R=5Sx6mHQP-OIq3^EJgygz#?Fg4}h*r+0D1`1aU?%*9><>k>~pjHL{3?(US
z_ygke{%P8Iqs+r*^5kNlg-<s2tIP8dx(4>!KIDlG#<jS-dfkV2u*MYUQ~G#kpMBS<
zMd*{r+m{P~DNe^(QQK_hcCom)rn~f4N4htR!a?`)yFab>7>Ca?Tq|rT3wMp+AP=mi
z$^aY15U2#9$DmZ+f#OdzB`jjEH~1`Tr7~o?fJSTFoO)O{{e5v7AL%kVbQk58v<a0^
z^!AQ$xBaCYX-bdv<po<!&=x8o^fUG^x>QOI0d5abmB9C+T6YEfb}lnRFpdN%Hqstk
zWy@r9)BV~$EK&p12JGll?WqwDD0fbYW*@DW+gO>h<t-gxbQ<QvV=bu|o!zjtD%QGL
zNhh?~WqGeKYGdBIP$s6wpbn=F=|QfK<vE0(=QEtnM0+LeAB@1>qrRAEcD02Dni}u6
zBQBjFC255#$(k522&^5iv2X_!Tyhg0RBa^M2Du&*6+c~PwYq^6sQaV4;H6fyo*YpA
z`eZ4P|NPHC{Q8G={X<yrxAmQSJZshMXRQjKnmZ}5I(jM=!>@MopW{GP@i{1hkILty
z=uP*BzKFO$<PBdL2p{z+_bxtsS!Xj~oUH-9@LUY(RSG-Z+sPg<H&B;~sKP+;$|&fY
zM&O6X@}Khf7|131PT26&3?d*#conX&Bnyi9-U)k&7S7+jQgBE-t?aOYkl?`zr>C&+
zV;uFdG(Ry>9OzR)O!&R9@V2lY$+at(i}Yw+^3D~;u4trix?Q4+B+=~D?$2&a8w`Gu
z6x7Hjrv+x-a=+U`+lkskyYgz6*L}_|yz7yqb8{c^4QAwr-GeBz`_c_3d2qvofv7fy
zSb>7Rnb$482!a+&3^<MC;2l{zs(R5|c8!YNe!%KzxZnDb?ZY}t<a@fnFX`r(HPa$$
zGBCiq(`6BE17&~M_cLi~%VW2s?naX$eY8faqiH`5^s!qnGpUZvT%$6Uggcd76zEx;
zooKtrgO-gea_`?C5WE2A=ahxWgiYy)fqXx}(1~R@WGY)_eHU-ZA%FBh*rD62)<k>G
zxkc_cf!wBQPbGB5*ATTA+vnZIZVsb+kQSC**QdBid%_hC?&w>+iIAB+byMK@#1*1a
zPZZrncZG?K!^4U>MvcME>yGXNod9cC^n4@%CygR(zN;_BdVsU~W&lex1n$Lkgydqr
zxvrpv(aj!5+|>IXuj<GqPW^%?XRL|hXaP=s?>0HB#t{TP7*t0A67#T!pvr|CZU7Si
z`RE*9Xh;(~3{z(S=ep2)GL*fHG|n@$rP~6e1qK#ZK@TJsGzTzs-mv6S#KZKy6|ah(
z!Q7C@<NY2K@2itPn#gXl$XnYSA$b<Roi@H_FT7};$|CT<CKUK*UzCDS$Xrdsrve=3
zA{=hkv@hHj>wS^N%&pN}e(hlzpCNoJZ5(CEdN{=WjTQXfz<1VxjQ^F|qfs|#pR~Uk
z^+DbLaY%S;GQ$V`-P!ux4swL3^s+Te=d{JNL?>hZ9B-Lhz1P6;`opO2>vNGGpKQKu
zuqnmtp>as=sI+{k7>@?LS|GH^IoW*)YoH-=>(lax$QuB0^TZ6wVz}cA@IJmo)9>K?
zr(<#do9zrt4RpN4aB~l6b#ik_Dg=*mmvF5iU0S=@#khF2NTQ?P>)Pd*^y3sWDYFI<
zFh9`vLMzi5fG!ZZBRCBAz?^2m(?oL&JmM^}Hw42p76-j>UT<JeJ)FHDjR`ruWdKP5
zb@y1z*-`O*Ttk)RFj=-+-kahvb3@Q!q^5V=w!$__v?o0+L6joiI+?f<Y-vk_Jtm=@
zb!(&Yx})h}XdyT}6RasT7bbQeab;QG?}=31t+lm<zMsvT;>vA<c-cqW3gI8yhn(ES
zXiOH5#nkh<(>e)>Lq@n8@K`WRJsZKnNeak=L@blj9hf33moSPx>A;o=q`_z^o$XZ`
zp#gtg`n%l$0!qPxK#8~WiFBda>PrgIE0$^q*r?}|&O|Ji3*7Mvrq=}qw4-?mQg6=F
ziW^N|AmDeJqyp*bH6}%_bG@bfwq08dZNV<>qhQi&;rH~QQ6Hcry;j}tEr_P4^@<fQ
zS+pV6;KPF8HBH%_BA1+sPEL@+4AgEFqzraio_xDGotJzmoPe#k8p@Rh4gc;i0<zCz
zn?{bdWvqt$x@u!<#y1+E#>G@H`gIn2YwCum&Ps-5_o9cijS2g_L&2p7pG32-E$*=x
z&S1)^ANAo1P9~;<EO}k*tspOq2aBm@u$p*WCot}~mD?QYGp)b%XiqLkTAQgg2u$xr
z2r;a2g-yo*7N=m=T649BG^yjh8dh9FJwr~hb0Yoe*!QhWu(Gm`Z+|Hi{yqfYen`d;
z?Xh=c)Y?I=%db9GgZYCY&{(D8bBg(%h6H0w?#+krUJ6j~7d~-&FroC`75{IWH?+S3
z;arB#)VGff`n|0MTL(mCJxvP$Mp4U`J@LoKe*0=ZXR7dGz>}cP|La%tL%D!_J<9!9
zE{r-aB!r47Nr+8J4hCCoKQ80Fx`?zeXyknB&VoBWssjcmZVW@zmC|$bDw!1g$|@WN
zIo6Qn!fa)TCnp{m?#rdSPffO&YkWysve5IHZNb>}^w{1P$Bf+tH_zS~fhttUV%fwy
zbG@otpb%|&Yg$sUS8h2vbr$%&kg;O1+epr4DG7~|JK4(IZ|yAR>VSm7t?YXSD%eMg
zzBQ77Y(>>~xV*_A_WR@u;pSW(1RxMRR(X-`yv=r%Zf^ro=a8PaCGEp*eH|}^P2T&;
z2~YR>ENAP&Y3O+EJP?JyD#V$2K)S^=dvdq-xx-HPCu`wym-u{-;dCT76CK!oGcsHu
zU_L`nx0Px%*yNtTxJ3zDn@*|#zaTB?vkwa_S#y4d@FOsvsmpLLFxo0@({z8ScDd3r
zC+0#um~P%K8p&-)yOl^GW!GW7f{5OmTjN27c{swI`GTbfNx1^^0M|jz?|C$iE^?=y
zShiU%BA1uP%DO@YA+^&kXXc8zjRHlCvy*Vcncj<wz!U_V>JU@AHY#?seBR%uGbCot
z#bbpsgRs2Ep6r9O?Ft72J^;(~L_ina>dJcLtcS(hNv?8_j@kgXZZ8T=6(V6=!cgF}
zF-I|pR1W6q!928LZb0G*nEoq#qwp{<PPZX<zOj(*o-MI?!-r<I#z*R?YC^%7PTr5)
z=7iANR?w^i-=)w+VD(Mtm!jNnAk`k|g8#By04n-tPw~5Qf!i|b5S!FJZ-A6_hx=+3
z4owch&FN3&!pC!wUp@H`<pOvje_t*jA#?5jt8zim`7|c$RkqR*OmYUSX{ZDxyAui-
z<$lHcRpBwf+D6MYyDc~iP{Q8LrH&$h<9qAU93=r{2Q1p!5Iab<r6=t?VMB08ajehD
z?T&>Vh)|~+Fv*}t;7{pnJTT2Fj~7PLF1L;t>Dw6{iD}p4Kv0n&GK{6Yv86lya66UY
z3Lus;vN;{W$bsU8I0P8SRGVB_fV*KhZI8wp*azh8+^>L4Ye3B~+UvVkw+Vk~gyK}p
zkE;xkW(8ALSYj=y5m`ReI&A5&M<*HA!3f#K>;nW8P{uBI)+wC<dQ2Qyb_eE%u52i=
zY%evs>b6f)k6$Q^khLC=F!F)(5r)t9zFEQzNG!_XxE$l6aNM1PI(Oh+70~GMn6s}#
zunbBmy2|s|(YURFOCAUXSKhJ6T%TevqqJ75BS4B{E*PZ8IwhzDQlfWRFy!3a{Rs@p
zQ@hi4tMVZ=z7Bz=1@8;HbUxaVVMgS=U8WtgJb;T4t0Cj&@j#Aa$W8h|6=%-YRPpZT
z3;N-$M8A(><77UDky}vPq46YksVU8gCD(*PFNG#PQ&+WpP+sg`4-jz=qh^E=tL;P8
z@IlfJir%XDxw&x$2W0e{@0H%-LSCv>XRH%HT(J5iVk)S^4Vq|sRij#U>-S|AJ`~Ab
zJOYGC*?fOq>=Jbj1ByRMfjr&8f3-?IU<6ML5$;-w$e>@g>u|E&9g6uzA4z=fnruzu
zk3W?Qe0;TU4@h_XR1AQW;C}<~`cFF#O!AzOSJnr}%is?a>O*Aw{o-uB_0+HX8yty$
z>1gnN)mc6&)Qgz&^h>Jd+U?)-x_`plLTLe*+ize0I@w45jJbXHM?Mv`dHC*NcvIBA
zB5rK}_0A4-noD2d(h>O<L<vI5m!cNDSx-pNOHqpo=Oghk$NLc<3<9&u*RLBeax3pJ
z%m?PM`r-$Fh2?!^<^8FnOuma~k&ljYe@k8PLpzCp);CWr;{AnorcFXz4G~a*N7(P%
znl;SZf><xl?qaUk;jw{Db2t#>F*xilM-_~O>`=>)!f@U?TCW~6l7yj7{_q=*UdJmY
z*Y9Wm_{9h394XWH35|+7FDD_JAmp62IPT!=C7ha@;SdN22_;LTcLyb9Fq^i|9BvI<
zIf%J?%Ea4P8ZCqFh7o8^t6fa+VN5~JX$p3e$EjKe=`@b`iAa44)@Q^V<f=Y08JLBs
z%^pLG`(kn4NB0OVb~(N|7Sz${S#iZGEOb)XTCaD<4C24Kb^(JJIj$C#*VmiN*5g6k
zSuOYIu=6p-8W^+L;0r~pG{Rdn6C|?w{cvA&s~*BDY6{@jAbd7R>be3Q74x$(;Wcvm
zm5bp=ae}PE&2kRiY)R{C1z`LBep>@YQo+ALF#zHig7^&dY9EMOt{IYPj_$Lm6CQ+B
zSpLG=W$pgNsvBMquG)1a)XfGI5$KkV^+2MRh3OS$jvo-a!*1N&fo#Ri9ssu)SvaG;
z-_46ut7G75`uE!R<{N-?61$4|5({?Vndd##ZWqO(ecYVGRPf7mNpmWq9+f$B>SMLq
z9Y&~aq5rj61GlAu0GDXzyJERLTn~`)E@a*WX!$BEr@GEnYYRtFpMs>j&3-6&h)NTD
zge$KjWO1xQkF+ca(o@mq?CiPRfg7yZ*Oq2T!{R}YM&&hab2oQ3<%kzES^Hfy#Fn$y
zbM&xG+Nr+q2ztxRrSGYS@#>KF7SQ>J6ny1z0IAm2*5q=(0~(wjA0Njw(NbN!p#Ih^
z{@9wmU;N=7e=eba_8<JMH4C9*us{x<)N(^*cdk1z?k*lBf?jGW(cOYyE|AY|&NqDL
zXNJXh5>iM+^G?y~3>^*&^s3MkyDIbxvd_W{D|tD&zK@)EIkO>C43;0jn)@fn;9LXE
zFlRAn*M8t}`vb2upqGNh(Mu!#a)~y;;d;lTzk0loU?}n&3SWDMx|Q4W(MX?ZpCS)A
zKSJ?+v$+0vm6Y+h{&BBFoKG3pt5*!@#CI>lhoQ`a5MB0*4D9>&<O7xdySD`SH}A>c
z-V)^ByeEHqOAzia@5#R~t98g#2Z9AxJGBxQGrqnYPm08{4n!7>4`1iH<LYv?++wpN
zyO|HA4G2jM`3zz#Ex6f6RSW6?o6Z26xy2A1MVkrLF^VZ!PHsq9u(<mhlUEN?6J8|J
zh|@I@@m}wPRT`399aHNf3f~KlYWdhEs76mUs$ZcZNca$e!YoX#Rk1w?UU+oWE6VM5
zx7iXaQhm6OXtTTR9k-Kjcs6aVT|;S?>_T9-NsY9@84KaQ(JyGWwTk5d@yYNxiaU&V
z=jqYkhbJ){L#hVIBIIdGVX3y<AU&22U@_>#QtC6CfYa8R2uy-*(`GF^1c$X<y}}w`
z$#%G9c`@xD$#?}pK8v&lVdBC|&;?#kCm}jUA7(Wphj5ACWad8QyU)D8FZx~3m3{A4
z+Z|UY9c){2+^<F;CbocY<+o@g!np7!k_7S^Ou(Et2Ait2(;y6J9=2e^W{x=5m#1q%
z$Qn5V_dPoff}6-f-mI_e+&MyUJ;Ug;In?7r?2qU*W|7A%mj-;Jh^S^y+T3{JWCn0#
zX?>5{!AMv*#n?&2!3j8p8pNvH4(P06SRZqf;j_uv<C$tnhJ=`rCuG#a<2JOtu&>2%
zGbdcOcUyZ$_*uNZ#&`eGqzM`WQH^XPc7M4%dIuN87<!ZF4&*ZHWedi!_y}jBJ%Qxo
z9gnU(tL~HwFWRM=GB<Cc&(P9eD1)^#LU`Bvl7+|OG{or@;Voq$uswV{Z@fi?8QT;z
zE5Ivp1!UgMgtFAv^qN~F4SP$SgqCLJ(Tr3Ckfu<x`hwDd599iZ{}=PG8{e|mWbQ4m
zV)GAZC!v1-+z@^P#eEL&@ZSvZKFaAo|M#zZ$?B=U+C9D%+v?4%voo=UDUImn>dbe$
zUt8G#YWS->q;I+g=;M{pm#fNGpYCI+|1xL;3-S2~b#?RJS^Fq?H7K^t4MZ5-6#F`8
zZ*{TE&g|8jiWFoUkmF5#`HSxxRrZ|DoqK2mUOp(gF<6Ze5#+-ljOzmkD^VN=beXS0
zn_=hi%QyMV;h2G<h<er7c+;1?<!{*>Z&OO}FFi8_pF_%L*YVxC@jj&dY2QJ<zJZ?`
z&!=SX-@Si8O#jsWY5(o}2P%TM45{ZE_}{*N$bWJ;`FQ{Swn;rTfghIYULt`M4Dh77
zB9-<L$pA+R6R?At=CnI3n8m@6ad*h_v*@_e@-*JUbj;jUwx!#)DGoD|FEiW?;_<$E
z%*!jWpwOh-t@PkriQs7_QpUHvO`oU3S`>Ad5o}lW;pkeN^l)zZ_(-<P<D~#!pA}sg
zka08+lX%d#0%h6}F99$`2z7tnP||9>xknZn*yRlzAX|PQfM&68!C+E_9-bIHr7es{
zY>>FR0)TM6M9AB1rE9x1TXdTYfCcDwo`{|#ZUl-JwsICsNXgz91rMD+AXY7IbYqR!
z(xuoMdL*ilF<{<$lC5^D8<G@2Ec7AJDY8vtPu1Z<acE%&<N{)TFv?*@S4&llkJ$sa
zw}27**rd_`kf=V)#@;4%ABl#ay}MuP0fVo3>c+r=aV5FB5LYr&XN3y~*Yu>PV*k8$
zdIb2{l|qATe}3}EoEwuu%pgbua$wW&SHH_W{5I?tq<=R>o*n^rh7A(9HNMeIpP3B%
z#}gm;Y|?`g7JJF5WxxOjJW&S|4s?PaREomVKKnp{i^1%{kkWpo;0s6;08!ZdA#M0f
zGoS#522he`f~F0l_~Epc^)=Oc^1Nvwa_Qcm5&%;cnFjpeW=mmgi)KvW!mQ_1munAb
zq~!uR!HqK0mvW!-EloATrmyitl6;ykQYI=E!4cat$DN0#zG(n@YXRo!4RNR12YnQk
zS_Lo`mVy9o*B)4Knsqh=(FTmVD#5>4WMj#$3T{gs%fpc+(y89ke%jG(=A5+P6eNq7
zJzK4t-PZi3+)~_CxwAlITlyJ|9IDk-7*_JAooYVw&RAlY!APTJGCxU-pyJs^j*DII
z&6ZoCeczl=&JYgTeZb+D73>tiV242#Nm$P!XLqk&YCklE+Zx=4>ys^t%GPx5mvrfK
z^JEk)r*Rf@B801~A38Sjue+n#{;B_D#UbD_-da#UnN|L_y)lpHQK_Xrw>L(oH3;21
zHa<!xO9oOuFXz=SDkey`6@S%B`c9?z0fPVNCjp~(f-&<1Lg*)&_j`$7`4$IwcHcEx
zPif0jehm~&7)i0u%>hE#*-_(O>T3uB5ua{s91{D`r}`QcEV4qrQRqMRil>MAb7;n)
z-+sbV1O`6<hTRlIalTQ7e~?DnZ$W78$EN%8t{}gYV*jMXK0H86aCGP7wC2Pe4F~}T
zt26f~jTC0A&jxLr5J2n#P^(&Cp4NNiB&`5Za9r~<MzN-04pGq4*khT4{*T`_Mb6jb
zn&_|F2G&`u)K(3Dw_qZpMC5|FBg`=bN><3<!kb7WAmv_rrA1mt8eQj)OTf_V(lq<j
zIq9-^O8Ik;ti@UMufuT<=l;>st6biX#R@px0>X1XDRs87QK)`HY}27(k5}F>t(CPG
z`k-D^dVFrjbtCZkj`1B|Xf};3)C|zU<HrFFuK{olAM&1rc$*bkEMcg-H)$)xHrh;V
ziIcn%tJ*JD56INm7rp9iMuDDv=(PI{4Z=zlhPCHDCb0!NxJqneG<N0ON<bdB4=bNW
z)j&@MkPo;!i}$#1z|?{>ow{Hwpd7(fNXpMCrC{tqAP7mWfP4DgQvKbS1;j&ZuD`N?
zo{Ox_jaep<Vo<1WZ1;ou#wrJu9fax4WCK}xDFl&1_^e>AUel2^WUB*X`~Ch|z4;B^
z!t@-}gdyzj!}Olm$)^<zIb6@;Pt7TYx!bGT<;XNvSU%9ERZITh?;QKS*`7D;3MWF0
zb<21FyT?KrfB@l8ru9GoQwo`Kp%vohf}tU$Fc^$v4G4=`Mj>yz+sWl}b87liq=baN
z5MiODTLTnEa@<Le6ud4YV3dXMAqxpo1>;_HjT_Eo*#ixmTeEGh5TghHE{INI`sK0K
z0fZstu~yepFR8h?lVlaMCFj(TUM1Ylg_EBXk5A4>tvFl+sK4{1TQqd>xBga67VJPU
z(P9~T(>hznCRB+5z|1yx5q4pDY3Lv*=g&M<TGc)m4<3-{eH;W%8}TkY9z2d7W#ig1
ztMlMu2T+FugMC~mXXT2wqq?(H+`7oZ8SB6obM1i=w#(T+HmBZJ*l+2oKOH#!y2|#i
zRrXuQ9_r_;(O^D&EmeZjMd+`!%`Y2L)ZxEgWdRp``p^)v-1{16Q33+nfA-4ec((<<
zm*uZ04J8-f;=gl{NaYHY>jer2i!})Hzk5aRfR1me+)s7--`4%#y`qm7gglLke=Bsq
z*14*+^KSg;`o?+0O|X>6styo_i@D54#A7lkGxVm9o7gNBX~WIv@sV##>q>8D?t)Im
zY%4oZTVFRxy4WgB4g-X^&-Y0<+1YUqdf+7|+<hf5kN^}klFC1fE;~BB^1yT)T{h79
zUKKEUS8!!*X+mozg(}s$ET5MJkGp;-<+aEO*FG}EBW>I%2~fBxvR8q_0T~cPd!mX~
z=Y}>EJrYeRiay3?3<k`8bMLREY1g7qo$z9LZ1=7Vxm%#Y^*o<}aM}g@TnG>aqAdfi
z!|m1cAq@&8ZrZ|DQZzXdehvxVhP~^!*{Q*^cA_?U>qzLS?wx3{$6fg*(U&w}n-Y-X
z7a2<A41wNBoo^aYR)0j7u7IU|{r?bMdLSUHTB4F7W7aDSCm(Zn5pLBC8f1m}2y!s)
z%2W6bISde8dgQ-GB0Qr@!}q|MH_`YTWlKg?0dAxB_;1L^w%+rx^U#Hy?5Ad&{Db3$
z%K7v{CJ^!F3@CK5+Q+j9bWD%gjo=2<wB9c>o8tkEezn7l-7XN#yKeUjpN0bzrS7zw
zbTC{7@;{D6!=4KKw5U&-u*$V_Z|n`)w<>-Jb^sG#sCahTQ_w&z<~%1rp$%)WVKy2M
z;*kcNyWF0@b*wKMLV;YWRmGr}W(ka-?oO*2t}PD^!GLmaeO~oCh0PV&$0!z<%-!h#
zqP)bB4ehDm3lkzT@G_O%9>5ba-$gv-#h_K9lNF4u>vA>cv<uZp3iNdJ)J-vPYS+}j
znDar1?E2zB?8guR+7VMx;|)w9<QCv9X8_5ORNF%t=AIc-d`Jg;ADG4=k#uof`T;QY
zOihTy&BZ+K!{eW7Sz1#xM*CkzApLC>W`(CCXE=V=x;7vSva3{LWCwcyhLS~4f^_ny
z?jU%Qsr1LOfVmRCIG*9lMfA+DfD{cewT~Q}u`VS<NR(mD8t>yOmKqhoS##YPI{9aY
z^#MiwN?&{qUQJ^~sP@X#Uy^S2Gbv+~X0rTPzJ)^7ZHHiOaO?bP-U<;DP&7Qld%q!!
zAcF4$8TSUHg|s9Xwm$!TAGnG<^>vehC0#IbHUKWhd!`emK-`X0LfQUQk*QBghtInZ
zBH~Vp2?fhjdTJN~0MTvlI9AY@DirfN@E-%)kz){hh3Y1|f8lZ={NoNcYzgl{t<cVE
z0s^z;_{&@H$3ZOwxqagueF)4y_2oa{j#ex!XB#;$%zop9Y~YfXx7t?AA@+e*4z=!L
z_YmiFlgslS62j^1;L_*Yz1v;gc!)S|oMV(O9F$tNwYcV6J_3g3@#vo%Zo5K(PH!qp
zU&US8w}$~41%7sV@&bVx-Xkyfg0#&gX)~px?*Z$6Z{VGK+{F)75*8>gRy^zbIs`r}
z5=c&4JUZy~%Rp2c{3>t}jP-NKtGohXjJ3pav$>s`#X7nhakbY#4Y@vw^giIP`VrgI
zn%HA^Et9T2|KMCE)7V^ZPnN`hym;G@2zub>);fVha8U~|gMjx!Zq&BZG{SXggF8FE
zt+M-j?r(S$GA~xw-GMrC05Mbr+`FFH+lsY;99<tE{_-Yw74hl);WbFdlfQ4k8}d;*
z$|D$c)U8RjHr<E;4w3K85yE}G>x+Qc2}<){2&Uik(v<Wn;mtk!8VNAlR6mg0OMdyZ
z_B?ZXUJ3F?qCkSRdU^f9BAB;(j`-=L@^$DB&z*u_9dKVED=-q!n=G!D9x#7H{JzN^
z!b$XiKP~{O5Xy(8)b_{gJto|Q=B#s4yG+b=1hZ9&jb_TJ%-U%*276jZG@cb(Y74r1
zvKMIrx)?%evB&*B4B|(9zA5{ZlwnJ;HNXOg3$fJq^S}jH$U3hcYKm#<zV(TfH7nr~
zF~>-`b|PP<^>Mw#mO2XoCup9z>3*;6*|3;TXN@~_yDsnKoYG3>xRNMBj`<pdb0RiN
zF3fi*z^j1{PUI7I*H^N;7FI=lCUndK{#0_s{Tb|E&e?6x{sJe%8?N7Qb9T;n6oI~I
zL1S9QVH&Pbhjg?RmkJ6MS?ohDdgz{Y-2hfL&c}?s_D{!&xf5a;_dGl6G&p0NjSuDQ
zuZbLLd>^j2;vWm_y0A>wfadrV*2Mq(=ec<A{r|u14h;VE^K{$K-ND^=bFh~}?r|2l
zOOh81cZDo~U+fN4_I3FLKhKL<_bJGQ#oe<jcNIbR!=ETX9FH)=h52u-q&AH!a-lq+
z8$QTy%P8nqqR;*`0_OD|@E-56N8hjd(6gsL`q1BEQlYh;!(#fiwf<higTwCw?eYCf
z{@=V}fR6t9CI6dO{88g0wAJ!I#iM@52W5(0*jfs|;nyN|qj#|RKFGSf@8U&rwN~>m
zhdWTX?#2<xwPSM6q*O17Xin>83U<1x<P1LZrA_2sPG%jl>Hbc)5_2*WLzTpITkE<;
z54@C?+ptfz4kKejquYq@i}_seBZSVn!W@`DK)`Zd&p57ZcE-j|*8TNC>ZQE~4>Yef
z7I|FA{eHd!VCi_QvGWKp?lpfjsgsR_H*(qY4M6)Z%RJO>pb)yjY=?4Y3+?2ZqtYAO
zu_x4U(MJZNK1hDAawuEyNH&u|G*SKTDcJ#lJavdH4KQw^7*p=C9Z6Y&v>~z9!ynxg
ztt$?~N@IXLj5qgUiTguv^kF6qL3!5BpJ`oQdGz(hD&hm$Yf$6E0F(z2NeD&0@x5Uk
zde2h=NR<^x%dWr-ywkOO<4DDiAPesmk^^{<m)8?~@!}H^qQ5$B02BEhCd$3}H-*8J
zAu?z%pM6$V<Bj9Dc(Z3MAHXLi1V0&g&t8i<5&(nfV_f6F7vPSE<{>_Rud{H!eiD9*
z20`8%)w^5KgYP`KCEM%$4jbI%c%$%x6X9C8-zMhye2RhU3<^feSz|(blJlH&*ECa+
zK2z_?bzI>Fg2}BY>(IrdvY~>$CG!}fhnO!fC>cVyMUF8Cvfq-jxz8B|A_33yfwaf)
z5*uT4Qt=oR0QB7oT1O?82+up+lt_?ZRS3&^e6m7kiND5F5^_)|wrtmdOiPe&S#BrY
zVRdfU>NF1ziCfA?a(O#w;5Ru3^G3M$$Z&|oJ=g_1NZz=$*ma#dT`-&6Y6>o!#~9dC
zbmQuYFk#zGZ<-XFp%L2X_If#MgdwO6KbHQ1lcI8DkSWUcY87?O*+Nle!ePHdVTc{C
z75%o@KMD^*#$2CH;T@a)RNNn9AWOdYQr^eAZ#tj<W?6q*-T$_@vp+a6ev^GlQ<pDg
z{?{h*TLFGm5Rl&hYh|2akP7+NQ7>|DfBtBBDCRKmZwtB1Q0fWtu{#j`@G#<oQ42*w
z?R*nioKkEc&%auUa291~^zDfs9cqh$JR}Kh$dHFx5q<P?*u&S`{lobg!paW=nkUm~
zhM?T%pfv|dN&Y01z)1M{7;y0ac8Sj`gx1ji%_Tms5P`G7e|?GnW~X=rJBx^8pqkQr
ztdS-q#f#f<9POeV0kOWJM;hl!QbCanig|aTC%K%}(RIHgwmAzh<(z4_t9oru`L!8+
zvVio4++m6(hGuuAp#q4H(7b%7#m=3$7j~s<GP*&smHG^!wwDY<J(0Y9zv%D{*`%LN
zmxu{FXbZ(UpJUc;59ny@p+2$(KO`6fj;!c3pt1$2t+&0r-`RU4klc$<CMQZ{rxcZN
z{Ho^jE|EQ~GqD>QZ{f_bJG;|CwmP<w8!OkVbUr+&)6wlc%Uw@RSYg<MkuwT8tb|JA
z{KPunyQpsurzE6ekJ}HPN>T|vE{$Oc)0Gqw^hy|;_GCzMI<~~RmCzpxX>p7J+d7Qp
zx1Az|dkW5Aa9{nJ91V#|PjD14o}LNtgWQtcm2UY1aOfj}+=4eFe=hb&$ln{ee)|ZW
zD{gv@Dqq2tn`jON38FaPdgVU@`=0oK)f=5EHI5WzR7hzS2FNZiB`z#CRDHf1#$r_~
z2OsYG?)jO7fCh;2ZsD*}r|=hk`}~%t5YRtS865oolaI^aJ}t<9x+4513j&bK|F$Aj
zFURLEkaBPQ0l9$aTg!wpK!jTZIBLtxkEc+bPlsJ)Fo%Tq&W;kemA1b1gX-hXrW86b
z?VV#Djg@G5W+Yv3H}0k(7V2(=s;5mEpI1n5S?zB%lpk<1q^xnYCTBq{)VS&a@H<r^
zpJ&OuZ73205+JT%ad#a|sw+%JH*i%54@|y-8w*vfGDlir%I^LOYqa%!8T!7kT%5oz
zvFqkzT>x!60wXXq^i8faYNeb52wN!&G}iQLg-Yq=jK&d*ZoJf6#2vz}6D9zFj#<xc
zd0~m{NJQ;HxaG#pAI8-j9Ke<_R(^FNWZhHE0}!(yUl9WFFhjO_hie>}+Fl==BuU79
zD*H9R8k6L-znY|Q*<<}N8Wu<O&-{V+?iNeaUyd?xviVnsnZIpyboT6WrQh-ngjL)_
zo;?|Ilw*^8T&=b9Atw)D25v7!u0d!nLqgd7k2gB<f7RiFg#fT}f9P;Mr)VD?F7aE3
zi!)Zr54quQ5#XQVn{S7nbWmm(6MsGQ%<vdM^uC>We!j%#6(ZlTh5!2|-t!WlR|tcP
zUpMjo?Ik{sbqMs)n&%}(IcZR7`2O318XI&VkaSOqsS1VxgE2Q<AB7dod8b_zTPAn6
zZe9b5P%xlT9Zr$L;)Gzvwa}>*MQIqZx~?S1ib<o`-Yz!r;m|vb$9C&%^m49UNnv6h
z2>Jr~JzG)Wv1LH-d)h6722`U8UT4TKuVGiXcDtQyQ;#Ztb!NU>j!iW5LpY#T`x+*b
zhLa?vVu}%7+t8Z)H|?JvZqz}&g7lyybC%o<Y7n>6CMKKdj^)X6=kf?gLgaVLREd%E
zP(2g;;R1Ph!NDUZfI3&nj1Ib@@e7U-S~4dcLQ?jYAI@<@@#ap=p;aB&x2Wftm_(c<
zK*=-`t@6Me2Oh`o0PbJzA@uxubcxE+BJYobLr~-_-|?Y;``EJNv;FY+1ch>%$w88y
z@k@Vj>N($Ymx{Bc7%z+dtJ)G-2q*?w18k+v$mDO4hZGswluJR?S^_Nevug{0`pB>?
z!Zuf?rWn}p%HlQ*nH^jfF;LaEY*tf9=;p#r4vhuQ0w5$1GDi{8^<eT2x8)mri8dfH
z6L84hnj^z?R0J6g0eg5IY}~b6odGcPrsjC&pN`OgUMYDoy=9_;=-note5tp`%Iol_
zON4q-gf7ef%_Tms5c$8l#Ai+P`$V-L8q?inX|8ygQ|6weL|^QP8O!tp0tWRYzOH=>
z1VV`2-V{NuF54-V_C?Tu>SILBwI+wl>gL~bM~NJ;`&-kQT#xs(UpIW~EAF(sga8DJ
z7IENAnZO)Eg+3zIbyv9IDw1$(IoD#l<7PaZm_?t2kL(y@k5x3YePXJ|V!KaU{MHk^
zQTPDr;5ojpmSDb@E=iojlyX7@Zk>2~+}!-rawgc-;)Dj=J_ykiyZU=Mg2v(OO90Xr
zBM?0SNg89W#(=us*>2BFBy8*Cu#zQ@3e^ZnZv543$kFZa;n4V#B40LmU6Z|4?3QP9
zV*{ddcGwc5cLKlb9}QIXt8#s3s}=-W$!>~J-q83+he`sue^`h%AxtEGQ=$HVHZ|dV
z|LQYaB>VitDC8ATQZ0qVDTl*o0i;M@t4%8h*^1|J44oOb=rD2m62rBYftxu0F$?9}
zbCF*>`R9J(laT{`XDS*1YTQ8XvUy$7M5WbsdWwx^Id}q#nFplZ6Zb1cn5BM^K<T*V
zfO4~mfUQWZ^(%+&2?)8aaSUtNMCtJ0+&4qn4k`$}7j9~RWA;IdeVz;fw*mIxxtk8?
z>U2J9o+pox+XZ6<YIhcNt&s{Dhwd^BAzN#L!u|FF`Lh1<3ghhbs{86*tkK{B3$+0Q
zA+Jjju($h5>DuY?hPm&{D5f1wM?I$+W`8j^;<AaFQ+G8keFXdu10cE+AWS!*hT-}E
zrXw}Ce48^D+@Oj`X{1jreSa|^kI&h#b=2=}-F8}mCL~=CQFl}EJ!Dc;ryY!^W$koF
z%@wg<r#>6RG%<Ed?rgK3cQ`yaWK~he<6pv1UM&j96Z(g~R@I}<LjT!*{$TV2E9fWv
z$4?UTzpeI=U<Rvw^*M^9AYS(88-bzHuZ?Bs(jFY<OCID!2@z-I!kA<Ai4=uW{W+Jr
zL(hS+C#-|Fkp+iRtI+9d0SAZD<-J$&tSz9vBJRP-eSj0^0@K4+f%S>shxz&VzG!bi
z95TSY{an0f<&6~a<6|I#{U*VCd8Z(_?Mvw9)fPAvsyadyA%XBE!Q%i<4nxx51eM-u
z{{X?eW760N^(nvtRF@(<TXfYD@gN)kCWcQ7XCFMh%n;7`t|I(K+kXlMpYN;n&EWSg
zp@QQf`8@l)V5#6BeCJEc<8sHvfuu2uS_X~+MA~J2OJces92~vP)9XsTuEIqdquME7
zmO#&y7gp;J7GSX33ac^=S-LLt&=y5#>@*6J+>9$|qio=hV7&_%IDJ!8k{+E4=paH|
z{DeYeHYC79c*UuKCPT#D&rCrz>&MlwPRLQHdy#Ev9~-iMF(k~QV#9bdG7k&mDP3`M
z2%S7%W9JAT=!#lri~;Qe-Rz{z!HcCr2sxhGT|nD{&B!^>G}?77Jvb7}A-Ua-oXgWw
z<DV}F04s6IXqFKPvaM*_ldBxe)pp2Wmf_QYhiHOOemJ!0-^decV@MD;(u|_u>|6F#
zCxP3^L)}7Z6>dql%OKP3#XOn3t~m3lUKQ`egEw00cK}qzJs^a$?sX!jwZn7+htetX
zrpP_()$bb0uZ8)?e)2mk^;bIN>t#`Q3qZ-7LY+Ey5ozJc{xWLDv9p0r-6YvAuJ21o
z3qM+$?51J-9u}tUSRolJXj^?Q*5g)e9;@1+I$EJkK29+9vY}6gpqqojo%nOpG!{r8
zaWuQ=XLGP-NgRPdF2^^?)KK%m!(_kTSGi(|#RWYsFJ-+iPepoAZ}s86!CY<x=vu%^
zx1c|racUEv4~Ml6q}$P686BkPLANv^(k11Mb$vuCVY=qYme(Oas(5faeB5+^?_C%f
zegNu{nE|N;`tfCjw+jqxLhj*y)JC`;eM4EWa!~mU2&5OgQ_LO_8&wWxz0?Sq&|DeB
zB4Xb5_+U$lH2A)>WcKO)X5>;QUC8Tnt20r*17dddwyynpUZ;srMg4OWt@sd6A7Z&z
zgO?`(q4T#z^RG+h{`t|qSD*7a!$zlZY2F0DGbD*%9XLAyG3YOrQ6QZEXvBFEbs#17
zbr5-y^eHrwU|E#Ur{asMqXSh3XoQ6l2>Q`AqRDUm8XI!$XcB(M&y(M%ob6R<yDX3k
zD*_fL^YXEj!vBNFSba(zK3$?vL?<85hlm>c?A%L$<9`$6y?1y70J4JW_#;{v6dRCs
z0&DQBGUh9c3JT_r$G|!dV*-)?6z4s=u8?#gJ;{Eb0u(r-!YuLqC)(dqg5e4**a?Oh
zyhi8s1xVtT{9wUi527Bpfg!H&!(%j-y5*{lX;D1X^^Y_Erkxhq=l_Gcd+inq(59r1
zS>{N5JpTlLue8=Kj{KZp6+xR57Vm$d>X=rG85GI{qTV<7$|+laq(D+J1KC&R7tDb=
zIGSrqprV~v*Nk@gM2)8FJ|4v)-7Kg3Z5!EbpWULCV{hm>U_5tB5wgqCxt0KXkxzEz
zh*h^P{7X%d!F_J-$$B6*YJafg`?<V4NIts~ri#KLvhX+9oxF_a0|Dlm@s{)fgGBfW
z@TGmhi1TG_2baq0xXS47rp-!ICz9)BG+iYh2qth??c%j}QUI#zL)7<U<!uj10upyc
z=Pwe`pOGb?jWdZ)F?KI>^aW&?X;7{#4X|A)x8Bu<$3oW(-~)-WXHcdj*9{D`k_EQ5
zmBB5c6*pWO^iE*Z#zf>R9}L6=V!GBxemf4G={YoUPal>CJLS9o9%To6{<D7iNvZiI
z)vkjw^OpPjPN+dRWK{PHzYqhdxisK+AwB6Zzz+gVo;o`4->3THPqJtBaaj(Kl6=$g
z!O=jeu3QId^_Yy3S~LbJ9IT>+w1s?L!OjX6j9`MG8nHqYr7)(;T3l?5vj7WjA5fi>
z6zk;Pqc#K4j1Cc)`{k9M*6Wd~qDXOt5u$OJ?cl8Ec@^SQLjgK&BbXSn=u_0=3YrD^
zU$5WQjt1W2?w%?_Q!D7qRgnOQ($Wf-52bCw@+8QB#N6KOe9lvwPsnYs8+5G6xRZ}S
z0=z-)g6wk8Dix9nwN$w9Y=1QQY}S*~jznuV8#qj-^d^OuhsAI*zdtg3%5-4;W?YVu
z*f^EymGzKoF<Mban%DKNf;TBY0)y4rJB!F$@V-@g&h8jK){Eq%INnZO_1HO=mhukb
zGFN}>UMN|3yLNH-Py9dD4Aj)?2iX7jOZndxaiIkvG}V0CO-xdmFnmWqR?&^P-p8sj
zhmIK`pa(SYKS<yQP||%GJl?8H&iMkm4nfo?p|BtgXE1j_V$D}zhzj-X6l_5JJW)!-
z8-U%1JMn!^2fc)Q35nm^C!`;Sbg<^Wt?us~kF$L&VO&9gqF+g22x`YwSqIhrb2j&F
zjV4!Mrhh6hfJgJTNJCeHg!$XI#}q;Qp?Pz0zr??TitYkgmq4utg&h(x?00mQudl=w
zOsFcsat<?%_iaf2`s$w5J5X5DkJc)}J7@!1ss1UK6?y0GY?@~<0?aVKaREN&Y=0AE
zKoQN{Cz5f~699KeN|@CLJ6If;KrDlyg|mYtcfC+qXS=v9L^{xgOhv`@<Do5WHkv5B
z+waQjd<@1Dk3^eibRUR-Z(faVpx5sx`|Z8%<OPQB*G+9<4|<+t?ChQvvI1F}&Sk5g
zPI%jr^<(oK@4$^y;}uk=DspWj6*6U)X*+S_gVhT*K9^)OhK<rren9s5c(ei4B`vwz
zJvk9ehJ+Mlncw2)1$h=|KVW0wbhphQ4-vy$O4&v@x|E$m)CZ>Tb}OJa2ZXP8=vi7}
zwK41lA;2Q*R-L=nUuHyenRsy}B|El`t1CQLfVD<(B}8wJ3lRE`+x5I9%v%A_o?YU)
z<PN<&AbP~@Y3nw`i_NF%8jk6(DS{Ub$S7(0^tb@21k&-qf(G#QVmOeV_*zR1M^H$f
z7hdx<YSywn$UvV$IOh&Is;}QNG?9RG33Im%%;UaT5Eb&kBLLI<7|gw87F+22ewPFI
z+UZ{v32f0%wOsVOp&xADA)Eo9&is^uKfv6LCc#>mrd$TG=<gOC{>qb2x8PIZ&l%5i
zDGlf?2nFUS(2sDJ@l)6LM0O)TT>=WwJoMVXiUkPA`nY^>x;`xfj&KzzXr2I;lq4gP
zx{y3ZE>a3Khbt0+2iBy<l0i!0&nIjRY9a-Lglw#igcLLQ@0QdGG8~wO1a$xF&u`D0
zpU;1Lo<{})RU;(Nl4p{+knq-H26ZL*JqeMg>GN@sq-;{SKA#BRIDCG+d;TEUn^&Z+
z{PcBb&wqIWyuxsxn&(e+AjjqT4ZRe85I>Od^A*CMUzY}7hv&hc9T1$r^Wjn>l7upb
z@7`bP@cc>T^Dp|(Zy<=zOK(W#`h5KQd`0k$hUcH<0p`9!NJay-A!!W$J_$T)g@gv!
z^E*QdQcCsDFZd*dl7Q)jR|Xkp3Z(s#@XHXyw?o>VO+$!v4!|xDv}qc|@UO*x++GL_
ze*b!Z1)@*okig-cUW4N9xupMe7pznb(Z8%`k}VEj7Qa$~23@cwR?o*_A?$3_A%6He
z7_mnE<9i^_Me+IM*Q%P0iVhZukIh786~f}2<tv``6Yb(!=hL}|Zud)^2b$(NV(uB>
z;x7m=bbpaC2aF1lhkR);q?s&iJ{WYl9~8@%4-kZt@an#N<c>u&Q&4*=P<y`zCo)v?
zbsPKKu?9%>_?QX9&h>4-7zXw8k~<9~D;FC%=3kxwxUa7piUke=*ig)O?8Z~BE*iB9
zzX%GX?8Z;K7Ra!9U<$+KK<;zu8(@Xz;f{z?>*O3k=NUj29lPq;jUh3;eV;eilV3y(
zHT}@l5?z}$!He;V4*2P(axkTN+rq%_|K&Fh;iaoH^olr!wBEiBc>ML-t_HSg*w&s$
z3OKyHpIsoNr~Kvn2Rr`jR(U=@y|Cmg=XDQrXCAP%8{RJwfgW^-UHi@<IBqE^GS8Qb
z+Km=wTYEbq3ysh=I;<T(7cXyvcU7^Gnm^}9Yj-#aB;@XBdp7`2NE&tLEYSwLOIxt!
zo25DI%A0Z(lGbO9db<>x4dA@dqz4XD2=Q)5j9zdo0`IqA%p~|Xg>rYVOG$Nt&4N1s
z_BR>PB^F<V7F@IBbZxvY-rbpE4MzENE!A_XbpoLK<S_~YB6F8MvfrjyJv%(DU|AXL
zR5Y0uM*_!vV%)Yzw*snMc`TRr<gRsQg(?Jd5biX|V*H48LSkN=v)o(`n<%~<kjXuF
zD1Nxri7RQUq&o}MEr3~HNda#q;)0O&0j9^|hIX^c0ldj>zf^QZQRkgAK+<ElW2rli
z`sKx-5NlhQwN<wAor=<s(K|rdj)R`6D+sc<k7to~{#ePrgu`j=e-^HPt5cb0ol1VI
zQzcU0h1dX`K__2+IN(3@suURD!K(Db#{(vECLEk$<$giy7_wx2usy^_V&y|v-nSS(
z1?>F*hN_>Lqd16$UykQzYjpyn1`0>O0o={cys0Or>AMF9^-I6?saAXC1HOjE&(<mQ
z8(?$*Bmq3wm`yl;`(Ts9AR_PjNhoChjn|G5eI0l}cYjs|r%HjBc$arAa?;ppXL5oU
z^BEX-F@qKgPs06)*Ds^YAB%qP1;+x!Pxp)<Hkq0&Olnt(^RVmU!0YhYltVIwf^Y)!
z@mNo(5$c*mo>!x=xIjO3<zlhW1Zv9901^OkRewlVhstGZ03>wdZ4%96mTL_@iLNsS
z!%QWe3(h>`nzkMd=l>Hv_8T_OF9sMeT7O$_M-dyUiC7xbro>kLWyDU76wE1><*T)U
zZh$r|E3(YD%RJvZtXHH6Q-?s=TsyRwY$S8%b`?qP_<ehz-NzxvJZ}t3ea2alf`sla
zErAhuxNHB9w)aSOTidoa>uAL(W(MO0Nr1Kp4-k$pgo|4N0T3P_VF=TUzX{oZ>_Apu
zWWREP+h|Ki=G_0?JP|sQD2h(tT64`g#y5ggJS!5!H!SRPqp?v+XoYY?cpzl-)%p%$
zFgpvs-af0oYY&?zY#?8t62IcXvK_WijRIy~q~<;&ZoIT6bIhwPM!<k{zV>F#6SL~A
z4SxB!^WX2w#)_bVeB$jQes|0Rd2yaA!&Ut6Uc^z9;mdY{D{*;3^fL^m-_%z>@S13r
zv`AKmx^)4c7?|>A1u+h#+TVM4?=vwZRspwFa;W-P#udtf3=UfQOCev&4|Cgk3y*S+
zH^8nMeEnRhhxlsl@(bwm=x6#A+WT`i&F*568>eQ0@Ws;uf!T7+K?Hz7#eh)5S*Jg*
z;cAQ~VqY+&fAhU9MdngvD9H}>cdla?c_+335G>ODrH1}ay3X<rITQMNxF6PBP60pX
z`0#0!bbyM=Sup_)X?w>gfEWc-TYG`+tkZA|B)Ochc3cZybho9#+ZtONGS*!kWF2LE
zj-K;g-}UVZnC#>guh*BpOrNHlR4#GWeD1jfgXwL#I-{IQa3-xQdtJigho@HJ4P=cX
zp$MStL#&ZI&@ODm8TSmgs2$uEfVm|ti(Eds)z$76hAaMICPsx_+kkTUFC)yutSUZQ
z0e}4}K4l(>{c)8oXSPxWH8T|cP`?wWpSEwI4Ea8W6!NTSUn1ZfF*jdh*i<#vTnC&C
zt~3bOW6j33TF5LQW-j{eeQ-N>X)TjnUqF$Y&l9lE85FF(*Yai5OJON~Yw1FZ*89bS
ze%1;zt8b<my@#pJK#G}a)W3wcZ(*w2cgqFbnEv@~{Nf2e-^E9>`5(QF_uj=X-vt<@
zjxz#-nBT@fJ@o_Qr6zaMdc!w&h&R5(-r+zi>K3_9W}rXzkc+^|`K4T4Q9We~qJU6p
zoUaUbE!a5_$Q2J~+wQXVm1YkKDKV+a%X77Vz{U8`*<J4!Z3UXd@lq~ev(T1Yv>ONT
z>Va^r1tB3fk!9NBkR`E3qY3=zeO)t@w{Fg1v^9hWKDtu0;dREnde5h~F=Ge8FZs>S
z=mVoVL<)+5N1Z@iCmQZ{x5`V_x-2)l{dxegYJELS+;(5Gj?`!u<^ppQw>9@3c4P4*
zsy3d0=;78NoP%q~b_h|**HuyZ>f_0*z1!Kt7uHr`;(~MI@x&a~bC2%Mz|@)z$j)o?
zBSb}{ci>J>9zGc+lI9I@v?@q6^3ZE`hs4!4ul3)>RO--U*|mF`#(KC#iCwh}+6&Ft
z1)zF+32>1)yY)+SE1b)I>dpA&%9^v9TN`<{9m^V_ms_ug*)eeO=@W>?f&&q4(A$3J
z6h*XOSP{VP){=DGZ+G~$Ztm?V-Ze&h*-aU6$u?dZJrCT(xRR=m$Q0iuQ3?_{2MR6N
za?Gpl=)yf)w2m*gF&#2@kDu;#zBG3dz358xF6m>ZRQNePcjSF7p8(OMRGX4NAho++
z0(CJfR@MqQx&35hZb6CpE@9$KQ3u}HteTq-JQ3*Y!=x=q{rH*Z>E<jj3B?$|=DEZ+
zu|1L-=sJOFL>v~I<XJ8Tfc(pyPVJQdJVnRU0$O0bVZoZA5vlB?>v(@YTG3TwmBa$y
zj%*h3*2bBl-=yZYxFBSK?i5R^;=4EjTF+S=xiew9uuXg{gg+mfk>eHo4=;c;n^#va
ze=V(_mFoA>f<LaPkO>Vd>gSM5B=FaNkm_wc#LG`J$$H>2LHyb~hcuSXDwZcNn)4gD
z<P9G8g-QnpDVGW@NQ^1pr}%WxK)n-R-@{YCQG57d4wrc2i@srdUID7F#EDs}$pF3X
zC(k}xn}7A>5bgShXa6^*NXSBhPMM~uZhT%IWO=&>2uVBCTw2)!T`EAMuT&Je?qERj
zf@!PW%qxnK3c6MRtsv5v5tt{c*ss0nW+AHzl9LYCWI1`ag0(;+zh{L)zph<ep9YmZ
zQgD1Dlf@3-ZvfW9=Pat^2Cdo(nust>QwntCAZ)6qT@jM!%@fFX*rX*^sRuBdX5TH9
z@Z7O&alno=E*^`Mrq&op`8YMS&MOS{_Q|=0vGCnUYb?86*XC-w<dYIP9Q{Xd<|^yK
z@5?ng16vr`KdInz@|_K&6kMSIa))HJ<cVqVJGCnA`k9;_b<hl+-zyg)fb_Y<)?EWR
zNE@H;<M;u8zl*d)Rr;I45f~l0!+pXXTfda|k0}yNq@hay@x?z~j(%Yk`(A2=l3XK)
z4UEj_k%cBrIPKal)`1H&;mfyf3T#3Yj>2`ON74iD<7E3?6Hx*s;=7-w{f-gwBJ~{8
z08*vl_U9GC24^$z985TGwgL2x=43*g5gk740Wl}^k-@_lAm5)K=#CL>!v5KD@XcmM
zSo^+8D@Ie095QDXd<h|n+@*pz*&}f<9{Vtj&G_D@WQ!6ONC(a*2-ETwoRnG*SaWaZ
zBcC?PO1;wP<Vt?8E88ocAX~|v#E0$j3uR+uRdh!;<mq%s95&b|!+IqB%U~<{w%ich
z*+XNp_b=zmZHdD$lfqkii%R(&TZ~`^Ly=fnrN?r`J)mM-9T(GcQ3!2=R#oBGm)62u
zh%_=-CCco=xJ1Vb<|zB}$ev+cp8CUcDJjE&5K~7Z76xKt^og&*{o7SOzS;#Wpe7d^
z8rw7b1DRR%dg0q<Kt{WQyxswToHETFEQb+uTDAFVyk~<IIKE2QtJFd)acuz+SE2nV
z-iVTxS@sp;ig-oZnr{o?&sWs&Eb#C3>8G>o-&EQEX`N-7xj(S|Mwn*&<mg@kzuqUX
zDeN<%6%OT3Q4-+F{8G)%X5ouD(XYz$RnA7er=pO)S%BaGrQfUBIf3gHSPkESx31S0
zg_OiwrmC;tc>2)={23bJZ)6DY8G}){f5Cmy;7w2P<u<5X17PuQzHe*fnmzX<-&>Lb
zv?M8ygV8@4V-7@X&$abz-+{XFDJJk;c^gqZkWcwp-^Z>=h`RpSbNemn0lNJEJV|h)
zU(WjqyC#&pb`NOFVXK`{puuW9Tn?zb8hbkS(%U0uaESSGujEB;5Z>KBM)^V=E#jVn
z;hyi*+jbd13Ji;hkhHovL(4mZ3{BK#TdgMUtD)sfJWf(>%rX5$YJ9<+_5!L236dL=
zw*$8nDG(XNQj_h<e&$O9;LRbVZ(T@>Lmi6-nXZJ{lv^1RFfp844p`z~(`>8h*puuO
zgA0&Kp2$NPHW!-|AgPbK=p06a&M(gQxh|nL3_jKdQ8w<)tp~8s?lZ7&)>3@hsw&xL
znX-RCgGFW~2qrztsg11Bx>Ke!6^^=1)!7vm#M5o^9}7F!OlzaHR$fp+Vp?`N6YIRa
zG2-KPuU4710t_3*Z<^F?<mJL$nJWdc)JN^zZg>-EOqi>H;@ld?+AcS@=ul0Z{<v*o
zNEBy4Zx7;fpUnxKFny=o;_R6@KxlUmF7#!0vAFdDp$%nFpFwgZQfun->RcY1BaHW)
z2bw#_=5!wpi3&5k>$VWi?QMHr5xek8=C~{BrLp1}@1o*5I!UEhfmjp;DSy2ufU-w}
z=qf!xn%;QvAk!r0PB9Gm0g=2&OrofS2IzG`VWUd3I6E<zoBFDnv|5jbO+1X_5_vw4
zu_p#4yKgsuIaS<+o%PNR<I!tUx&#!a$z4u50rtvm8BS6HTBMVc`&R7$@Zi{nx?Wz2
zl!_@E3aV<IQi&<AwE~En8Fn9qz@$Es1my?B41I6BCN@xDyXB!{bugtfJ>;2-rN0D3
zWCFeNmmEK!d-XG6S5Y0>9V-qTA%jF7V8#cBYJhisqlQA868Blc@?nd_VMGSjDM{oI
zGIwDV+is|ny=9p*agQO)?Pq$_S9m9llUD`$@PKx{o2Tx^65TULc~V)QP-2Cw*@r7L
z+Xf7TZn+eSMprqFtscX}QbIzkv}5kDgY>Y=;#&6A!`(Ugxp(hod8U#AwzRPzOiX3R
z`}9BrCEB>7Qy&6reMT)~LAgO%p|VPn92?fp(*Zw1Py2kb+v>t;W{D5(&8&{9tLB#|
zjWjSOOB&2o`@lqoy3t$rwz4-0)0G^7TEOB5#U7k!m(l%(?NV&F==%1dCl_UX4OTfl
zvj&eOVM)Q7Ux6F^TFX+97W5gB?GBU{`JF^hkxJrdx?Aq$Zqt%<6?Ed_YR^D7t&1Qd
z6jT`dNb0qlCsbv;z+4El^7mIP_27`<92Ygh%EfNm<-#r}1qx*iuH#!a<a=Ot=(f5;
ziQU14luC&2VMH;*f;bd)GUo+mTOZgOq^;eBs89*xK@Ncor&cR8TJg6%o37RFKE^=M
zD>yS*VNb`TbA^WE)#&$Ifsr94i`pkwJ(Klis_t&r^ezMg`tYNEL1+UKNlqwdpa;ts
zANJr12Fx%z&gm_-b*x>oNwxF+j!u-eKqvMNIqnq-#ibr%W~TN#EZAgN3VbPUucRZ=
zoOe1GI|BQh+`}>|+opi9L#=U7#PSHeZtLh{b;c{0r+QZm`nb}?x^~RYBlkeFJji!y
zJSkXG8ihh31WEJm2Dmy|sai{PnA8Oo%j$yXZA?#Bb$MO^)ijrP6}O??{<q@+<6nc1
z{cAY-{~9*vzXo#iuYp(@xbW8nNXqnI0^9_DOzeO~@>Wn}-d|iY6=D0A?!Ddr2>9R;
z9tQsEb^Whl2ls#bb@TBHK$Yw74@A}<zbO8~esR+K{@nNf+wcBZ=rD7OZu#jZBw%jw
z_a4Gr4Cy~c!E4Yuh6t@UjSYuT_z_ZhPCNG%LD<f7Q!vtJgEkR>y?C2f07Zj(awl+<
z<T+2D?pGL?z+ATBA-)jWb-4#qJzxYrAWEZ&REVS++~aBxnel4Tkq#1ZSSQvqDBF}2
zUu|@7FfpP8$KF@k8F-o~>Im~TbDF7#8q_KfvM@p;YzvMP0tL4kp!M=TfB^6-dcQ;J
z@dyD5fd146zgWk!#u3Eo_drq(K0tO}PjoRw^>TUlLdhnuj2el}+W2Z;BIq)XpaU&D
ztBrl^C9RjCG=p%@`(1)#fc`ccq9~fV^xJ_PMB<`en#g>6&<N2F3OW;GKI%CS3PPRD
zcUQ}VeNT=d&7^Vcl5VUs-PrJB{%n~ue1^c`WPuDjy=zCO`4FDMzF<f&S$Y+Fvn~nD
zks$DYzik#&2tNc*nhZeBDEu|E2LG5ut%BSs?>RERcOAqej%7Ar)!-o9`mGxa3kQ7o
z2WaHoL+b)@qB&3u)}Wl~qo8;-XX|V6_W8m)Bm9D6kP<WC<)2!pAjQBwt>@29glh83
zXENv%!M;y<rSsYZ3Wz-S8<;C=)~};mk=OHO0^-sZoB{-NAcrSJ$^Gpz(X1^PB=S|i
zA7J>@&rf6mfEkG7UyAU)_jHtLkwE|Y;WHCV?>e4d1G66#y9Ma(z)$}6sw7D@4XUI?
zx?wazAKxDo-ci4IYxu{?K@<QZK6sh%HhIs!w=RUQUr-JvSVJ88`rsVkI)fkcXM2I`
z%Wss!RmZ27(&dNFILNQw)3kpul_kYmTn;3!G-vd7H>zDiWWKNgzT;*WP5?C0*7|K-
z?hfNg$~{tj(y@Zj-gDmqXF9B;DkWVOg7-|7590D*!Elb{fF>X)d#%_t1HM$0$5j+X
zJZQ$}QLJNs5mTm3SyOELV!TE`t*$XZs%6*l)mt3ahX?2m8e6%Y^4nD`HvAnTCl71M
zXCDhMGMDtWhdB3op&d35@4RTTaeEN9i!js_4cRG=a7^QMS_pmaN{?-@yxVu6tT4x0
z%-v~4%W)3UD1>FxHx>7cm9xJ;0V_*Y7*fbDcz?Mcq~e?`DiMi`i%SP*h{HXHkgQ9M
zhU4=V>`m--C=AM3@6Qx>O9=6i9HR(At1404grJ;hWE@;dUt_6)H^z;+p&5X&Kb?dd
zoh5q5EAi1indO~MORp7@6r3nO8-4wdYXbG$3zqY1<p8lhEVAM(D+PFhd8>$^vjd4)
zz{_v?@rq7$G(08a_9SZvJl)a`2En`TDXUo)#qo7{s-Lm5qV{b+DmTv<S%nzG{jn-T
znVo~J1Xw$^6SifD)@f1r7lKmFs(sNw#-pgJH>r>_h#$9lY2VqJItE4Wq*z!>(N%fC
zcjd|PN(2H9jUe1IIWHjfOaM&1AA(wTY3dtVZ(=7#U?LSq;wc0zyiq2U<78x=DfdhV
zh(D#$I`~>DapSw5lqRw7_f2VU!J9Ke2H?tD@RSDT<4Rtf)(=!df`?@ETP+sQr+bxX
zj!byK_>9bn_U_MgA%?t8-V#+RZzut#xBQ{3P~#4e9d}gGZoI4M*tQY>+sc4|`*{tJ
z$MX6A=$r)nbNAPu_E7+}!XBA+pQA4Pr~4=&er*mwZ$L37I*<s^$IYXzsJ|c}$l3ZV
zGl9X2xQr|Kd?fG$NE5{B4V-~rKb9jf53eEQ9X6iXL_fTKmfj+OKRjK~Fay^uIB}r~
zaOsto$^e3&zP}BeE#G9fv&g{64qu+Al0KdYc|DUUwTck|gvOa5Qy7!9RnN|I>=^+=
zXuvtid&ibRizLI&1wT{#iwOCnZ|ftN?K|Vvr*DgI4#diKju2@!93H#$a%Qv*+C#|A
z2HgrYCmf75etgQA6Ak2ELVO!vRjFafBc$pYcT7A`wGhVss<!T#p>up*E++X|7T&_@
zy_8dBG-sTl8@bWO56^->i*~*~43&PskX60!0I1MD=rQCimr{1g;m{YUBAf0>;z$ki
zhJ+bOW#82I$gop9JWqj1a5u8+$gIhkQzW0%J=(#hymD95IR<#}Br>@CSPPb;C?#a3
znWr1xS$a6yMiy_LjEh2&II8TF#c7H)K<%ykNIS{CkSu8YV&A|^jC4$3?tOHBX@s6D
zQ_?WpYJ{Yg-V}+aeZ?%<9B=WFhE^_~Xe+CPu+RMp6gb_n-thgN3cvZb5DA9y|3ras
z-k=a4@jQfB=L7$C(@<rbad5phFl~5W`(U@?mN4L-p}Xv_`=z+amK!6f%Zk{^;%O2_
z^~8;Qz+~|ih@-6t-Y4Ef)&Rkz$fA*ljep}@Jr5orRZ?QX4)W0uf(N;^9w>oMvA0U?
z<EED__u|M)yDh*N603X2c&=4b3ZBmv*~eSXJp*Y0Tl)fjWCn9c?1$%NE{k0Tzr%!e
zRJv}aT-T*10$;~(<?Z@#GdEY0KlRlDJs8T_MxkYoA~!b=xqhl1D>V~r5DgiwUT+2`
zgHh0)rJK+LjLQD9L^I6g!SVorWshnlPZzQU=u$4pW<03Sk2?0-9_c9?Q@m5)w#43Y
z7nL>$-f2`aKme7|mXE5yH*ZV2V<Dqe=MK-?l8sr>ipi}WISD?N1O{V@#nRy8>UZr@
znqdE;VW{8MRhj&cjyV6pqs*VyQ)WA_r-#pT32z9;NsUJ(!2f3coWmVKX^9W>jNoH=
z6jvsxjRXUvt1ks4lxhF8N_%0p-jtaF1w1J=^CRR{?!uA5v)~K@5ByQ;!jp`i{yJj3
zmws^U1+NuP)7{{1x*t~k%LNt<gy~na<90JYlSF~GfXe4C?d$-9+f=v>u<URf%vW&9
zh<f?$s0FC4<LlLY9B}6I4f1atBOnt{qCN}yPP?4Ax8qpJfZVDDDCM`am<PuwAL{J6
zy8k$vyuV#SY^)FjrJt^i7g3Fw<pw-VgjT@hez2eZO;CW(IWi>m!ankIh7tIJsIKC*
zctzbGZj6s8MQ?7q5VB#=cysIQU>^^87s1@xqm2D^gaO3Ky2f&JS6OY7L+-oj-vU(D
zmgC`w?z83Nk*&8_sUAVC*(Tg$H!7E9QY;snLZevSBQ8}v;Zt`o&?^!ZMP{564}jbW
zOi_0BRy?qzFFb@hy+}4qdv4hxil_(^>O|2VepBYIcI<>`5j2Zvzr}|xZ146mxYbDL
zv>~XkjUZ|p{BEUrXNW=0X~GGO+2=r48776$!lVpz$*c}Z!cfP$zBU;{2-bFCs$MNn
z`;!26B^Rg=G9+|7H~#AM0Q4Z*s8zuPsF8Kk^<f?_d@g7{vodyB+FT>KL4$R(RG^Gq
zE{bRnt_Kh5-2=2}DRxAhZ65ob|8P&G-YNNdQ;zUySKgzMEp#BgGcW6H*;(yw<bqwY
zJrur})5=s&TcFSLXrtQ<nbHL`-Akl<Na8KHVu`<oG<AQ+74G@?AWVidC6SWkd&Wvn
zRUN#T7|Hue5lim~4iYVKx(d#9)AQKuhC~Ip=7~ohStkpC@ie!zdSt7Lv9?^W(fo_8
z-`KD$xI%)iq6AH7*<myNI3D0@)rO!PJwSp)n5g4vyNZbh1Gp{Bjk!bVr~-{Wd*w(S
zs=l_9@L}DmHAJ0mGaHmy)nhUt%>yK_f^cAO9>Bb1hNH5bv)!Z<<g+lhaot{5h+JjY
z1(`R=9=;GqHWMF@1v!+Y7UsP@9~DlR3oW@GG;}`)q<xu(BFp}ze4E*s-)W4uhdpa7
zd&}KkcltvDn3`UWXdR-N|F^LBkNYA&yYtTl$<Q7?bxx6^i^CYDpg?Yp*s@Shdt7cu
zlhV)@Vf+h<%9rha$%ha)J4Lk?bgSMQ@4Sqy{EVxm42UEA&XIe}J=?qMW5NbVA8+^3
zczzm7tG_P@6D%K5flMb^ufd&>Z7Cr*++{snCRUq0o(82Dq5zrplww_$o~5lH;RBk}
zJYPhPiym*#;1|=b>C4L4_+YBrWQctxZZ%%@6Ay<jXg~C~?sPhXbBGzSGk!1S=)y}^
z<%!}<1-tep#f-3HSP0SSqq8UOo1+SWXHQ6Tei+D=EF9iT_RLec)A@1LI-zs8UkYq8
zo=L7~2}t~NL0aBOakJheFg_nG$Wqp2R=DBB_F`U6<&do*3a7asnz0h{*i&eMwTr+u
zB;J;=S{vK_D8RJ<S?ZYYtMGCG%)RK)98#hng>0izFg|BvLRyyQVhKsi3dFY{>jiCA
zrfXW-Q-kVCL+L$)zdWCrxWxB7Zvgn)@{`dps^dcl7R*_e<c+`s%xN&iRfyfU4FnW(
zXhIJyf<7`r+sN4o<<$tfEszpGag%-@U2ajY*KM4H7!Qd9tW-tYJ$_zq;BZ#>iL=zf
zE77p}zI8xmXSe%IMVw=D(rf*JX>NM7b_++~HZah(Kx*X;jTH%jQu1E9wo<qvRz<?Q
zpanDeC@%%iZJyYwa;Y0BEb?b@L~t+Wu<){1&-q%|HnmphEX2mF(q+R1M+#Dyc|SGz
z%RPdBt^b<;`}yeChyJY2v+`e>q8L9jo&Tw>pp?0nnS2@)SuhVUOo@{p+tv4??;m@f
zr}KxCZ?<0pS}9KF-rZ%Db4E(L65+oc<UrB*@{%y}Nn35AOLjb)!1(<15vDk7z$&!p
z<}_dp@|*h=9I3v)QKEkR?7w?*1cZ>kc=oxC`?vDx=SMKnpE`TbBd5zlX<<zehz=CE
zRJGDYP?Oe1wn#5E@5U;7k4!Ya#58*Fa+M@*%UiLn9(BIrpiK`Jg5`5DK6l}b<B<9W
zMZl=533@#}_@u7aj}%vmMF-+H3zPRE5t5O)O+<kvfo^de^aj5*#sM7hHWfk)V6zRh
zz}ZZOg(6MEsSvs%JUV^EO@OGa7cNNATXYk}yDh-eDr%AhncCh-7!F=4_&QwIyG&K^
z{UTir*J5>js5h%8IfFu0d^EM?WC!xnk7a0Ym6t8V`{IIf-;5xY6VoC{`L<)?{&}i+
z73iLMj#qTg>@nMXZfHeRyYenBjvEYE%eZHaX$mN(>6{8&A#kUQzO=DNKQ*hvckGN`
z9Kk*bYM>K(&$jY^$naB02^y`remxI2{%SRuz>!CyuH-WN*~L6dR^Xh5FIJ#l6jS@K
z;B3IP9rNU@jfRvp>*=H%jw?63gvS&dkXe6JbzL<)P&WD9;k4SZ@n%KBsn~a<QavC#
zyE&Ju=jAx;jcfx-xKoh&Ic1vua>t(}vbsMhek&_XZlV=OY_I+*<_?Fac<tTICK?t@
z=0)&N2tn5?WPs8Sg39c)KCPEpLy|%3UNOP&ZvLPKAPNx|TvM;C@K!vwB0PSu1}c)|
zw+77-l(%B<<gtQGNGOe3<Oya@YU6bWuUQN`98Q`tzY3|VtAu*%FP?rT_sOz^!EJ9m
z%AH=U10!4Bdq_c3q+4-y*VsKz<IvvA6Jk`6330Gt5EBUxahjfMQaA;<PTj7zvh7J@
zG{;^E)bG|UQkHr0wP<}gL;Z2BGH0=0ihlHzXx)(TpZAr3<nTpw0(~V4mZMLo+YDy}
zEYDWaFJMfW0Il_!$3Raeo5jsJ1qR`>u?8^sE4vl&!W6WLD3FC2^gCww+il3}#u*g!
zn`iu|PdJB}ex4LQmfrV+0`YcG_y{|t4Eg3SeJpHuOgk|LQM)eI=9Aq&c7nETiitw1
z#WF7U^j_!fHOzTpKE|8D5f!vkylZ-k<8kskC{4R6HnAo8K!?Myi?{)sGNRYcU;*Ug
zfq^*7HklBU^LByAVz&US{b$!5E*SsxphA@En?<X+6a#G-+k)SB13eL9uVHY_-|!wi
zNY@}K_K*Uvn9J-6)E5JRBj~fR_E$2}eHxI(9K>)ys*~^tnJh{<h$gtT_j`sPOVR-L
zd@RGxirF>EawVY1FY9Sb)lWk)%C%A>{`4%0NsYt-dp@*v3W=IH+Dp%c?s#!4$lb13
zT+#>Ut|Ixl*CUjK-s<`|$=IRT!jZJXmS<iFvBOk@k84|sAHrmR0u4L(6@TH{xXqW%
zYz3SrMiHUqhm>uF=MBS5!6F-<M0$1qytkMAhxzGmjQAg*Y|9@CV)w=mwO(V#<61}Z
zsOIp7R`$EAsr(!n@aDRsn>9dVv$|eV*9Qu81!yF#n-Cg4_X?^JMG>|L`&@+EVl!ge
zuG^&+c61fZXz%34mu#TCf&Hj2I!C~E1(ub(QQ}x;*W2ir%TKP0s1lM63G={Un!!tm
zvSe2rp7cwTg)R*WxquQ{6eu=K<96fQF0+K<(O8)kHSW3SnT#iV`QRPM(k?{1vPEOv
zicZXM!^J!xgnlF%FjRS3=jI(@hG4&o@5Kg!*I9*q7}eq?UOW2+3TQKNhvUs&I7gK{
z?K=f_pV4Z4lig+4C^lhFWDgkx4RG@0P!sgx4kKHN+9giWlcDbUi=wPHntRj9RB(=U
z0yaRr<ZY!oz_`H@BE@F%%L(1wp^DuWNHP3AHV?nwA;S9wPlLYsfBkjx?p_%-;a_a^
z4C}H01k;z*?w2F(KQ4pST(VW4T0CHY{cJ={P(Lh4GfVAz7l~dlJ@DCc9|evNdPJdg
z>tNK{eFtF}Bw4kI9!Ase0J813vh9WAd1;<|(4xN5mY{~489i?|&bLXl^u44myQ^QI
z@d6Nj1Tfsjkb^LXa&uP<hMIn{ZCz{om1+Te&Nbn41dLbZH`-A4&j>mQoOxb=!8?jY
z0lnkW-G#-9BtbWxtD~yy-vH(3b_ztENn(3?C<kWOo`N`a9*g1r@QGSyFQJN2rM5wC
z`9ipD^=!TJc}R})%ucInGM-=-cfSpZu0o<vd9yl=jwd<tLYM6KGI_q*<Tdl{;=Bf`
zDqj~y@7FTYo7XmBYrCSdm|*;-dRhl<#hmR_1xM=TCRpoGf6B67ZwWIYxuue;+R^VB
zW$UMmwX+<QTaxE6JK{8wp0Vc@6un_n<VtZLrGqO_>ctikSPs0Kvwo6`BL!c@ByLrU
zPJ=R~I&dl;){hPf!v`7$4pzI1(lI-v94q0vJy4LGW(x#*Xb;DWaM^0oEi2n3DwdB^
zy3L?Mp%h2^mWvcZZ|?d%Ia0eIR?`$`jFWQOfhzI*fHt+7Ew?1FDnIMG&$143k*vAR
z2zXWkT4xslD6Mq=c?p5wVD{<F*U+cfLZI+vk?-uH6sEk`V?{15^X}e%suW%#g-#)>
zPNDLJLOtNuO$e%*ClKRWxZw^zniX76KMX4WIH#P$rZ^W<sJcNaG+?P^Arg;t?*ZF$
zD)$ZAvZotY*J3mt3qvd$(y;pFa@|zX3AW$VRM_-hgXr@>G7l-$OO8W1_ZT5dDw^m8
z8)7(Mm||BI?B?m8=yP!eMjwxw!0uUuLSDd_L*VUzq=lh4!H5zV8%KzeeeSK{Mgt;<
zc><s!sTL(2)6%$HEkmF6mzLO7@GnZqbZrkl@Y8n7o@^GR`&=Y~yJ%0my(E=ObX|A%
zgC+G{qiMl&qxyw~1n|k)Sf+amPy?(h23Vx0X=HCn)}MUVyPq1mrZ1am73~k}V$!Mf
zltJWIOW5%ZAJA&EC7Y9ipH(3c5diK~?M4(>{8qH<cD2)S_qOM+iK+f}P|1ds@$!(I
z>!k2%U|0TJ`V7l3`)^8){<!`wU*ux%qYPCUZ-7Y3VuW-A;+TYR-9=}JFaJR{)G%|b
zUh&1j{9^i;rZjx-_=j+S^@q<S@0@WnC~y*fD^VLm`bH|-09WkuzMunn-Fen`0I7_R
z92-KEfKUd?iyXjA>v@rf&Uh0Vb1nM{nw;szt%jP<kG$#&2O1=c84;kf{VhZc<UtVQ
z^d7hO5;ghLyYpoO4+t|=5LdsMT)@2zx*LE?%<xP^hA34Sw$8fu&#&?8Tb!H557;75
zOdUV+r%%^n&WXx;`E8ymLZx0I&X`5>GbtqTw>!^565!E|Jp1`dcdn&Ipq<IVh}vDj
z@Cw6()knFHh?j34AkPp$y@T}0a)yL{2$|lun&i9SsrfVJq<J9orF-!E39dXPc5!kz
zFw!aTnJl`e7;naMP1gZYr4N_i425vnWW>&aqA5j$CBZ&veM~$K7l0}1ldM^y29y|*
zju)3AZXk~HM(x%cJ@SR)nZ5f<A~Vb}Y)6{IfJv7EI2caact9immQOs-rxJDyoI|Xm
z_o&?VKvSi5t5Fk)!3rtcI0ILKYrK{)HZ8l83lc<0ZS0@}@N)Mz^`wT69wQN6kNf1j
zv`)p)IJlY{xw*K7i)BuyVLM=f&#xF8Dyd>AlKZaKK)2I60DLf`D}%2f50)qo;ebP?
zrqCflZb2I$T%5BEGNDlC1e53!P(oX|JMBtHyQ-!Ad4HD#npqN{sZF8Baf~j(4ffS|
z?%iMf&+ftRCb&P{X1>F8O4@d;cCi8J;+#Cx2s$y>oZ2iUIGX&}J|O$o<OT-3cspq4
zvsn?q*G!J(ZqpRu?TNOFiyh_n`y%K(emo+ue{&CYXVY~o94y070thRT4Pv_Nhy#A0
z2U4G807cv%Rga-@>>qF2w6ae7ejDa`EC@+>ruj_=eGMAH?|YBpf}F}!IfJlrIMShu
zn3g%GSBOpi5X>v)KaGHK2Mi{`He^5>D8DCE$+p2u=J{|G@F2W0tXT@dt@xN;*f=Lr
z;$%*T!+A&HRdAqG>?}sHuV=V??nVL{#2I?1BPv;OO}e!(9$Tx<h10=2L>}7H8N(W+
zU1DNtg(vp$Y`bBX0-A?s#s_IrOfZ8np#^ab<eXVR4pU=49v+FZ%{@ytK3xynX>(XF
zswfs~R-@bBbr0Iu>)Y_JkEQ=DRB5ib+26Ud|7g!9A^rrCBtH8Ks(b^;vRu%{Av>PZ
zwO#D?w)wLpi7$)#Jn}<ci@GuAw7+zo2E}QWV*|-ne=`@tj1ZR0cL(AiJq1>}&$eXf
zF+{M$#sMILMUr@}fG_FcpH{UKg#&SFelp2mv5zMMbXJkKki@k8AfC-Z({HO63|wFH
z)(HfW)a7mAeyiXo7#o8~GWwdAO8;<{FfS8r0^eSGYwyMqFy_90*<S{<B0V0yfka0F
zwj~A&ApYy?mIdJtzumJkiF~8J!E3vIJlWSZ`@;*Q!qx<vp+8FSg#eHPH)HjtbOlEM
zhY>b3==pidiEz4VEk7zdv%E2S$ln%8>@A59QXu=&nNq<6&9LxYIO&K8as@Rl>1DNi
zj6S<q)7Y}a9^s~D4#fj4<0zlvBUeIb<?im?yT$%7!9lA`1b)316t95>O5YuZ#Ut4w
zhh@VLcTNM{zgexoZ}%&>*q4FXujXY)(MDi#ph^!IIIOZ0+h!+F=SbUJo1UCmVQ+68
zlr0kFfhYh}^#d+rn++3i*6vo2jdDCXpc|YnGQw`MgUP#?AU)=H6Sg5_FP(ve1fj-w
zZjjIA(y5HS3zne)mvkAxAO7Vi`lTINaCDz|h_TtMtHFkcK<avB2p3g8t-ETACO-WP
zob$LH@jOh^9ew!|f8p0(@!wW+f2!SRVXoZ<;j?ygaMo^DUb(CFRnfYlxC~VQt^Txj
ztG`V4ztwKP|C$4n=JRWQ`$<_bOPx`WcY_WC+!@d#yZa~zr+>LVM1NZ!ek^+6><kU$
zAF7^zx;}g^5AQ;11l9t8InFE7d=0%$_|xlzKRf5X<thIt=y<IUKQ0eIJ^Z$;eSziz
z1pV)kwX=fQFWozgjX*pHB6LXc@Ia0Hv98SXrtcOOC<uSLu>6yC<prAfCkqR#D?gPd
ze^n>H!kRyN4o!V1#q;xuGS497nrm4GMdUY&CpUJ7ol_R6o_Z7mQXNP(Kg)w#7T_GR
z(!29CQxsOR^GXErdNcNrVJXsFu$7kw$DI;?&nG$Pj#*|&xAD=w;-D5FMPAPO-A%By
zY2!T;A)xVCUir2psCh;r3*f<eD-n#+i=DR-5^xHYdc6v9TfFL+Q7PDQd4u5L7MPG5
z5m@N75B`DFZb(-LX9x^Ft`4f}$751E$ev_Q_UVa+9gqwRbCJXPmu;nT+zR{MW5D~<
zk^~Lw*}iEWCV?*D;ViCrYm7CjPjLn_6i2eD_&)T8g**XsWG%0lcbfc@X7Y)^-5<Ba
zDHRIbdUWN$c--B3vf`FvpIrcb@SQyQJIw(CC_>OKa{#XTsR{c#%|UIAj~qlUbB^62
z-3Y+fO7UfgTGv;+%nA8IWfkA4`UDPyK=fqhHD2h6?`)3*^Z`_$BCOXN>{(^m^$D7A
z?lyo3ne2`*+vgra9|{PL)zoe}2qME<71;vj-crvPd6!1B)u4`1#@qP;es)NTxGAd%
zmnQA1I?2)-+q@prXY;hFM{!kPbk`y6Ar)N++%AdAf$s&StLMSNx9W6(ToNqU9`}y9
zq0;LB+i!*UKwoxOsD9DWl=Ikl?+X`J<I~}O-!~+3s9pcmE2krVeH06Aj5;MIeeRAE
zrEzYx-rL`y7j}2*r@r4=Bf8&!NP7PShrVk$fh61ynr{G$pRoM`3G1aptog=y27m_2
zwomrp_mJ@nbAad6TpAbG!h~`=DbR4X;2}rwu*SVz<LvPN*s1@wrQqM2uYcM($a>y6
z($5476~2xu9k1{Wq;s4o*et0%<xM}NH$lwU*AxG@;_y4h-<Lz^+adl}Cw}bfiGQx1
z;5fYbkDU1D1OJPG;aT&V`vE!sJV&p>w-f*P$!jy{9<UZ~8^a5AAwfw?!$QVS^MN1v
zPo4PZ13!3qAkTCBpP%^W13&WFUio)V{3(HgZGk05$ZnQ4$64G9<BAU0t9XK=K=LB2
zVRv3p%;lWdN(bjiCfOeaO3qJ5&GbdRTJKG&*<JI4i8q9LZw|SLVZc{UxeE*7CPG5<
zw^N0p_;az+uZ%^}y+GQNZK7!>dU<;i+bcQVcE>f9c>F<MB^n2hcxsoHL+*~e_ZC1g
z57B2Tt?{NqzJS6I5{rL`q5nvs|Lzz}m^axBH1pQv&0|G*x1NBi7fj2hZN5hYz1}<g
zWe}J8xWuZY#OjviT`Mx!iQV(=C4Aa*Th7oP>8NKt=`7pEp3<wBvBpE)7f2cO11Ad6
zeE`Qq$Z;~DY?a)57Nyo4pB=TK&IZXW;!-+Y0j*@UO{Wosk`SFg!TMQ6ny8NZTFs&k
zMasMPAn=9oy91*c)`<>8@4qf(0;3q{o%|1Z3E!_De_A`}ZXR6kfUo<;{Zdt;rd@Ep
z(<4p6ATy)mngK7vPpi)7mxSJLRp<1U-ZV!!P`L|&)ZCb1-XDj?FZ3Fyu3YE>obHgj
zU0xI7qrIlU*ImOv2wqDDP-#B$4kp-ZzqQ!D{KfZ^YXXit+I{)}y7-7<4+}FxcyJlQ
z_~v;n1^|P76X(+7@ymk{CIs2cbEcOg@#Dw_zXzo3w{-`OZ(kOk_dJMyScx8gJgFjZ
zRJA4IUELc?idld**_44kx%oJO4$!WBRRweY4)f*n{Kib%<^dNxp}1}a1@y?D?O6RY
zIAX&yfB%R0u{!@+oFgAS$uBx>LtKA*5T1dYA{%~RzPx`QAhr+5_GciF3aws1qt3lv
zZTmC#+}HAS{_~IObU<mCs?&rlx5;|2uFEy>B1kDk)&;3#L8unIyxZoHtMC(Z>`d?_
zK)rK&0&K4IAYK`7rKt6HHs{$agvDovCR4dmiIhdQ76y`o0Tz2kq~sJwjRlj9opfUk
z77syxE1Ni*5ShR!XDJM~4vO!|W@Q;YIh951*I|#tEQ6jxg`PG~NoC6gCyIL%Io<+t
z(*fxK6;5_fh}Z#CI{*WeGMMa@Ob6{PF}OJV%CB`{Ypc_;+p27vr3^@!2uTYaqqQNt
zUOpXo&=N_|*93r1TMwLlFT8CWxevqLANekNa{az9ADWbzUdL0{^la3pqlrs5lw=>5
z&RqJiPz^ClZLQJQkoMeLhq|@A3eVGz@^r(FFcN$Ia<23<vmCo(k#C4_S|T9U1aOK$
z#WcNT0Dy<NSfmAw9V^I1G`PDd4j9!#PxR2>=r<G|S2V*f^VRTRYaUe)?{GdxY<&T2
za1#z$2b@%kht!kzWMVM?22ZhkU3YNCf>orL`71}%2TSQ3Q0xM7^kb-pP_Oi19R*|J
zU8lp_Wf0B214(M>a4Xm6*q^Oq?>I~SnQj^!rv3o#0ih6vIESG#&iN4}<^CvU>kcjp
zXIW*7Vx&?!3>5S{6@Xii?v~cr;h{b~77fJ7^S(#qrt48PN3KP-u)Jx22kSNvY~yA$
zEUAQ3aWL4V&mD20BIznk!`X*SwuLfnM(%W?%KK6?A*=>qLjfWCI{=?Ov*X^2Z%%6r
z$ax7pYsRlxpggpaTA_E{z5_|5CeQ^jaQ!-X23u$9*9IVx7OuSkvCZ=VozH0%w1K+W
zwP9~l_p_HS&PdO?P}|mb)Kje^&jIdORc>JPmgQ1UF-Xy}JO5UrjuXU8yiL?0>{tl`
zr=Ck{*{>}=T0NhH%d`GX*dPT!0rnC~IyZW=%@%e+-n~@)J#_7J1AH^~VSl~n(SNj1
ze=|MX{1eT>_xs_O-~DN81j!j}js0f~s48x#1#oH=K05O5z*m_QHqB4U>hVi6=380)
zAJ`V>U2#sL{pGd@b?BEJ?f2W_yelG~+v4TY11k094Wy^x%Yy*w=;d(4C+KK`fo-@$
zY{$3e?FTLyGT@Mp5V(2G|5);0o6PhnSA*s)q!|4n8hHNvz=r#ZEvH{Tm3a{PKluE>
z?rly1Nhp6REB?vn_lg3Z>x%EnivP{e56X$Tm-I5Df467;`1u9t?{bIr(k$-><kH?V
zfFQQu$A{=M{<w^|g1uBnurrE_0`x<2K_ynXGbIoD+*?y9jF5PGvYvJ=O+lQLOY#I?
zMiR?T2&jUIJN683qA)jvW1ROo6k>{2d;xhJYq!YQ;Yo)$cG`lxsz{?Nn29{-Ow5<j
z5A{9rs{$W72;Wj8AO*%Th+Ac%ku?#n#Q43bIj4E@>7o}{?qR7X=^lZKi2Fv;Q7`Gj
zHrUE*QfFKDhXwAn8l?RWdsg4{C(TJg^tJM->M6eO8Npm-*{>=KPN{&%5;1oImPL65
zg&i0uzi0o5l9k=C?q2M(R^Oh|O|XHT5i<YS^^d0x1cw2Q#9rXo;lvaX6%JEv!Zncb
zh1)@{G%eHdCdk&**;j*@Jgc?r9%S{ZOjdEsaO`@eW0F&M4;7l>2!861It8iVX5=y2
z$@T90=h`er&g7fIh7;Ua-hj;$hmeVcJ<rfx0g1w+YVXH{71~ut<G){c<**eiKl=->
zy>k5P|9D$|*(wkI&%c=e_kXos`QQIf^WXn>`~2_!s~rFFLi?vhm1gGi#`ANK#p)NZ
zy8|cbXM;>APs$%^=2t$!T(Z3Sq6EOxzosHE-#XrK;D@@H9sq9A03PF&rtr;4*Y};t
zGF`vFPyKGaLV`w#uVr3)79_ds<r+fQ{w^~3()JZ*96a?lVgiDDV-)7fV2|m{(SSmn
z>7XY8w8}TDH|W<Im&_uqv9#dJv%WoP_Acd(AgO33fPniO42CaoaT*PCdjV|V?>9oa
z&iw}qIAmdb0oX&d@OQyW7l>hft?Hmd8eYdL>59VJquT;%RS?AC>-Xb=B<5OtZIm70
zQ~=C)g;;Hf8$k2em6XXF=XdkH@Bo1W+iOJdHX#7{f`L!(gFZ?8^5M^Lm@6n3gb$xN
z9RN8-1$FC{0Y>fXNs+J5`fnW*z~<%ex8hYZsOn~<`IhLA8MDr%Q|BJ@Y(qTY%UAVd
zbxwJSOOF(wb(w>AW7_sP5$Bk^q2+KqJtfa20c%7p&850gF=x$+;lu2aC3klw2NJ`4
znZ5E_ZwKH?bn$daZ~BM5k5A~i+*|6Q64R4}o)^rWT7n1RxP~D!EI$;vznYfkg=ULq
z2!|2d`q*que4PitEeARlQG*8VSYg9*x6YOc>8;mOy?i(~z_F(Z=9;y(*@FnFi$}Y9
zJPx;^&{$p!dgL}up)PmUW$$l;h3ah@--_Y9V~vYFt69u~EtsQQl#d&~y6*~*R-7T|
z!8w}TfD)&+UI0lHrwm|r<ZGl1Egs+9<pqA$)$$54HPFQt60urAWiY+PdVFYvzrc+F
zhWhtLHu9r;ZaUw~Y)PV>Oy)hYpTKr4U5*QCy>))%Y<z2E3+-|r_ZISS`@=;A-~s_;
zdQdHHivy7`aNRuFEi0^r`nDrfP=txgc6}ocm5%Zm3>d{DjQ%)ww{d{L8r|+mWwajR
zw93Wxjd)w2<bpipzR1U?h8Zu~LJ}o~&Lr7O_x1TG1o=JeoYk#in#^4@y8O6q%v9p3
ze(i2<f;N%J6<(}mqN+7@;1#9LPi(!Mo@@~9A(5ciME7SNU#?XwV%_4dnQ0a1X1rv>
z?xvg$6UuaAdVSUwWCKDTh78IL-BqcWHX!mu8YavVPXV{O7sfh!dNmO?$DMZ(<wAsB
zu6vWK4z*KuBRt#iSaLUw^uj@L<m{?~c*V8wi**+z!Rws|YaAdPXKJ)GFpjRa|GidN
zZ#}R5D+ueq<PiR!du0%opOE=KDH+}i>_6=T%sBU~_n*^F8b2jg`Zv<utV{!)?)UZu
z#h~a=B3{@%8YHy6dBL>U0fdhW-fkY2p801eTOkJ1>?FPEg^$=C@?^jY978RbfuA&j
zWgC7zHJ~)kTx;`<+Br|^UxWI$2oPu=OQZZ+o})otzal`Oy$rWO3h2xG4fJG)^ZCwQ
zjetIE;8NUwTNHuLVahAVVpcp0g!=@lQ44-929Os+>T*ig68Tp0!iJLk_>PdTZ|Ora
z@i`KmKg;A-ZZZCqw2Vh{&@%W;CDTQ==n?E(z%Uf8O?}#k+jHrTk2Zi51qjR#?Rw2Y
zqlCtFwmzdB&8h?IK4LYeAFE`CL)hX$zjw;zmR#Brd0Gs=Ut*AoO!C`n<~U?Z1tVk>
zP~k&9ZZk{2v*tn*yIX|a<O{mpP(=GYU)*hH?WDauM0aDwwx|55qzDPQ>H68vw>Amr
zUo|>~@-i^%TtCN3*h?I7evoh$?GDtFzDqatXse4m95=|ba6JVgTH|VQ&M!MBNpd9I
z^L@RHupUHX>oh$cl_Q{rLidz_dj#BYTkTq6Ctw`*%Sw?@SBC6(`*umLXwA?upm5jr
z2$2xQeIhjkq`n)@ap+jAyptV+KYS@&SjwUyZf<W$H&CY^`}hI_ry4SV)qBc`4in2q
za+M1Q9jaqas|N!M8sW|(q_OJYZOd$7yTT^%ifixO?#u3GxrcL&UOJvrCO7IsiFvw5
za$zrsY7a!=#?KndPV`56s2vtm3+8T%Y!3DAwkFy0gI}?*=XfNfv@1s3nf*fBSL1m|
zC2M1>0mBLrE|8NKDq=-H0IcpjcaPN?qN|OBKg9=vPms2r)U{_c^;+jox_Y=n9;csV
z%X1zUx3%@WCRY|-r=&82rt!28%3Dx2TT3pm-FAHi*!*=#V#T`NV|(Ob6&lRPtB6oh
z?kX?4v2f(aC0bMvN}DR#UiNmxF6e-@b-BZv&r-vQZuS8vR}Y)mqt$A)@6P^qcD;aA
zk4+=j2d$?1(?iB`0<`JN_JY&;M4TiV0GkYeqj9_k0pamWY9EQ<nlpT!3ouBx8@D}F
zJI~kIfHm(fx$Umxu?KV+%$#zNKpcVXbX);t8j6=Z4fQpkq2Dy(2&W#1uOjj9#ulI%
zlq%okoHA)|Y$GQTNX*-VQ>FFKYF=dKE<m<JrC#o(a2JRuC%R2=bHFyUKlHpwnuN9=
z(c=lwlyUmpw&JLhuD3llBy`kpHgtN;$T>Z9@CjLL*Zi{^2!LbVZdrhHin$}D$0Nb4
zmH|`Qj&%-E6qeg0B&{JWD0J^$sGXq3FafDGYsolj1x%Z#+Na2Ese{I*_b%6cpk3Hy
zrt2$i191U|<DPem8?Z~aO6K_2b18@0QJwSz@)dKl1B!sm@@sn4T$J3RI>(a{U|6w7
z+|Ny^x6I)rL?xtv$eA4_PHXG<E%f$_HHTZw-J($H0?9C^Akcbi&}4I3-K4-Wb-ms9
zjd05l^6rR>9Y$^vi*8yhd)iy~dV2N;IW5O0)Hf4=X0Ll`0ig(vaU9X<p`q-BV8K`q
zDt?<R2l06t27kOW$VDrdiQgZ!9dJiAxCXtW!t5CkkIx>k;5uB;a_epw24XInd=cqd
zBFjovT<Iu$6p2oRFtcAjt{5I%5t3aWVgHo*pgb@n`RFtIc3${+lXF@BjF`^Oyp%Gg
zGJj_Bx~=v0(3qP$Zl&H1+^Ucx5s*GUEih&iBbA^C{fYEF`L1rWi>hG>YQVxd`Vs~*
zE7E}xx4gQv;pX8#d;8&n*5x>5>Yzst-mb1mqh<@^V2-ECMEJ95I`?aPb!GLb=#Nmq
z+_Ass@y;!JUneTe-&w>)+*bgG(DwS}VMv^{=nEFTNLPE~tF_<VhNDza1+v>M1pE(;
zzEjuL|D_8bN69}`2sAkD04uZmm`u^iW-BiLRv}#G3L$uRTztu?%?KvS=dUylae__f
zT9@wy-kHgK38Vw7PYwvLZ-hZm+UE0Q>S}?u2XsCjW=;~|{h`PF-X-e;A+XI6vTDWm
z!823>K-;3|6>z+J=>y?KMRVowaa(U2ZwU`yc`GMajliGK-c?Sn#lp;gD0mnEBEBNd
zU*xe^a)Frijc5si8rL#MP-?9~q1A$As$E0E*k{l<h=(|65Uu@WZRkHJ!$C4H!vyAU
zzVg}ZN#HyKn)J;?9IsIKu8dphk6QCQv)5B3T|7H}>W=m>NcUo&S1aNG3cKV%04w=0
z@=c%+`V-QTvG#@-YSpczg~%_qXFS{Ocl};?YH_;baX$@dKQ*vuxo&T}!pPwZwJhud
zK+3ZTv@FJb<YI`7vZA4DYt#UtkOuNy)&<^z1476;IoHh%0K!Im&Adh|=|#WNj!)ZK
zXn@`%JB($x#fE#L4?2M)@b)Y&mYdbZ4*`ML!*_i(<<I5QGBD2wXj7JMk28eWg<Bxq
z)6)5TjTj)yOpv+gjM0)*(c87%o+G~EyXVua1YX(pU``B2%0rC%&~AWBoMQK;!Lwuy
zUUZ=0npWaoDW7_5ElMqyn>99l_{tR$_j*9^u+@)Km`@BL{6oCORO&blzm{J!vbl$%
zLONj5Kv=6vOt5vszhWQVAmBBCEbnh2t|Y{PQV2JrsDVxmr_Fp$GKHj4U^Vh{;3(vY
zI=Ba?&{wkQ9El2l4pFJy{Bsad*;`;KgttNf>6~<mypl>)BAf33Pj2AkYJPxMC@Tb!
z{`kPF`7gen7@iRB*W&CP>N-E+1-_f}FF43d?cc&$udkTc6)w&XOgD2h>hzwex|;u@
z`$|A<=I~Z{Me6*3@Cq;U{c6Zs&E{xSNQ2K_FLjQQeZ4ZcgVz(&94YnXvw?tH^Sf*c
zv3&dbiKyHo{&?si(01-+-Uw3d=lEuTp+y|;0NUieqxe=iKz{_z9QyI`S`c1A7KiVe
zhzNtj0-J^D7IMzm<{+qo@0`))b7!NAIpG4n*xv#!es^CuZNJhms981f&Z^n|03?M=
zgTdd&TH%3qaz;r=*mX`t=w0COMCgZFwiC_Z#tNY|*vIJHt)&!;Ez{TPW)D7YvOcWO
zt>2?UHw4RW5oLaXX6W${F-9}=aZfq>ebn_2hl}rWdesJZ6cO)g;^RrngXPKQEoVPf
zCf=rU%GSp%_Z)mKkb-RJM2Z2?p_}zm9Gx;#yx$-ta|!CbqMH_SxRb=){($iOwb9`V
z6;eEQuV4xDjvgIfA!zV<xS}g%@bEBh3NR&(PLv;0iU+v`#An`{NlaRA>C&AtR*$h!
zvKW!>yT=1wk_^W~-smbN=f^Vu`{j+pTCwE-s0Kp&HY;BAsUuCRH=}^4Y5<<^Jqz#5
zy7N}90UZYo08Yp_KK%3#DQ#R;Mj8D0>s&SM=jr+9#7a0b|7ka*q<I(>e>!VP2vYvB
zb@GC%LCo5Br%)a6Gha<y2z7};2>}UP=oXM1#B)xQg;r1a-nuiG)!+X7TcoWHo5|ae
z7&f}g0RAZ$=NLe}d`lF1-3Yg>WP=}PmDQhK&DR$LD6{{KSM!U6Avi6-;`#%Sw>}>>
zZI>5#R;E>L#MwE=Q>@%>>h*RR3CDe+>zpp4Xaxg;+ulC+<pp!uHTYU4H8tzl^kp@!
zGbjbc6!oaVQ>6-C0K+G1>pr!Y+;kF4J3uq<5S=YFk?T!biA>jC>GC{*P(6Iqh}Xrd
z=KrJaP1@8(lP}&qpQ86X|EmEc&=MguAqmYA+DT}h(JVgw4Ki()?Y7I+eS33jQICyD
z638SJZ^Svj(~G4mkVp)OpcE3&*y)8y#WA@iSsWrN%lC$FRrknX5BcQ;*snHEz570J
z4W+srvQjR2A$>km7>+?eMr)qiYTaouW3Q+Co)|P4gb4?e!?avB81Z;9Q?{V$+C~Ah
zdUo?CIC1H-QcJyG>>%z{JM9+vzOk`m6+z(5xh{7T5>3O-Ct7U;#S5xI@>}{OEZxP?
zV}mV9f=z1=#;`=_rSz-IcD)16;&WE`PC@%P6n!@=WPUX)T>hqdF-2gj*F&~RsWM9L
z{kghagd98@_%My)GM)*%We>D3nR_QW9>jUFpU&8~V-P`F6Mm?#8R_&(@gfgK{FH><
z0mguv@^BxA>QQ4?PecHy;<<jLWynLQR^LQKpEaxH0r@djlny-PAmX4yZtmg%dz5sO
z$#lr1b~bMbsfx0VaME^oJ4@&SRDNOWMy5WcmjsL)GJrDmOHHsdCD2iNQP0;945+CI
zBNDv7i03F}&xfPvO}e+e9*=FmF_A{UrIxMts>1J;K(a2z!KQ0Xmq9=`TG_L`#WJ^S
z$=Df;$5cZ4(w^RBs6O{RYG#N;>W%k0U|DHy9(zBwsFY#Kopaxd3xB8470&OcYVw>v
zR4*fXj9v7hOM6?G&L`}bmFtgHD|mM*<u_$2{c9u&egCQ9ee(dmjl%#?l5IYD0Q+Cr
zAHQ2de6uipg9$*rGT#Z4z%qxV$q$v8@e#KOv$OXy`on1iCRT6~fG~`)aOD%p4uQWv
z-X^m@IYi$+Zx}6ZzRVV{msR|P1pwIhXVuvEj)A{)aP!Q4Jxsvz#B&r3R!8=SSnhMe
zrvuIap5<>!+^;Ge;zp2lu|o4SNgGq_Um4a|x^LD69q2nOHU#}#u*fd%6qq+{%^+4|
zK=LIF-s*S6#@1Mv5qD<mkanE?EF4p69M_q<ZRSa-mIeZCPZD?Q?5uwxWG!tAVlYm!
zn7^37H*>PKa3)xLb-k|HnzH5%(5}$vA<}qnEAIZ<B0C_5ZxS-aM=<OUFfh;>M-|4$
z$W2%K90=@S>9>%)Bok~U&7=iy{|=)yj5SejL5yT;A^inE?_5vEt{l>-o7>^ERJlFE
zp@T!`vwnoYGtXKV<#azqz)?Mdnbdi}iyO0A?$oNfoVn$J)|Y0Iqaz7fK1+6ogNT|9
zr%lirr?Z3OJx=X{mnB$ge@>6qHv6iY`t(KppwkD%mtr0OG&g7@I<a~6gw;aqxtjri
zw_mxtKVo-3JHE?j6Zb7P=GSaMIEY=(>E=dzmGlf9e&rtY)P#qi$QCoH@C%_Ht_l)`
z1Sj|$KpBtxS$wXo>`ct-dOz==N(X)r8!#}jZk>7FcblMY=We;*U_st*QK{lY3|ca{
zA~7Hx+VZHB^bdsvcHE^q<S;9B3F|R!@lNRMvfn#ooox4AWRu*!)i$>jr<h<l*z8Ff
zt?Oj_r+eaKWfg6=gt#J$;B;CbS^+heF96-`8wfklN)aH)2dAo>Qw9_gTl7d5#WpLL
zfY*9UBj74Pkp~PO+<1G7iB@gbat~WR-__(a)x)x|f)mmm*zG9?p>cA&v6V{6b%rVg
zA29_=V!)+|0A_d#a=KD9BNVmUAtEy6atqf~7<}@C{5^C0biObUrz3y#Z(j!&pBs)p
zl|FWOmp<a#qFcD%cYK=nFom>!-M9f%3ntzlu0_}YD4M0P`lVzx5H<DLfH4>OtvL7?
z0?adGT?SFsKCB>b-41Mcy^k+|`~N)ZFu$7N0fnN04jydwxwgvX$07rk8<0lv`gYr_
z|K^$e{6s!h@aO;ZnfxL}?5zR)Y}k_H3Ci91Lj;;!P6%<e;mXd1oz|~xDy+bew9n9G
zttL`%PH^Gv!;E*q9TpqMG~)($@oy4eKwe#aK&fPa|F6p7*s~749bzp-)0>WK(QwjS
z=XijVcuAcTts{J(Z=9}4CzY0Y-}6^}Sq?7kxluYY8?cg-YUMI7jgNLu(_y6NZ8sE`
zLmE0lJAx_@Q`ltOMY~Ae8V-g!xOV}|6xZbynHWZwqxuNcRrMYLi{|c3k^OuiV|a60
zQPq&OyFt9xCG)B=5@d|XZy1}Eke2`=4kNtmv<~*nodAcFi;_u;X2krRN(k*9Y^vjy
zi2M@;T?wM)RJVn5S@b$9uj&Od{d?dZsCpDM57XT#I*yu>v)LNV=O^mL6|(nzQA0uW
zbL-*&cGFsk7)`QQ4@8ZahC}JV$Kp-tc01U?b+Ql@JmoVTq^976x!U4aa6ADh_z4TF
z<DR909sW!!d|+ME+$>jdc}|)YM~`)+A$lTaZ$OZ7Pk9Z}L;HN|A18Gzs{6t}fsuSg
zmMrXt@F2Z9ubifaZ?EG8a|PUp@eRaT20$mF(@5FX-m^H)v(9Tf0eAr8a2J7Ja|EB~
zaS}51{*VVEkW1<KES{^n@>lBNTWV61XNuXsp40|s!v%Sel~qo}&e{z`>`hq&Kw@IS
z1=pR6VuoN~7||Mst#yHYGvrqEGmL#7;ww-uynC?<mT|@eVV@iGUXt5Ry-Q9oD@}}D
zG5L$>{|2a)w_4*j%-NsM{8#_A*&BUl1cCb?xo4?bG_g}k^SXQ;<b~g#`O}|pTy3YO
z*L4twA$$QY*xFE_b_N-z+_9%w{|ui`Xa<Fh*Aqb}I>5nB;FEy-_4zvm7R;Ky6W;=0
z!7f1EF>M0ly;Cjo0LSV_MPZLK0e?m(4<s)?FEQvZ>}cRc-=x!EfS$LCwYg6A2ulsm
zWaIOLT9k)p%gZZF+%A0q@R?{<BMTyLt^v<@J8762<W|3)(>^u1*K3zd4Xc=Hs`#lp
z-(uhb9FgKBs+)+rGvOTGNm%E``9X`o^KncENT(Q&<*gtioB#^uH4;dj-9q1gIh-X<
z;zKyBg)J%~c5pa2hk?h`x!AlkV=NQP#-6Vm%>J(DqeJd9DQRKtNx-!&q*=sNsXcGH
zK9LA0xg^HjeQ;;su99gsG|eME_R;CAfuO=n&rSZqxskVY=t{Dz^$teiwg<Lmr)p$~
zk#9Q5n)zwZ(tC#38?s^vN4SU`dl!h+!*i!5ft9AmcMNwX*YKOwy|%dJkSm}xB>sBk
zu4iIUjBFWF^C&*@rQQ$qgFbW+Yc%*LM|BEzP+9y=#dGg91k9mX;kl_HrO>%ak2W?^
zwt?};eWX{&z}XIPw$;MYwkd-rkK~qsUf8#Tngcep6p*jyTE>ic^-u=lhqN4k8kY?O
z@XZM8oLXdn22a2|9ISdaF@dFA(=M98cNt)sr_u32Kf@Q$AAzX7RIk8t9w0Ke%KPZ7
zK;wAhZY@Mp&Mu7Tj2d(JO9{DU2?aT=ybOAct<gS*4m~%azc>YN>>F_{BL_aodMEgr
z+m2Ji^neH4!;kpY?wZX|+?AjPcyAS1bKl1`AHq{7S_r3xn1B(-(JfPidJ-4?_WDYA
z+e@SN51KhSdjJ+Fdi*Q;KjQ_io#xwQ2S-ur#}45;sqsH+87}Ad#$)=Xi4dW&-fA;s
z@4SxQ&mXfl=mYd|=buhaANn{L9)C2Lmw(g8{pIb0@eS5O!0rS^BLA9XD!xwTR|ywi
zeEvJEPv8!mHCVeAj48zQz53t-cVk}Sw*>#a(fGoJd@nRK-ytBCoh~I;A>Yz{P**k^
z;A(w6a)GSv4o5Ipm6+)prVmH1w{iP@V*9-9cQ5bjtK0l9Uf$PNxA|YZyzt%g0eENg
zQz!p#`<0LN%B|k#h>~~Vlf$oZqD_-pgk&X7nb$hX+<9>H%9Yb5cEIxcgnXsjA%(zT
zODnHFEI8Qkyf#b<8^AQ%<?!bhGCUsOegP~^6E{d9E%^H6A?H+zQo@|n%4#{F=gMCE
zIH#9%wF9d|aXBhYQyx>ZJcaJwiiH!LwHaP>qCEh@Q~T=mW(pC3+MXvL6)A7yE#zM|
z60m<Z?NVd)u-G6^bfY;G5TZUP`=(gfTDE|po<E;0*TK}i2?+NvE*9VqQLU3Pm)*jn
z!RgaBgG`Lq6`=4Kzl0rl2w_}P(i-m_?D#kv3W2DdAPP78c(9+A#%m!Jw3jDOaE*FM
z(vj%J7!8yAOGB#!nMZ8fS+L&+zQ)N1whlvYf8MQpBd5Jj-#V)Pa5VcN@7k1@V-Rn&
zLY39Wh#BWw%GJ;6+Vv=;H~$3L{JisT?)#Ai#Os&KhJ^uHqSI|qd6NbotH+m*l~ko5
zDE{GNME^zeVMJ6TB582tem8GdWTucAg;AC#Useh6Uv9P62xe}j!*m>vX(hcZj70$~
zMVt=|-8a`~yH4Dc!FBb#Q=SkCJ!<?u?m+YzMg4eYVs|4TH@QCS!#?7QFdUP8sabu{
zJg&ZWr_w9gq#Cf(E*Llikzspc!!o9MwqCaqQXC9Ba|LsbA$Ybw9Ec6Do^&vY({_EX
zA&5HSUI3<g;t#HuFX|0F-@`j-&z1Pr#OoDd*>r1Box=0>-k^LEqI1*==ipj$exWdM
zx1?1bLh>F3kS@D}@j@+D)C-~pO*eRDz%|VwUJYV%;n~incgz=J;AO~neeN;Z4{(=3
z(?c1|>m3v>lG|F(9mB+iSSIz^`8>(}8e-0sP2iW`=3wv8ssH;Im@xjWw*RvV&$#dX
zYY~0jX-i<weg8)bc50rl|EA!2iy>!FyMg@WLuGjZUbw4+jD2&j*#7H0>#a)x9|8UL
zT3Gq_*x&}R!6A64K!D4B=-q!p3V;6F=F2~R5hBA{<^>F(F9zyQA@WbR<+Kc}ubKZG
zUxAGi1j8Ja-JQjRHS{|Y?jykcBTfo{WClDS96>LiN7q!ozF)$(SCIepZ&3Sx__s~)
zz9N1{GW*@W`lFu8a~ZZYLDxyfPVOl=_?I=yB1E(C%flHq88036L}Zr3^$H`^Sd{zG
z*s@7d1gQmR1AuqUfr1riG-1+JR+Ai+uO#VWhdZ=kG%>z<GvOood}GB8TD5Q(qoQr=
zHM@OLmu42z0IXggedTD_3up=gY20tg;&KOUlmP#htVIR6$zFRy0%<*GI^p5hyZzX0
zI$pDOyqVlaVwG0%v8B$yFR-Z74KkXx+4)KmW((A<$Nm*++CI7UXHfL=ZC`9PkR2za
z9ZMq+z<n++H@CBwguxZ|2wR=KZ6PGFK6cj0Hk8X*b}pXVW$uB*b^~tsQ3tJ|u1*y$
z9UwLkf^rJysi}T2R!p&D@y$|IEU>n|k<31#sD4#9e?zDIUJA7V=ihAXApQ?18C`*C
z+v?_Luyxt#<MH>UkCrrLI$76lyPhDEgzzlgstz#QIf<IT!mp~&T~CkKT1{0bRpbd|
z#_g#yshD4w{S$fM;WasimuQzK9Y6+yS#I-OCiKVcMo6U_T)X4u(gwYCybr`#2d#*n
zZ>h!fxzJIY+y3>k()n06NPqC>^7ed*@bidiz=o~%&Ta2kx|#U!pq!YH&Vj(R3Bu64
zisAN%KQ4Ugx>@T(b4bV{smt6>d4-eaRu3VLl0z<O*>Fzh9(k>ut?kvq-Q(m@agE@R
zN*Q+C+=YqJ?29Q7yBqZqyYj>}RrKiP<w)LE*GLQIc3W32lFL+D4Dv7xc0g?}hr|-i
z0Bvwyg0MYN1KDr4DcXpU!|F9WTHax(`0q;JX&jm^3;htJz<(huA6ZlXi_N;qZ@VUr
zdiX0i*3Sa_uM>%ryFE%Ce^(rTQ!~H2`A;)7j(smdk@ai&6s+xhI85Veu%ASf?|>*D
z%TVh(gZkS@rh}18%jwfcVkAgm<(u_(bAvnC(NiE%7T|~bet4h1u|D7_Z{L}-4_=%+
zAOXK`dxzF)Xvsln6eRcM_m=|50T>~D?QOA-#+U)_glWHf+qWzTdfN+lJ3eoNsJ?$6
z(pk_sYF~qBnUV$<`vk7+;tt3pkAno)?<^P~N`zew&H3-JR+k2GBA8^~x50%0E(W3w
zT*!@*3%CsKJ0(wk2zvSn-~MyX6R=yy8#(2Nqc324o|5sAVZdD$QbWoJCkUU>*6w2B
z&vi|g@G7=@pdhu54I4v)J`+N(5(6}8$<2@uOD8zVm1~Ft)HM0w_mt)C)eTKiIc>k%
ziO{FnsLr!wg07IERcRa)w0*1-D=EpocpeZY<zwo}GH4&goJ0wObH~VYK{~!IoS(T5
zxF7XJvVtrt3nCVIh+zB6Ks*c(8t+l0icuib>L{Y0UKk92Pt;83`-Ee=8?_Sx(OJ!q
zHq2&sTbSpQ-(1cR47>ti&58#fO8BL3LqxfEeJdOV8A)C8m`qO^k9R?Npr72`l()og
zL{;(xLE~4#ueL<GLT~4+!lo@5%CBUJ@??iwDzzVR7?xEUrCsFRQc%omzZZ_hl+Q2y
zG(nDC=-p5OJ$C4}I3nb99b~q43!!-o?nVxb$10vz4*^_UE0f42zN2pD$oiJN&P(ii
z)o|aJzIdh~DF~sSC|L<l5T%TuJ*~0k>Z^6u<^Cn;-Hqf`;bql1>gZ;;2ov^Wq=p>k
z_uw<3{MRVb=RtxZh^)A;*DMb)H}N?00mUQcRX8q=J(;U^tp*mNae_8JXh%yBtL>GN
zCaPYuCllK2agj+Ltc#IA(Ew+QO3oOgd16VWpfjfMONgS-LJMpj3rI{5Zi1lZ&Gaf2
zOeE-`**d&)a|)`06=<^f;(kZx<0-T3^k%3$hZ;E>S&ywK%;VP$G$_45@dVQG4<cW<
z`LbdW<kS7VjRr};<<~2jANqbB_7r2M==&3Rdw^$rdJMb*x!Gx3NWjf$KQDj@e3MVA
zw7}iJ&C~==1?=RHI-^YaQD;<nEv)+ft)z?}W<(Gl0o@;514&0G7pP-!_08cxRP0aI
zTOIW&r<&gKEARBF=h}S8rrvt2gq}{^{w<FJ71R*~T<`^OyK*}1KJ;4ezk?3z{?=)6
z^QX{@yMTP^G@qC^vF`-NLx_|3=>Yq#ra<rq`v4tMjas3*hs%;ACW(-8czRA@drz-@
zLfOVTxH&VGa=6<rjI?YBY$yQFzH+y<g#1@`j~X~Qd7g<>z$XI&%&6Od$g0z9J;-T%
z>3cJYl59dm+?lq^-CoYblHME)L7v4%KVXI2O}KbBG7*(9L1NCA<)p)h#~uq%RSn{0
z@7*i%#R``Red2w=y#PD@v?e=Ub0RVgL(PZ_sru^11gNLFI<H%}<kyylU})rGrQ)qz
z=VvG4z>&_^1`mRHxI(g@SGB~xWVVGQ6Kr{gs3_dljGd1#X0nerV!2m#(oH!i4*gIL
z27ETyBfi`yYZ0MA)DcRMMC*+{;4VvXosyHOqC8O;k-6?jsqZZ?lBc_Rkl1q5^v3b%
z<><3M@%MwV$MZqn7N<;Rd%c}bEYsi5Cp7EQb_81G6p~L7Pk4o*y!o<R9ECU(kBm4x
zv7M^};FXU;%dE9}m)ZD0?ZoP+-sOgD&Sw6wW$(exNdVJXfZ_K*<W?@x+Npg;_W)zn
zL!M`;Y7Aw+85(`l`^Hf?*2UDs$Owe`EsG|`pvS}kYa-lTS=iUzp2a|hX(wJ+*SY9?
zXBQBs&6e=C+!PLk?LeSbuMp&Oh0Pk0J>cMePT1G0BlB^7JpFvMTFFg4fs(GEBwdZg
z!{d$@@QE=_3OBYHA)Q7#9uB-`Qiew75(osy%W$+Gez<+g)`&Z7P!Z~--d|M_xBsrq
z5nfT<M*q*FslMx5{?v%D$M>Lzs=hWNkcoEQIq;O-1Cp+QFvpg<iho6$U?4B@J2w;5
zCy@mTjP<2pUBJyG8FtH>C8&y68Jhb2_h&Y+2)zl4#OxU#6Zo$&FmLbZ3nGN&w|R8`
zab|lxK5hhE!wT6|kp1=JcmC5i-u&r1e;*wAhkgQ1OFKwRT*$uN2Mt@<&;B8RV^dtp
zu-;zSN$$2$8EFuw@=&kG(q1S4FNcKrdOWpWSF)&R!8-8dYf(I1H{evhEnxyt0+2X+
zyFm0~bA1eWUg@r%_*b6dhf%B}(Yfv%a$Z*$4!?qG7esYRlUfGnA(EG>yAJ$lYqNc=
z;iEqvr3pC?6sm-yOkzlc`VyH@(TxszmJ2Xk*<W^?%{n3n2kdjlx}$%)R<hdPUO?rs
z*fy@+^F}>0>^*38ojfIlcuR}R^fGJsRM>`(YF4EMXUqi2@ziU~dbW>%;RtSsCQ7ze
z8`l{^B#mC}bEzocAbm3*`9YECZFCa#3#NE75Gm{@FLW#TnLy6jK_&?qA5*$+!5}IO
z(1fLe`pQBUrJt}apn~%*i&*O%TsNR26Qd?{B)XybYTaH)n8QgBc4lUJ6GC7vr1PKy
z!GlHRn8cpvLjkh3mBK!T>q?fBezOAYfu!j(zHU(N8Bdp0p>xG=8n-MMC%KmZfSH)Q
zfiXM>>)J=4+$E?)XR7y^?Gy?6kpto$)64Z##jn$79i=qOw~*mnKljwLoYAnoj7$K<
z$;-AF*lizYY{Yr)oPbtMm&2_tWNF9Uh|Xfb@f3yvt+7uf3bxV<om5yu(wh=OqfDXM
zvIKASj#MV2bE2;r6E`@5*V~LCLY0ABzAY8&gez`&Io%=H=re~1U@QPX^AqQAi3%3_
ziFX1Q-RyZB8VVq5a`T6|2u65`58bry&<tLVUPYbW82g}xLHC<Y1JSaDd4|y^3rUFI
za$C2XpFY%YG9F`C_XKk0VEGP6`usjiwg|nam`kPCex6D$z`$o;AsCV(z+=7P8457;
z-+g^zy^=+H(e&rMzL0MPi+kMV?SX^?0QTP3q3~eqJ4Gdcblvv@Lc%2Y9ATVfLL?Ck
zaNq&KVSc6^pfkNs8Yw{NAe|f%W!>gC9;;u+VthU3Z)o}sEerbaG_pBn_F)itm5{iF
z=rB9vW{|r4a>5CH8NF|l8N7m&oH5;Loyl?o3AgcRr7<TQl{+jPUwYYfpf90Xmg;J_
ziX7w{<bDUG57A^+2z#SNxJ@=2OR}7oELa;YbJqt{93=jxxU5|P=?4c(0ICgg@8f4y
zjBFD41nJM!bIIz_zm&Bsz<@5lI5&IP9`$+LIkVV7*wK6!Z(}eKQ+IrBQWH68SDAF;
z#WP-WBPt`H<(-kaXQWiCr^(KE0*00(b{N3|npKY;GCH*sT!8&lp%syWEhjriI-sRC
z30BS$<2v2zjt(+<fMo3EuvPY1JSw{J@@pyGYV?v4FERtok96!I`$)RC2%@eLS;W%w
zY})G-rJ71?d^48nI(^cC$;xgf9{6XSMjR1=S$s+fB9+PbaU#4qx5nb!EO-Qwtp%i}
zrnf*j5mkg%ohQbc-YdkWb^$>}`f%FyDTCYH!CpL*46v<2pb6d7U_D9CSjEqHF9ufb
zz|iP9jJab9<Rdl|6w}*my>NNan*15cb~1AJi?-S{hTK}8?e{viJxrDW0hsX-(Q273
zFRGpA%k5+~UaM1_S;v|iIbclWcm!@pmCWoyuK`$iwt}Psa_}NO#*AWUPTSg-kKks9
zx@shtJnu!=7|<=58A{#ONAmn&e6z}@xXk$62_7=uUUo+n^EOU*^H#3xwe?Obj*{4S
z7^>9neJnc9R*`YnFn1}u>@j;koQafDs~4CYefXXDD%}3{^!S{Du@jhn|7zlYotvtP
zx<~Z6F#NxN!M^a%lafF6A-oCmWZL{TPll_b3ZDhC=>3EL^GEHyn1?d<GY^@yO_qUx
zg3swJq>7OyKrr582Y_IJv&c&BJ9`CT;s~zk)Z_HkbD^Q&z{kvg_$KqagMp9vb4c>8
zfgl&4zTFW>8OEDT`Bp8_x-4oyZ&zO0OU{#>toJkbCAa&sh_trS%=ZuWZh$`ekgIvB
zzv|OrxY59!GN^6vET{_pJ_R4o0FpGm%$=dOtw0&tf0aGG#ho&27hzOWVDdL4C;*e+
zE62~<prr4A9?96UCC}RDGvqI?R=GPm%!;(+3ihx!Jn5|~1&c5No^H1AcoBR;GQ^ue
zt_L_n;Jrw}6});?xx%XA%jxdh{aRTU-W2$^c#DNbu%w}P+bQ5GzcG)61CRgbOK!{y
z2u|Q-e*a*JwGuS2`dj(@nPUo95tt99a32ves%jYmRp8<RaGUotxI@6oC(#gW?H|bO
ze?AX_%iwuB$pb)6$CFbuWWZ-u>?;s)kKjt^Q24P|bY6gd<f>U+o8Q*-9+1u}@Qmt6
zVFkmwI5oyAfRj%fcVx4rySMv0+aEimj7{-rUr#GUh-~@&<f_h@OAUS6!y6F3_DPvr
zseh0K+&zt+8eNJ~b(rp~WlfiFAmfrl4>$yqU%f^jIc9*|hOU?HHdultE`kva(5VrP
zPj+f$o+uvTsV{8SV}sgYTS%YKfLaYe1R_z84V2c>bW&`LDmdWmy+;i3r^GVc=|X~3
z<}eIbMC@_3*d;{dDv_uiuI_CBZBi_G7nXmtx6YE)?(BNS1~MdxVRzTh8iwl+qwR9+
z*&S@kGqx5q4!z5-Q#@XAPN}sDC43`<Jwh_cs{Ce4L=QevZ+Us!(jk#-7<?+Zp-1*x
zSlSvjeT;g*2@1i<(sh;ZOYCTtR;h!$F}|`H$U7sQKGcnXwB?kLzJBWHUMcM-vD<+p
zaty9X=(|VHU_c!+_YPPMbh6)Kc)NIcJ3u5#)a|LJ(8xPhTj$k{ICM(^sW5tlLW_Av
z7vL?^G_TqQoy!=HZdc{85Ke)5-~|UG0XKArPKcmEJ!#|8VXJf8pVlolQ<cgkD%wTU
zb8AUMX;*1D-bD=ft7*Pj!3*xv3APtu`H?TGvR;XJoK%1_r)>_=yzwC5`GvY}0R<hN
zvB`xQx2zdiAm>$jTHvw9V0+xlONwtmc><j6cmF2P!Fy`#B=hBSZ1aUY3p^3<I3j>O
zWtbnczPE<aGlwg-<?GY7Itv@_XnLl0jo77_ljwu`18yyY#$gr&Y+dW;FZk>JRhk4w
zkp$q)`%+!LjGo^?VGzafH|6D@RF};^EieD1x_tj&e_LMuTwOA_+H2<=FT`ytaU=f4
zDF3rhUVT?q9R*$59klnS_1g^X+|1Y0U%Ipn;EWxf`$mQQbmi{~sHp4!*MaI|2>#_n
z{M*Tucrv<Gz&7(!CXVMrL<Vjf@ry747((~Nhzy?#7tf9=Zb8B6NB$Tyqm2vX3Hppx
zR(F|2y8WWI4-S+3xj@T=-|U`}>fiN5-X<EO-&%n>v&5Y^wC$dPbmCP;oM-&-a(>QH
zBw-OzK6VnTiMvtg0uQA07HG`!Wg8WNwb7lHg9v9V0p`jR;&*&neo-wtAGu?FGc=e|
zF;sCbuw$#n%GKp4B%ksVl|dB!h$6PN6>z*NA$PqI$c;?G_&UjVN!LB45e|Kwf$6@?
z-l+N^753@2?H)1c%t(p{NOOX<48z`@i(7XJ043?eRwu5sBhYP@Vw<}d8H}^vnr6r_
zo?ye(ID6Tg*PJ0gjon`5TnS`35+UdO&X6**Ou^DA$MvXGG&h1%CH7CO%IK0*u^TGP
z^8S&P0t|02@DXRaU3~THNnw=$-RjTF9;m3fa9-H!C=Al0&)74Vqff_kEgnvd?7jn>
z`$a%skL%_|3lJ7i&8ns#zBQrBnWp-e`OJ#}4z~xYUzHjHbuXg|;#IBGRS9rE@wpVJ
zhzW0IDfKmc#=|qHRk=;wF#z*{fGCWK;uJ^(mIVsh3MJKtAI)%%rOtpDnfOW1An^5g
zOt|5sE`>K%&Xs`r?kX~lZK;!$>B;8CJuY&zg69mPX}i)szQBoOcaK3!?9RQH=L9=Q
zO`I{;J2##0T3|-7Ma1?wzn+$u@j~E4{N3d0Q=9OwuTAEgA^j0I^pASWKX#S|Ot@(C
z8y6b1XtjcFYdbbn2rK^a5cj3C?7yFEz2{5+Kj<vKQ@ok{SFHrZzXKm}xC18;R%&lD
zweg{`0DZ;hUhm7E@7-YnS03#C;Dr1up=bw-zK<K<w?W7K%lGB)8t%=1+HwC~!@c=W
zJMQ=3?{{wtpq&5ieSz!jKgJL-@x}%r@U0zi2$C>y1fd7eEh#7%mTi$Imngj6vzy|~
zoSYH-eNsePf=bQ!%L8aE4kR04Va4j|R<~C--nhAc1JBCiVaXvLc9W|}2bgJ*!vhxC
z$K}$__$cTWGP-S0z~Qc?g?*-1*I~~X`&^vtcGa;a(ar}2-FTvi-c=QUEh&h*1rUNf
z4zG%ZsUdP^6hLveIh-cbTXQI`U3E@i6ldHIZ8bT0vJLkBQ{WDK!S2IQr8h`I!$>By
z$aADAIws2`?jqu2tONs*WH^yha4d$1<2#aHD<QqJL=Do|n;YjV$Ah$gnyz{(K)$h(
zoHizp!?`K+uWobYz?R;Cqb7P5bZi99RwHjt%u^6pf?dtty`dZ~ggfw?z;7{ej5UlA
z$HUS`ZFGnHPkF;5<HPbv=^>$Yz2hP0VxbByXfO;VnSin929ACr9?s5ALnTTlZt8x1
zC3pvP$T0H3=%|DRA;^9bN1NM%-@#PV&!EX_$Ww{zu2`7vq97`IFI$XMq1mF}KZwYB
z@!`x{n2|pl;ZWs|X{X>o@~&Y>i;oCrup8c7>s1S)RHZtS>wQ&rvL2NV*$i0WIW-ns
zx?m@JnT~ovKni!VO}3X7SkHron+CN|4cbu6t-VlUx|u!hU_jBzQBF3{SK#ZgmllnO
zASm8riyfO-T99zJGeU)N;G%awHWbLwF}ti=2I5d(unH~KLj4Ovlt3aX*zy2x{fQj<
zcdx7bl}h~uN==*}&j&wDX27&vH)7V&9!&t{S)&U+f&FJEl+yz5!MG73yNzFd!N>id
z9Nj++x`Xbs`!3$?gTSxb)^}6Ehj}Xio(j>K9>g+()D^NO73@)qFt5}SILyL1-+n>7
z)0>C}CsxS4`O;BszH6zbzwTHpmk$u_JI>8A&lc#!@&_uvggqoLLAvfIzziIlhGxCb
zqo3{Bmj|Owski`-?J>ht=@T`)5MP92AgB0pcm<;Tn}7e)c{MwqHAZ}eYP-Q)G=Ken
z>?prBl$$S{@&B%)6wX4tW1=F<J=)W`4MNcFGrml_rBQsOJu>McihNPy7ylWorDrwi
zd>Ok?jh(>4;E?N_v*)QI4!6OJfvJ6j0XZ47HQ{f`jSge~UEf}qqu?@ir0Z_;a<Q!=
zahWt*ky?p81#S+B3Y~M&Xc!vo@f+){vOVyU-?%NO=;Ykcos0Cq1B6yXN0xZGE@Pw@
zF&;a@+My#;hSe=e+bu__mf(jX_yX~?2U(2xLXZP^hb!q4yNTI)CkeDKxSVXLoMq2U
zaie&8tMRn2Qgo+1OH?(I=~Q8m<!*^^Y|kNtGq;JW>E<q-$HkReA>%WB;hp;mTR^A}
z9nWv##;UeY9pH0qE-N52ctxw1Q4}4=H9&0dCF7oOvuS4!Uo8+X8PC#h`+_-acPK%o
zW7g?zdK(LsV#duA1RM^o0;FB+o)G!=s9|BSMG1n1VGQ}?F^ht9-vgj@k4Lc1HKzoY
zo42g)uVPWx_1;hgux7og^<s7#X!yM&sGA|F7#x%_%h8~3?)pYXrB>Hk5C&hl$Us0I
z2PU$6dF%lem0b_f9INu)IRtX(n_XKaht+;U&S@q(uV%dLlW01gjgrxV{Tl7Eg_yJn
zsYx>mEE*k}sW=~c;O8^#GI1cF?zxFSCvh{(xo2)^@=1CdaJJ1a2tPL$K&wh?ZJc*|
z`K;{lizh|S4$p5W8RHOSq?#=RL&7<eMib>CX)QwkZAaPsi}7L3@x|2eEw(uN7l^@9
zP=gKYACoM<lXHC?j)L%0Q2up>OPs(0+swTD-}}FJ78$d=_o&ZrCP^B?v3{8}0)f*%
zP%YjmmKWx87eSvAD5O9Efjtb!pI`(2I_aOIvxj*u0Zy_5S_=-qP|jMo9E2cWqAxF2
zIO!jAyZ1>FFwkFPEnzD7_C<Y`s5RwbeR6r%Ku99`Yy^LVc9%b7g8WF1B*1YA;@Kb5
zh~D|bFnat<Bl-v<y8jeJ{}DfY`c$fDnsAvz8YZ$NIM-o-ugml0>r8TPO?c`Mm!N=3
zY$;{E0W9XnGY6y<%Lgdv+Bh9h1u!K2E8%l_D^x!0n=m$Pz~`5Ic^|%!KHyaSE%yW_
z5pOA41LXNO<clvMo$!PEPMG`m5yKL-+!HwEmIX>JlgM4wU2-Y!q;X{~WB^(XnZE7K
zg=lQCUgPJ6YTz{VJfryTd|zGy-D}pC1{5HrwyJ~`voJzJtnGFC=ENh=8ibH729UaP
zYdUy%A=~^en1pmko*D$HZ=Y8P(6pa1wU_3@-D6MgmF#Z~4HG=Qo>wLYK2Ziu9^_Ui
zfGpvp*zD6K8JL_rHLn{=VH{0?l3*OdO)RI!?cvl}q+!4|O&2vg<q8-CpI2bA-SZte
zlZn^dt`F<+j3Ah$^P?JaS#OWSk_pVgz;?1cuI#;#%Z+jp9dYLLff?p&PK|w(J+c!^
z)s#u#m;%;D9FcC}{dq`ZeR|y%x@Xi=-5Dkerbhp@zdXJPbN|0Z3^PwZ)gGF+0yG2h
zsSRzhqW~mxJVBzLPkVov63!luVFVUeuH!tId(3Wc^L+^oJQV~&_3jZ0#M;MZQu0;;
zhqyBy)KFMn1CyTd#6E0K5W%V9WgkW(HN?^7IG7inyy^qz!f;}|E?iegr#eyVrpCHf
zA-3IyP^iA&TW)RcJn;f}QPXwY+3XVG)>Vlx3_1gHSB9r~<#cd`vbk3AK_ok$zIyi2
zS1)-Zwi0#$4V}GFj$A-6*Zg4S4bGV_iQiiY?6m6R4x#bH#n-ELJ7Z)XvM{Hqa~Fgw
zQH~wXqk4_8BSA#)3VIvo81=T1VC`pKA#6B`Zzh>vDzvre-bUik5+FF@fz*2!UQ<L3
zn$s7u$Xoy%f{!RD5z4iGLWZ(<nabCWOkddFRcYY!c5w#3TmQ2%{b%~n?p>$*?>HX6
zsnhxYs!oe0Xx_fmX`Tgc?t9AT*uYVzEg)qyK}GdY4IyC$95G+Qdg07{`dHF~ve&%6
zVY`2Q(I4wJ<jVhm{3ZTb-9G-fGWt@tzfuCf74ARQZNr*{&nphp?d1(Aey@#6;G`!T
zo3nGDBY-;!Zb-*|eZ5Y>6=Vio<Cn`+`NJ3eD!=Mo$2Wf}<bUh1kajRbQ;$@G98VoP
zZ5)CUa1SH)&+G1lvknIx+%Uvf)wIUUQg@A01-qv45nWDdNf*PH(ApDD90XV;OzpZm
zhiSPz;7()0PcyBE%y>OgcCtft=-i_twmNNOmS<B^it#rOAy{KH5a2MCkUAeH?!&G-
zK1+Zr=7fes^n+>?z|MgM%&s!#+rqguk_LeFyFU?ypND*WJ;N=s5GN5bq2*(h6m$;Y
zpw&+cr2^y+Q@b*7YNhPKdhq=b_n^cZ&t&%PAymS$iNbOZChE+w+v?gQG}6jI3ISk$
z3_^!Jd(^kw_%!U7UU{b8j7sjzB|g|*wzme9_h85?4jK@nVC{3---rQ_uG={-)XTR*
z{^79r<j(yuBM)|SKc!Xhvn;3E8aHWV;jyN{uV6|hI~wcj-`DYz?!-;GF=u5e#o|z<
z+jDBxR+u=JJ6@Us$1+!|&&@TrdfwU}AwI8;j5urVl@;B4^l=}8)=G?4pbt`Z^>+QT
z*;kJ-7cEB22&T50g-S?_k;dH~^w~n!oSK<9SkusrujzvBO5N;njp87q>5&}ux=~K<
zj&*L~Ks468VF~aQ(o90DadNB?>zQ;2z#Mi5&n!K*Fa-XwM_<DBCa_pptE4Y%Az5@E
zFVTyv`GX4b!G&INxSAr@O)DW<IosFi*4(eBt2WvC{%o)dO0g6l7ETW&7c~TX<5@Z`
zyGc4bI(vQ%rz8%kn31qbZtr=GJ`XMJfjQPBlv<BOpU1q!8sk@CKaCDVCJ<8U)aXAe
z<VMs@*(2)zcNP6x8~&$3FLQWX(uQvmg5S2_)pw(Yuc18uRSW)KmG)0;!DouRT>enn
zzvDmsx;_6fv+utt^WI<Vrwx6A-Q92c^Z&HWZ@#bdzc2IseET{ud3^xvzg-&uZM8!D
zbPiX+zu@)%Y4854k`I;T&5H;2kYdTVUW0}0A)|u5S7O^+#@-;75RqQQF}dJqd+QT5
zHeWk`cZYb10>59gz%7C3DG8I~BKRgnic;HyHCZ2ws%6qg1XEQFy9G!f7-5AUm)7aB
z5f5pzdm(+bU!ElI-A6wk?xj%jKzuAgV>MfQ)!(FGmYeY?qPYYNN5lzs_Au<En8KWD
zdjOJ?JQ#<1pt`)PCx$2VI5^)y8RTd}m&k8AEci8Ax(qZ)-AE!=0eL)1^;C>G@7BE9
z;TgCPR6&xAcH7Go#ghlN%MEZ>97!6rY))>yIi>b60x6j^@cPGLD9DPXH=SHtwgPE@
z$4Az~TDrzc%NLR3lr_`BT94486LxQ%ZL^Z%+4|d(|8_9{vs_ehiBHP*c?uu%^_2-t
zmkvhzS}z_4yuHsw&E~5>_pjcamXxP|3Gq6-EmaI-BO#c<w-Q6_B8%LWhF2h5j+Q(M
z;_n{54>c}mf;2m;>IS-Q=Cu{jel|jWg`jREjaCJ~=n-iIFA+Vq+!^u^&szN)uH7Sv
zE&LdjmNBR9bBe=fX%KxVA}TML>jr|m^E9Ma^@+Wn^B9EMGU!2cE3fMt50#+8(bne!
zu!7YvM|KlvWp=M~)ZQEGb<7{aVaKar+l*Ga2shb<QDZ>xBg=u`KMsgMP29G@Z1Y*1
z3G|`s`2j@&>~J;w!_#3i1<pH?(cu@tD^{R@gJWQKb!mdkUl0|tIx*s8JNdTE+nsYA
za4hi-LbOUXuy4}o)C@Iap_RSkeURPKMtN3xBP0K=<e%agymMwZ0u%I$A}sUE@091?
z>Nfo0kEPpu3mbCds|p=P{SOXWnYMQU%B0=DqhVh@oWuZj`nM^{+m8N4lhDX#l&wD0
zO18d+fD->MtCeAB7;=89yuRkNGQfj*@87{D$AaZ%?=mnrfq}>6jdB2@9&GhL5L0iQ
z!+YTI?Y1{!>X*;tOUwW5iEM)Z=9&BknD<L2D3uPXHv`o}cQHDXB!1@B{Z3)&LKh`z
z5JlZ|Uj1$NTo>5%Ue=P0$;%VS?jSl+`Gw>-KY9)ZPPo9zou6A)af?+7`o}$w*WMnk
z>Z4O0HqvSxX_0asZ4FWI<e+>W?_J=dEaW}doAYt&6EVat3;jVSK}xx6$t@!z<8AAe
zTR8#3O_M#g*7i`che$cG3vY$$Ns=R>6RRg^J;AR8OxUro$qgOfD1i0aU5(wqjdW4m
zJ4w%ys34FOwmmdQ_Z~hLcP&W+Tlp4<5?AQosRz>UE0a{J#B?TIYArsU;ip+SA(5}9
z3-Yw+t&QgnGB-^Gnn$B}es=aOgRTqU{O)#lRa$J?AM)LaAmpzCobP~nfSB4}*hI5x
zPvxxr5UtN6QZIXz0AJ;FHt|yUb$;Bh;^RJg-(UQegG~vpTA=cxPTKADI4Q{7QJB;?
z?pje1o`L?kHU=8qneMTMV!^qV%5BnYzkM8lRAg<<C*-hQ+h-?u533JJ86`)NeH&U}
z;L1J^_m+KGF~;2vdP4V}l25^F52+_U97S$g&19RWi*lli)=f-I2lyF@3S2)h%Rez7
zJ!2*h)@w6zzkp*Lj&j*Fct_*5THxm8dOOpu$NRM0z_R@U8`1<Ct7m52RQ$G00GTKS
zUY@lw<_Sj!4dgxK(=$|}5Ib|Nu<aWf3W+9owv6#(!VJY?*;9K55c<`BpnyT{t&-QR
z2t;!YdU=!zL<V~UbqC;Ud-DRv>e4B`1}Enj9AWPAAX(hNXvxZ<K3+?6bmQCcOIiB0
z;`RIbrbNTsbVXJt|Eu~IUyeXQd0CZ#{#@Ta?7qLND*tVL`<538D4Y5h^$nD)pIXxI
z>f6U{|MU9x<1_hB>f0cgvb;P-pU`_BXN~4I=C564nKw}OinU}r`o+6bHRLFBdu-kr
z@qMcH#%052yY()jv)QkgG!pvPLPkjLck=Fm>45(&(}q&Ulw^k7)=Bp1-zN`|da|oa
zAB#=r5sZFuR=TQ-giR<sWWL-Ib+#j)nA`#33s+kOkXD(@KZTrE^e2QOla+QE1ILvn
z*v#7oWT{__g`?thZrsx@Jt1m@?>EpnECQxJCfsoYswOm^Q{n)(F`3*p?7kML1Jw?v
z3jB;M`C7OmgC*08($Hb9G@FZ9$kd>(%50NKut~S*u&t*tKXv)j@kQ=2(AyBZC>5ON
z>yTUd9b~OsX=U04ips9>B+G1vqwL?+HyTK#1n~k}0hwu%0mj#pX7Qjkmf#onKRMg~
zRNubXFyGM(rQ4?-`2=e*HdI0NGS4+6&$Ac2ZWW}Tp1|I`s6k=Iw}BDHjm28NViC9o
zdesA=?99}jL43r_IyUU&5fdp5E>^5zz1j;}s{)LgS<h~>qJ53p9bNG410~BM#Wa@O
zV7RgOm|9sQd^%V?V_c)80&?Y!b8_WLl#Z3ENihjt%({GJds5OMfs*S~WQhZ}Sf^8$
zuoljT4!+wskZiegcE;(-pcR^J=T}sWd62|H#FjvuZHE2Wlq1zGIRQFo<-yi_kR<Ty
zF-quHOQFUR26RDsolR{JkDU@&`JF{IU@!5~<T0;%dpO5OlXuO`9pbggUQCV!G;22u
zL}>=g(|5c`mdEz#4-?h@<NBs{QI(C~+XlZiZ-46CSpCg4S-)kQ2*G`}^p2~}BM@VM
z?A(9}dYXX){L6mnZz|hgbYX^8e(u~Lz}Fsp*B-uBwRh77jx*@cAjR>sYt#O!YM;0L
z_F4SgwY{4*7&N~7$A9=Negg&lKpX=T^t~(m1_fPcDSq0>Rs)Sa1?22K;`gvT+E?J8
zIr0?QF!FOJpb4=1hF}|+xCxm>Rj%eGoHT3*ETbW+M&zjQDoAvR#dqUen8V0z3u_mr
z+G28I6Ny(*&5?7KYLi5rO9XxD_d^B=Vg9p;h|b&QIT&Xi_J)Wvh=D7GeS$jTf<~(V
zbeJF_pQR1S6>?^*xSm5v&3vj_wl#7EV}GY~Mt@XOUtGb0)|~9&fZCqjZ?=cFfRly<
ziDmGi9J1Hh_p8JByf=zGoIFyNAyaURa;6s%ysRx(%8n<<v0;hTv)8f;@k=Uek5RO$
zaKiyh;kXV<E;-~)3&jw!38nU`(cm(YcH!ZG-Vg<fG_IDfc73fV3ySp1zH;}FC3z~Q
zQOzIpSd3Lp=$n?hWXtJU-!5t2BQ+BHBZRu$NUX8<;h|;7B1Nbn0|hG3bEpd*u<7Fn
z<|NM6d`eRdz`v!rgdTfJ5jMs<3WnbrPu}AI4ppkaqn|aPG#7VpZpN>Ha#q{mr*T;1
zlNcXZ(W>PG(r9SNp-JYk)#*^x2ZI>V{e~Q8n$H)(E(&=6=iNyp1*WH8dxxq>SL9^|
z$XHs>tX_LQaWA&yiF6UVINl5X8LKaj2{MbrlKA$W*dPJO+}D@mF@ZN(`H<4owl%Km
zwM);G03>SmrbU<iz)d(Epr(79S+p#q<a858f7`J@obSg2vW2jTX0ugZcwd9Uta|)#
z)c<Oh`|!$%JNUWYWnm~P%5p$2U>g^>2J1eAwtdIc*+AIGU!3S)0(gGZWHJB?_!1@d
zvQbpK7kd(t+Ma-VoHlZ%3O!o-mbc(n0VAFz45BC(vnGG1-x5KWj`vg%q(t7a;G;j;
zu;olq{xKLvqyn~YRfoCGsD0VM*hv|PBZC)acC|63O@y`&;!Zx~j^gjsfh3RH9@UgA
zOJaMku87^nFRvN&h(7JzE+**X%~>B)dr5NTydlrnox*(io^MI51Ire>Y^BPoTvmbO
zj<}DXaAh`dWu);N*yLx#VKn&)?ej^`o)2?SLV_%}=<UXm#zrEPx%I0@K!((1!M011
z3cXF#Bz6(RsIEkS=wChVf#&7w%*Iir`6zgKfStgfA^UXCUz#&a!nN7rubb_D#sn8X
z-_&bsix*Gvz!K{;<Pv}yd0%k_B#f=vk!*L*eG}7r9gKp?ij<H$`r0A;3+}T`b3PXP
zq_Q{k0Ir-7+$`g8q*|FBx7dp`z;18WF0q;cK)qvazl~5@BOvs722`2*#(x}&C#<?*
z=8vDaL)zy^!zLSHY598tK~;2tCfxSb(Oq66^uf69>{D3Q!?RqAZ8LbCvyL~7QYC4E
z^Z1jTBjou4jLO|Va#tUFvLw1m@TK!D#;mE=GUH`8M(l8o&E;9r__=PWD2-Cv^Ujrd
z0f6^O&tVuBgBco@wjrx@JiFD{vUy!Z$qF{FMdLxMvu_B3PP_Sna-*}JpC<*fZw|Ac
z@9dmSO>~<hWOR)H+;l1%PM`1EAB=V1A2YsOhW~XrpMP`C_~RgjM&6g>`?vL`5Q>iy
zvKEo3)aEl}$-cQluN=LAV-~{oetju^Uj$(|{EH#VU!*h`qJ+P_6bl!dfu!9xLI{$?
z-p1S<LY3cctz3UIgkYfZ!4Ud(8~BTVv3-OT^>=yr0D=$Sz8P4;>{s|5Ne&R>uI6|j
z)(M*zF^GXlEHJP^;29PeG)U&<8OSXfojjgtk>6C<TD?Qp6HPm9gha^zHloPvK41X{
zi(Xkm2_B>>isYJFkC!98>a+KtU%S#)pE(AjLK0}u?l!w|em_Y?UUhWJMAw%FOZXeC
z^wFLu;*1-;d}Pp3+5r`gnotJvEKC;CSO_aL)AU@rAQF46`o0}tKv?B(nA7({OYk2X
zeg~5cK#YJ@+SBC|!}t?%!u*4yGQWW_|02Q|5{$3VTqR##yT+VmrG})r;*_7<bA017
zmiD_Nyg?nsyNKQUO#n-R5G9z@#AX}eUS_^zKo>}gBOK4&4RJ`hV~nT!8Zc}Z(MwU)
zvcfHi&c*8Vq!5=6s%ELPk664ExO38~gyU?%_14~rdmn+iZ#&$1C77@BCi+9x-Wjx8
zWA(3(L1Pp`knhJ)9Pov|`-uF5Y2)}_6g_WWr;QAKF@YEJ<5lOw3Bdp4qb%RFKw$>z
z%*U_g=CYJ_0BR-=Oy>ftmv=lJA!l>)K5{nyO(hC=LE;0-6*?%WME^a7{CjcI@{Z~H
z>_=sjvy}2t>+j#^&ygi(7w>rBkPpG_-oNpLUjP~qzRn0>4~31o0%rU#(8CWA|F_$6
zAc=m5?f`>>2580~a`|kBXd{S%2f>)N`6^TRqEh&3sRA&=7c9UcM;|YTw;^lz<C(TF
ze9iB!-Uc9m)m+H6eY4429aLn0y@B93D8IcQpC4Vl1YN|p838;P{Z=Q#Tk(^a?(&ff
z29vW7%s%~r+1D<oJu*N5`IVd}Mr4H;^Jn(9cZd^xXa^C)5DM#u9_x6^-gYf>t?u^-
z50Wj1=BM+fL~_@+qm%<FsbA^DlCWOK%!`qyM>H`FW6c|ZJsxOMl1*0)jwGQj7-XV9
zBPWZy*AvezVefh3@hEP@!`cV+mEjqB-EGeq8A@r$Ti#?mqX;;&G(EZ6s~T)TKkd`!
zfdZ9Am#5&WTJRL&NA>Vp)RO@J%I`PvO&ef>4>axGsSlR{4LZ57M9{LhV!@;njxh3e
zzmH0+*kTTSBND4QSS4{mN77LGY3oYn?a^&{DnU1o$w4*vHm8T$@6yWvnMm=-G9W&!
zYCqfq%RIkYqD!=xDcORj>3&TZKh6l$_wVI97HRVl7HfjoQR9*C%NKBhg|OdAG;%Dc
z-c=642-%$1i-LL>_4+k>;b+$HUo(QM-*koqUV6N05Gq42S(stO+VkmUPr=B0t|<Y{
z$;s$0qYY^i`Z?d^MjV~@Vva&m2E&AeX7^d&oqf$-tH%x29egYa)H}y%ain+Fy|(k7
z-7!sU#-pE3gAmpZ1ytj{Q?X5DU=Z5Sm0D7};YpaNeSNRMUMo^Mp9hVir_1iO7zNN+
z3i)9?uDr%5>hZ-Heo2C&d3cfTQU{Gl`1!^gc4;?>v~zsk?M8V$ZU)B<!lDJuG-asi
z)ZS!{=`GRrhyz3EhQ7k=JkgE5XBJ6&UNfKqZBK+PEduap?sW}aDE$n?15+DMFeq0^
zU!QUdNHb@fKtOWFI)J?noZ)3+ala?3q1Yd79wx2x-_8i$wKuDq|ChV>*m6|ewsq?~
z#j4YG6CfdpB35`OtZH}=$N)l^H@eSpU+fMdGIM6;O0)Lf_f$?{&WKdV0Fr()M(<z0
zRXvDhKPif@|M&m-mmMY5J^TIm`yVuXg!|g?*?nvH{4P%E`qMada=&A&f;eA6z6_%<
z>_TSf!Ur1jZA>y??E#k$(vKF6G@Js2P}&43SoN1tRTDo)RR9=MP(pwW0H93PA$SFd
zVB@|94XFSfA+q_$eGluH1-w$%0tdA#bcrPJP5p4&7=96seTY&&P>qfBosf40*?x|Y
zUtV<*7TWtfqTb;^_XXp30p9t=aOxdoH!a#Z91;18I*McPFIY!VQNcoaCFQQe>xZIM
z1J|Y?L)Zb(TvnOeZyZbziTYELx=cVRDt31Jjo9B_gNCx%3*dd;WlI*2{Y9`k!97!V
zxiru*;mK)t4Fw#0bT=_VMxOMxLmCa@(@1xTPR7>`eUdQ99r$G~v57QsK?|X^;*fP>
zbTUAjxrzi!<?bPFx!gNb#vw_kcx+}w9k!nK%&ntWZZV$)JNO~KJsbkqF_5frv7UFT
zP0p!Qa(>pC<P$>U_QBO{(C&OcJKnZw6T(MDeLF+$B_ZJ2)2g^*6xId<4ks3UJit}-
ztccSJ8r&Fj%-tPK!D==m8T@L_7sKL7yl!RkLK5w=_Z6<VRTQBkSB(Q5I{AJA1;P*s
zZ@i1cAvg~t7gZMI8y_t;-VIK$@Dg|4mU}WF4ui5AB-wPxlb9zNKyAom^LkOBp~s)K
zG<Ym=L>ZbiL(0$qO0E*L$o5A~*mJP#S{D0h$G!s;V8~7guMC{T)xFlTk1m-?Rg{(Q
z8}K$FO{YE}8l4S5&o)wKlWPqwQ8`A%OGfG&y--oR5{Wy1u*jjB_wQ8b78Bai2F~B_
z`1wJA6$8X;qN&Ho1hF#%!B8o25j{zwA+ojC`qRT;-W_l-9emjBsGJ2q@dnjGk-@67
zr{aLckJORdU4&}`j9+Z5aA98{<QBWh*fp1-VtcO2WBELgx9NEqe1F5F1YtD*lLhu*
z(d-IpI<ecn3-4&SwdXL~Z?D@hfpVWvuvUJLkR7yy(CkXN*NULRd%jwj)*y9%VIWfA
zH!)PdDlY&Zas6(lpLQlll2I=uQJ;61ZByfcq=Fh4mi@EsIT}p-Fwa3=?J5t_=n-;i
zTP7<YB%%P$;0j=b3McXttM!vE$XaMfUr-Tpxy;wQ?WQ?WX<nr?A1B;;q6D0|f^^z}
zI_={PyXqpAKovM4etS=)rtn&+bNVct*-+wQJ=w5Yz8m5P9mjXbGMDBcvNuoe<PL|6
zW{Ag!#s!=EtUE85X$8p01=K}^01W}loBiFpGNrP6ah*vJ6?=s(M2LobySJV+cG+#)
zJx&Q8upBz+T~9aHJiJ&f@I!LBDtCfmgHhaJ$X!Mus|0enjz`!a5ule#ABm<xYC3l(
zba&T4gi4kT8@AX<<k5Y@UaC+g8z{G8!6fc|7g{!d==`$883gd(*D5t-7Oj-w-d-PD
zZL5b2#dpP$V)Sk_pP8ns^Q$!`yj8qrcZ%Hse>x^ie1aK!LeD{QHt*x>8<AD~eV;^b
zf>dR0F00bNWT*Quzfg40llr3VtDOv9YQPCUc7Z3^LVQ?KT^g`V2BY|s3b_->+m?zu
zJkr)C<jD)r1BwN>a{cWNEb#ogZ8oScyp_PouAuS}N#hQ&L8l!*qoQNme)~-M6tbzW
zd)5J3ic^()EJpm?<8PaHkZ#8uK26o*Oh=F2jTF-=X>>OLMV^K})1`P_;<Q&)1BMjZ
zWUt(Cdc`%jCH5?a@zEfpDUX0rA4gM3J(fAqonlW`WfKF4%?McT90qKlsE|&O$oRoA
zp<*i^;{pn@bs&_XqU_@QI)ajB@DIh!+#v7bMF<V(1JF23<*d579{DG|$A=*spz){d
zo^??Jv9zy$c~O(ylPZF5=eLHzFMs)47vTi;VU`KL?Z|io3mV8J_-WDqk&yoQX-Tj`
z2R~9fRxS?$YXaXUyaBWv;F4)TI1L_k00{%xX4Q5Kp~IW&R&TvCd<IWz^q7rbn4i3S
zA0lf|+X0H6Pn#I*)8I(~k{O_agbO^e2R54q(xSfZx_H=Z(Q6C{FnFvlo34-B*3ia}
z+YtZC{PEELJ4e2Leg1gp|H+Yqx}aJw1@c|o350Kd6k8c?q=B={D>Ni$K>pE5X<q^F
zBySL9DfHG#0n_k$6aYKNWm>K?6w>c*##SN_0b3s1Lv>KMacS`v>|y~+r>L;azS72w
z6Du#Y499#&s+|YW+?4hpFsKyJ{=}oaVS*vQKAo$)Wv>*2Tuifr1_7~Y&TS5OjqYs*
zsU|7G#+4L7k`}>Bv0?*Dgi|x@Ps@05bkCsY)H&Iwn=)_jLA7A03^+Lp;V?da2~Q<G
z;1K%9IwZcob4JDhPx;&lP<I8iy8M2TWYS}yKwBB0A0jFhTbw#Z+H<(OJjVmG?-sl=
zM2@a(i#qfXpOf9;U})}Ciw(GJ-aeNXi$6am@7f)TcU_Bsk*I;(7>c3O`iv^USMhZq
zO=zVS(|1N(4d6f$2cILLhkOI(!o|04+8Wxu7VpLo_Ej!mb5AGX_6;KP4!8QK-kB~s
zlXoP&@$l&fW7TgMG4xs?yndwfrI3&UtHgdKY6C3}OrRr-6zhvX(4_}uG63kpol6bw
zAU?i3DtH38J$+<wyAQJ?ovv}~K<Q0BLd}8v{UufXE7sQnN-xB%|B%yu{m5w_A-5gA
zkuX#hv;c(_i+j<~NP-mfuW9K0dO!HLxTd2+DEdb%I|8o&fB2G)4)N)52Y<aI@PmN|
z4BX)LiPwLBaD#U=XMqlyAK?Yyg?MYtafbs~|IN%0w+{5G?~mlP-eEl=NSH6Ul0JPT
z#={eWOW{jE{P(B!{h1+<)Bf+LCVpI_&&c|ZYm|Jxz<Nsb8W<n0*APg@+M4tJm3;hy
zBZH*)^$&20@kg@!mm~W)ba)_4^>ESQntVjlf86im$ba#stk>)#>;89#{_7)OpLjjO
zkBfg@BktD&f_KweU-<*d{BeW;XZ{Lrh8K9Q*W~jeucyQ<>ze{^(sq5KaDBId$dS_U
zv4Jx{VB`l3G*F&LxPzF`YQc>3sfP#Ba7~gg^yl6Bgv)wAc(X=0<<)-u?OU-rp&<bs
z`|5=LaDgX)TT!8#7((tFuokWfJm`n6mDuAL#+^m6xLgJbgXF5jtkF56og)A*pbq}A
z0@UE!gpEbm(hz9<+Xp=SxsLuY%ztbN|2yOSpN4tx41Ebp`Qr?A0tCoQAORZnQ#byf
z>I!R@i1YR#4M$=Xkocw1R)`Z&zi=oZ*nMgA9GXAfD!w{~=ImLmIDzdDEPc6+Pc;{H
zIGdW0>kkArO?|o`Y<)!~QGLb7<xvp4Z$IHn>AtAz^Lb{^8zP>T`Volcl?8U<*5GUm
zGxc}RJHY<hm5zKli3gl<3QFE@5nVbeaz_M#X_H0`rqE=5IeRv&I%{g?u1bJYld0ET
zB{gP$R_qJ^#$A2K5ojQ!L*iWS9(u`p=1v+SQpsD=Zc>m?v+o<Eu}OQYs6%m&!FW1|
zTf3t!Z}B!#dt0R19Ee$u=mMR?XkRvgc@gzCVL{N5IM4q6LeNunz4v;Q6gN@PMLs#&
z3p5|C)DEZnA%EnW_Q&!H90dKdPht(F_-0Z5%c|->sI#mYnk)^zcpB(nuxa^SohAQi
zQRRBeKd?g}Im?L03$lj^hU!8|*8XPbRe?D6)sY5^s1MrJhmDJ41=qu1{q!|J=eu$@
zm<2u}AHFYle=T2sV4pQEnb#7?C;l0ny}+e`g%oZ$E(z<XAC7hfPygHD{!fn<`PJe6
z>S(_%p8m9ULOvHyEJR5uyC-e9WD6xpz!XDRAzkj%qwia8c5rQ8-=c2{xPW+EmFED{
z1Ia`M)3P#wrxOJeLHCAOK))y*tx7vdfuO=h_87AaBlJxc1Aj^YiOLIQ;aiYcX60gG
z^%9f3P*GI2ftt%GSbk3_nANUNYGRQ`ML?<RWsSkGGt0N^6=zk&l^9m`{yq^m@Dun3
z;BC#bz|O~uxAa|%!k0pBgiX1fP&vC+pga|Y*+h0?*xVZEjS9}jmVqNjN%Pbj(7fG%
z!+5iF0dWosIaX2)e~?CG5q1&8>Rx+8EnT5>>(r4GYQCin-XSyUn$=S1hoU=f3N*Y!
zjNaCt&Jgpe7V)my*;7a#TKlxC#jhoRkKWlYT3-Mq`Awhf$Bcy55c=CS-K+jvhGp5A
zJK5JO%lS9Z(tX>4Ak>Jh%j1v61MK~2Q}?!J3I^7=6*;<eql&TWZV%cbf|PG(j;%oF
zp|OF&X3!?<<Qj4;$bCrXyM{#|-@+z6(@>Ag4E9`pGsqUCcAWBLiSm+`4N}6D#+Wxl
z-!W;^2U%<#D80Qrcb+17`#scVI%W@{UyYAg;P!Uy3iaf6WAi8QF6@fo)+C3n+%5^(
zi+h4=%97$b*EK?CQVgKXlH@_WKWG18ED%2VewmBKwq6J!&2f$rIO*pKFEz@*9hXDS
zz3)*Ez4=7nqq+ef9dsv2PN_GRKVe*E?4HXOIGU`x<OjXm;DXLRlKYH=B~}&7987tA
z39GEXoYeqZmJ=E7XZHwjrfwWJ_ukM*?r=yYx3-UB_m`IWIZK8QfWvA#817XPz1C54
z-Jr5G@s~04Z#z*gT?fnt{b#`Zw_Rv2K=(BLX~0xmj&C&fgQ&>E6rUh2=5cGz^tYDu
zwpsDlFJRPg&b;d#h`0vq=?_gbFhN~`!}`G{_#mDJ-T<++<<|l8W2A*%-~6Q@4Haz{
z911iJmG0$BLHg(0#^9$mKXPv9=_l(NG|55EBz$@$Bta)vbao6SaSJ;R(6)4K_Is38
z?}&}wtZXj;;@BkYHdk;h4z18{J--jz?%KXvcM9L(6@F#W`s&a^Al<_=JY{k_V?TK}
z;O60wIsN1=Ap=|^#eeHP(LmV7MVcQ0Ald=zyWNrr1z}MH#5R^M_tY(MgQiauMNnr)
z=n8!p^Kjx}E~*ziD71jAE_cYbd9}D2MAn72AQ9QO9(VS=0ZsqY(>yS2>Vqv5+upsT
zaRTmk8`<7>S_=kGA>Qr1_7cV03uNy#`ysh@(GzOpFZ#vadR!=K3-C}av<goUqa~6e
z88{27?ul$yDd042KwAOdBtDOCV`<l?j6KAm-R;ZVQ|bt4s4w4mT<>|KmZLZqn`$%B
z!Oh@INo!BfrduF3EV{mr@e~K*23HW?cmeh%+=IkOb2q0Kem;*8gszWg>?piI{>ws^
zMBe{iw?(%?8>9_$DUS%G-0nEo02?N+9JmEGW*=#NKLj@ac2NDPgN}S>(b9}o=OL3^
z=#$LJ`$A<Qga<ro$)Ra_R&C{g=#=x*&$7o6!ZV*j@{lL)+MSmZdVeYW6&W^ReTO{Z
z&11eTy?Q2&v%V3<BR$`l7>W)s#}F51kYiw`W@eB)>LNcms8OR3;9KUy;vTLHBYXG|
z)W|mPQ6^X_2wgut4|CuU#hwDOE?XN!ho?z+E03=YafNyvXmAx&8E>L)GA$hgbJh!C
zVxqUptoKKE-y*NOed2GkeYiT_L2hjlJEMRKo^iKksX}ciX4B>R&G(}F<{)+C&CNHu
z_h8$?QLAPUlk?4fM`tW62nxBqy+DGkjbtaA^Yv~sZmm8K;TA-F=~V%4(u05GN=nIS
zPj?S&`=^o}&;A$vO-l5)flu(@i^_1{guK=Pd?$O#D!1kN5w;_Lq_#Z$Z$ThyNseWX
z-@y{Tz`L~RAEWET(Q4uTS%TwNmKFxpw6%Kn%PcT}bsKA*fD7IT1d_;L<(Ci;@bJ{m
zun_%J1i_F*eYj;WYY$9d@DTb6ArPy76)pxVM^HP390F1zm7gd7;-^h|W<HXPlI8IR
z4e2(W$#v`k|0%@r+(22uQp8$B{6LHWax9!bk|SZH{T$=h5?p^?3xk0?rG8U_=xKnJ
zB2*l0pn2PX()msZ;p<OOQgHA$C>{!HeP)6kF{~{<eZUA)dz|VDUsY3(TM4HPi(vQ(
z{Xph3cx}yBSZ}+!r0z7pV}vk=A7vbCE%dN*dcNjU&K8_%eAAyHgfx78b@=@k@o>jN
z;mI{@%~qXH8l=92FXVi#CxxSLUptf{zMj?vaupzDQuHb)vR+|I`_&UpI&?3;K28-b
zz!?W?gY-eaWfr44zQ3}?)6@E-)&PZ8Ox;5K+f7(G?LB-Hu0Tu&!`<gQ_~i{$);9)p
zU;-Zbd|`vo5oRG!SN!_zhc}{x%5A^KM?i=8Tlp8R=E-)om*wg}jsJtn<Y$pdKdYjH
zDuMO%C3^=kY$X;P0UIYXq<thS1sO?@a$~H;1gscpHQj<A>0;NFVCSqwOgCG8=nS_b
z<GrV5?eqHDJ^6duT<g%j?byN`Y86BUhpoV7kV2a6ZLZaGt`oaUE6>gZ3$L38U8R@F
zY|gL(<RYV1F4IBY^i$-CT2AUJ7#C9i610r;5D#~O@OO{9zeB{g*;se*I^>3OKG|te
zo*e={FPnpIC2yI4;M&w`7!s2$G#W?x{VF@I-n9qd{Vaf|J)Phaw#_o=k4NC1y>gIY
zD>d70NDSgm-j_H@j;{u|Lvb48hPvB4LsvP))q|iL92{l*e%Pp7Utn-x)s3EaJb5P0
z=rSUh%-!XGvd^yY>4pa1N1;CDEz8%KGsEA!Yd{Fs5QE;?_uti@<ZuY{Rgv=R=_~yx
z>cDKjI`|%%q1wu0`*i(HvtuGG7j}>5`x!vNC;u2VR<5%M#>FdxF$FBg3#hGu0+_Zm
zhP^Vd>Vv)@+n?6G0R!7?q5A#Eg8bXi!qNDTg;n}_s^G2r4~8h@r!ngH6&{6=e87iO
zH7)@s&(kk6oi)Y<_qt6V@P~z~Wd!V)J~5yuw2V@0I)s}5H<yTE04nOd!T4Q1$8G&&
zd{JKjj~_G>1ajyps&G&rkzDiHKCX?vRyA*Wgg~+)prnd@39<Hf*sb0#3sWA-eP~^u
z>G;|t%4@$(Z-IKUvJz0|sT`H#_HjPYV-!C$%{0o{er^tD-OG>$jiyI_tq)k(G+;C!
zh<f)Pri5!^@#2Y%D^=1PQ#}PVnrPf+YG*2xXq@R{K%G<EN#>rM5B?=bBfzO|V4;C+
zFNa1yG1vFWX$i-F2x!bc)u`}<t0|SxSG7uFYVI=8gtEJL?;9M8dMSRJgTJ{+X0YFZ
zu8l)TUixE&N6Z@NkX6aQXjp@R$0qp~&_@4dXMGlRN%=47-oL6BxFa6-PhbYrg9r@A
z|FVbqWs7aS@o#fHudS+(>08*>F9;c*^Lg{rmchoQ$>tdLb1;8Eq7zmE(Px6GrMUt#
z4uUgayWqp`>*U5VX9(74U}C&RD~Jwo%P>AeIw=%=AkN`KvBUnO3jEkYXf8dPpL09M
z15)SWPixQa8P4D>ARuYN$JgIs(E#FJAbnnZ0kR;9;_Q>Kt7<S%?SRyfT)q@=Y|i=!
z&H#Gxr<ofT5g>?!Xfd6G+>B2z_s4DE<^FoxSX_RTAtQ91o;BARfiU_Z_~e(@|Ki9#
z76cZgu&zn1un&SkG;;(RY#krdHOxobGejP(-Nnd<>jWD=h(LO+zr);BLum?fQN1#`
z=?`d4HUsR@V5bPuN{DfVDfsK=a3wp#sr`4hUEqiXeZ%m4M+SQ1LKOnDP<CuR{-6(`
zT{c1q=tx2(d2>m_o9<=~r&Eej@i-}V$$1O)*7Ku1ldn`la0{)Cs^Rsx5UE>{dI`33
za~3FJ5LCJ|<vfG=0>z??!~!%nQbVB4XY(ZMkbUd>{S8~h!|NQ%cV~)_11<^*h5A9(
zgNi_n68_%D#UkCVnNDqZBINs<Dk&KzUKh@S0<*x6yS(n?NfHU8rA{X$?1g%p$A<wq
z&Wvp}Eq5Rd$dU-A@JZx1=f*zb6jL8>z|Fo6>V}ip*MrKeHmaw&QAByET_bu@f?h%s
zp$*Qaz;l4h*Y~~U+x?5V#q>>o$)~evt5UP?$uy<G^3u#YB*A~#b{R|sQCj71%29|$
z!}PIAQbwpN!4v{ptRLp6=Tl%ewqBkqBV!foNv4IvuqI^cxg(ei*64wZfk&DxeS{~2
zy7ElklPe90;2JD8-H~}nyn6HM)Pc530Sm3I3=}1w`F20+>;3qD?q^v8@_EO9mqvpj
z2Z~{Ie%Lnf1$zzVLD9-91XtA{0h{*vrGpJv(wKa4EJxZ1%F25o(7Uaas00o%0d`jG
z1Rv`CbX>UG)!ww@0oFB3i4G3rC=%#<cGgcvHx^qU_mjre=O52oN0IrZ+Ly%<aazW?
z$I||83@}jh{;MX~9X$$XrRk|;+07&_zf78a=J-gXU=E6WV3uCwanYAIY&=t`5A5?r
z6P_my+1yfUaq;VA)4|GH@iRKodq00&uPA!Qju1en$?8VeO6u6oqz8@8x!3RQ=5g2e
z)DC#xm$roXSQGLXbmTCr$>l!qZ`TD}EOwl`W~)mZ=$RZhnsc~ThC!dtK=eEK>FwF*
zJ!I#e3pT#!L~gimTc)zzpZ0Z-;R*tb{-aNj-Cy4q!2YO<|Cl%aHb)2_;`sfyj)4T|
zY`ei!es}yeiX-}N@3+bw&mI&queA2X(3yZB6MwFo9__4+(kJ~#;g}}_I<Hlt0oyli
zNLa`Ks*bSl(Qwe_LgZ0R0EhWoaSRwTV+I4_x^aX^<xh&|npU({$G+6(|7k;f9rC#v
zHcKnO9-Jp%@f~m**ipXSIR5VNKc~9CA3aPse?R>H^yohnCm)UY?+VYKb>~mPI}H}n
z_3s~g_kyQ^-OE4Ew^}tIj#cwVSJB_DGe9atCxc$7*s;BoJQ$!B#lASW^kxfWcyX*a
z9O?~~!tFo{X2utY;&{?N?j5CAx(B;xhJ0vc+DW=y-5OwfULqu(H?%G8L78;67w<Z3
z&9^AlAR<0zB)AdtXT_%!sA*Jeftg`jR4rD$V9$4bT-+<Eb4t1+?R$0ASYwYYIVinw
zP(-B)7h$C~DMMZe&)Dby%sNY(f(7HvxZhUAlrl>q&BSufAPLsy0%6g{;raq-Sqz@p
z0?(1q<15d8Uv02?0__YBpIzKdj^yt}t+L<nk&KTIxzCFTB49_Sdn~sz^F%M$VdLy;
z%DOXmw0Qu?=@byiV=CaPg%%0ZSj#;LnMWBZfgs>f-PI(7L#b~Rrw>k~-qT5J`C`x=
zPrRqF^f(`jELOYdXbyxBXU7KNbBH#n<&whMU@r|5Y#a8M@H9hx&!m=Rb28tjqt~i_
zCb6A3oH2gqI(Q>c#gr)m3ITPLupH6%4LoK@sq0%Gn1h{s=I_J)IP$G|ewsU*6)O#b
zq=8^=5r&56{CIikJK??8PWNMlFPW9&Q*z>h)&uyb9x>e!;h9%;k@43~hQ=^p2fPt;
z$qp0HLzR;+RNaAhA%_JOutI5FpdY?TpIs5u)oORkPR0=eZC|%}fc6AWB3H;yW@pyj
z-Un_lH&^dLY$18_a5a@_^nh$hzchuq%_q~x@7T9W<W)O?d}<vKY|Qv@kP2#uKxCqH
z-<G!nbbhj>92>0$wMlIKPU0zkPJd9_D@Nf#yY8!!jK0dVQ~W}z9FlrP)}1Yh)n8wl
zmNsUtZiQ9W1d;&|T>Ln@eW_YPjw7vRTLmaO?CtDA0=jR6X=Ntvg|BXeP$*n|boBIy
zi>dQsuA)>pWQ(WR;dE7pn!{|YI$+lx<fRqBNlqCWA7Ywcnxbtb43RH7?!NZ-dxRO~
zt~n<@Tw?$~LBTgtz$`^^uvy8u#eXDa0$F}DO7Ct9dO!zybP_PE5UvGXD!Un#2~oB(
z8!2<~cFBk+t4%~gZ_Xs5ceC@(8U}L5JFR##wwJD3^BN5=LZ!r;3gKLZCGurVbaQQ}
z&-(#_caWgTa~ulsj&!QA8kSUJ!?Lz>q?J1h(afgWGc;ifkdFM^(>>?x=Yo-&5*xuI
zC*26A$uTmI`xeaXEV7&DAi!Wnc5(KbYQ7R;7!5oz8$24LB`$?xecQcFs!iEwFS6!2
z=zKE96z<i(jyqzrhXO9}U&JO59|+Ov$0<>FsVwoQ9rUAE*|F59-_#SaYz9~BY5cwo
zH}T+Gam(K1R@tfQ!B%!4G$XFkQ1D$3{VFjo7syx5!Lgb&N1D;Vj*GsEK^0AtcI90&
zWVE!sUT&KuYq0p`Uf4yJ51v}z$!0@ouU*_2ZxOAn^T~ox#~Wh6`sF?tId;0d8DIgW
z_x&cCqGEr58oew-d}eRUv%Ags$&Na1S%fkbOaY<~f(_(w5q*e(seiKUPNR_bcnY0^
zeDU}MFgWmHUgSU$XM3s2-C|?zd(WDX$GKA<hgpz1Z6A9_l$cXa3CgDUx(oa@)#k%S
z2TKRRJbzIh{L)AFw?zVXS-WKQw=S7=n$Xt^GHGRq4T?6ocns&l|7Ee55x%ckuFZSV
zy7HmMgkNix<`*Z}I?rnn?E9c=R)5D|fZDJA#RLb72mWHYU-mwqv**X;zm^OosGdTn
z3{20T6bR|lcIv>~zl1pOZvM>xYh%)M;$T?}W!cpL>u~sIM{F&2I;RFqBH&Nscvuxc
zcjr?<3ZwsOIknfLJ-GrDMQ_v_s4SfI-45(~i8v_u)&P3m{Lt2YWfN+im65xHo^tsF
zH1+HA4N)&Nam#IX-zI1wX;5KFp+$4E5Zl#J+GqjU|KQ#Zw!K07{e}99_lTVPsZ=BS
ztktB(npf_uFOMn=>!fr^>Pq?V3>qbQRf?4(CfvQL5s|!~{X>W51iOJaySvF1&#2lo
zwe<or#q{=d_cf=acDbJ*Z#dyW5^xjC>-BxE+17j{J8vMN&cmG5F+!4Xdk+Wq)erAH
z*EVsBcZ!w^n)KELev`3ipp%>m8s}-DG21=Xm{4O`gh~_8!lfn?iJ|FE`}=vfypRVZ
z9dv}QtGnPVaDJrLdu@MJLRiU{H54h0&M9{rb$`@q<xz`in?YLFuJDX$aZgVtxD%iS
zx?vwXFa*}~4jjYpm^JxFMewZg>)Q{JC42PFH(B@!V24$f#cSFhz-adXP|ZT8pPH~Q
z<-_6qoaEr9dKlkGVW=kAFkk4sD*4Ip>v19(ZZ;SW3yuUi@o3hw!}0nU-OW4~M4iSX
ziZW_+1pT=!`5O&&id>8nf2ohhO7~{M+`9K*y!p!-NL|xf0F5nFHdth??#}?;XK&Z<
zbTE3HYKYajkj!A<kY#(WEv?%1irYm2%}2et!x?EttpnGm<6QDqA0J;Q4t?G%@YW+o
zuWFtJrgMatJuhUM`+jLM`Mni9<FMz%LvrDB+`%6xYeLt~lOD(C>v6j7d?@{bWbuN|
zFG)UX;Z5|wTuDI<dCL$D_EaA3raJ`#Y~ut79LcxnKBv^pAm&3zim0XR>>W>&{n<0$
zg7<3d`k8iz?qJ-dbjia8shEARu%|?^c>R*S;3KUEP4|+PkpE*nZUv6NKsNBN4`*Bc
zb?5Y-3$Fe)Fp40Wfe`9%10w^yly3@>&xQPle1Q6X5QH&M2LCOTxIa|6t9(F5T-Y6f
z;}^h$U_BLW=m_wg#(rt+Oj`9tDgbRJ8$xEF?c}e)Gv8Z^-NIcXzi!QdT$Uf>Rlk4o
zA0PBPT~;^(QU}r=Y78Q~zu1bGe{O`^2oPrd^|Zc!;q|PL|IKOrd{#ePv%j4dNS;5`
z@4pamv8(ufHOZ~&cUT8<Or`ho`vjSaVwE?+fP-8G`ctYuIw3rboh?JNJ=U7?C@#>Z
zoDKb5?8wR-;~N3#pZJVCc5d=WHR{D_9X6xy&Cre(gfhK)>y%q<mP&!2xihVAJMj8?
zXFrFv1a{$`x&liDl}mCR53gq!R-PL~d~Ak;<5+x>`HLq}!`1Y+u($xUvvMXktXysO
zFFQ<<tGAu6Q%d$-DL;B3pI$&o`@#e^Ko7BBRP-iAGVSpa$1Koj{*~PCd;1kSN?U=o
zZs5-LwmA-)l%!9_iEhmh*Zk6l?8dU*FRFKY_HvifkcWc`gxT3IgfK7sF3>9xloD^K
z&v=#ByViZ%wr>TfV-%ohmU}vEOg}nuLezF-DK*Yh-R{%S)$3gvHSxaIH2RHRkUd)V
zPQb*l98R{;GhLJ;qh+JErxbl}v`L17U>BIdP)m?-f@F8gCWEky!_LA-6pDZCX^cG8
zPUHag!Vjnu0@>{Vxd=f}c)EYuokdF=_>$JAs|;JQ&5<%YJ${N0y}qHRx_nkg#Rv5H
zxxdxHXtfspev(pMZII#qdZHpHqW8y^&<h;{3gn7$dZC7my|(Z1#%?c5mu}iNIqGLM
z-%8ceM>&`ocipZ9TgCN)PFF~m^*|?do#j_qx5>yW9K;~;H*U@<lp*%|u$jZL+4$-<
zY=T#dQT_hdO>s)lZ??g!Au3C&P2TZU+f6C!_71W=3|h+QAOZOn8q4ylhQEN~&F4<(
zuPFAfBzEN6zKq|ghpDYCVQ2&bWV6}>i+GSHMh6fgeLC-?(SkJ`P{@9BU*zM?&}8*L
zZ3*D|lNiX|Vrdt2)rBB`>#T(*TnAt<z5dN9M)f=2!>aI>!1wT@!uxZeToPjZ%)Z$&
zkT1|&voi^4X#j5hE`D`=0_C6d)enU@0*kkGr2Xd!z&;C0P6)d8YOoApmoHCLB**Xj
zd@OV*S;(yRK`yoCGhnL<3i*=X{alA*<;x#dC6{t$wDe>R8V#gCi;qBU{%vh5@t|FR
zgn-qrltHu&biPYyqF@kF=oV1%zkU|D=+=R3>Gg@fEJZ&HUv~{6`XD@9v>y<Q@DZJm
zu6itJ_^(=q$iF<HfdWqkU4kE44`;>#Rt2agYRLJ%lo3cWKt%z$YG88u=^Fifi4gUt
ziwr3~awuKEzY*^&pq74au0f&C6(<&gu+QLFL3s!$wyaKobZY3w&tH!6guwYi-ns5T
zh5Uy@{<kCk)m%y&@Zsj^>th4BP{2#%=1Is&N{$Ke>g9vBDI8(+!EEFz%3CnD4Xxia
z4xPxcg-B?|lLLy#N=pabrePT&N=Q}5%o*1P{tdDhOmXAB&rXB|mR6qL*Vo3)7w%QK
z(^ew{sfJkXjXVV%9Zg;9RNLx&egwOWM!8LezEm@>pKAZYAkPsox&mF^mmR1;MOkaF
zx)U4IF%oZj*Q4ldUaBmvG!67Uq3Zk9dro_c!?xgE(vznxzw1tM8o=ZV<K6+$5m{9v
zvuPn8Yh&oRz#m)fMa5jk;TfsvKpi(r^wcnX@Gu#rfzRoP#i?^7Xm@ycu*4UR;?^sL
znGO&L&pue^CfUKB74oID8|;Qf``$z2WrW0T@}zizi8UOASERD;_Kyva;LaINX1&Wi
zpAZduIv6cBt_==^0H~xYQkOs=>^`6?4;>V!@hP}<d>*t5J+i0-X{?N}%`us2%y-_E
zeMOgN3X2G~q1x`6dB#|JA=86i!5wEsv&dMzDL}LvA$;O0S5w(M&C07^Ly~Wy%BSQ<
zHf7xIX4MpE4Y22vL4}rjH$_stlq;@*j*>^@I}|%Sa)OKBoQNe1Q!UyFPP}yOQ|sJ{
zy6X(<5jy~47#w-C@H&s;@v#r9!oBR&%Q1!U?aN!$(h8u-yJ_om-lKX$dR=81sy~er
zcoeok*-GGJgl&748jKxIH$W^U@shjW^4V{WHAqVVT}NF1%%A<+QuSY_ze!#HvvPm9
z?&PuIo8A>vTfbVcI{%XeYp=1N`()@xUcX=Z{bg}hARwXoGz6@1V;{OF9&}Lj^l1pF
zKMes_=oP@<VBPu0qgy?p`uDH?WjXynIy#to>S8^*zpWqtSw6Bv<y>OskQ91HumcKn
z7;Si2do+*-9?0^gQJ~80TVZB7XQ6m9XVExrss6ghE~DNaGTuBN!J`u}!j4xUXs^EF
zG_OAxh8&=G^v(^)q!Lfaj^3#E$HQei2pw!L)iLpA2%m0<SY05zK;X{PfL~x8EA_Nx
zpV=6r0h{}Ha2q$v=X~3!{ML+fu?Dg`LT3xrNcVC%9UO@7Qcl5bz?U<7$*H5&4aC+O
zZn{FLp-FBmtUgt{i5rA?CW0e;P4o7NRiLBw@O?Ejh(!ji5S~+3!ftWn$1Sy-4am<D
zeXi=X$+yoA^m!p`1Uj+J6)NBSWbS2;cojjaSNi>J9$`cKIFN1JBL+I&c*mBm5X7wj
z7L^_bMTAPda)0?kIn7FEv^y9`!2itOQ28yEcIfW`{(=|B@$fVZ2&URy!p<CyBJzeI
zHAp7?K5omc(jIpHE;>k?@&!zfERa*+W2<7CGd9)wvcXwE5Q_pLeDV7DRv@fVe(w~K
zxoa4=QxVWM$m+X^A!}8<8x54`J%Cku3#6vWM>1oBAu<iC$Y7-p<F({$MmC`i`_|Pq
zmFr#u0C&Doo<ubW4?={Kv`>B0%%Cb$T!-hd+^b2&dYV2+!hUoP7}i{Yt68(?lc*X>
zj3~^2()p<InR)lGLVJ5>m7LxuEkc%kRzNBB#J)_5d`2EVo01NC&RlsAom@nGCl=07
zO~xkh;|)LJUzd-c$n$TwpOyCW-x61yf$eRb2FUR=8WD62_BE2r(bc#v{HTt6Ul?1#
z{-5jwQ4h>{@ft@)b!zbIp}z!o(jhUi=i`fbI0V@tu#+i28Rv{U#?RVWE~tVKJ+^0G
z#ojOKO|U?1cr?#ICyU;f+0u`oizr}&!FwkK2$$KbE!)U`@=s;Vo!GUdJaBXV8B!N~
zvEv$%cXf(t(tI|z@hxS|0JpJ4nP83(D#98SeV5yhDV44-DG>8G*rcy!zabvD8TiiL
zj??TaU^jLq#s)>?@O2m(<$S*$9#a5Wzy_lv#H(C(abj&Y8!6vigR`wrss7>!iDx`u
z2;N{TG<)QECf!r`)>ICdG${|^tY|TVVjS1Dmn|3)!5tzpcnoOwgp<YMAxlzKay^Lg
zFoN~KT>7LGljIs#nD@Ha74a|))UDnfc+N?XyHq-dy)^lFY1Y0fjn6h(Y%evGyZ?R{
z&Wd!8a`0KU%RyLj3_K<Hg`h2Tr(39d#?#A03kMb51Wk&>O&JCY2fc#*TLZ$mbDQ_i
zeN^U)8XE>XK`%p5NU@J@+S=f8GYzWo1vDHtdo8+$q!N(;%TYX*l@g>mk})&vv|R_^
zRP5|fuVQCgZh6SguXa)hdT+C8wdGmo!Ci<Oe;b5LEj=D(>ZPeF+FZrJde&(AlW{{1
z!3$P$5Vol}!|mnV0BrfbY4u!`{8Q>in!O5{YHjE%6d?P@abR9oh!oCG{M*(x4jqg~
zF*1brwavO4@`#`0VQYr3Mh9<IY(Y4-^rwbBGtAGZ@BhU%>;H*ARo?#-8JFO%dn4>y
zqCEp7zhBMS#!qv$;VgWrF&|PB9!z~MV%<G5ISFfSGULt|X#gMzN;wNeAy2S@%s!jo
zfC<NBJPVjI0fjXPSU-MhU$f*_lLr*tku@>{=IU9$f_!)>MlgE;AZcYU|8U#rTz|w)
zIyK-l96J7du?JLXWgPkTYM%B2bQ6^G!v(^bx7v>gN;ZXbo-oj%51{uSKLMWfLj@if
zNI!GMzEt3UVj$_c7`$_aq3kV+fzZQr@F;Tztq|kKF3d@x)`ua*kpfr2tud8?TOC9z
zSFS*t+saY8@Aaa7M64njiC4s%GvSOJ``Df#Kxi4K<KZnR!&VzFx#_7pgUAQ-j!>6<
z7n*nHcp-pl)YYOK_5O}~zH)0S_qx5&QuN`Y2JoES<|ud?DWNBp!f}N-os+@qaO)K7
z#fjhHF^miR4R(mMthP)N+$>vQ9+3Kx4=SXeoEs*LbY`A40#C$A1CG0M-(cm_>`UD+
z-u$o$Xu$x^Wg<wei?FTJLXL*&W*{}1u6ypycNoxfAswlt4i?Bf(<-><#RKr|uUGRF
zD0kms&g<O!Ky-fGbHl^&eW%J<o+HuN#SmucI4!urf_^wQH83X=eSnWx{^)n7<$7bA
z>|$oSa<7;;$HXb{(=lj1wqOtUgdODBQ}TG93JF<k;ejVj=k{CPer%FgZie1g%k~!w
zL?jz$3qb3aVD^zcS9yOnN#-?nxl*C|dn?@RTXdlSNTh-qy99!E{oLUGj61Qm>`r1A
z#E0jNKF;`rt0qfEw5Ql?uIBs4He?`SuPw!T(`(y%X`juFQ!S=k))Z`ah<AXvEwufV
z7j5mCK=<_BQ80I*y9S}hK}^0hgT}95Cwc3Sp-RSwPMueF_fy~Bff<#)#bCm#@R(EE
zi`XL?dcfn@9<<XHGKs=lFM3&autZ6&O0a_L9Dn+VabFoop9aWZGmzE}`#+@|?MA04
zQ`HX@x1Y;0yi*O_&b!ARBY)%d`L~hfGkga2+%A4Bskj}DMj};N8DpB0564|`7$GA3
zd3ZiWdjslTaMJh$I=QjAA8w;4r@Sgmhi)==d!k2Z0vi%Qs!oRSdK@r~k1`Ax<#RSc
zlYXBDR;<$9+v@%FBJ$F8kvu46YtNztabG)aUq&;GjZX4AjF+IwC-!0*Qx23jZHji;
zV|SZ7?~rauM#T5rbci(L5{)JqF_upi(Di<xd0!=I19%|(^gf*Vj-(TAcRV9eyLlf<
za^DrQQk6-UzW2Q+)wRkWH?K~)JCFHMU-E;JTn+(CkTVexuYDQw%}ID!4Omypanm+!
zsD>%R-D_==bge8g^9UHQBT#g#SyCo+AUNsFsuI&#;X@ih%@$GUh3^Li_x5l_#X>F$
z;;aRG14yu&+2;`O^X}hNM_u3hX2#rJJf0<YPj0!_ElPVInlHq4cm=)I9_Y4Rwpak4
zG0GBm^>r&f_Ty8g+U2pI`x}QLk6E=Oba4Iq^SQOqjU{A#gf|uu8JC@ZK(CfS%8<pr
z1#(ITm)ajv^$eDjlMyO$7MVL;Igl^v7*{728U2EE;&bSqF4e-fR?x#jemqZ^?P0fM
z222~5$u~OL;}`qcnk}hlo9Biqx26y|{y0mRIdsZ(IMk0b!a7r$Bnq149a|Db%5Lx^
zO2!i=KPgYH+u5n3Qv_0!+JuHKctH&54ZHFhGL0<u#b$3RW7u4jr#aLSSmyd(OaHMw
z^y56e{l8?M{?Z+X8S)qN^a?WkNPhWlp8mY;|MWclQ>Xkto~M-{V!^B78isdl_Gx+r
z3D3Qak2V9)Z8my&5{SiY5$WY)cv0twdUsxyR)CEkE0x;q%rq`qqfS))wd1p!5edAP
z^!cKJO64J7YG5z{Ge1Tu*@7=Wc{p}uBV@Pn+&4a{0gaNNxvK_qROmOBJ}byuO1=L2
za5id!%1Z|PeB<HF-8)o!>b2ZCt`J1<HLVL+0BU3<a>#G#yA%LX-8;b4PlZQ2>eCe5
z-C$G6A-Hqi3bTE^OTr8=PHyoAhqUa|5?}8X$Q1#6PTvK6w;PS|f?OWqc`xI$Q0leY
zXHR`vsN%{2@4rYX1xA`JnD&eT>~i}4OkF&b_VqIXh>KWz=*{R(g0MJ#uO1ysM2ION
zacX=(8VY{jXv63p1EmDLiq<{$;LdO7eRQ7*HZC03El!T=5T}iJIBrpj6{AacI6(Z!
zJEI;-<{_IXzg1^gw@FZWG^DFa)#p=nB0WepVM&2JSu+I47HVS1d#pHw5nmmOQX|1m
z@FduWIF)Mb>+iFHyP$a(UGGB3=m{3{_Vp-pjWq{=r4NB_df0Iff@&ZF<FWN2-xWLu
zTl{pCT{)p<X3K4Nlb{JPXr|8H<!#bP@PuLH=)A}6)K$xq1kbv>EZeHOpoPW{U}2bD
zS-hFKKRj)}05=Js<;q2k;;VA@<=3o(|LBDm9A34^b$>cCk#~Fa$2jd2{-O$+;i)u8
z<;E-Dv%MN^e*_i$C-XF)?UsDg#OPty598Z`s2e#3V*lw4QHa0v<o<r1rdQsR<MNQy
zO@}w`y}NoEzM%c?dbS2dc&G1&FSNF#&&#vc)Y&u(<xW)$8NXtpa(fu~JtR3>GX=(7
zXh5|fHR$^yX9IE*FHn7jK3*H6g3N2<gzwk`h#A61LDnAHR#c)qH9I&<r>@_y5~FD7
z2^l0hgNB#=be37@bn%`qAb=M6nPS9*Kb7-CyRhu^IyU(BE<IeHvItRDL8%ySw@Q4g
zWjJ_z>P~=_Ml4Lu&4L$Q+)gjSc3a(z67uXLac($)&DqjI@e1~w26^6!K6`&%mb>CN
zf&Uzwtk%ZUXa-?l(vXvOL@tJ{;H2QtQBhQ<eG*`2nAllHH>NkX-9~3&)4gQ&s77_d
z?#Bd=<j0E1#NYFDozcVj%In~lP4<u<f+*)0Ctg?nc=y1lz8_L3nPoN8)Vq8qK3J_B
z{1g5-8XH0DL!<CRyRFPxKP(OaAWt2BeAGAEuOO93S`dJU@pUZOI>V#;7XyrnBUh8O
zw!wMI+t^Gzi0VT<bh8f}&t$1lE@~qgh7pel65i5b_l$21UJK#@0#+}MnwN3MpV4bw
z2a+R-m$!DixDcPA;+*+*yeGwB-cm%3N*__`aoRWPn;Q;Plj+JS%PHBIT8=Zh49Sc|
zI)ba_^xCsZer5bu2i1=74gwF@IAVz#m5&y4u9)-!RbX)#s7c2OXLjSmt!-b?-Vq0;
zo0j*aR|xyuh{GlM<2>#Bmy<Oaa3KD5Y94Xae6k5YHrLxfdZ+#--{8iz8s7d>4gYt;
z^D5r}AJk8KSQ`Qil!bnN-NPbMW4+veh!3E^w87!Z00JEnc7zWLEwtr4*te#y5C%;V
zbVZ*J?ls}Z$GV@12WxW-1n-wX`VQWpX`%o^EpQE8JaAmij0n-{6^P`&is8R(iUE36
zCOUNEAdBI{QB4Cj!X-Y->A==mZ~G{xuY8@qpWfHA11`k}(Bgk}dLMZXaCU&1|DT`U
z&paO3^Ft5$moP(?N-yRZR`Od`4n+oFlN@LcvhMF*n?*nJDunS?lGp4(I%B1dH7f}9
zGR?%a&#$y{7(+jmcLCDe-=k7Z9!*`O5ul`ov7G~M+}P^unEnaqt(g9PLx~+P3|l<%
zrw8EK$<Re~K#iZ_y8%#!DB2nms_mcbd6``*-~|2*F!+F|<{tRN+T!W=eNuK?38E+}
zS?F>iZbqV|dlsTt0Jn=&CU(iUz9>%UbQ+(FEDYo&qkB&)3ycT@6BRWNpn|m}?E#*A
zNUyg_1eLI%?D1LFA;^7Cz6`IppD7=~GOi1nm&V^|E#|GFoy3r0?$A%0&N`a!U%fOL
zSmz4z?KMpu9(bEQb109Tz441_zkEN<feq=`^Bngn0QgaZ{R5BZjDs+RFo+Wyl}dXO
z%nqUFll$HrjQLlUuC)aWzA5jz1-gb$5qD$u)%{Q_L@1LvO2?<sE!drtS+NBBNhI&N
zA-DnLIieE0+(y{_4)yuu2{B6Iwy-bUM!4Xovv4MFZy6v=vl~4dYR~p1WCf86(QWx@
z8}!2Syg0~^JVmPSW@^Y@gIjFgOtc;SEw(Ri?}KUJ<mhKlC~CO4=+^+(Pi2!B()K1>
z9+r7e)Z+H+c@Aw2Qg9Fx6GeNZaXb$k3LFT912c}_Mm>78%q;F=N|dBJR~HSWFc<Z5
zGPLSl)Y<7uM*U!SoM7*BS2qUf)*u;`Rpuo+Ev9tSS-h>EM05fX{I0<bMrHD9hIqgu
zPc@-x+t0_r6sujl2XMWAt}<r#be<;n6fD>O3y+8Af0EO&U`pz;zwMqr+LQ3Vzm3T}
zy^hJlw=voBB#wND=F1!zZGPOt-B!Gv|Lt?}!HUJduYj&vW5IU5%eg4Vi{-&S<g08B
z$PE`rGg>vrP@D&|<qHR3z(Zp{waMV($!%EouHDO($aDM7eDpCkCpXaCeUtxwikiRE
zm&vpp$}5o267>$;c;FnDLJpS4kk$d?FSmvZfBy8#!UPxR@{yC{tQd$d$OrL9gDPEn
zkeT*RAPL7b3<7fK6^-$s{76N;f#07iDn5egCfa(ylmCX%fD|7}uzymDe9A8L2BxeT
z0ZGZ%sRXcZV6iB^4`Z$wp=id4Xy}CKhl}q*Tl5nu``u}O8-|gept6#ras5v7)OJjF
zO|J8Ycdv3Lg<Wr_L+`7tL{;6xcjs&v3;{!Zf_@(ANWR8erXdbXlJ9*jY)(zmgm{I*
z$aqL}J5K@adRsb;f0u8O*tpNP5#L|U@mfuzBFzUvnInTK-6Kyl`KE8w2Z40YF?lYM
zNd_}E-6(7m^x<Baf`b(n06rLj{7fy;_4df8@h$Do2{i=I?O1l#Ekn7L7_%95pFN;G
z`6EJY1Hz2#I-)Knc_iu1+N4m$K31jdK;uJ3CpvY4gPxa^&z*&_%)_FXHpEaK`78R|
zRvhra(5LYTUpitET5M}NkApH{?lK*pNab>K^g#2I0u`ZCQbh-Te?2-o*wR8PC=%4K
z<8To^fLRGT<S)uEh)Y?0FD}R%|HDLfMxFiZq4mmnvsh7o%=u=96(J4|Zv;Y{K1{1H
zM4}q^Jp)056~s4!EkuKS$LJij?=(~z|Mh+d1EF~kT=fskbFBX$pMDcit<>AH5nmV{
zJV5R^9E27I%#1+oD>8-q(uIOj(-pgswW<r)!GzPvdsCdKKn9Vy#UA|x(W8FHcL(O2
zPS>(|mEbZqQ@z=xPn-cPgbUTFRzn);2?cs+@3q#6Cr{wj$7q|QJZvW&PcJWi3l~kQ
zU^lya)E2v4fM9+|n6_@ZI_h>QbGki3&bpY>LJr7qZgkGQ$$Xg1K6+6Bk5vc81xq0B
zMF6~hlOGPmLEuT}Id)vyk)5_A)m85Cte^_Xujo*HwHH-luP(B^5^Copg1ZH)Y(F2Z
zDV>7%U09g*8D@uVS}aTQw&xDE!Spp0?c?Ts@>n?pW;>Wt%(3s&<BcL}E!U7VIr-BC
z`}4ZnwH3#9&(3!Pe)g|(*1fAF>;GkTk9U=3-RH<pVA$c0i}9}$zz;wFIvbd4?`!@R
zaP}LZCCj`Qexq@ppl=ws2~RT;w|=>9<2fG72ng8hyHu3lu*&N`SL7h1hTnmXBSpaz
z(%^wKK#2Su`dwLap2FjKO5nfmr<4^v35g^oEOtI$iGh?*%O^X~1D2Y%8-p)NQJmnp
zBRQSVxPA6K-Ie%N?w39#?L5MPU{T`d5rjgeFoPi0t?9qsU>}WRdh>=D;wA48?s`7K
zy3~if4-hCuz~NJ&-J(_>cp}IF0p_`Oo}!rs11-J%(v^c>zzd^0Japxf>5~$-O1Z%w
zbNzPPSj5N4TxjZR%TI3{?k>ox2iQ*+w}7%UH3(!X&OljsM&C^_5y5{uD!7j21f=~i
zxbwYl@_hjR{lt>+1nGl)^p8Kd79hG4a1Idtt_kC(v8I@ySNp>|#)3fc9-TS92ku0A
zx(j!rc|mUl6!7vB-pwyJ@W3jB8)M)>!I3;AzJwer93C(MRj4kk^Fg>*yFxi>@Hco-
ze}N<IS7rhr+~K1Iv*~a8g+DoVpqXbDtu_M>OvobTFaw6Ja~^VP>e^8%w#<E~I7&yt
zH<to3#|)#oeWlJkK(>J5^ly`BlX$BWa<;>O`#tQte4OY`S9TH1czV&h5ctzPAh+ri
z6jZf+LVzoxxHoV~yC?Ku!{dQkd5b`jp~>^`oUecc!(pL-<G{w5l95vP8}leXx&=gc
z(H(1}{K21ksd{H8@ij+?Qo1LqHAl+eL_xRh{D@)%LA-4;3~t*L0yWd=ezr8~X>Ce~
zJ#h}*;5;DH#Mzrb63rGxY12Fj<MHkwU7NZJ=Nv%$cnC$Jcta>9==z{_Rx9VDZ9RsZ
z%IsLoiauP9q;}YDeTG28V}wp)&;^C#@7glMc3RJZ+ncAsu>;sl`bUQ-dlzb@Z@NjW
z!x+LAcV$^;p|N>_RqWSgG4s={16fr+L{<G{`UI-_nlTDX6_ER<-Od=s2L@l&0xBe`
zAnd^gL^!}#Ld*%}@4n{-O5xKrw~x!mwlt{u9)^l{!F@%UZmLf2y{CpTv;SgM)&#P=
z>?dt{V;0B|tVlbc+J-<SB;s*}slqLIpNv-@`^oceMB6+Hucz7r5<&nK8=~0~;3SBM
z%N{Rs8e%UE34>~7)UuGJVqbLUVlErQ0N*ASbhhsiIx*w6H_-eZEXmZZS=;Ba3&L$Z
zs8D}tUvFA1XUo=mu(MV5q1*8g3hyNz<z0Nc!XqU{7~yuBPXJ5a&aCY}DB-Z_?}4jz
zt^W33+WW@E*wJ(^?IfhL0%g&rY#|$qdtap6<tpicReGt3YBwFi2cHic%0CVDCy502
zU!N1p_oW<w8Gu*<<`L>I^T=N(4Q>5qPv5o>92=U;uYA86kNucKS=)kUyCc6Ypdg15
zfjSAVLMHSY3H`NnBQJFGb<`AB+uWBC^XpCk*asl6pjU|kBY`|&9cn<gm?4maxw>j#
z<?jOZSi^OvTtRP8nfdXyza7<jP{{v3NA=$s0Ci(N10v8yDf~8@d!yR;F7L~_(?L`2
zjS>JIX$_h~cpKZgbd@{q^U-kr;!-C!+$Rumv&r1x+AG)Xa^_3REV|rod{rj4-2cbj
zn{+9vZf(1Le#O^!oFRpRG^r7(K?-{U0#YkAzy3@kGIQlxd+oLN`;Ifd#%bip48t^#
z5T0kwdEeJ92&i6q3P_>SEVv%<!I3Y)Rx$9wq@%P4SIcwkES9?It?N6c;mCz8ojI57
zM9_pmy4sbeJ1}eR?bwu=khVd_)y6Lpoy1T{7gd#23GSvh=CXQ$#?uDmxni>Jbr<Tk
zY%C{RuQ$x1sVlQSZ~Y?mS|z@`O7E<ClJ}i6)Y{!uj0+E;?7IW!P#e<i^5*&)J$c7U
z=I|`%Q+J01ZiW;fAms=!s~ji+GRV;4mru0R&;$757Fy_WjA(D+uz<MlRSIrD@%85o
z5c(Q6?BKz)@b8)_f3nY;#2u=^RMQVR+`~B|UEkrVE6Bl=gulN#g2~{!G)P(CwS%GI
z49t6`l}X5~zmbsy#t0tph&Z@g0lk($gd+mBic$@}uYeygGu|z*&sFiYvr&Mv^!>>P
zipLL8&gU_0{rZ~y`uFog`1sVleA&C(*MI)Hq(0xYaqQ`E3vXU+(ZoNU(yx8l2lxgn
zxQ9ph58?-`@DTMngV=&?c^E2~!2VBUuwJh8m$Bo<Vd!^n@dykY3sN$aUd62eHhmw%
zZW@0ob&)Q<u21X)##lWMHYNRBpNQcci@5$<;>pLn%QEsVz&HL(KrNU$5OyW=hR7-%
zh@!kuu-Lb%b{d%6ubNULxJ>Gr>dhLk!0vIn#d@EKVPx?sw|g@p2175P%;5V0XQ;+B
zFY-fq+;xD<-_pvGz{LBMcc|(dM7o1{J+Z$M6c;kDWxC^iuQ>MF`oXAp@9J2n&1rxs
zd-~o=sVwl>y&bV7+)Kf=q9|?89WEPped(5a2o=FY5*K6ltCmV}-jfH#WiL_ba*AGr
zE4fv}Mgbx5DiL-g2Gl*)@cFQw%(GwL=eFj7WnN3%nD${=9l9I|j*@Zf&Pu*T`t#Ti
z=5(_$k-tJeiO%+_c991l7ZRb;<}<c?*_#Voww|*h@cahQAID%Q)$ab8sP^yI3Bmhs
zmWuzT;beVQi}g>d1^TzEg~<HRSBp!*K$6NA_ES4w|MO~Ld*RoLfqa&XKd%_qS?6zm
zSuwsJ)BZ0X+4r*Xmw<`CH@Yj$lJoXFK6iw6;fyJmg^fCbi=v^kE_IMc)-_d;DRj;n
zm7{ER2zNDyXjq`o$LobCI*1D4XZn2+`yG8dDp$ZbXiJH=+3pf$CYNX^;hI=D$T>*G
znvTkI(-oiNA*Xy^>^&a@!lPz|H<l_i>ViFk#NgHat&)1T!cPu#6K*dH%;JwFyYtb&
z$s<8WI~Sq+aLEpv6uXY>=IQHw06DY0_}2aK)Ug9`%8485pt_1z_9y7BYg_X2CS>U?
z=c38&j2s{_*tpy$2q*6WaNAD(!yx#nJmz#E8$QsDr##@mD;AxA%(0U%e8`^$C&QH!
z)ebI1<8~gYzCtSqcUMAC_TBh?5Jv>VWa3W#VvRu&V|`YQ?7vzybmRIgm#46{H??0U
zByKxeA0BsEUjPwy_-gU+pMKFVHW|kf-_(}Rs^P(^;lQdPw`WO!KWyI6{Y?_*58oBJ
zN!1GJ!Jl}_qgwyG_x=FAqPk)JaRpRb@!e|tECN47rtlcEeSHn?dcZa@>(>h5g`aPH
zYTP4;v1psazIhM;deOe;X}txjitW;0INzTZIB0}cNlkwt${{`F2>W0CBRI?K3(Mu@
zH?{K*EbU*87T4E^qkpY^0RgQ8SVcl;4}gVGPc;l)**mQP4g*(c8N%8&ewLiSyL0>~
zHBe9`-_472eHy2`5I|ZsW$qBuY#&!AdguGFBSwC~v=H(s8+OC4(ZB<GBY*GMN~*;>
z@aiaV9IQ;MTw#?gZiH^+C4pTQAx^eOj~<bxzYGP~o=#Qwe$yui)bytwfL`LcJ|?IB
zrA3qpxE?b?pI5ECO-Dk~+RI&#+*m$ky2v@BlmuHE;8lPy`Y_^`+j=es3&t6tH|z+>
zq<IVD#z$#WC%LjQ6zWCVfK)ABajE9MbPowVtNc+=hW(p=f<f3&#)Lt1?v8Olo8lzB
zFt`Mh3-cAdiUb%dW+LEF7qNf>egzkAo~Bl0$8g(|4F=&!l#XqLtG+aEx)ilCdJ{6!
z>w#J2SF=QmDmvIdl^UkQL8mH)Y}?<;ivLtXDESAEjsL=9v)teMq~Jzq9~QG<PCxGP
zu73T{9wGbMw?C%N^3zfO-Q@_Wbl}T`)Yf|d?r|980!JE%mSPFKXrRCzt49s}6#qE=
zrSIg!B*TKf)7J$mu!|-GR~y<HdJZFto5n)0;0I8`YL*R*HQ4h2M&$&f3vq_3FOPTu
zu)n`>0|Sr(wy$9>biWLEkq=8VkKl0s_GkqdH^W5@8>p{z0`02J);HKq6}HYz0tov8
zy5+}3JHlJ~%F6pfV*eU^R})Ci|GO6l#32H?WXc~Yn=p0AZ#i6BaLzg=hLHwL($@sC
zFmE{hQ~^%?1NsT`l=Eu9TZQJ=tbNJczJ@KIZWusT-0~CRJ7<aTBNe!O<>gcQU>Z0}
z@I@S4w2civ$et{jg~r!sGy3D+_4y&z>8n-rsSIU+Zt-{N142D5pRmyHUq48vQJ5vw
zpeT}r<&UMmch4TYAds%1;-2E?_o+aBPHjHG(+H&YFlaR`)N{PQ`E7kB@Bj46Lf}Qd
zKU;tOZT%S;{`nla8Y_H#TwwZixr)%ChQ8?6IsI72|7&HQ?i2LIxN4**y*w{>EWFJL
z{ZTsD4DRrIrMEQ2dLR-_eeGSh!6|$V`!nZ}tuD0HKzi}f>`1U`zO%IDQU+)`y8_Tv
z{5$gq{nqak>HuM+X?8wYd4ka@F)|3@0dtr$4^dz5s_Q<7wp0d@r`{IH)6{evu9ZEy
z>+jO@q7KAKdOohOVyMhWYcPV_S?T#Os08R`$~H$pf75Z7ZJ~34!*3voZPBS3Lyian
zPK2;3tgmP#hEVfM*W?eip9b+{ny(4-sXYXWUunpL?jKNKnzaGJ3@7PjT@(n))uJxV
zpqUb#6CO3Qyo+>zxbqOxLfu!r-Nh7yCPG7V-_}iFV8e=Db}tiK>XL+!RSi~8^u^s)
z(Q$+O6GtN~;i@B1UCzUNz8I6%@(`dn#(ViXfJ$qsACYHF`R>KuDL&tsy{xr;0@9!0
zW$vyryx5p4H@L({Ckxy-G<<>4FigT2UKq6Ybh=*(bJD=gRL*M2Cv>-qmA>|Ml7I%)
zWlz>k<T)4bx`I()NOpq=VL*u=70WtD<#Bx#7xRMq^GU?<n2;Cuz%h-rMMq&2PJxQm
z6RC9P?Rltu?#5R)*XWiVWZKRMTkn^=xm)i&K7HbCizB^%9?wMso{FU1d(Wfg#&tto
z;KThi_H`#xHJGqozL2W%^>pPh>da&ri1DTldE?_nic{#{Mc5&EVR*Xk<E;Si*JFxa
z7IDbp=GUkUl0g8h2<%}7xYgg`_Me6!$1?!-miGvl%vqrCe{F{#!(i4plH&;-i@!_M
z3PR?l@Oi{<_m7js%km1|^os$dH0wZ`!C<fGXTt(pc4kBLCp!Oxn}5d;+bq8TXT7WT
zRyLN{`UoNBCvZ9F<bAF|4~_!(u;&@_8m}>I?nE9aMZ>tB6J!BU2XilIi12XA9UjhC
zd_+kJc7gZuYa9!t1-THuW^|u50WNavXJj~R)XLw(3J{opz&-49|3APeTkIa6dlMiB
zY}hIR<avUg1@zWE6379-Vm|%3{;6)zE&CM>Cd0rzfzW|ZAURMQ?Q#~s+i1W$V1>`#
z>q9{PjzwT$brO3p`FwB<->7BqWPF|j;Ae<dIOcFa12zpLx&D5*zaQ;qK=Fx={>&#r
zTv)7)1t#F9(CS^If?w2UV)9WK0d@8xHG-StcmBlJ_#fUjEX02w+yJNIuP^=`pYy$5
zd<8>&Ef|m%5BK^GBz_w-2*3880FP)Q{0!1Rfbv@H{I}{C;V0(amSC#ta6Aab{<_+<
zkZaEZNg;p`w-!v%T*ZPk2|h-jjC=l)NpmOyf#Jbq7ufmF*0ro8M7_2HK+tf|4*>O6
z8}yCEzdzk4A=+3_`uv;CZ}YDoY}mNKMua#|Un>%r92`i3vxU=eEk(tDVy+_y8)W(M
zkZvv5-uZj;%Ng1E&ZLm}{HrK{j~4_(`4y@D%}=?x4DoX)3^QS9$Xg%m(X7mW%!WVl
zrxO0F*)t*E`KSMS$T#=<Bk)P^Ha@O2K9tz{a}Hq|P%vKx@IR6J^J5DJZ~Lou`O2pF
zyCZbyGM2QzPN)BHuX4kUTraor;Jgr*Caq<Vq}VZl>|xk8!BIea$|!)`9&tiKcaTL(
zBo3CURYHD56#8`<YKrT6o5E#u>B_B6{fXgg9m%L&4Z_TwD$a{8(LKJ`AsqDis<?$E
zY+qXIzH{@4andcwfEvYO(D#Ge$an~Y1@EWllN%OmG{N5TLdIoS3aTz!6W}sQJ^9h8
zx*Q9bj%!oMQo?J>2|65lb%*WLw`W=mF-{^FY3Lfqc!(N<Xh-bfk1u@2kme-ive(4<
zSsq<0eLCtL`h>6;MKZY9jyrX}KJEs&ACGKh^aaeigb)<(&ysX9?XfHosI5D>R{Y#)
z5K&}A_hK>L#`JUK!DN2k$*nUPSJgZwOy4`si9mUt3h2Ul18-qb0M-+1y^)@Wo#;+C
zwLZ<OKR%`wId<yM9Rx*Wrc)KX&4Lb5cMBatq-<RdB)es%%5t>j5D72tuDtlutv*Q9
z^=@DCSFZ}zZh6;$>Trq|Lb$qbfQ9ZI@-9XI(zIT0A+WwiRK&rt7S;CiGE<ooh>_`K
z3r!lTNkt24@23%6660Lb{vP8u6)<No6;b^-em@4<C+|!#)bU2QoeCoJXR@<EtyLUN
z@Ta~2_#!JAzppB~&EL-hMALhR%ea`qF2pv?@CzaYijyjY++GeRHZK#{gqZh-mkf}y
zdUdqZAt}4bzDMSM+Di^s3z$h0yL@s^2ew;!{8Il@y~O3u<_=i}!{gmNhX12KhPi#0
z@IczW>+kHfkKzmF{?LN^1FQyufX^`Jw;JnDgZ}^Wi~m%Y@${!zH~a#w3664N{#CX6
z9c%FXi)z<G4<LJG=Fcth!+V6yuinBM0SnF#1Q7T(VDYIEIRRXM|4AFDp_MT}FZyeR
z1SQxPDoBGh+5WD8c%DD<K<<HUqw~o=_%y9QX+o@V3v2@Z^ZWDSx0(jNkO&&B50vLi
zYgIuRAbQU~6^Eeq)j+KMX;gz77rK8iEBcsgKhfeaBPzV_V4%M|2HhAyru3kE`C3YW
zva3Lg9#ELDm-&vM&F0R+VHXMsOz@WArzoEedFH~~O>n4HS2*(e25%Yr3D^FmJ+?Pi
z{9aZy4FR$G|LS}}jq88j_V2xY1S-NWsoDP|il-)y8}&x1$E7PRREFT!-SzY&SERVU
z47+ljGcfrLSK-*P=CK>?P!BN3Zo_pB*Be}M-OTLwJVa&-(dB@~LqxFD@y?6o;Wm*8
zI1)#P^AH+xZ;R`r_p>Mo@6(ggpQ*Uw@mkE_&GCRY2?3JV*r%Pv1-q~bB%E@(ayKX|
zLg*-_Sgxm5XsA1IF(73Xq7mC#+>&~Jz6So;I^}FT28w`2Ily-z{%rAs>AuPm<}u`i
zOGjX(Ptj(PeSNxy)X6uFV|STw0=aB?&Gy$Gc8-l)`b}ZkwKYLcX}3F{rP|a0DjC!5
zJMYBe+`ZS1o{VQWzJoZv&R##?^W8EW4`XHt974qqv{0MT%LorpSb%VgVOt!Mvs~sc
zd`$Zkm;G@$@lp8l%$jUN9Xyh2Yj6O4{h~!6^1n>n|E`tz=a>WV>ZwMYGB^%_6vy!e
zqYtO;Up$gk|DXor-u!}UL_y;a^?cT-VZKJ;sqK_pS?3i_pgElJ8<xfaFu{`}f(9n|
zGY!JO%$mpTB`?=ZEor4Aex+ppgyW|y0URpB_AdbIh5k#yG_fI=>Uzf@`Qj<l>5DC~
zqPX2PgxbX$+#^Wxu(6Lg*OP{L&~J&z9$dD&iwBsX-My{=Gadex4Wc)aS$52%*k|`~
ze%bN8>er_h$lAg0kz1+AbCs3#L$1ux!_@PU$13#nV9B-$&)rp*ZSnn{=IBV`5jd;Z
zqh^mH(Yw-na^B~yrGf2)?HS^VBF77^*q;5EUJz##_4|k1ykkm(ktM!c)5L$9cY<<u
z(b;KaL%n+o<HG48XpwtabyUim@*e<2^Rw?`U}=7j>|s>@H@Eu#bo&<ZkDTlD3uR!*
z9e9z2Vm&_gd@8gLKhK-tU&Os%49)WQd)!CdTikiU{?YUPKs3)+B7qU1UI0V$H_rrw
zSN<b2^CRwsa^REECT+~^$IZ?aDxH9W1g>a$_>~fjLX*Atac6cJTq~^4B>$kf_Rq*R
z9NN|}lmUC;=Rr-f{#-IamWyWpz|w-iK$vqkfnFR`l~V*iQu~=Gx@a-<@tRLo&ZjT;
zdj}B!bzg#C9OTHKfU`(rzcHau=sCZgf<L6mpJ*Hi5kSuWa$^5JWuUq$My_BsSlAU3
z5Cvzvi5@ZsJPhGo%qYKgriQ2)bAFKhR+w%FAhp1~jE_->uUZqY1Z?w&JXc}Ho$mOO
z+6$zD%bR%H?cV6SL<1bihPZ`q{bk-3h__k$LIrsC(*!#Nv1@w>g*e8_P21yLSgUaW
zo_^;T9At*YQhW?|u!SS;VRme<Vy5__TDesl=Ig@EKI9<qcj(?g@Y68n`)HYDiIaxP
zfF67KwtEjULdJa?O{%%$We<FqkY3X5G@-n9_}NPcE?<x68QkQ1r;`!?J=RT%gC1MD
zbly(-jHgy|zjvJClo^=zNWc!Q@e}hnT9%bzXM%aZ>^oXWq)0;2=eEhobqyn8#qNlc
zDGWq{FvD0bkjv5@eVzw1w^iiO{szw$*td6{oiFMoK>PRXh0`fMP=^H1k3qjJyDdB?
z<wIA=b#V|~<t~uFY*cUl9m~6iktVhEmfmksy(c9b0GFI5r30j|jZ7Xs6JATXgyQZs
z5Zgzybpz??8ug>BXeL7H=pkw4kl`9Nq0F+cZuFF+JNeljQ%+uA{vPI7FUf8&k9OY-
zB=f9ub$K@VUUi*{j|1W|ivW08K|zg#-3zQK^{f6A4`;lNl+ZiuqIG9Vd|0oP_o#p<
zs?l^L22b=<0Bp6Yq;?{}16hIJKfDO)<zdI3_au;)x!$O!2ibxUA%m`FZe$(5M=g)J
zKp1sX&A^jC&Zl3&JJ7giv+cVhD_f^;8Q6!y7=|<QLqCy2{J<>opVkjV;sq9a{e<xR
z_HT^^T?8~>gzrJvS!2S_3qeUA{}QMnUl|4hTA#U#;?SnoF|RzLsi)A&;>n%Ie`RI<
z`{#ZsN`Dv8{}xNIa_>tiod$YpmEajwuXGPv4(DK}QFb$%ns{CERmbm<G~&GgbrCY*
z>fGqi_aa5tVUFwrB6zwbj)tb_Ek)j}7i!JeHtlitmY#62J@mB0&J|NEO`2X;(mBmX
z7y&(s!a`1X__{$^ryYz-L@eV=?6CR`!F-T`X$z0mH9O6*0IQUri5D~$=TniRVysu&
z1-44Bw4(K3$4rl7pl(P@Px~RsZ+&AXt^>u$8{h<SA7iO^T$Os;YoeAI1`4=wLbR+_
z9*&~j@co<W!1dd{wv0W3&^49(xJT~jG~K04wRT27`)7qBqvYmG+52F7jv_j;7FMP6
zhG^U~jOyyiwk1)vgc8De%zl3g07CB05X^dZmIW~^$||spMQO=6H;Ux4xc@HcrGE(G
z<#xN#>i9@p!UTwt4s~)}NJe+2>4q(q6Kot&YZU822nC|Q`$)~usw20h*7$ZqZUiJY
zf>g+>whGJ9HJNcf-8ED_wcODkNWHo~Q0lcfdl#~%NO5;zZ`1g!$(W+(7TNWXtG3g*
zHDLl#ft%~dP30Qro^!`;@QqWtt1ZSCS#P_rAbW+R($*F8{rn0hI{~6Md91=4=4OP;
zA@50v=R6w9f)va~&z|an)7%D5B#cHW5M<s2M=+X?6TUjo!B9K9T@s3_Vpbyfca@7H
zx(#;|U%3__M<dn20DqjF-f3R<hja>R9$0F)_0#9`_pt=dpY{@;GV<@Y6aR&JLp*%?
zrKjglprOr}DtR*$G^7oEZ^MR@rSY}O{;q#B15sMlnD5o|E5{Zh<`&_{mi7|}z=?ls
z0$_2SVKDXSRzf7fBHaDttBd*9!sa6`2Mgy1!wTjb`!Itb?em{_Ve$T0?0@-ezAB@S
zQ1Ax;5o*2S6J1cEaLYS<9Gl2j0a%<Nix%kYp9c#-(EYjK`>*#x$bZ@l{oD)51rz_=
z3yIzzr{U*b=p$4CFZO3c1it_n8vVT-<S*1E%_U@q4#+C$X@X5cie0W*mDyYd_XeUf
zxal2|^{QU&@b2A@SEY9~+e%n>VUK|RgjArz-T<X5qusr=n)KOu@z@%tnsq3|cU-N@
zijE<(B0>I2%Dvp064Q6sF@SBq9lYn>%f<t4?9}XSoks0kzE~5i*#a&RU2t)pjryh+
z<;HJdkMFJ*w>;Rx6OnF~((X-uGWIA$Dn!DOFq><pT<et4*V2a3lLH_dQIhTBQP_#1
zF@S;cve-!hVo!bl&Q0gO_WSOnK9QG%;XZ1l)7we<6}T!3y+5XEdwXWSlCWyEJ)5I;
zQcySIIf}X)s2vs&wKQ4vhA9oa!%<$w<<XdT8M)m7QWsm`!>EtzspOk<=qphQ1a_3T
zrG*wy$fXTr-FYCS2227<uF#S$+(fl+O)S4dSHIf{;QA8i6~M?Ylka<chu5e3FtZy)
z91FUp{gXfj2kOqFV^<&1*G1~@Tr)#7+F3mY@erqXP7kGV2fps(_};xu>yFff9h2Y_
zrz<?Md#gwD!~5*Ncm+>9=j9eF<d@t!cEv}~6?4b*Kmx(FhBoOI=n$XM7D=+Kx>jmn
znMESAdL5i1f*B%Bsz(pRdJ*NgH-4$j0M|la<k+k4O{b3D-7~k*IeQtKo9eSr*xO!a
z-&@$=F)RS~pUARha-4P5tllZbY3^c2s&GpO(44n;2Lm%@DLtUOeVUVOgG#}#nGx(D
zvJg7Kzd2f=jr55XlBA#Go%2_Atu+f-D2HbcWjxq={QmFa?*~pWZ-P|fKlx&95w`ej
zRN!Yf#1^!S*r>cdPA7+H{<T8?m**nCJ^5c)5zG&aa>aKW+tH4*u`y<nO8|@q>~Ngu
zFe#fIdM)0WP&{)Ls23eo6g3KTYWzWhJMASU&nYR6kC8S)^?D%MA)-&NsY0|!b%)-h
zer-;e0OY;X!994>9M;Xlol7;kB6D|`)qFoK#9O#(>FWWGgOzCOb>>k2HDcwmW1gWU
zR3WaXbL~zvq)i%j0kqKasAe6Y?Dtng5_rs`kYEOI!``QB`*5Xu3iVlN@y{(S#qcOO
zzUrh{wTP$;2@*A!K&E`{u4b<wJwMzFb%Eq&8Hj-%Hau$?Int{tH6)5_jSJYVK6NT#
zYF{rpIFM?cEhSaG@a~=AyK?N{+L0OfSZ2ksF~uCjpzc6GGhY!OgeU9~oY2g-WrtjX
zLUP;nVh%b^KqRi*GBl~-<xLm1<p_lO5^z>?qqx12<nzI4-Y0Xlt`^a}3Xsc!*4lp|
z3Q-SQfTnxU#ogl|8D09c#dmauax|?;2P5b_sa_C$>**6Kj^|*x>Bk0?soRmQ(N%l}
zxBI=w34Ub|xn^xbPp)U1x`Yh~_lKw1uRX~J`MUZ!>zB!8ewlA+h%c{gUcDH?!f#ue
z=+O+D<z!WBb8sW^z0aP=sI|$7zT=X$OG98Tk&Sy!&wFT@V^x;#Z+zI?lR%}Yy(r%x
zC4SX~+mnlHh(bK33stH-O;MBlaKq6D2&Czm#2lS>c-b$gAD22b^u$}*tL@vl@7w90
zQ4z3_;Gg|7`|eWM{~E3PA3GHHPlp2g)uB)rP2=nFXqD2T{j0W%1sjP5{w5FyDc@@)
z^v)_B_<P!n03-c}RHTD<3%%dqgDkw|pKt#79@PKcXENH|AnI0awDGs8DQp)2pAI3|
zpYJ2^gl}Ii{Inlx7wGSTy~|~w`h<9+ue?L+uaC{(R{OQX`J^>o(SAW-=xu)=ZtQZU
z0+IG0_=W*5^hK9Hz6ZOX^7DArfYJ{FXAg@6V9cLe0^#FsfW8Xy^#t)FjsyQ(A;7Hq
zpMy4j10}y^q96$6S#((UNS5b{5SCr?xJlQ?P|5TFuDsycaQCY*K5y}1O$a(ov%P+B
zQ9e)UOM{L9q}ZDcMAHz#1{kx;zBSHNRPVeKGM-jYFnAEQ>FUQ<JV*`|Vr};pjs6^F
z?~s5q<HlN%iZtvYwN)6lfxi$;Y5}&XRBwL6LsvKr7<<E!nj8&CA(S2$=HV66NFB~F
zc$Eq1RF5wRLc_S_O$FHdXef_!)`{!ETHO1#!)Z|-0Zd+P^>T~T$L`pa2=}}!D&RtZ
zuiU+Mmlqf{Pvx^*$q2H4>{HXmxG~x(1&TT)nsfrsP=~GvdE7z^^IZS5<DWLzrePR%
zM?g6+DQ_2xuX_cf&k})~m;EZY5Q>&@x-W=uL0ND<mOH1^(y%RcVcZguPUnE=9~AD9
zVCs$opi*cgGIcdT`!2;hF8u*QQ-iKLq&YMJ-YUc1T_$_^g?)rzrfS6}DUFWLPKVKB
z+>%F;rdAau*$p;ad4efaM9rg10DgW_N_nWBYTIaS+n%u-gi+eQ09Vv`>~825_`$(S
z%W*wgdzsWU&9Mb424|_qkxL1Q-X=inZ!pFTd^&2G7BvkjAlqkJn=l}Kx@1KuWiq_p
zv_TY5TTNJ6LO>y&W+V;bl6#|hfw!T<1e#V6n2S9EVZhNbxfJ9kNAjk@zMNk>&&1Z9
zK)ZD|;H;XwCbf2NgK~_oZ41~#zrSYyu=;M3{{~9_U$wrUP=XtZc0azRu|5jnF1?=2
zBCrs!8UJ-BE&c)}^B0xwu*(`9uE%ig4Xq`)#EQ1|z7db<IzFE};dKN<xQ)J9-9{EK
z3vZB<iw{9uo|xtk9ZwB*c?ea6V>g%%$bP&lVh4mgE+wWrcHX@)TP1k%m<JW)ilxCx
zGyTY2XckeIRX`7od>C3T8Nri=Z|umfyv#;q1!~w9Vyn^qnM9LM-xGT<-Jq3UiW%y$
zd`Z`}C7mS7eFrAI8UjuHF4fOB_o_&gN&vt-0eM`$>%gfRX9n+!#8hq_3&6Hl%D-Cr
zzLqenda5vi*efTt7|&K8yO$$;1`*G>-poDG0z73qJS1QX2_R75f)IxPjK;p*z&N0`
zCYvz5)~@g&=(^zmtfJ%1>5I_uOir3$3L~%j>4-qTQS~uK#SS9F-d0F)eBsaKAVkM7
zGah)EI!)wgBuD5LSJ)^_+~2aZXZz4ogF%$J%-;YNJvMD}age0nf%!EFJ5zzk#~8w<
z^Fwxqi$uMqWo>}@|8?=6^Zj+X$AbqM1vKO^UgUS_Zx(b|CSsmSkUU8ZzzmRzS;E3#
zy;m_5(e0&#YF=JCRe~~iXQ2iGv2oBG*#$hE0#37`rQu<&=}Wv55K!3Gcn83<@{lw1
zM(M~EhDPK_AMI7b_K;b*oOC;)NG35seX+eEjk!vRrMy6Q&b7#S=~NR!4(AZdr0%r7
zOg*FCg9=?KbW@N!waZU*19LfbKJ2ivDL_Y8DP|@b>OYQ}|GZlN>qOyyzFI36Q0I@(
zUl5DGuGXKH+W+;{8WM93fJOdHV}GpH?sfatrvA2Ce?RtreYO4{#Tq=@lf`h!T?`xy
zc)Jk@I|<{V$I9XbBnyl(>1yyS5F-x9*8rXE?Y<}G7{GD?)4*&D6$BFe$5W>DprW5{
zy7Lb1x@`JRMEcJ=vVi!*q+Ywe?NR1QhMwJ=3TgL(EKhAKQT}{4L$q)JR2D(V+RTUr
z@Zfvmyx?p7PPC;1P37?+*4Us@2t(QIEiNqpLWp2FX6|)4?-R&rHBP!hLxmEU=)no@
zP*On(19ZsR>4ZKqL6n5S3dnvT5@W^0>kcCwL}GivhkoB0yX3GK84M)s3zhn?q@J2Q
zwITC#mzRf9twQY~;19;W)E6>lL3DNsb6m)ENIM6=VduO&Bs+mP*;+Q)8dtmHZL4@2
zmPC5*6u>4IY^)BBk@~yGvt@M0$F@T=I&yDrb85cT%$sm~GIb&pR@J^E(AjNP8pCCp
zxg}%Pvylm#7B{ShCC>q@coeL&1|;**$}8+bu6jni$Kpxyh95QKsNV+sw$XRRN<#rD
z+qS#-*xSe<vnxQk*rK-Cc+l(8B#qClbA8K(w;)h`Q|F)rffVWXxX6%=bLF4=TjM_<
zlsA`XxDPQ119iUz3^E(z&-E?@SLalM!Y0aZ?<Y~cy=7mLwscdEymCzqyKhv@Fzv^)
zFDwUZ$n_=E6dJUxrIJqsdYS1FGSP&<q)Mi%Eb^=a0#IVS>y}20J_5V@{qi`Em~kOo
zr~ra=ns9;Y=l`SCdbl-*H^p7)hn8X4YuL)8-*LeAutxN+tMzZfun=r+d~<Wnb5>Kg
zhK)(+GJ&i_#vlC91<(C!k5Yid1+s1}YRJG@(15}QFtQMh8eER=^`RV&niM~~2_lEE
zEi|GKf~sDH)Yb&PYxE+emhoYMur++8QVR^H;Ys1H=OMi<i!aOgM0X!^MFVy>#nQRr
zJYl+aqqd!1f;UOq+dW!42Ca2`S4P#8<RgH;!4!LlRI7;gJoJQ>2a?#kTXlZ|-}IdW
zT7}xtqSVuu(~Yv9p4uujZp<Zl0i|u_sO^w}n(RdTtWPa${5)A;Tv>S^w~UqYPY71-
z*_2<#!q&l0GQBmOkKOxT#DMJQio>!#fGIj=2ey3X(x|@B0WJwIyYCN-ku^a*oJu)j
zsA<0y3sTFn05WewOql8prk2K;1~$D(Hn|P6$zP$<*HhgsDuL)IS0PDK{8YV1AA(=W
zQ)x(v<ajHWW5j}D$T*ydY|CAP4=LE)bUVK{P_q|EfhI8g%0A9;fft87oWy-d2-<q1
z1+a(!lGNWtt+2yR{98I^W(>+5q;KEad<N$;ekA*Sa&N82{&fX50?s_Eoppt;Be#i|
zq9N#x3&2-Q>}A5}LE=@B{^?S5Lj_UH;&3-W611qA7bf}XAdP}>%crZAL-&$~h{2u0
z2c*ozM!liWbv!NVVeN%CF~Rl(0_`4YpX=uB=>oA3a!Q`12O5Uk01WBjwkNiT9WeW4
zR!*LFogV+=YRzz*XEW#9Z$<TgTAyVU<XV8L{Mns92faabrM~YB&<#U~*dHzr;j_)1
zAsg5MvF5isVV)K1s`Fl@u!!Te%Wq<1m1h#mK`Yimbzig4nfva{Uw9~OzaRUPAWFP$
zzsB|$_>jSY?`nj$!E!dE-`p)wS^a!W6V8iJFxHP^7s_T4Oiwq*x>k1$Zr?v`INUtT
z$W<7=4--(J8uKe{LHIiKzm(s$r~PAr`~#WwGl`OmUiWbyUSGj`-9fjsLj#(>RTE)?
z7qO<6Q)+J}XJ5>;9uDaMWsDdsoo<WM2=ONf@T)=@T06LXKM6vH2PRweNA@0Y7J-qb
z4<p=VOa(S>ojFYg9ftNNuq8tZ%M0gQ)qx<LR>GgH`6y7tmR7orRh6Mg&CE3>*XfI+
z@x?g<R<QeMBmu_n7Fu}CNpO=gfRT2!>K#+b=6lPIfe9m7qeCyPmw;aGz623O7Ca0e
z8~7=lgjK&>q^2bwBvT$L)On56{lK5!`TNYA=8Ncfpd=W#tgD+uWK{p~d>_#}=$>x~
zTJM=tw?8~6jkBh~dO;{7*zt?kDPsAfx<@Z(U|ncRY;SwpHiYBz_ShNRS!UiR@qKUO
z`F=P_hQPdd3GpF#C(-Ho{W5sv@%ESFLoUe_I8g0Y;S2%uxxlOsotWHno|eukdr2em
zRW7Caw(e?CT}_t(r(E&wDdYs{3(FoR_!psK(dT;&D|A0U%}m`hRUK^9xTtp(Pap<v
z=}~?mLGWOxV|>)bJ$RMEn%GFp)Qp<&fk~})j+$3HG!1k=Z%$(t;XCZof<a#((&9Lr
z2dhOPCJqMniuvNLUXpji73`&eKMuO`4tbxUr0MwSs30JVP0Mq+iu-)CsKW6(Lkgih
zORV?n;otD5KraxHS7DG2{pieUy!L^`fOzcPU_`l(5311{3kdwr_64zU@&+UFF1Fsx
z?dO_%ICJYqtJ%jhh~fSKeOZDKyX>S1zY)YX=9eE6`NNP2?kXY#p`tc{?jIazV3P8y
zQANS_24Y5Vc5dqLynv6hYNf#vSCJq&F8~xif>K;TKV$e`s_%_q^IUI#ioQX0zs1}U
z91A)Q_Mq+04S3hdDHmuJaZ_Rej(FpX1=>CMRuB8>$B^(}WN&}X$JD_EZ<wjRK3p%o
z`D80sdcSIjT<qwN@-hD0w`}w)AM^g1r1>yp$iyd`*&gWFna^|l&xjL<%D>*le1<g2
zGYB(D@E<Y^Vp%N4$QBdL?8-<W5{GxmNEUNu*piX1pT9v1dS&8_MZdleu1ZD%QG>-6
zpRZ3d#5Z#cbIsTsVGNXcWFRGm7r0rUzh`(N{4EHKHNy*uA-o{de!jlVY{f_iCY(Pn
zDE$5dub&@?z&A!lIx!K$PK*>Ylm$lyKk4!Hf*Sk)IAS>RyT#U@BZ23`k!$cZzkOa9
zj$Hg406eciK2MCZ?{C0=fI}A;6#|bKz+E^ND}B8Mi$NJaWHg3Zz~02@7K1awl>veI
z*IR^lU@@Y`An?jxhYo+YpBGLTc4qjNQD<giWC+fkS%Tr<t$qC<jY&ZF{dEqvkAV0W
z?hbwdVTs&j8~s9O0Er@5|M)zq_)<Yy^9Ia`FLNHusz!M3po!lsJ$bebxDsHB$(~!`
zF4+*o`KPqguTS#xSivV7@RQ(p@$KR6A0}Xb6(Bms!8Z8q-1~v2yh(6Xd^;t-KGcZj
z(Iqg1e^`_MR+IVF)0J&O_!t>82%`C14c5op_<i~PTBwnHigiG*{?U+9;)Ut<&Hk!4
z4lH;_u9m6m+};Mx)RRsRqvTSdIHdJ{>`Mn@X&Oj{Ogf%I1;nQ79ZrGhs+(@_SAD}0
z?W8Ft2Jfrg0}aYH-HX_HCyid7PckERTpDr?8d0b0s;pdh69B<K+^|=@0CfvKi(pW>
zhBOe~RhUv<Y2i^#w|gtRE&UVL3m7MEJ5$%Eo$|XHINHJ`&|nKnN|H+^R9;^s+MlWL
zKqC$JX80tdZ=(j1SHi1A@UFdLOqwl*F6evl==cc&*yVe}_OJ^)lqkauZ$-hTMmVxy
zYrJl4<qd?vD2VIq79jsOiNz=jAeJsxo;Fq84-Gjqm^Gg>la&E;wV%oZ*M%Wrov&}s
zRM~ufbTY2<$)T|x&jHVU0_iz3xsZj6KYf{c<en|yXV?+!VEA-+o3Y`QrGXV%g|)V2
z)pI7~=nJC4PN#-rJDJU<%*X2EL9kVDSe?nO>l{Y|C&AT!U7q8$apq*RyB)d+&iat<
z(RlEJd^X-DAG*+pdOXuh@PKw3@H#gp#})?fS7@6C8KTFK<)e~l5Mv}@LszOzYjB~A
zpq-b%30}{)cL&iXa<jfZ2=8dy0ZNM40kOAT1Ls}j1_cbNaH^h6f)Gi;EA(UDrPM72
z=u2}rfvYdKn-#1jTQ(5m)5Ka_?o7|OmitXLozi<-=qL%q8Y^MOj#)*@CqE=QL`Wl>
zM+Z+A!q@W<QvcMDa>2Fr12X<o=lVA)*Z;_{6|FC7Vf}O4%kk>r*H!f6BKp*c^Yvd~
zdGT)>&(=j?#|?xT&_6se-uiv#2?p?=z$th<*4SUko31A&pTW@&H2C~W2Z~tVNRf{R
z0;H?)M`!}(0{!<b@S?jv(6gfOm$8-&m&+MIZC_KZ3kYzlE9{2@>v|YK2J)K=x(6rH
z_dMwHxXd&6Uqd6z4nC3*8(*gI<L&z~H~Q#E{&8RYO_Ke!EuPr~sH_LjzC!#PbvWvf
z^m>soWu|<bLmDN8NY}-%uZ?O67fdMZUWyg8<Ha<$E&(L~AFMK+2F&6EB>v#*<lU!>
zaqHdjG7EA!I2+2S>D9b>U@8rbb;rsfWD5bj6T22|o9%N-?89)BRNr2Gdz+Y>P~t<l
z4_uS*(hGzTXzcuIm1iwR#JIOh{A65g#5(NBbqFm5tuzYsFQ?F(>(l!Mk~)4b;Zj7H
zFi!D1xa-K<ImhJp(6iGnx^IVolB8E0Y?6Fe36Q~f$lz{%>&uvs#$vP(v=C8Hn55h#
zViflo#Ru}zmVTSgohUxVt*X_0>Fn!mh1e+}UgiY6eD#9cwu$-3<-+gA!|k=2UCd!c
zyp#jJUiv2<Xl9q%ccv}(?(y)<M)H}6+ROB`=5AwsQJni5!_?rw#NC$;<0YC#`W5e^
z8~3d4s<F`kuA5_2x<%c{Nv}&l)`m(Nmus8fnlkNA2FCg60~;*ap0pRJd!IVat63D9
zEwk^CRq^&<xVBkvbD4I)MD#&{g`3AV7<>p^1(KTRM*Csuj!iB<Jq6Ua9I(X_73>WB
zmX54rkovAG04Ir*g_g`}>HtRLJZ1G><w`M{m3K4lPSg52`6t#YP;rjK$LdD0@Cuzm
zd+kQ5?t>?FS@!jom*@5TQXm0vgUWi{Ytn|ndn_-DoF57GKxW78Y_QWTo`-XNa78<H
z)tLY3uUMfe00AwZE8=%`Dxi8HA@h4L_}9etf2-aIHiu2doi9Tw-jI}Bc(DBWPG(3f
zN)3%JvL0tDwgOEukZ%w`&62qlMI1mDr3GA(WP11d3{vc$DD?GGB~gorhpq*rkNTxh
zTQ8@3^$kq%Iz5Ya0wyz$_ch(^udhVg@#`4dBF70AyskiXf7Yi0W(MbRn3n@2!CiQ2
zcRehmFUsJ^(2aS2F%KGY@*a}gR~`)}8vNn`ZN8hZ<OJqy(K!45GK@m-xOpPmAJa`=
zq^Ga9_8E6+$(~ytPcBTr8v%C+n?6JK4>uyMXFe(inu|fkux^KvxUmpzn;kqD)K~yh
zJyN^oSv}~JXou|nfoJ2Y20me-PxrXL9hcTlGO@~#%3Kch`2@ih@?a^9OhD$f0vPAm
z@^kEH-JiU=oEz);e)%;H<S$p^Ad42-g1TLl+3=X2G=3)yTzRi1!qC~}1T~@kMvowC
z@X5`F#s)0RsWec#eVmE;nY@k%vi5s2U6~;Y&URY&^KE|{A26snQ^M&w^1_R#ktbxo
zHNE=ImS=i@_mLtoTdZw8?DRk@1tN;a4oWmiIiIwWf&p9a;<gLIvdO<~3TwVP@OjR6
zjLsSDgb4Cel8B~QOcPvR*X;d-oK96E8R@b=7GZht9^|r+R+4zBdGzwjNuaiezuhI-
zvY+?v>?xi4=*ML@5Xkll$-&Y#mx$~mwp<yHP#>nN0AI-z+C!i^dRrNf_x@F{kQZmM
z+J68JKdR^5X`9FU{&V^LKhtmRrazeo|5NWtG(Kh1%Kcn~ZAS4SC7dVJGFl#6&BfD8
z-EQB(w14eGf#2@>Lj*1$ke#W1GB@r%q=)0M83&!j9|0Z^ms#0WBvttpr=eU`@}JAv
zXQ_iS^hXv)3}!yz0POb<E9g3ZXK{d<10MYKSOM@}`%66bWs2vS;d=9zE(A>vxbOj)
zE2KY4RJe#|ANk8s{aWf_`Ny@-9^}8)WnOj&xu!ErS+=`#^U{tGJ^OA4CNZ5MU@w!~
z0qrC$+~k?!+<7JgCvIjdN-bj81$8QeSV{G{CwYX_ml-%6l81q<Sr1N<4BGPQa*$ek
zBbD><I*T2<C&gHHjQ9qWBJoBE1o+us`++*XWt@kp3%m?8yr`pEv{;F-SbXq&P{bCf
z3*Bn-FzB`Xa!u)-JoLi64$PgF7@1BWFqRPltPqLzOf9>85#*G|QZJX&r9@MT8X_Qr
z=ya-<+#T2Ixy0Qg?q#_GHIsVhP#G_I%6LDorl6T7*Pk$LI01=?^5>%w7>wBG4(!48
z15d?-yU0kZ_Skd0g!l8fJ4W<M4Jz3^P=`v_P`Z?L0_Is~VKQ0Q_hJ(dl)qzWw1y(d
zrvkr_F)hc$VyF1ven3Pty}G9CJcDxQx*)rk<cE=iX$wTo2TfMw^X3*n|E@q-qJQvH
z$9@w{W`KAOH$B&y)+-~$qdCJj3ex(;^Lp0C9&YKU?Vi24eHr00ceTo+33Ra$+yMBn
zxMOg;=^e%&9X^9gH_IKx9!~K#(x|7B_MyuX3dyb^&YmIT*bU^$nW_Gw1^UC=WXg{H
z$No;{Z1o`CR(jlR1B9R&1gF;6c!V)o-&cZDVOr-8?kwbdjJWfKf)w5<9IXBQs-O5I
zI{EUIeJWTemeXT|&5%7y3R}PNfVx<BAOdvR<vdZkoIc-R?d#eH@97_1CWKQ;#fM}4
z@7QNwePJ6lzjd0B9rq`Rr8^f`d0<qiI*%_O;q~CXYrx?ENgUAs{ma4&rs5y6$!WA>
zxK)-kGoQgv>H@f=EYT9eS*#|ah5@7qHQXFQ{F`oAokdl$I5v62*vAnV5=OA!b21yG
zd?{lOBDvRjFrVu^oyst7dAR$6kcEhOz{{)L89TA{f}qmSs4CYUpgFDsx-u6R*x=FJ
z!z7#4!Ynewg#`GE)IoQb=Di2XilVkOFod^S!Q8^Nx-14#acQMK(nb4o?N7tPHuVer
zv>?1mdnlt5XZxgOYGw_=Z4m)niFFT4GVAOk0cHH-zT059ct3zfYznTS4yV;aB(l5l
z?<y{=+xZ<ha&Ssd8dNR9PL<CV`8co#{SdR1b%0VW?82qKULIn)E>F#55+Iz@2XlA4
zh>6q1NPNJC2kfZV@hHBt8>FgjZtp=jW$`*2knd$zWN^V;<BIQeXMFfOP;2w$MNLh#
zZ9$N>!bT>$Lh;!cE~50sn};M-5REVj#H&u;^iBez<Ibc6GU{V8S6vr2r=H*}h=noH
z^+H)aNyl@aDbKf6Z`dq9?S2ar9C_Ife>k1=avBZkQbW4E^w`&|5JSS9^dLCel@}oE
zolDNfJoY|plY6+^c4?$zF{<5cX>lW~d%>$r64)kzby{XH)Oqi`%5@3YRD4KQjW^na
z9F|w81}@Wd8;&c6jgf^vvKvGXEzSi5q(gpt>W(!n_YHs4J>vT!M1ZUi?<rpnQJJ~t
z6ImYWs@2v#y1!wa?E8@y{v+vM_<!6ngGKibJ3r(e{2yi<^7j24*1xNPe_xCL(^4!s
zpD_?`{9KC9_d44(q7<#z--0!OPXAX?rS1HQzr}-5GpdLDvlN2>d-q*owwjJ!2Tu-!
z26|xNI_K<_{ea}V<wkdr!@_DhjKH#)zy$~XTS!XzCd&UUH2w!A><cLa)%NFwfq4NQ
z`d<MuJ{$l6n3^Z6e_rf;3p(&Z{W)$0K)KlM>uCPufB)tQ1P}%AfyD*(@rO<XW>KIW
z1&sg*MoRQMtMmNzAh<w&KK5_V4~*2lG^f9x9^`*}e*W(C0O#!&RR4c;e*W(CAfGsz
z|N8uZgY};btjNdW%4HW&+Y8KLy6(%_v*p%%dg77f_ScXF%6s~vPc-dZ?xa^(TYYr)
zFo~`m#k^UgcZN#QEoBFt5+t|4ba`6$;5E=T8E?!88X3oY-kn!bZXBbm(a6%J8vsMG
zsDF0%<D1EOX<AFLAFVw~!uI7pOA%%rF9>((qOx&Bi#EU<DxH0j+(5Iaijcr-YtNj9
zM7wPu&1xM<A6kjLWCgS8yW2xv1Uv#<e=~tEt9@a+3=yjH951y|>|iY;RenU@Gnt>f
zmvI2coh|Ly{t4(`u(P$jq86emZdcjNlN-qTN({W=+o?H)hX=ZL7Q)_j509tD09S0S
zkeVl5ROQVRX@~){|Na1nx|fw{(^pR&_z-f+9bo^<#3p2$Dw@9s_{gy*%9zI9C?f2Z
ze#h^hzP<CP^1OqThgNGn1)xtmXOm37ABRMUQ}QMe#F`1`7k}k30tjvUS9JEJ>He%Y
z2pgDp(uTZ;!|D2<RDqb;^k5l*CtaId(}pKmxn_BcL7Q9kEtE*Mx5RHUuz)L0-^xOK
zZH)Em2CI|BJv3FX3}kk`025wF;Sqw_Hg!-u*b<Eio7Xj6MMlK0+Yxh8((HB@W${x;
zW({4p5OE-$Z%r-&?0(o&3vu14TZ6pCE*{{n8i;}6Mb<O+hKSqqCUIf&oAK%}JiCNs
zn(^)~14Zf!ZK&yb=j*F8cx8SKuBVZ>Wyjhziz5QwVC%Rj^Md}wx$1=m1dl&~zB#JE
zDMdk*<qCT)Qd!@-O`sf7I9`9>_8=d5s|JEcaH`JocP{}XIna%5nX5XQQC~JzBib;F
zZJWzPY3`u5Ze{(+(DHMi^iPIVQQsi?Jxatw>0t^;^%G2OWj(V>AfZCmyVUWA$K|mk
zXu`cy10n%RAwN)8GGva;?jmuv>@vwH=_f!>U_<(L!yu}$EKyeoxBy%!nZN{8gx`ds
zbD^-P`;lbE{q?Bj`vQ^=F0W9U0~q_<E?LV}+M$OXCwf5XJzx}?9Wa)XFtkmya(n?7
zT(jLV_lR_D>WW#kavwJgjFn&!IzeKM2T5!_(?omU#Emw))IPmC#E0mLcVJV{ZNLGa
zeC+bvLlmhYuz7*L{83?uodnqy3d4vF3AiKyyUvdmx#Q0fx*jLMdFcRgn>W7PCvO}^
zAaTuv_xegJnL#Xk3c(SRa=p(E#AmY*yxGZUT-BG+i%urxBzVKc2)H>k3**P4U{^N*
z$9bAcFF(CS_n8wOXHtCnjSn*G2-Dd{$PPB1ZD^9p1#P3}#O2u=OYTpgty2N~eeuxZ
z1z|6(tw2R*1)Hl@D{KQTlAzU=THZSqt<BbvS3aCCv6jcaMf3dt^*GS;kDngcg0^p9
z^=B@^F9)UZag$L_9Nx>MOmrj%nz&<|y{h9*dG1jnz)s{=0JpOorx;{XxT(zU@w5s%
zWA3oD0(b`x8hI2s0i}nm3G_1V**g*4cLYi0?&e%iias@R4(>GijK{h_1R%=snsbMb
z@C&oK^pxo8es?~`1X?ttX`vFrQEp_A#yi61WN6?HcsQe!b&D0|pAD)1kTCkMqSXJ%
z&WHd1Yr^OcyXtQVqu^N;(Eq<AjDD1{e@?dhr-V@uskRY_K)3P{+ym2j9?P^k;kg$w
z*MxaFTa7iF+`cCLLY(+my&(?71>M8FayIAumBlwIVq>7<%fSwX2>aor+LJ%NYi7=$
zoY=bL?aiE*$||yltB==mPnYxv#*5Zq_98NkNgJE!u|f%FIX>;{<qZ=)Ft)grOnS-_
z%P(&5j+d%Di#y`p$iNz%r>%SRUjGV2lkA7XByrP1Qz(MR<1Q_XD(W2m#Jl7fxDQ2h
z$xIOWXyP>GCH8{ZCU0LdOSB=Q!%M~51V`EP2{=_=95AiEUX7JoBoFZonc<1(HbWVk
zfe0d2j?+GW8Mt?XIA>+j<#p!>uw^Kk09UX>)9oZP(;ree2<;HaKdqNm!7{KS{jU>7
z_r3!%0J(<cLU|W@06@aza!|FA``08%fX)0!7}YF*Cs}=(7{FKPZ^Q12KX%Vcdk95;
zJK<Ebd%a7$JRonfJ3%blHoOcSp!NnNpLg93n4oub$H0E?`C*PsOMc#kTfMR5eY-s$
zBr2Q_c6fWJz@mCLil+gYPfw9aZ`tr!Kf7nL91C!{^dcl&dGi^#63|6@?H8XpM(^M;
z<WUmL;^SG#;CBJ8j81G<h8jdsemoC0@U3iw09FHfo;**A$P5c!sdnz3_S69e+&E;u
zusWyR<Lo`7z5+i`UL@j|KUJ&U=G|%YeE=IsVf!)8S~@~Et?ZrRXEt94Q|T+EK*OG_
zc!<zAzi`LoU&g*+$3)YoJIa>htN1WW3|Q)2{cTq!>pZ}_Esk^c&v#>HX#4-;zQ=xe
zBww^49L4@k7x+I_o_zArAXUG3ap!<H)$qBb5XbVj6IqiN9x%K6DM0kB;3F$ye}pni
zUim$B{m|%v$o3cEA%MTPzZmmD%==4tNPIj1KwE|;@W%c=b^Rmolb@I0jQ1cD5rL(z
zW!`s&2>@5W2dgl_T|kKVr8oSEcoV)3^*5ry$L|M0)&FuQ`-iAx4}rdD5UuV1OWm7v
zJ8CA;+H?Jinrqz!2t+ePAo@liL^q@FK>{QJzy5|!x=E+YcDb5web<A^s*~wcQXcU}
zyb*ivhtlfT<^Iy|27FRP^}T=0J>f*`UTm@>*v)bp`o~L?57|BD6q&#nwb2n-W8wtl
zSyjf6-D}((i4kbTq;=ExET;Gr2@_)M1R`HJK@>NSCzn4=;;HfA-V9oEJbpt~RegMc
zA^q{F*++6v_t#J;fpfjz@rC|i0Q-Qmvdg;!tZcWsm_%#|v4->3+6xOwp_90+<}#Ip
z3A|*u5i}N|P?H;-arvQ20K=SY{DAHxt*ulq7}hZ_8^mi`)(oEz`aEZ$Ob(cZ7=5;h
zw8YP-mfR1}R0mS)&;$R%g%on1vHlu?Ui(=Wu-4u1Sm>6cn<@(-#}X_U&%kBT+_o%o
zTCl2YfD8y2R(@ucxpSg^JA=ITcMF;<HBbVRvUlyieUIat61^>kP2pF%+Fz>7)T}*?
z=jjen3e*gnJs*x6ehPQ7F~KGpl=7c2Vb=PjWqm&Nn)(h-pt5~y79iIw$fNqs(*BF1
z{Rd{+?x#o^_<goZHz#h-6^yvWRO1P;5&3hi!kR=Y@~8iDPFvsX>mxPQm`R0n{R(ri
zXKpX*g|YW}U+x<641t^5OEykN7V?^zr4JNnAmCMX&(e$DYn~j=^ksvV8#y)~H1U|)
zaVh6ppiHSWli+oblqDa>a_5AuWYXO$jzsjGtqL<*7gO`A(0I2S8OP)iLfIop;0%g6
z&ZCR9cY)Wh_<OwFD9W;tH{s5?j~(cU*^$_zr-h(;I7{p5f&0)7Vm}@Dwkydb5Ea6D
zm1KF2A;eRULU2PgVrGNOA$zJ4*Pj`3Z%21Si<bl!cX)R~p1E<cf!m;tNpN{2;lY}j
z$w^Pv$dSPbUy-7*rwb=#WS0J!IK_gnw<xW?<!pW;Q?P%C_&*=b{nfE31fSg)`b%oc
zEfdNdr1leFw`=2_z?DKD<}kj6;XfbH0hjU1SNwJOd=46`LBI0{56)QtK|dCjKm-92
z+VUY1Llm?1{psci?stC%PPX00L;P8q>7ZWGe-><i9d+_zY%*@}UpZ^&x8v{Euj%LF
zw_=q>fz*^=6R1`2QroEqDgh51&A<-)AfF*ozxg4a<@-PL3kDfN!i$}~{kd@a2Z2iQ
zYFVp(_G^6R(37$XPe)R54sXs;bKYG}v-^gTI-CQ|tY2PbO6r;^oi6E~J2v~-Cen!#
zX61bw-6w%eCAlEQ{UO3`hworah1Om+8sqy$x83eyTqoEsMT((Q4^|OJ(n<(W^?Fi-
zGX-h!vH#xajoc@peoRczRz3^?Coq!s^XHpP-ts{rh=M4+oZMZeY$Q%*a<Z#nJ6<5f
zSw#`<;%tn6R?xoaYk;Dt@Mv9!lDNE&kM<JfK2(UJbzTB0ob-Kp9rju}$i{BSYY0}d
zaqq|;PJ|d=ikQ3svucR~7J2yhTRC6n5bke<t_0%&j->E;pV9VR7#*Gi1rK~h&hQi&
z5ovb?$-*RtI0H<;0`V=gnEN4$-vmS-H0>-@hpuPYFnX@EZ5!sGg9D|k+wO90w<ZW5
z;1q8jnT&y`Z~zCyLka6go8XgwY&a+dXLrFvkf=zH`~r)&T%ZoGG8_hhd~)lB_KN*O
z?gM~mo-GI}aBR-BDg|cr&)DBF;bViGkr#x?$Rc*F=CG(Kw!|gHO#`T8Rk-9s^RT-o
zT|Sz|<J5`?uEhefx4F#YaEV0?&2E<7VEL|s0L#m{$#udF`Shi6zRSanf{Ji~p$+8h
zfa82vDY^A~Eg@BySE9`xqneX;TtzbIHE<|&fY<cdrr6E+{`R~f3vVx3b(Z6EocB%o
z%;Jmvz=^lmvF5yS_3M7iH^?^nowND-vh&4~{^xw4--ddESKOm9vO1gMXooZ~0W6v-
zgm<t1N>}??NcyW+>Qs{+?$eiklzcpw))~2g_4vA+PbtQ$kEb~Ln0<5Ydr{xB$nus7
zC;}#E{Y?%w>XE!<J+-h`?<!BHg+O6tJ3+7s{78E%b@upa9y^rxAP@|ve67#)wtPS2
zLphdidPsQPL7u|pUK#jKybMuoR!o=xnGJCif!W4|0@*l6%yz}x!%2^8gqkk%MQt_a
zz*RTn;ZK~Tg4zV5HE7epS>co~GMRVIh#89k(QB0)pt?6fj0Sfh2XcV6zJ*{wPQ37`
zeLV(TF?)x)5FTw;z|e`wX6L|*aklU9{`MLxF!t^CfS|6L=3Vdcg<N_xU4}(GW4$ZB
zU-7{|9yWZEvn69LSD2%nTGtSk@Zucr5A&pN!ItD)lL?GD26_oBX<Ml?aN{ZXnzqA%
zE-&b~2Zie4{D?1k(MK?a0Y_lGQqNiHjfTJQ6COiqz-j6&i2Asb7IR<L4b7w78{B*b
zgi_M21I?Al-wUl$9QgK*Jo~)~DnUp`&<W<cmyTj!r5HF4x3dI=*Wq7jtauLy&o5N=
zW!=b~iDn4}iZbWo$+vhv$T$u8<g7*oQ#W{y6WqqpSC9-Abv;YzIqqR;<dydaMv#=W
zK+l>`<&DgQxI=fs4K|_v9j4jsM0zXCpW50Vt%>GPhH@d>Ky717SL7}b=DwswwSHe`
zd4h;rxuMm_!1}ScQeo3}CbY)7qHcgPQ)I~bHGTeX)#}54uU6Yaf-1uE1f)ug28^G&
z&GCJW-zjXKe(+R%iA?{;wYvI!t^TCl>zneq;`~ypfBxG4xK_)5yH-QUy0@}+ow!7S
z*w5DyW}%~uOI0@u<u0s*{PCb6+R}ZBOMKIO%$gt-?&TL`U@Bloo$2!a^2C9hk%e}1
z#&e|-MyBNz03uRujJgNr(Rxjv!467vpfAZ_zt(=u`Ri=B2InivK364z?4vvqn23@|
zC29e)aGmzr@wsQ#E&<$p5Cy!z@#~Su>PkJ#yFuQNdoi^)@OBD?<gOHWQ?0c&5xyky
zCGr!JO+1{qM$_J;9t1|P;L`R-kGcqc$3FCkdl~zpFz@?er&n)ixSA9_lBORodIxfB
zu`4M`Fgwz=uz^90LV*+cG9}DKL_o1kV!^Uprz9V9NEjMT-p|#u``ouX7Xmue=%7C^
zhz-5WF`J^b9DQsf(7}j<12F6QTs>J_%)2($45?C{>Z?3q7flt|%&6Mib?JKSDTl*O
z!;EY))Q6nF{-);Z5c8$(Ku%&*y7%TLqMS}1_Hzp@iG9vl%AxNw1BUwqgcPg*3KEs5
zA&6_EGt1gdSMQ01AU^?!Ho^DN45z~!P0;bnrAg@6ag*|_5}r?9mUA10Xg3!ElC%qf
zXXVSd*&+^2%DXNs>&5TL1>UV{g@;HT$GzO&+F)ddVG=m!_(EmwwZayAK#I67ncIrD
zR9I~-=oxdQ!5CvN!yE@|rV}06fRt<P*diH18#e-s7;ftcEHR?>>cxRa*KZ&{Xkx+<
z5N7dqvHw=Brn2+}YX&4R7saEaHl^lO=*4*9f3-FLzE&#|EVm#s$3HIbS3XWebs=1F
zeeoE!zm^aV1ySqxoIW6%hp<MR)HJ7OuwM5`a>qua1T8St7?$j7N)g@p_X<CX;MHkV
zK0np@EjrE{7D$a(wWWS#k)ae}U@56xy0^J;sA>x<kNJZF(6p-?4+ln>S2heHdonKz
zuRo=ejezIwa<=7K^?F@9c!$o*Ru-l;V&oxVC<b)~>4o^snhBpBvH{hY-Eut0Fz{S6
zRt6HW`Kp3UUT+Up%0?@$kEK>elwlEl`@%YdhsA$?S%h|xPOS}U(R*+nS(PGuW{M%o
zF!dgebkd-#9AA=;DAIb-N>R+EK{@Z;{T#Hnrfb!Ja;1KN>$>?2UeY2(H-eqYkQ@!0
zkL8wmEH3HqZiu|HMe`o88qRmgO{NH<4_fMt+zxgET%Pu*x4N5+&%*SFC&JJ;aHgRD
z+8|W?RPT9+*UATbwD!h(*BNt~#<JmAnIK&CBleQ=$2GWMO)X7iP%&VO_3rSp@Fyvr
z+!w+Gw@@||6gZu7I$aFUfF?$(!H>u9w!rJ~D}6a0QUDlEK(pMiYk%n%hhiZS5n2L-
zp5KPTwl7)9;>zV~EDJJSa?KU`aOzaol6I*HP_7s|C3ErH%i^VT)!hWhv*%AwdQkYs
zcJNCRy0}$3Id8h^hKTU4<H6og$PvSbG7~dOzRA%jOpq6tQs-jaK@iX!PC#MtIv(%$
z`roS6_<yffYXDkdtExZXBZ5Wdof=Vo(CKSb{`!P{ZwUDx*J>D~{cx{_Ii?J}px@N$
zU%vJ~uGPT0_?!3YZG96~NKg+!$$GsSRJRV1=)6hOfhKn{sO*rXMefOCqDJG;`{TV@
zo?`?lPy1qa92F7n_WgK~?>a(mpm;x;hd^FA)XCQQsgJ7h6b7BTceb|Ehn^D30HS<m
z9`vT@by6y>*H!%1Z7!p}xxq^lkZ26_i`8HH$pp@$xcT;tG8by{oG6g@!gw3(hm`He
zp;4%tNt|wxTE5(mPA`r@4GD8X=8?+7TCztj860EUCu4-Msnm>YO``}h6p@cRR3~B2
z@zU!Ci|=G2T!Y)~eihmw#<NZvK=grPuaTzC;BU4C3FeH;KHtmfeX^+6n`lCEKO{lE
zkg+@_SfOhGrY0dc;kI-6Zd4xnM)uz6P{D)+MH-foUP73}(UBL}JLiOjTuwJAjK{lL
zXsugQuvDgbPe0^Q4#~eWYoAFpr*_A-Ix_DKMsUvsMxJ*dD^{8_%TfBSFXwv9-Yv*1
z@#w*G!A5}dlc+imjh;#0LbkSEh7pMq_S#O@N43+_V|TjbENovE{F5h<te#S;S{b^&
z_?XAkn$H*eJY1x?ocHP_M3I)_7d?<qcXul?)_odoR!gvP(|75rqgUUhnR{WAE1pre
z^HX4N9GJ@YvN>T#V7|)}C?irz(M8l^3ICI;tP7)mnrc#x#S_xN(aj*iyc;cW`Mtq4
z7@U|F+FDeGD2wz?Na6&J#@NDbup*>pf>Emf_}LHrU1He3b+2B^+yegTgKaLW_>MyV
z`m|)N|J<*B>H2?Pt9iRTxDj?9(ognqZ8(DXBrPc!7szVVhCAdc8qf{R<Aqd1AW6Ku
zW{o~MmiJ|+0#GE`3)jXN-B&%fGvHa>0VW$5-jU8DJ37<>0EUuUK)emOkYwq^+#WhW
zMIab#>Sf_}D%T#UN9uOn%?O(u<={?3z<eZLtoh2Rf+t4#5CyZ|bLs1)KWo<8RHIya
z%Q<>LfnEFuix5tAfH4OxT}jyiaRgvc9lz~Ja%318OfEfp8-acsE`|=C>GGxJVdJNZ
z7zhPO3IdosCLEGo<o4FT&k&L0JYlk8rQN~kHHLrg_~hQb6#H_2;4@+#>rqIM@akWt
z0VJVpjXb3QK)h6|?AI~VTo6tN2(SPl2e;b8@2DlQo@u0C2C<5FGGxXo@YJ0YVFBB8
zj;p-$bfca)7IRLRDU4mlo>xu%?z|y<yi)HB|E?&$11+#Ay=cEUaT21hO5gFj@)EbS
zF4*UZ)c_D(^TM@tI0j@uqdj-Wg7;o@L_7sWyyBHbj>;{KFOUdwjd6(t6u{w??)fs2
zQMB@Rs3cx#mOfr;IFTegpeyzHJoKneUAwE`+x%yiX08WrJ@lpH8wf}(dPXuX?Eus4
zhADM&2~#_4Lc!amSLp=K8c2T?iXON*r&EpQ<_zh~jB~Bz`yOiWr8$VBFe58`SUryq
zJkfLOn`1LLWR^hU+c@yp30)7n%Y9i~r8!>_EZBz-HSvEw|NQghaK0+4k4W_&zy$*s
zcKQ#Q|K=0{X6E<0j355{U(Lh$_G7*8f5H7)+)+==IXH%UXnQDxACDt?{ig`v6M$iV
zjT!-4qyni#XIDSzUpqu_0_q=<2MFCiG}#Y-ss3&<4N2lB@Qfz_jD<h*0eS#&_JbUO
zNPXboe%3Ny$>H-?+&J*3RzORg0CIvQo!bCf>io0W#D;|?1TGo6qXU!VvmX5TOxT}%
zCdC`ZWM}M`FV??(*?;#{@BZv%|J_%;`%!A~#|uh6Jg&v-R?~;1>Evc}w!JuHx(kuW
zBzCi6SK8snXu#AxtHq7F;XYGEE~8vV$SHY4_WjN0HWLwV*a-p)VBoYE=L;dtS(3yR
zY~VcrF{Kc*TOKQ#c|CTq!08vl(%Wf|ukTB-cA&gJzX92X&#ZR3=IY*#wKH*Mr2Dz3
zS%@P@zz7(nID6<w%FKd1DSE!+eEYp~JqY|iRQZcJ9AE~!_C9+Dm)eMc%?d0jESNZ9
zGjCETQ9Pc<1Cq3NFL5RW1K%U}A_j`QJP%>_Vj1(Y3$%P~9^%JG3qS|nc$2{s(gERl
zRB(U>^@TOicYqRPae<sPm^sLKDUys`Pxz!=Mgr;Zw+UwKYH#<u))Bwl6rTm<M{4n}
zD=mHvzW&BFb<EuxeH=U2U!TTYyvTMB=31=Xq8|I5q2gDi#R>r7+Puh?cf2cPik%_m
zx~#+34!z@I#usT6Mpe@_puJzga^&195OwQLrmg^acpNsPTGj{~p&ryq_l9dYz0xE_
z!*xRuMHH?bEHlHc>;SxC_cG&EEX|OvsG=Hde6;#nP27T=G(rTYU<cSn7*g8uem-R{
zB8%2e+!we3E{fF?A499Yu08AR%1I_y?d&7~#oG21%GHkRp&N5LuGMpHPD#6M098-<
z@tp99UB4bUDqNllr79=B+4L16t5CqRJSG&9dWJ6SybThoxR+@#FK{F8RrpXpa0fZB
z3W49#t80_|>{<o8C!g;#<KcC#zj1*6av6o<839lVKUdxZ)F5Mvb-H7M;eLGoxzys1
z&(lBHEPgv!|LQ#D?9X|c9DWx3TO27w`x4Iw%~fm_0eskLqIgz?Bn^OY!;c<LId=Q~
zp|tyNou!{cY4@+6rTU+qrJtkpCnFymu>av%`Z-E>e^dhg-dU=@Q%J4aor>hAztj)a
z<~G8(ij@naVp%IZQV^~$`tt_of_HfulbhXb4i$u4dvOSDrOF-)S|x|XyBr)c@y_-e
zZPP=CyB`eS(x3(GGqwd$>vb?|r&Oz^7hih7gUly))BB$Ds4B-`wtr<Wo)3A&(6Q1S
zo*6G?+Nn<WpB^w8svYh`SO^edDi{VBjkAO_r0dv}KYpIx2EfE20jFl$y9}7o4p5@F
zi#l{s6Oe-rq2yUE_wVf}fF3M7W@kkS-Ld1Bh3`$X4U{@$u*^csy5V%`eYhd)`{*gP
zdGD+Fp;O)FkkUL@O)q4PZT5I%Kbz^j?y%Pl5I4yStWkB;0eaznJI#L~LH*@P6ya7E
zwSU#s{M8f4*Iyk?^ve^-!X67!u<b|0QL1=dB&#Qn&W|3qFVb51Kk8^$05JX1(WKvZ
zG(TolNWW0lwhzYl!@m4rkH3kgzI8XZ1xRb(z6gT(->&~#7pNh(_xYoN|DvpYqqY3h
zJ^||zP;b9vnNQAHO5q7ykIe_^0@N1hv5sF5IarBiKWA46sC|bhpDz-dd`_qMpnS+|
zKgk4=4`J=}vp;25HkHW0?e`Ow^G!`ie<eA6`BSVj``mKFFcvDJyERg{`P%!4Jq{%Y
z+dsx`Z)^UP4~Y4M5Hn@hI171vOrqGmE+Y)Q{Gw@W?Op*F-5bbqPY5<9v)**mvLsmU
zlWuI0v~&x$z{NDpixsS$21!j4YM49RpC)t!kjE8i9&t4TsU*B>K}&}7$ch|H0-1|^
zYXatwS&ze6>}6zqK06c=zSfBHH+K(+Q2L-ls3<U06<49R+uS}4B~CG%KxXAiT_@Md
z0B7rg71X305fkW3bf50G_k6)1#_-)=+vJwZYP$L66<tE80+5JxGsKHjbbEo6Iyo7c
zVt=z<ve|Nz-NWsEgVZ6$?qj}Ou8Qo~pnqTlUJVt<=B9cb+5tkpmp|RMcf%iCw?Az6
zf40Rg>)&p%ol3dMr^-D(jpy!IAmn~*uD~|)>+N@F%AdB_XOJ$PdR4Jsd{cgs#Z$6N
z0ys>Wi|%RH9=X5fY1|3S9)i~mT&-ot420eqteS4z-pE$@g#d>EAkKz*1x)OEUgj5J
zWWwQ)0^w<VNj6)F`4+<VZ;WCVz3n{5W9X(&=Mqt-Hzuj|A=Icjc~bsOHpS4#3{gnu
zp+ophH*R}9gX^_I)zv&kZB1Zj(#`2h`@TshX=mAjFWxA&CJAv-ZjP|hu>vO`J&I-*
z0sM~=PT<~=)v3Ikxf6gV^_!jf1!yWB<?*&QZ-{(AV^ULa1cVIg5y;t!9XZ)ugLch8
z<J?vAx^4MN7Ms3`-I&BvJ_+zVL*cLN)utG8pz76F1Pme*v@q<*&2aK|r;mSbU;O81
z=O1=3e|2_pmyc*P@qRfwPp<2|E(*jx-LN-lU9e~tlKOYu%Wuxj&z0)G)V;9p&$0Pa
z_wpC#=Fh6BKRGu)$0kHzo`6pIpP!qbV{`W(oSU021)oqM>65^))hPE#LJ?G0g0Y;;
z2Vn96h9tn%9|VU?aOaV(v<6xxM2Au9gEq`DTkcO9Rr~|3Zx%=0OAiyo{7%gMo-`qU
z?(I{M<;EoYv(|7>*9*k*P-;xgMi81qrl7pFC#>^fS@8!c-$`MdZx=IU-Zk>plVLCK
ze?Ci~uY%B)e}QxR<6#21x$^zel}lPZjYiVX9uPf1;!R)a;LQ~fj`;1pL#BhyM=bk>
z5XFU64>aAqkuyLmVS&bTMdv5TRW^ndNO!NUk}<1{(s~~fz|H6)hesrRvbqcIfj_L9
zk%5)prWeNV?rJ=hqXj1p5aYCYVK)2f(~zW7%jLkBp9If`^0-%%Z?hUs>|9vh9^ILq
z*2MqzsQkq&<|nE4Pb~}ahl7~EdhUegqe{eoQ6*ZZaEe{nNXw2^nHFGz@D%g;AA6R!
z09{`7Kk8Y|(6fANBcKZ@zwcRYN$+kS5#g^B80&A{8h-PV|Lz3lPi_t19{d}x_-9_r
z?tA(q3t$@-5<Gv9H6q9RX^WtjI0~40;n_+Tn++JlJyd<D$@mJ=>+t<N-*@*73m$kB
zZPAeu3Cmb`FeNUKzkZ5G2vYH5`!G&@zpkZpJqm~ifVV4;$@y|8+`CWid$RR~wsn&S
z01z9T=KvImb>4l0!6_sa0)lVebM!kqL&!H3`%<M@@0hB1cfGP{ur~7sgfwTWoINL$
z(5H{5B+MYRE84xxmjdRE$vB_#$-j40pUYRZ>|89%U7>=|m?!RlEb`#S&$XXZg2WXE
zqgo?NmX&!p{=Aws!AJ9540e`sIs(K-nn#GC_J+WOB{L+C-VD-uo_xsWgC)>e5#FGL
z;4DY;4{v)@K~TZ^0RA<u^jJ=O>S9DVH{?#gT_YXk5Y9-WD`1F@kFE%r18j~~FvNA(
z*)4o#wG(>)_YW=HD=7-n8VmiKawv%ooJV(=Ho59t&fTu-9Tjh;?D&R>_A|ag&iK{5
z6AaGsjeYf$`F=l(*LnX|t|;0GGr3*DC1tb|3!GScb>9Sn%dxti<BZzni00N3J6Rt0
zm^?kO!Mm7t%AOgx_a3tE@Abv5th{{1d1sMU@Ki&b&CZt={7X`-K<c%p#OBSswSu%u
zHKH4}@RX)UauoK)h3;q+J=^E@)t&`@)DvG#09ebVz!Q^hNU(pGJ`HdoZ+=bDQ>;Ld
ziZyQDgFk6E*p%>=_i%O*s&enKqu9K<6FM_#cb`hU75_pi@qIrC;mhX}cxd4mmO*v?
z+lG)b^17In#rJF-Ujg{F&&W<C5+40h)PEmjtwXXOFw6`2#A<VS@DD&NApAu+$8S=s
zSQ)QFvADPZK!jBQWKiAJF7A~~fWu!vnqz6*-C{aiT@0`=+}tZcunWtn2B6lFgnWLi
zbqKWHAmy5SN-mGxUaMtN&!F?#aiPWjAy$pTtm*!mK}5KBWV+Na>yIe0scxt<?p6VX
z^e)WGdaRPI2fIo}hei<VDj+}(zxE?p<FVr<x$sc2hfH+2K~#uspcRyd!g<!pEET?N
zyieA;qqbA1J*Xnd74IlM{P?I?aTpnq^oXW9xMfAWDq5qQUEb$3k*NH~Sw0AyntAmW
z$%U2Xb}u?1Sw-cpS{&awb?facEEk6{63zr)!I*$ESl92@GjA12yDXQtc^qZQdZSPF
zXq-4fwr?lSEM0v*F4wdf1PlThj^oileGkTKBW1-Ue~&;r^LEdGJwJ4NjRb`WA!EjL
z>X_d3ftGT6BXLzm*)_i2)v{_CIU}B6{iR?Eg4|M-PnkXhHi3sqmbrIO6S3eY)W6=X
z#oi)LRj1y}^vxxQ<Sc3za|GP)h?v(7^81e686rw<>=9!>*oLRu+g<q#*QKM*tn@@I
z9%Z;#Z`7B0<oES`xdU8YdE780Xd&y+tt3ag4%~dB!n^-YU*uV*!<m@Xod|a0xyszz
zzJgRxEPjdg4IT*X7D3!q&UJ+fy4TU9`8VUF*X}`<27WY1y|{|=X*o&f=pbFJ+)m@|
z0<7O~|7RuN8?IH+|5V`rvYe-%5d_46ex4_`kS)Kfy)#@f0_^0$q<OnckbC;O5d_5S
z@F(}ja?U&TIr_=U@`1hE?tyWj?}U5O2SoO(!9cCq;paQM{Izs}dpq#mKajKdKmfD!
zyU-0;WI9<$UpJtf#oxg05CZ=hITUZ;bbp@6FQ3+rUjt*-*TmtgB|Nb}eQOL;w+HZz
z{WNk|1RspqFfky$91V`x&XDrah7jqdb6<v>Gwf)}uaU%0*}c0|{K^tJrM?Y|edP`+
z0Lb_bil5)=C*k7f5yS5$48_-k;g<xA-79$*qgL618~NE`PO?eMBBm=RUAJW9(GM{^
zQ~9tzM6e=V0)`M@#1m@bVe23d`JoNqX@{L|1q>l`$Hmo9rn%mRsg@d5mcLl`$dby~
zG=qH_tj>`yN>Dm?_OS7bEDOcNC)Xs&_fm|$us+f)!M6s}_j!MJP_p7B7^m(o8jM05
zLU4V$6J%cOBVd%Vhr0}zNN$lhtUZJ^hrMQMu8}H5sp+Jf@k+m`LQ_<hCj8@0l^-Zz
z!Xhe956H<Kgy*cr=JY%d$ZbOieE}yp2kC<Pc+4iYd04ZrICpOItOEs_apTiO&k_<y
zVq@?2*W{ADqn9h|LAIAtSGwDgc*GGsLT(MnLLTB8E)Qxz8GO)=V9H-S4mH;F$gBOU
zU(FoE;JE`i9oe=!Lw{n`NxAW-_xe$9(o5ig#VgNl(cQ@mh)G-9H%kIZqLJ<dzFZg-
z<8#O^A!WN`ZScY%H$EDl(xch6p7MxCUX3atit-Dj=VEU&^tbzg?24c(EB!_#$)mnK
zN&bl(iWG60c%O2I0|sG51A5h9=wMy};h-LNG}Q9T+uD;yy2#RqICVA*NhKd5j<0*d
z79y)*)p>CN-67h|i@lbp;%&`<L>)x=L!^FFYRWTsJE|M(*3mu}4bnF@v3;W9kVT^5
zyjCUre8>Ip(p26{3PhN<?!f^N>)M*tAYCxK_)bgBEzijzw8NM|_|M%Bu7VK#%z-QI
zpLo(Bz3UGwz`u#hFoh0YiM^v^Lz~}00S0I2&in)Nn)dn+rnV2k@zdzC7B{T8!cKjB
z0$wgXdQPqPFa4sy&Kt&7JIz$yeo{OSY{+!RS(AaL=RxRrpLZ|L2gCp*4Pf~~gP7nr
zFklfh?^lRB7Z_k16V9OM&!99zlbJx%uyYP0$^;Xl;j7_slt2J;?jx+dLIJTP5E355
zdr0@2qIFyB-X^bvi>j}g9C>&wm|FoBm<lMP1RqMh17=;szq07m;2FjlrW}T`>4JIs
z%A}k?C|S<XE#(>oNf_7@!mBfOLzNDqBe?$ZLaqa<e;yo>Y%Fef=mD2#uz+oEq~++B
zj6-3*53H2CcTeTA^fD7!+rfGEDq0KZq1<=f;=BRT3KM@WenLa}OfI9}$Jw+yCNArN
zMJARe7*0ho1a4^CYlvq~^5;VYp`EvPg}n_(yzHl!J&!YR4H_iq_f5}^(0|R*1xAYU
z`OYA_Gk9+<&0?GP=#AQ!Wnl`S(2r`9*!MoAPpW6z@u7|c^DSFp25mUE78mXEt`A7#
z)j}aA*&?!zo4Vz$b5r_xzg>b`L$19?uhzr491`>Vk_V!OMoj?a^cxHTiLmFFy&x6C
zAs6PoLM`NL&&P~$zrIpJcO`b&H8;~;HCW!wsFK$w_@5x8i7c4YVW4#;T#DikY$;aa
zqPHKzjRzSW{yCyT;N9BKA-k$Eq>gmhOAxDFbC2W46o6Yb-f!Fq9oEBL>t5h1k!?Bl
zoGd$@OCU{3Pb=I1#d7akwf+ZJR(<rFOZi_`?l0m1`b%`C1@pmiOA7}%&8E2-D_RP2
z{kzKjwLB32?aD1~5b6aZ7T7-ntQo4h8Lb6GwRg{`cxKgWS7g`Y{qIz1fI<BR&H7!D
z{!G)r7Wju%I@V9$xbTp(dHr}2kE6<y0SClJ?ntWp(^2cXp9?ay0g&tW<IMkuij1Bc
z$S3ycl0HA)k1bAzKy;qDmGVwPdC=(zyjSu2W=BiN$YabA?nw(0FSW;KPPf&>j6LuB
zxm*_NN*79I#OK!nk@vt3d<>%hfYrHYrsl!tK7&-&k7H{(?w*!_JYF$KpttUnagF0@
z^nrmJ(jw{^fn5Nb!jCUnU>C!`@%g@1!t43G<gcT>z#Y%K6bla@VnFc#8xbO-MQFTg
ztPPG4o9iXYSP<2-9B!68L6sqcZaN%ilSupK)_A&B0@l-;Ldh%cnh>9fl5pIE^A0le
zxI3qNu-#^bNa8-U-><_u%@^i0)CF`W@4aTYMSNa16tBIKYjKa9>1#~$RAi-MGOs5F
z6j8$dWTY9a?_!Dt1nH|J0<-MVdpgo9NIvCaF3T!Co~eOx+c8}L9hD^ZS!}MuYixb(
z&J=BftHs>oHbLrQ#d&73qewn>jtBR#o-Q-R<0ZCsgK6QWo+iY01D67LwN!1=TAT&K
zt9DAXE;&$#D)lk6iA=J?YH=ae)jxE*y|RtR)UC@KwuN0|BU1sqElC1Xk&{2IFhZnz
z)^>I5Q-ZJ$`sGQ-hXX?MH&kR}2)^P^qU(VyKm>Et+GOO&ZMt3VfgTqi^d5zu>JcBj
z`<PeQ><NpC$;bYAks=JVrV7&7^wdK1Czas~lPv00ESCU)p)lGL)thl-{Z^%Aa3ZU2
z17g&GU(Cv(`rZWL>-)_8s|6W!PhWTEU-PFvNpv=9o=(A8Fi*ydSz`-!ucQd-{(a*D
zZ|3xlf$8)^LH?M`ci$@V=~4A8;l*)OEw%irMrSU2H>@B7GQnaEK%q)Q)&R>J=e0HN
z&#pZOW)Cr~uk-nOJyve&93=QVW23#(yk5vlUyv!J6ZU@cjB)60PjB@rx|7~;8$}_R
zDNxNp=G3RAXmsg`*Z7&8U<{mRPI#kw`d9_Kk;c<Qcw{U7JeZH^$x+<Ia}JqRQ9v(B
z35wKIK+-+YS3b;9ffL%1AgEbiDY``)Y;Wp@;R|EEW7u(5qn_$TN60=S7JGKXI-%hr
ziY`im#DNi=Y)J-wQhtDRS`XO|_rul;;73(Oh`U;$xzmF{h*+??87RAJwEG=sf^mH3
zPURzcfcCyC`JR}O*7w25lDUJ$3zQonH0FPz@w&B6qyBP}FHU~3JBCn=i?uMM!BD#g
z3jjD}eOeMH9Y7^03`+E>-AO(ZRgiY{%^!`dlM+}XACWFWx0N`DoshW4q^>U}Xm09U
zeeC;pjs&|J#$`{Ycube!ro^xjB0|2r5|X@th5oidV1Z7-b_%8JXDz)549-4Ug1(!v
zD89|luMtCA6O%d&qP=&f0eD;JMejCgC$r5wj>!hf&pjb_2NeQdxByMioH->;AtK7~
zJg)q)Bekr`4o}Fqd#!B-|91PfOzX9CXysj?ysrSf`S@;fl<_6#+0nzBpa7|+Szxe~
zgZN$;Hg-^73)^WYunfCFLOa+Ry}>U8xeXH_;3sp}$Om$;{#<UunYziF{{U^ozZ<m)
z`kTM3-iOZ^o&L3YJL-5`XMo1;*&GsP1Mh8${qL&xXN(RXo$rVgyRQTh6j*t`RA$J<
zZPFE3OejEQ3da*btXP7P+J*qwU%6<f51ry5Oku~YaO3^Dw||Q%fs4C^GKK>k-0v>#
z8$6ib%D*Ew#>D}Iiv!s68x-H~ugRZV@(v<5zrQB`G<Sy!yK@E}`sd>QtMU=PM(DMU
znSi$N>fgS_&pX^fUfO@<Ekeo$goXc?-s0yS?!KD@{2e}qO$yg5ICi`#0{4x93mXb5
zzf@^P10HV~XW1lP9GrGYwNx)xoVMKYkOQU}55TN+ebfGZ!|RE)OX@}2EuDh|hEO5T
zBiV<S&XC4%Cy1}U*C3x%2EE>cSsEjs5DO=JaU*ex<Cv4Z-^Y;iTrrpY*dlWds%~6m
zcow>xY#ue^GS|l<5!89VJ7&fcF{<IH$jt%d<^lM!k!(D@#Pv`ix><)~(;b$3mzTtq
zKM7cbm%OtG!3D0+?YRdMIlorqA==-?9o)f+K|AMRdXzcs(KwHOEzkfj&DXHWf@~c?
znqNI)U&C~9tyG^=s9A;2aYjMRH7M|Ip&3Ks*8YBjSp~?0T=Y!Wt~7I0n2hUEi!<Y|
z{GKLV{M?a@&}C}vRaf-+XuV4Uve6f#cpCM7ToClB;&%h$3!fKD+nKL^uuwUx(h6XC
zm*o(@%F5%2nk&8WWTkNF>67erfzTP|5k}=zLurT;!T779zfa<c5(0{u&5CHBfC4J4
z>UQ_o=Fksyv$S-0<~1MqUeAXXF1LH+%G{I5^P6~g?Qvg#$E3`0-g=YjLjw?pO%p@U
zhC({-xt+*mC>D|Eit$~9*m~KY-((c$!VnjS3&Z01%d&<N88)5N+IMEz;whksAn~S=
zQX_TV1OQOpB|<ljJ1-3i@qw(-y?TJr0Adalt&bs(M?2|O=FgIYWvX&WHf_J%(noLa
zJxD3~P5eOmIS3?+IPZkQf|&-hXKS3izs@vp6K2pa!0oqL_4%J?SN_qI03zIeL!A5?
z)PK$C*>Qn(={%E1`*~Pb=LW<u2-N8HSNj$qKoZ(Q_SG|!DjUp|3oa)eTp$7vQH^ka
z4z;@m)I6xa>lCA*EiR=IsnL82u95rDi8R^YPvfAy!D8Q?)AMP1+~t{EazeO2B^Dqm
z`(6JaFDi7YU0w&tyAAA)?0O)mD~t*KcE-aP$w~HHWgi>f6ma+I!-&R){B;FzvTV{s
zZc0pbPn*LPav#Elu4DSy9OQo9`bb!+=gWC3(Pe+`w#~(0I!9nxCR}oxL=i}0PwJTp
z6$|eKUmM4!c$Ukv93E#t3{3VT<esGi6E_N^Zb51-?L&HLGL=N{suJGI3qd)+&XP98
zfweLupo2@W0mgf^3|ERmuGDyt#4=Z5Ey&?Lqu!eHIpfy}avS|Nn5-xU49DXTq9^QV
z!yv&>!Au&HYIC~0Ks22+8N{nIkgTadLw31_G6VexyW8|7T%G|*+43`TmZ^;;=22o)
zPDFdDAXfYJqRs?^<>;zi(Xd7e7PTap#!(%_eTQ$NxPBNicc4-|cA>J6#8u`W@Gax_
zv2_ls1mA0mAN&#W1)vcxVV#(z*Rf~Q6`1QpzT)GmQ%j!DXrrOI@&x*4#5c#oj)I(7
zCTqQ^KOoBiRU2Y$QJp(TF+n=_YXv~GiKA6DzYGOUJCJ6J8V~(xH={t88Gxr@A`Ae|
zeoH8Mr$wO+x`&hDn!>5Mr#*R~M2z`Uo5bx3xnrLIwg0fy_*c3O{4?dJzY`kf7=+ra
z_1cPqNG%Ot-+1j0x(#TfzhZ>GD(9ybyRN=KyuAIb5hxt4IkxF~QXxKi3;kvhc=r13
z_Y#Lg?_ifO+6<g`UkTba0j9HW^LH>PxVK2W6)wbJ+*<nzzux3Cd^0J1CEc`WH9P&>
zZ78nm*sk(zp%0S2A3KhJHkkjl<M`wg{YA&||M1)VSsw5&r~}w@ck_(`rd{u9+o9?L
z!~S^Qt~nSlF{?O8)dPakxhurArJGrWh<=IsiJ6cSevzCh4Jl(SY!wnk%p{`CvO|3`
zV)F1%?*ddkq%ZYuu1|Y!DlgQ0d$PhRm!~$QdK1iwbPGBs&fyME2SpIw+3KQg`!j$R
z1nG4+p7thY+F-O7`kh^r_DTdd7izeU9QU!ZfiDydUan{wQPhW131nyRVUK`V!wVi7
zW0KUjHx8p~=R&?tVSq^xIx-`Ox!XGjT+`x%YvA&6kx3^y>eZMg-FU@i9#`{KM#r9@
zA~87G5Uk`hAq;bjf;nFIw=J_0^s?!b;ZV%;u|s;>jFNP$Mb5z|N8$~n)E|`IY_pvy
zZeQ&Q#(fbwDmEAKjcm#_9Q6Z^$4yScH)`{OY{T(<&w2H{3ouVvSp3ugWKgut3B^$y
z=0~a7y)oaIPZ+RgAGb!snToB7&Itw*3n#j*^uzh7osm1=Tn?5BI3Pvewf?q6Z}0Ax
z5x;3=Pdp2!PnNit4Sm{UF~U%7LcVKrMcLQpx`5+W4(B!y`#q?SLc1xbP3h6Zu`_zN
zW(%*(9Adc)`?%R>M(KN($g<ms@m;mpBfed0lwgHc-+5>GX)SkszObzgWcbsGbTFth
z*Oog?Gr8YoT26<HU#@O;mI|iEo-Oz<!rZ-54_^-k811)aT$n0=G8`pIP8)69Uj9t2
zdyu#Mw(Iz&@PfAE({<eN4)Gqtc%7JAVM;+qb16N_d<etr-{<B1em4PmrQf2s|FG+T
zWcHlJFU(tDtPpgO(nYx45pA#U!&47aphkE5BLQb(=Iw-waJ_Ju)N=JKZNteRl>;Ie
zz>mDfOE;U(0Td+=2<3XJO&=Qfpsg}NUmqNTzaD0*xz|C2cq$i0fZ>j6>9>ELnzIY-
zzH)#1)i6@<qwH}6XPxRMC@z%Ao(k0dDXG)7R=LQ3WI@!nbG6a(o~~_RN<$3l)!Lm4
zW>Y^)jLJ3<p`PfAEg+S;k?Y<>=T-<T|I{A#LkcS1ig+JJiZ+0exzt(I#S1gtd!bs=
z+A44v_+VZFN|H4>qswX-dl*$fJV7s|MBIV`;(hf!AJ%ipb%t94viCE`3+x-*O2+5g
zpP_4Mf-Sq|RBHGzOBb%$P0|@KXhu{_8lX}V(&G%F<m+(|w<kC5S3V^CTPFzU6n9{A
z?vWfXLSWpe0Y_df^{5-`06Ms|f-%xDWwtP0pYYMPL=?9NsRvH%h2aw{do$SOL0#xo
zz&ZwbTyYrf!IKqc(d+;S;E=t6#Cca0qOM;|vQr>lY&s*`X%Fz3*$zv3zFMLGV9t=H
z*4ZbpPeDS1>f)Z7f?@RORV7-gUOXRp;3?)g6EugE*F04h-@A!@75#bRB_pPZLx<UD
zbkazbdoV_AOOVT$(=%kl%{BhSV4?fGl>G(Z@ByM%))PB~5D$AjwzVJ;#j=B~pe&`i
zSl0_5m4{w5j?Be|1@Ws_8?|QsOdarR*Ky{!|6suJFLWJT{gLRFze;ogaT_k~<F-Hu
zr#n+d74X1&`cb0GoxU%{Uzh*?R>$#aIClSL$01m#{x@bEf3xHGf9!33EKGmmV1k$r
zJ9^~pAn#KWyV|cC9Y#>G#swnOc4!uENkpZ+5$NE)jt9BnpDd(&3b{!Ee4VJoZF}Xn
zs_%<|F!ovqaPq`iI)DK+o;l`_hc`=nwUFm+*I!ITYZo98?YG=fkgA4IGZ3+SStNhM
zk+Mj_>O>4QId71Br04bk>|bo@jJm78!5(+t(;oi>_V`;z+}jvH4AX*x=@61<k2AKW
zeZyQxx{-Ur(@)#>d|0K>49ekVTC$zF3nk_E=9vL)X*7{)l9k2Wn&}2HNfr-EGeN2a
zMZKbMcfZ7*ZG=x?!JB04R*Yh_-qjpmi>io404d!~40pfTSRqOhbum9Gi$8@3eKRh7
z1VaD#JpW(G&5%<1fBs?ru_X9kc+^kC2V@rjlKh*D$X4sBk(59^8Aooas+QXqXa{2F
z-zur)_l-X{Iwya#VJ|vZ*?z*#eyM$`)7Mk?3q<$I&M+K8bz6U*zfpA5^ZWjjF=_dd
z8h!VBJS;qZ_{|gHL+sQF)5ZldR=y^Xx-h`kz?cHsVUQ4Qx?{Z!Z215Bl?lzq3iXS^
z9zK0PE6ySEF#%B>dw%NS5ha*OvB+Gjn^0oSX1tH>&a@ImnC>;yM{zHM_RP&;a(>^2
zJZ-HPoxLq?jeFfh=%hT_Rr8qJ<wAlEB_Y8h^jjszSBB)zfRR5M()>osAza@Xi3v|r
zVRK|s<JWVq(+xeq8WGy7LFBlHc!gv^SQ$f{ZNcj28z)<Q0?z2-3<<$3BrAMSrwyzg
z+!Y13A3>uBN(JLI*nhR`3HWV=tc0Di0V8XQz=@Mz=K~P~RNjAdSto>2w!b@5_I?pP
zVf~w*|Ko4A@9>%b{-3`Z=X28Mf*oCd`TPe$s}Nl9d}98Q>oYaUTM&%AfB5|C|JJ7r
z`|dOT|F6&gfAGqFaG3mPFtO2_*}h<?8zXl!J(qih*5=uzq`OS+u0oU0qP%Fyi3Eae
zbjQ|ho>u>e`Zw<~+UWt~2Z;X+9SbB+V@huCZAiS-iR_mZ-iVxTGMGHuebZfc<?$-<
z{6n~QQQo_q<AiB>8bb9ahsp1#z<*LF(%;-wE?IEDDzC?h9!iMB>F|!q(BWeBSo7n@
z;#Y%->E6#%`#jLWvIAcFi3oFIkAN<xM}hv1(V+np`m3k8b5|U(W9k%7z8B_$8nKE}
z36G!v5>={VHMTuWU{rE}H>9CVfxviUt|o1yoH}r(@?37cxx82Ueh}_-o>0Vf)b`ic
z7TzJv=ug5vu5%R4|46g(A4UfM*+zYu_21Q%KT5y;{^|PbZCU!+mH{RCr}A=Um8ycn
z`Um*qm$n6DjK4;GfMdgSzM}<vFtP3px68n{icoad&$#LHYb**7@8Lf6>-1K?l@^}o
z6eksh64E@HuT2`HdBBWo3cO<zdZ1GMO<g)C+b?)HC@;pBA+|vE_ETPf;r>-}K9o2p
zJHh~t^k1iHU(KJ>e2*0vbip;y0Csq?2z$HgWQHF6+Ch?B2O!(}47|->v@g|<?`Ze^
zTl(4h@)5{rU!mRbcloa}XEiV`4Xt%aF4N=-{TS=l<_77D>UhzwwX;qFvhh=r0Bby}
z*m5|CG&3e@w#iC=S~Dq<&SZ~!p4<e&WI`h5f;}^JCKHHj5!)!oghLV>S4QF3CE50|
zZkQOchbUTHoK{-Z_A{6{DdFKS5DuBJ#d|O|VLD)t-KeafH+x+RuMCSa!!t!G_S2SK
z2xqLFVuRGrRYnGE<HAKxg;iCw&|8f;9NqbZY9sg|%w~q5jt*|QI<D@V`hXo>HKt>N
zI<_`o3xbS`CJAY;%K4_*%z7mQsffBuuAFn!dpoxXForLSxwK<LnMeG^FJKM0-JavF
zYg}!xnG=}I9l;kGjsSi6mE3O+D}G<Q^Dz1v#~BA}l|jUpoH3D1me+!)hXBGw#}=Cy
zgYk4_UT()7);}k5Z_wn8qSG5p^#Pa%^gs?carjZ<HXK)8UY)YAvl8w`zHrQu7gJmH
zxK*#}4iYl15^cjerr7Q~lycPg)Shld2fmNf<3(t%L^zlF{1(z-gScd;VKMGiD?<OE
z{gnb)Ha=1~#BN?d{0<6M$*q8!@Y4BV2ZU4UJh}<`n2Lr1h2&tojmb9#;xJc$iFn^k
z3BUlAp^<L2`7<B~qQ?m%`z^{Kj8PA5)e<~DRGQOyM_^%bxA|LCZy;F$UtE(NsAF2D
zHzY)+Y?V@HhC#sg@W7ffSPtf{komzpApyZod|!4C`xlR=&=>r~oc((-DL-;&?c11-
zQqyM79e^tmt<(T@3<67Iy}AD^W#}jFtgdwF$<buLAgGQ#lfWqGPe@Ac%`C#P6TiD1
zlgA|)X!}?^3EQRUPy>4D!&2<9eJjk(w+P<EW*k1h=ealpzySlr*oAwq)#-hIgHgAy
zR$V{Q_2scsA8-&@H6D7oxPhQ}uAcoX*=)W)<xlQnuSdb;UXGjrLO9ss6J{W^e9!Sf
zz#r-r^sOTnyn=)0EGKODLQCr6iqhaLs61)(3Eu<j`Z^&UShIKz;%km~qFU}nXemz!
zE_V`+uN51-p+N!Kj2vbhn0sqLCF-y{TID>pptr@BwjLbwt(G(9tSiMkC=rik##&?1
zCu!VYd_T)TDVw1DK<4fKa@3j4JBz`v?B75V#`iINGc}4JX6yRIPQeLGCDCA89R=}8
zGDx_?7CeX7lzE&YN*(sanRtfrL+zHnCAwQbZl!DZP1M(~>xja$Sz~NILE>$yJoks<
zO_{6K=dEIdL!=qq);c|i$~7EX>sH&}nAo`k<fBfGTya&kBV)MRgE+xbahV@GY@>Xz
zm&|Fbodo0d39*6Hz-gaxw1L;4gc(!$NSIM!G1U8J7$9l%&GQ`qv|UyCj?+^_c4cxg
zjx%|uh)mH$SFUtt54K*Gt<H}#McJBA;pf^2Ch)D{rwi7eP{a}>%t3A>mDB=B%NeL~
z+-yqs7C3d{W^#A)#^(>{Kg(M+NgkMKy}lv*mUvBs@bRWF#EFu<q}PS!Pg~}HUg6K~
z&qe;v`sS}H|5wek0G}<W?!Q<{Kh@p)1LbP)A6EC~t^DM>1Rc$vS9crYXMd^iKLdjJ
z58t(bSR;_I)a}Q2vg-9OUwi9y8R^20po^mppo?GkUf`Vhv;XHsb=dx=wU;h~ZtTY{
z0QjyywE=KHcB`M`EdU4dH@qhVo$jl=X?E9bgxi^pX#h~c*{&flaZs-F@yQkpp@?KD
z`(hb6>TznGH_}m*6UDe5O!Kds5Wo#?W_s>c1w^Gf9t6X4-Fre2U}rfKX5BmpYr$ZL
zQ$7w&fHO&nMCP%|uHlkh;h3IqBkWQWPeTM!00XRc-uoC1^EW)0;|?upYq<16zJ&L}
zGJs=#9BN-JomlRc@jm$n+#4BjITIu(D~&9I;vn+Jle!8BHO%Sh9i<j+-MQ*T@6?(U
z*KOrN-=Xm7_z{HA+4;Ny;u6s?yuB->#h?6?0QyV5l@K?qcG)Ah^%&b%Z2=Belhlur
zb3YdD4lu$U3ghHMzMMKG**_{D;)mtuBu%;x^Yw?5A3NN=J75S`($S?$<!!rujpIgU
zAh$ABIMZ-P1zj)jHfe63sS89cmfigp&IzmeV9V6p%aWGOL)vgewWGJ~jW^J~ps(tj
z`DVgjp3{kgbj<L|owbO}uSgiXJ2XCu{jD#+ZBxh%L7@7X*0tqbuM4&ryif~g3P8GY
z$c4bQ==m_9Z#^ejj~5Eg3G?y_pZc*q0$Hp}cN76>$?;YtjS$NrAT^xqhNiS)x_B>>
zk4Q@H^MieeW$@r$jVxg=nJD^8RAR68Vwf5HRZ=~q1y7K?xXZzu$iDW}kohDl{a@DJ
zY|BkF>)Ji%D)L+hGy*k;^dJPH7jQ-=dJqDH;9CAM_BH(1u)ho2OAd$AcDd57Z&!_x
zqau|N4p|X;*Lv5Q^O+TRFao9!UYQF}G&Z(&Xz&2YAf&bLpDIX^U27f4q!vjZ98Wdm
zNV`uMAAGf*AYZ&E?>DlE*>8I&WJb;K9~pH0nB@ID$opYnq3<5_iS^sUKJ@V%BE@rP
zr}JTs`fUFzsq2T*mD7{uHMtM!G61-lK3Na*P-$2jN%{SdFh^)^&x4Gl%JGhY33XfQ
z!UN)D3i@<pfagjNLIz=;{O)9@xOTbo3#dGsUGZo!VJmIdNM|<pZaWdl?Q&&qU2ioL
z5@Ur#QBNf*nmiCdJ%pwclXIvOhmq|%wM|_Do)+xn#@p+VHsArUo@7zqFz*1qGSH-|
zYBtz=1`T?If}~Y%jv!Xbr}$*_S1;b5)sxEMyO2cLhu-j!!`keu+b{7(n+gz)L|Tb(
zLQ`Y;C3UmLdFMJYds;(anHZE0nBi{E+;w^0N&!rfAjOPM%EI=V>VhUyFlGM<D(|T^
zC-fi^ooM^LdvqW7V;llVk&v`DM@J30;p-H}_m$TtrN0sB)AmWrX`~}fI)_UmP28rn
zxA(AYTc_Gs`%qB$hq&Gk5cceP^0Kk5hk`ZuYB2>uara$7TmXgo8Xv$;MNdb<?XoEt
zj?p+Y)u6?|j_2Dmb=jX_+g<HxSjp@*lEaph9qM`~hhn{xAv2FeuBpPo!0b$7W=sZG
zA`IA25I5eVll1Vi37W?ar#1q*(L39G1?nndvT;ezDF|aBG0M;hUo85O%Y{HexP>0`
zY>#8((@MMhc@4b|5_!^7;V>}N&2HnFxD|qjFX=&W6ZOtV3@NKA+$1##JBVU>rh47j
zW#Di_QqYCtJ^lpItU&|m6#_3%BH9MK+v(=;RlJ{8@BEt#KN+s-9|{z<Ynt>C7oSY4
zcW~8j>iTc1`cLsy2Ac^WcZL5sPPy<F{{9e#=D)hbE8kc7Ph6)3o7zuY=gETGa8#SX
z2h%Xsf`ak&&d$H6)B7I(5yl5eRe<S?uI}O!^%oKp5)nu2i>j3a1j82;r|}k~5~>YV
zy_DCFaMYh3E14gj(N_+Hh5R?3(dQFFWOzpZfs&(jm^Rg^>WFiic(@)--&dC2+2{dT
zLpHir+goGq;5eDl-OdGBXRwLf<R*iNapGT&WSg0RA65`j^q5{e6g<tF{mS~??OH0w
zZkn7;n#b(4qQIk9o^f=+9O`(<H!u}CgYW@RrUgHE9;m;a!<nG(sYYDo2N>ONez(1W
z|MPBPHL0I8puwIGP0|sIsuT#^?HyVKb3l7R>AIZn+O`JGUXu=wOzWLMczVZivzc^^
z!y~uj(U`=R!j3^oA9*(XsOl|GTOI*obQbXfG{c<~1eZeup0$&e6u0YqUW>`7s*73J
z7|J}z;gT>AAmAx-*mjH+FQBw-MPV<64MR%T+Z~sEw?;s^eqEo_<9=n6*J(>R5BB`@
zZ&bMqC0l631&voSuIK~N!d3p9*^k$zS{tZ3keDv*Hw^x$hj2ATbzKlGbx)R2D(+AL
zOzPm@{9!+vI%`9qtUZ(4ow>;Y>wWW!{aG^tGqqr3Vwc#an%cr$Vr+s@h#x44v_eFy
ze`9rwR6XI0={Lxbc}jh^g8`mxgo>`g!W2wwQ<*{tU*hc_=N+?OAXWgPW+YG~de3Li
zjp{r???7kBOH76XopGRjqi<MnsrwrymHYfSW+&3tDuvKJwcgz}G4uA-JnQ{pH_s;K
z?+lrCpw&r|<-Nd7qylD=lQ6}!wX=87aj4oJU4gN^V#k{8Uo6I7I-SxA(S4-$1^X!y
z<6kDJZ>rGD{kp1u6PvzOAA83LW=)@&?m6A1X=6+P+Et5-*233+m5=@_W}7CaiUTfT
zH6BdK4}6ns6%|k2%b|W=AQ()q?)Z*^VW{ZPLis!w)W*RTb*yIz*zm=!@kav)BW$J}
zbJk$(hXHShVGsB4)Jf0G4V2bJyf1VnP|4NDHp98x&+6uNnC20pf>qdPgv0KhAz40*
zm&1kE-BhU|uP}*9sMC1H$`=D8Dd*te%`D&;4ZPBQ=K@;exNYgEo5R=#?Tx#S+vF5b
z<EC#RVP@?xOsMw+z#<Oo72T(UW80GC>o$l#(wMBu;x#zi&FmW+aOMP{AyAS#(c~bi
z?154AsLdmpl}*#Qh+nQX<RKF{Zm=jSQqUndz_{~GABIVU_biMx5tKMvUPXP13~ggb
zb2pwbz`{dxIp13{RTD2A3@9MWzNOP9#|KXqOtZr7;M9K(Ufiww>-lKL@$-6o`JGHd
zfAjD_Lgm~i__g8hY@*p*0yNsy2g6zIkbCq2cp3ZL*LcRuBT3Exj*a4Ky)>N?+M8^-
z+?SGkm3eW2yZUl%?_Op(3&?_C#YD|S=9S~T$$@JT1g7Mq2;p5pNQfEv3D13XOg2GW
z*gi}*CJjB-)BZ7R6{`cReD+)+brDlG_2#s@1&vb!^7Tp+i=#`TRun_EA^|ZRcv%-=
zt<bGzcqwnRq(N%Vp?@BXo5vKJnbvUZL^(TpkM<zA%f!1mDr7Q~XM2tN*TxV>S83nD
zqOrIFd)@RlOtAwLu=G!p_rI&RK0%;r{Fjx+WB2&HmjAYZ!`=M1H5{z{u(SJ{4)J{r
zhiH$(Kh$vTKdj;1tbFJ~;fgK%P{SP<I7VV!!d<xb0l$pEIQ;wM$DitWjH^KI{jrXJ
zeXL}DcvkN^{>H!kLuc{l&+7f8e&{a#gIZ3=cZfvd3mD5J4;HQ?kjA_qI^d+!svA_R
zSIdaK&#Un0rG&`Yka)<DGi8(y*Ddzxvp59acIiao-1E1)-})(^lJIhgi^)XTZsc}|
zR2);{Fs??pI{+11Q80FSTzIgl%43^ad9#_rOTM4kvu;qVfiu)59@W~4l0jO!4WvY|
z9AsTc=g^apM?AT*V!RX~@voDmiBu0XYfL?4-FksTfW61l0oS&Z12Qw;K5GRF9WD_@
z1TI^mq}bZg#=@iZ<gFXhFULo=JwFql3z-s%d|ZN0F91O1t5LzVv@HP*z(x5Kn!7z@
zTn__?OzSAqfcFS94+=;XR4`e7+IJSxoL)iD`J2a0oZ(Oqx$Sf2onhPY=9nI)u9PGU
ze`_j+75wA=V_Wmy5G1O@5s)NX9~ciC<;0<%Gl)>Tn%dv%bvZ8QS49&xvAgrmHBxV$
z&w<OoQ_2iRkKh6%@A)QLA#6N4#_Tgu?3HA5Dl+or44MsBd@VV0et42FMPy$t7M!*O
z9gBE=Cr;cvkhNe#G})5s*7e{l(TIlwz(ifz7Ocncumj>&dhrObT9G&gF8V|Y&gp{!
zosJK!K5!T9cFLI$j4{URb>G}49213pjqhi|Squ~Yq-!WhE?Z66lHVlg!uqZ30mR{A
zrpp;2OZ9j=0^tD~rg+eA@(?ZVXeVHzTlQBR>}Zf;fNlk)=ZJW@@L-MrqSg<p%x|?E
z0fZPpHsPxNhqavCN?AJ=tueMy90-pZIPmih<cW$p|Lg_%JBpIO?Y2fg^8rsj96UL0
zU^|Q(K~@!#D=&R;YW*u9I_TK-0W9+Z3}1<iM3>0It){OhMmVY`V(Xow+5)%{3E2<4
zGLJx_pIhU)m0gM<Php9zheZsj+}%7}pg5V_cyrElx!Kc5la;QK1ThC2m#Bj~dG67Z
ziVBu_veq*ROiW|C&|a$Dup2PX=<t*Q3Wrk+(X-<1iWkR9scqEPVIzfzBj2BxHpnB@
z+i&x!cS}<~>Jq=fmy>r)V4*T&(m|}}hF5$h2lY^jZZAW3fyFKJdI@oqs@}ocUW5^8
zc4)gh0TY=A>PI<+>*FAZ0<H=6d~Na?QZM?YX^wS)pKiL4@e1qFK?81bcn2p@751`p
zjq`MU8R`qsOo2?5bGEZt^+s~Ht9Cd{MS3MW8jI44j}b=ufNi`YCGo3p^Ap@(Imott
zaq{yH{AWtf!06;(U!4Mh>Uj%7?wZ;%sJ4SQFIcD|0`sbLg3z?Po1kJlgX|}-J(pr(
z6-Sj>Tq`w1;_l_xGpQRco<_K?OdKa8VOes~eV=QusAsMxV9TDQeVAeco?l;ByWU=>
zj<p4VZEZ=ry6y4Kn_g~Ci(pN^qxIQXZ*eHxfztA1kwCK3k@13~U^-%Ph+&ZqX&3B;
zu-L?boYS5}i(Sbcz<mEaI4OSf=7G7qx;8RTCc|NIHN?T?cA%QcK(^QV7~U}2?XTO%
zIUL7Bx1rNb{7<!<Z^sWONY{=3PLv5Et$z;C_+h2~+uF|S@7lh6^MvqLeNmNv0XM#P
z>MZm8X<PjkWC|tiyvNiCLw7dkxKkpX@OW#00lA+1g?zyu)_lydK_-uDbzpss(YR9B
z7cL`9O@U`TBhiJoE38mp506Bi`j%mBKfh3A^AafgFm<OX8KMh+`X=v<b`xy@!TyaH
zb*}@K!DGV7pd3>}`qq1KACm)cf!xh+-~C$?0S=WP`BOjOr&ug|=XM}prUhIzYbz|p
z?0C^^2m9zU;Eyf8JcfUH4D74_^m@HtF66)Udc9vR<iGWLy<aZmzx8^(14BUng#oV`
z^j~rWwniunA1EK@gIUQdbO+Qi2p-7K!9q}^8g7(eXVc7m8};g=a1HZuA{yaxg$OzX
zeTHlr9arD*mAeHOY8<sqhm7MfxTw$5Qe@~h0|#peegc|z(m}no&EsIfZ949V?8$cN
zF>Hp5*LW);NjDdluYj);v0;U*t=-vuGvtS;w9Lc4TAY@@>~84>H;+Jv0V?1gzPY?(
zaj$Sr*;UQ8_`J=pbF4TB3>ia;O+#=mc&}+SngtxAhn0Qto=~2W9n%aw4`x-iq2vpJ
zMmWO<#xvEKJ&HT{Cf1(@kb{Ezg9o{VEuOZ!$E~BH0ednHC>35NwNEyqdwHqLxsKxF
ztTT@tc^})c)3GwVKK<&YkwP&9oB?T!1yKfOc~Z<5^4^92oftyC=V!_H2t}r(_4ixt
zUy!*!NN3GHE`l9`F%VLi#(R)GIFU@;D(2Mm|85lyh{iNHBrp*nQu|Pp{>{#3mG{N_
zqbaBWDLmDD0E*#g>S$?Kk6Bla_Lfz93+;u20T~o~N2Ed@hAG|SsuZ*oKCagq=HK|{
zb*}j`D4zTX(OGjBAZB%2cW45716nW?aI>No-jpHQ`#Dk2rpvHXoNo`Iw!Sv!&<P36
zI(b8X1cu8%Kn4?j(+Fs90Zy6~NPC5W$6y?vq#OnIrr~g+M=#e!h<p42t|vWV`Q4=x
z;|HT%>=XH1NwcM=stvmQfU*(0G9+=3;vCaA$WsJD&fu|<+Dahu-}7}QQxE^Rm%71z
zPFy{ZsxCH?D_uaurr0RX8xO{H=%hR_k$CC=HjzQ#3Jz)NNJ`%2h#ZCEZw9-6SWbBM
zcSGLqOUmyqi|ppD0q(vb4K!K_&#6^?xg$CnYi{g+mHI!|A7xPQ!h(EOmj6a!K|U+X
zzb`C2sKEZLuppn6<!fO9H5dP7Bn=gy|4$1GBEKt(qrSe%#z2)Uzb+gEKml+=`yv}t
zq0oFh28HGQ*qlX&A8|=vVif!t5k8@YGcH8!$6)--GVdG(5Vd9V=_AwuvaF7FaCNc-
z60Oe7x4-rT3pNPtc+~+QA&@gr5M$V<s31zyKKA8jaC~+E?aiRny;N8{|NKqff15W;
zLWX8D5g{+LfxLpNkpUk;tAXJ-MN;(^UB5mmyJ{=dx9|RK&kpiBEW8E~@F2kd9|0Ws
z?GM9XV>}}8lY%68hImijndL)ROlQ2vSQ7&WT|pBv&B%rg@esSn5%y%W8av#*fo~{0
zfk&pe))Kid*&F27BlK{+b$SrtJ_I-Tf$qGRky(tk3&DEr#DVX;j>}zbcdeoBiOrRT
zS@0f0)wCXOlimK*dN0@H`ohQg8*+q9WL&8w24Ksw@}sEwV?_z+Iq*#XL{p6j=cJ@w
zkEBd1`CebS`-SNNxLWhh>GYGPnxl!BRz-tVMS&tNE^8hS@uR*3!ajIHh+K|5U!sE<
z7nm1WS`k`G0<C*=m7es|+m8vV0(b}ooKth3<@rjrUT(+Y=nXRDx3>ibD}_|r-AS?m
zdg}qHntY6p`SaAzIsGVtae=A%;%1MLk;RD=|EbGm1=uUSg(H`X{lj+YcSnm*zK@pe
zHzyG<g8uQBSydc!{eJEQ6P`DebXM6fr4Ps-8Qk!Q{2n!YW;W2LLeyJ$6`*H*$GZLd
zF7Ugk5)OjfI1|`s?_b6%s4I0F+?yY}B%rx-Z!)R+?h;>Lx<lYT0N4cV$6p^SfBg)R
z&!_m`c!tR5Q~drJ{`owL^StX*<GI6)aa-u`9c&93r9;C`7s1@_Ds1pFSl#o&8T8fW
zEDR?%bcl;^C2mi^N6r%qp*+6&P!GkUKW<(2ezsx41xhRq*~7eQ*z&-qs1rSf#&4$s
z^nN|bFTC$x@RLEMb+5o;Up=_%R(?v?i?=(qfMC8<uFR;;Xj=V3!Fq<%kV)8V^y4mH
zDlEfU53*Tmd)$k%grYm*&feI*8Vwsp@MZ>A#UMCy=1J@`5#JRu285@yJi}Z+*P>#%
zB7yf*uD8cRj@wCP*Zs<e!%Ng3(e^+ntGR+`3*8>(rhN*F=Q8!xKi)y2kg5>eA<8}Q
z);4om-qLI3?2*?R<$L000_@GZZea!D0Oss9i<_<G!~JMI(4Hmk#jM-MQ4BoDFsehm
zEHQta>tb7=8(6eW!g{Km;|;k(Z!C*XJv?p!K5DVh-7G;t8j2ozS1=w5-GLD;e29R5
zb)|zUn?)`=Xj(O5%d1N(ZBt#>z8gd3pdi`R*~O|D!Q5nY)viA~lOXOdG}STyw;N$%
z!JwCY5mU4|7p7cS$4+T1NC9s&uoL@Y!|#sS4oj|bS0j0HmI9L;!FqBVhbX_19S@-A
zd!o~u_QZm0d;cm*D(s;ENyE+66@7%b%v94ma${uR^k5Q&)J+!xZjrGE?d$@#fF?d>
z?GeNDXFg_bHhElBW>1R8p{iWr3X!wa?C_`kR=z>x42e>{P<}?Z66y@3-`j`s-cosy
zqpulb1DkexR5UPhMm^CD7RvM;wC4K$y#AX8nEtz2Mm~Y#&@KP4h$2xycUdJwS^)X3
zKl9HS)UEykG1&hC9RD`ELTJ_#r{?*{56=|#E?51=@ONV3<5^EzYt*NLL7!5wtW-LS
z)Vku^YC1Xvfx-Hu0P0OzqNDTD4YD})NE#2cyp)YpLYeg&er^L@1Mq!t&@YA&%0RxD
z2&O0TY(yP>10gWUxxB2S)3D2(j(U4~V&SVj?ocO0NY?3_ag_`^+AfkIz=G1fB>wrb
zGullCbY^Bx0^$Zj!?TV?Q6Mt=^xf$&p$QJC(SR9=YQ^Tru*4uDJoPrQ^1{e&wBqWo
ziWfMbI3`bMzvz^dtr->_R>+!<4uxkv?dReMb}CxnMi1Kc#aD-iWAOEm$u|#SxCf`!
zGS1Zzc?WXhONML!^Jp)x-2!#4OYldAt52EP?eR;oyJ-6x@V-Y*@ly2?q<U6ANGH9X
zI;L!qPy*;`u?IqMe9=S5nIC-SMWV9zfE6f2;1z|nQ#bvMv}Ln>N|sG>gopv%??jr8
zRIhw~)~IGl&T<YzdYFzdFuLE>%}C_cBo7c!?Jp83h6k7t+~liSW_*cvu-McCTLy4r
zVw_=q_oS|uP?A@Cmx@<Y<h9O2swfsRWjQy&=}7I)J~I!*X_Lt5l~o`f#rE`kF|?zy
z95w+>cQpGFU7`*In#`1zxEZ+Dym}c2B;l4t(u&(1@MC3C=d-hK_Scxw(S&xLS@O4=
zXF#>l6NB)l%dV!tKF=yKaq1{vw@R<It23R%n6z&EVYeNof2!*C|NDPT1uo*2{TE-S
z)%fVzqJm}AV=Qh(75*g^Ym<(DS>gZuKXthL&;O18=l>4>^MC)6{?Tyt{l(*lkAK%H
z@Ta%wc;0+kMI>4#J+v^g?3Z>C02F)&Q+pdRTtR+uaH;)rf-rOpj8S0<31RVVK~wqR
zLxPgMAU^gIJJsko(1N=x!$R6+8$vexAinBEY=xGnM_CGLnLpi;=<@;o#y#P|sCnIA
zr{i7b^LyIh@C7gExU0;t_Ci~bueWslYon}QTpK(+r?U%|e+M8Ph<Sb0;nrN>7IF_4
z#{TS@!D0)<_<9F|7Yw#PIede`_ga3q>Y&tse=J7UK9#I1jlC_92sA9)uL$)89>CNx
zdp9v(9#duD;r?i+0r3v6zsEXn91JthqH%@mtapM1E;b|9$7Mm`c!+ib&cB6!Qe95>
zNKEdsYRxe2!@~G_3$Iv~G)`tTFeCW-@8;ohNZ?iiFBgxrMD^{j17Ea+7Q>u}YkDpr
z@s<BN(VJ)Lryqm>pZkC@;WKJ#Sd)G9bqL_)A3ps3A%J6qp=0I{65kT9JOw_Yf3RLL
z{3G*lh<q0j_b;?q<0Bysx<~~U96aQ%I{|sUraLUl9lnz(Q<|?ptwc#JSSzDHqx*bq
zXbzPz?w)SZa98YB&Je-!g&(JRn5gX@Dc#eZq-HoO_4Q7lg*)&-#iquiK~0LI=$FAo
zf`nC%vm~jx)fUwB(sQqe9dlr;bDg2pNjVM${TT2DN$->ZV!<t0<z2fwhlf;R_!Iij
z*@l6{=Y`d*48{t;VkE8r)!v+vqn#I!o~&EP!(4BY%_8pR!%8`tj=(q^kE|wu^C&cq
z=-E>rGG&*qTRGsZfnwn0&sSZ|GS;?1rr=JkQ#TMbV(nEdza02Uk4_QT%k)R!!j?(i
zdgu%OVyK4S_mX@*YOj7q6Za5<B_+Oz6&(grDZhY`@mbA365@WSx$eD^5D5f7oLE;Q
z{W45&yPVHtHtT=J{P-F9@sm;nhLk5tNZP@(xbt$6hx;?#U9D=zh-XkTES!cxG9x#%
z0i9H*btr<ZnAJQ;8V<eQA&gEfY#g8a9IIV#cWvr)I}MP*mcs>WOCED$k8dpUe%{=u
zNXV`(PS?khob>)NCR?%Z@6FYOY~ga6d&B|VuQzP_U^e0_$Kk%*AMX!bqk<4adF4c3
zpF|7+1rRSk;@2oxG-XvEaU@>1XBBN>P2{^dcbn(usYdF%Vk9U#9#7nz_e!rKTUs}!
zopQ4w*%n_)(4L+*JL?9)geXqzRDu=!Cn<rzmv^ghQ)hk1gM`~YCkKQ9DVSYu%mnfw
zT2oTDY4}=NLE~RBbsC(zY)!Uu?XR1DWJqTsG~-D(FXtbP2LIbavispK`9rn%hkyBf
zA$jjd=5N{u5nh)iNEhYjnf9B8`%eeu7pUgXC*}Y4phSLkQvPobN~lWzdQ$%3K?xI>
zFJ5|>5xpl=GzDpO#_PjO4=zV|P%CM`e13Th##TS4*qN4Ud})&3>7T$$2|(gK<}oZH
zZJRlNExw&<3P0Aqcd-bdF1!Hp8bDgeD>y<N4nx3h>0N+7|6Q8*FCw^7P*^}=0EOf0
zxo^T3ouKIQ(<>k+)sbi8tLTK|>Ypl>I|SDei<{#VUI(Zb5IG9SH^|W&^KVrfJS0AR
z2XK#4ud`6XKkMrZMVRwJ%E=c$|Em1F{y@I&KaxQ#Axsf->RzSU`{Q*~Z))Y>uT)B@
z<H7^-4>qxPu!zlGp`vcrlpea}ho=gr_kF6c)D;a8V(N!i7U?HYZ)CTKxv<J{1Hzg6
zLuH=a(GZp=vAcpHgS)KINpiAO*sdAOqJ3_PB&ncxmK}XdU_8$u5CFPGZy~(`^1k2M
z*BbQT6cyWQrr&X8<gTx1RF_kwL=l{g*fRoL*7)=juCyntCGM_q$c{R8pC}=MWy@AH
z%6v46e8g#Kzxst|c{}nz>f{(?PoVdlErEi~+^$sj8jtthBA)C~Vs`i1eHGR-(*hhe
z2aDt)8%{~ms0+P~!@CfcWIoYl-P)4aXyg1k05eHVd5+51v-<jB{qe&}!%kc?wvl7M
zb`~os`~CHa8T7Ibb}G2nzUdkMK+f<@z>IvP$->6;Z3Tq(Ai^`P#xA{38=L2@1-FfZ
z=Ek&lM^$x(Reo-R)|<k9*xjB}xL`bEXo*A~BWUiBTf&BY=qVyMfZCAkTY+{8l=5zM
z67xoV)+*OJuEZedIKL$)Pd?h*obDZm@-}Dwv4JS3ej<nGIUc~mOP&{3L;VA294zTP
zfuiVM!xd_6c6ik#5MS`bs%E{wE%g+1JMCeH`0F*CeEV6-iVT3r-FT~c8Xyahs{S0g
zMoNx~n(71#KVi~*));@MnIN@O+1}MH?Gy3N)F&UMZCjO;=Get2cZU=A950(EWRAo+
znJpuZTy~J*M&rqCPnM55aBGlW!sSU-fIdsz$Fj`}7-|gSNKd}C<mrRV{pB~A3hevt
z(RNKSyul`4R_gRGYsBwrfKa__z_$OqOfd8>FA<Q^`D2qTLX&K(%*Sp3V8aUX6^=ZC
z0_q;Gyr(kG$3^0$UcOkui+4M&QWraeLsNZ8HXVHq%7w49$*4I(K9QOsu=D<=fBX9w
zh8?*LC$1g8fcX1Hn!qXi!(;#AIX7$hh5Su&?11F>(Jg=dPA_4lK}Gyddl+o)OOP`}
zlymBOu`3dm&cL|GX7Qut2a|%y!3^y^zo#I}pkCh#fxrE8@H+mf=UlVF(%KD(PJtid
zWu+{&gy1QHA-8?($HPU+f~dt=VK)$3j{y+v!R>B<e{DOHtg7-#uI;o=W^g^->T{9$
zyZfa>04k;&F0e~uw{RS8a3W6WR+QVyFK4T3ldW7fg5J9Hf}(XY9or3<ICBG>ZjSP*
zl0#?~)P08Lt_|<VdPtt5Knm!zXPn0gSG(J=x9aACYKqXE;b;7a;(K`qN4NlO&|^$H
zhV14kkd9S^5|H0;IWLNQv3v|e4h%W~TzArx9&F~@cBmk_<xK03GEWAjiXg}CbnT@e
zsMx-Dhz9m*V=nC$`5^6;A!v`BMZg5MSFz&@G0)d=ma4V5u;@O#WWw2s*@#w1f`w6O
z-V`bx9lTf|&+Zxx!qJ3b34kOZebG_S)YWcb*LO=RU7L(Yb}+lFc4jdTUUTG8aT6TE
z(m$O<|DjwRb+@CjWd@5JKWopZ_tDDJ&r7R4c@X^X2bN7W7(EQ9)BM_whyY9KBO6Da
z5Wlv9H}wqsnkZ{w#xKt$@3{MFv(YUA%l2P7r!tWlC~K7U8EJ_PT4xvt)GI@`eFbc8
zdhmc~Li(tM){VpWxKo*TP!a96>hZYUuE9%LM7acqA2FHx?SmAEVkT?N^;zPn4+=+U
zaA`uJ>(C|%DJV&;Qb4_VAqHs123BbiZX56;F42T}*?G`HpMFyO6j;5@2zH%dBMT3J
z@c`~(+FOAM#M7^PcEn#{i16L)@Vn9BA1vDKXXUJLLa6hOJ;=yz)LwFLS)r6;1WYyG
z??vhFq2>>t{L|-tt4g0m3Hef$GM>JbxI+{tXl611ZW`_Mc&N)o4a2G1Y|QfmRFFqU
z-DkTB%!|X3a@(CLohlSlL{oaZ!yvT*xw`OPV$Xg7{00qO_kk+F1#b#{y6y2>mwOOj
z5~RQ(2E_4X4ChC5orRrmm6}<km-{uGW(BJ_P6y*2qF017Qbz@Tp2rh5K^BmGfotRn
zaRHSvO34bw9Q+kn8Fs+yvue5CwkE6dgE}oX8#C&4P9K2qH9LX!kea=8qtRR3{<NE~
z^l(w<@LoUBL$;QEeu%f{BCD2L>Tfhu4h}mi*zZV2>yvc4dZGenBGo$k@U&s%6S5FT
z>pE*oUb8M-oi)fVaV5&B2*rFh!wy#d<A{Moh3TM`&cKL0x1lyb1`sB0eCtuh7tn@2
z3*tIhM?}8e(7xW}k>pbW%%K4ck<;8I;;Fn`uBbs<z_Z5BEyo`=YqSD6$Zzip&DJSB
z?O^cCm<Kl+!<k~*4B=+53A12H^q(MaI40Jh(e;*c@9oN{GstXA`~(r+XEC@WQJob>
z5b2xQ{3Z%gSbHQKkf<$+^x`DHKx8bi6l+^_7%#a^<HNyydW!E;K+;N5$OMO=w*gn&
zV$(Qu^FFE`(rXH7WnGvjBho@nw>Zja`nFY(czJNQ`eNeEo;==k2@?dZtfu<9@dIIf
zUNfVOWypRlA1%y9^*q=Q>m+BcIiO=`duU?d;^9EO$k=@#c?DZVhyTZW<3+Y?{;#WS
zrAoT|^+7oMDElJw7yIkV|J}v#cTE5{zL$@8;<E`bk_VqTtv%;{A`WO)iqD9he<_?}
z#_t=0w~>v50ip5re0WDBN-kDIa>-bO?5Y6O>}OLD*zmZ^JFuiY;(&a~--!wkHz2^r
z&o@vRfy{<ZW&JN-02}$+=NEkY0|E6CiN^x;51(aNG75$>z77BZPy@O+{sjX;NE%;H
zUo#XxAB3IhDnR-;_M_aO(qz^An)~(+XPOhwa<$qbTuQZo#sIgecWM*J3~C+zgJW8Z
z7;Rl(>wGTs-ya{Bm~gA_WC`#*j-Q|Z?GKdsWB>`@>5Fmi!zP!oZ|~lp%;-KxbOe)b
zCBCZ^Xq}KDeiRZ=q%N0@e~e@Z{3P3n5t8&M*R@=*f)(`QD(-t>$Yty4bJ*^7eTuK^
zWKwtu1!W`Hu#*!!Y1ENNw4gcGoh@LG_c}0@Bz(`sSk8=_RJh?xQ59QRxAHU}JzT&I
zbboIN_7Xt!9!ehfpaL7W!*!e{wXTrkMp6vNzbKw$OcpRoec@(5ljnB7;UAt>P5Xo`
zo0h;OKe^qBb`ny>xh4yKw7Ln#!iJ=B$LWFXP_jed4TiRs8V#T`${0g}m}32SbQB>2
z`FN#nZ-vICu}0@lS5=b!(m)q=H!P`Ob^^JnZ9I6&#jKU-<(VNO7;`3&=&OY3S3CUv
zBoEP^gnjL&0vzXOdj(6uyEYkL#ozv~ja&%$vws-oX$C55jh2$VK5Xm}(rGBunJ#8i
zV(-BD6PT4j)*-Um%YpxYtIFYb@7E45nA`CT%r(=c@kQfy-wMN-ZdO0vE32YKk1UVf
z!n=he2-y!C3Qd8bwI$-92pHwgA>d{!NvzKG_M+N>oozhGBh3Aav(0XJuSgK?Mr9T(
zrCwKGwz&-q-{?}4@&&!8i@qg!-rsj^&Q@lE!8_oQI%1*)qu8S8Hki-Xt|K3p;OXcF
z*asCn+51?~_XTAUx;>ItVpIZ#w9XI6PUy=5`0_m!(IY<<h!mu0`aDerNESm~sO^}7
zttqH@!lpi)MURna*MwjsTvJf@3_K(Wx1!_ydTP!N0jcup^$;J}Gm8;&jA-D52~8IQ
zCs-fmlBsvNZ^5B<(H-sHp6))z8vE^Qr0bT%AHF2nt+*=~^&ijQFW0SG5r<>ZwSUpm
zn0Jq2T>k2Ae|=52`$OIGZ%^Xe8x+!fqwE>s;)O+5Bsa)yV4}+lSb&20>y!8gaLAXF
z_)o)Xfs6n6yo#NEArtxcVYLM#fm#Yeqp=Sh*gLS6b4bn3--j*s1rhe$V{dQ>^6#?n
z;80cmAqsC)*^kW>l!EFri3es?+GpYTK8fdtz*?D)mp?gBzb!pr4b69~LR$`i==|7(
zlRls3&C!=O&toTFq~kJ|CE<|*eJIz*Gbl%ap$2F+bq{i~@rnrx#Je6~byP^~_8<;-
zyeN;mXVtr);+W{BmVqwNb35A%7s!3~>$Tw~?#A4+OD3%f;JfTX94<X;Zs{X8Ay-sa
z!hpvwZX`shzSW&$TCLFK<>ehhJuUjURV5BP#TO%^ba8r|$2APK@{%0|ZYL<sa5nwj
zdgk@)h)@;TjhiQ5M^N0|-D6WpXD`6dTirao?}fLS*_DZsiS<j9ClMz`oEVNKvBh^}
zL4iq4u+<RoCgX-tfT+|szA=!Md`wQ_47pY#1u>zUHY9G1!i%T`NDtvg?mGq~Hovyi
ze2$?0*dDuS*!zy~gstg2EW|2tBoUE%Qay<H9}Z99#-(M*wld+9{Ydv!s=szgwD;Cl
zzXst2$|7X!UV(vfxJgqz_d!GM`WakHKB8l-=Zi_JuYqn2&zGRz!+go&y?<oXA#^w1
zHUyk|G(YH^B#+`2o|dsS)y>M0!WJTDo~vA$H+~heEn;T?0C_^RwpDPENRQ@av`PWE
zO4r%K(eCsPf3Y5tP=Kr$!_U5N-Zm)&kC)qqQ1NERfq0l~fz`ClCS)@dH#FRsfpI4<
zj91sM*xBCN?9AEZMPqq6<7h$PePuq!ssPR-S~zHSS3c&ldhGg(!k<p(Cx<t-wL$jA
z@hn~>>%=X4&q+akqG*`W2j!%}{sx~(c|$G33$XhVky5W+nwo8F{xgsl6mee<$xj~d
zFD6yLI~ZBOb^w0&+tS+z`MV#>%;O*fT;>_2yEAK4+|Tk`=7${6cXtX9ygy6u{yS|9
z9Qp4Y(9eMG?~uJn!u<z_-#_GlzRx}g3;BZ_(3kop2lO+H`>hr=AfD{qJgVOjMv=9|
zpqKCk1?o1<%A5~)UGhd;CSI+2xDn>D0)}=%Be_Qj!ARB+?c~eQqj068_R9=$cY>kw
z2_ET%gN)Skpg`~wNSDjj49?C$YGviAJy_4W0a;~sd)6!2b6VRMdVe)<yWYVxQL)s8
zWh|I%xj@5c2jWc_ML;C_pA4UW=%)JT*1h&Rk&^@X{9uNx))Q?8`bzJ!e?W{{5Eacb
zxJ}H7vN-n(V1NubVinN~(MogdB`54BhH&9@Z}-$C9(u5owZmwi1X0{P9)>*M;@dX7
zs?RPJ{XIwqz}1Cfw|#%v!cggH8H4GV#91zc1e}$NKOrG1n!}x`k3dqYE|l`|8_sLo
zUo52J0OXmaZT<^(6hW+j7G(2_Y80+>#Lw@K-@Q4w_&vm}zYTF6y}KvO=H+4~U{f~`
z#yU%nbFx<W1W{E#rcb?9TeI_3A@Z@v1Br;k&)=DwI#t5P#lPO4$y<Nn-3!neSkd-A
zK=}m;G<A=*NiGZOWLV_utw(!z6X2Rs0llMzh7+6|@DAJ95yEfaa`V|S*l-1V>#tX1
zRafTA^-TNeF;)0jt9@OUf$*(n)bFlU3dbDJaU=9!o&jG&_A6Xno8#_nr%pj<0UGNq
zJX`x*0bLr#0|KbT-}g7rxE}3mC5V<QXHIq0LEr*6NBNJoT3?r1F#Mzpr01WSarQ2L
znqHTOl>=|D7~rr>w?3Vafa=&1T|Uf>lNuRBJ*#+KmK19AV1fy;kZktTXm54Pz;g`<
zIw$%tqjrb^7QQMU@eoNUXuY_pz;D^PVxLZ*i1P-U4`(Kr^<>LH&~<c~w;7%GoGDGz
z<AG*_^Rz=AAu=p`qW5UoRaaTTDE~q}lPGFyYNhlizOk=I?Y!B@-1V{B-uPowuyw~M
znC-FDIws47Z0tM9@)Dbzkhp4abe-IGkko~ih8+#z4)>@XcA-Qi5yilKrpB0AKlel8
zLWsLVBy{$G2-LB$Fkcdg>(~;DOlxz26e*Vn8fecrf%5HTzf=-rWpwny&16{@&(C4w
zpYSbz<az(F0y8F|uNCp^ombt3n~4fb#XO;aj=cP;hlZsV{9!yjdOArTU%~YSB7sot
zQ;P0}YD=rwrf(pQXSx_Ix@#a4j1ox^KiH7Fey4*OGDjFbol%p-(73x%$OR>;XnAkj
zIP?<LJ%oNeBy%OYRxuV*K0;MKV~+_!N)I{JKpdu51$#8zbKKRK$;W8VD84f1;Lh##
zNT<(P?H=`0aWG|4yCep<swhBFGTS@stu{-;>Rq*vj~LA>8r2@zL-M?7>u6`kzdc;2
zUtUWlrhA!cgxow&c}UkXciWd>(9vW&*JW`@FrZ8M`7LkKTu(#H6GcKFpPO+hg)U2{
zF@7ke)meM$HqIqmE!KU9Y*0;;UlYKgtnB=}-q{s(yj0S22rCd{aLkT=CQaHp{=rsj
zYX15HY)TWrEsrx0F5pD`>tBw<@1|S7^??f5`*_@cJ01-=i43^t*4&|5G<4~`T+zUj
z_tQWb`{E1De$EDaAB}QJXCD`)&K6xbWV}1*+YXs4x@9QB)L%RT{mX@%$B!fNw8F9V
zb;~4uY@l!imipYMuI-m@5NJC-g9G5s6Meb!NG{|9e0X2p))N~%EeUa45D>||jd2Tn
zfzN@L6MB|+9vA^AN7!XK#5K?0x1Ywew^aak_ZLp3G`L`ctb_i3f(I><fdEhX?@3**
z1}>cNaF$}x!Z^WI>tJ~{L_~5AU{wclX69V-E!2$9$zJ;HnZjKX`A5%z1<64lwjG#{
zjbC1l1VnQmY%B9^bG?8DVjFEwk9Xw&g4Fwfmj9jX0HT!W_kVx8E}{Ejj~lNfsqsp4
zqLt+NbPdE9mCFvs_QDa_qhlP9>uD8Uk3Nu^$vK0R5f+&DaX$q$BrH#~bxe7K9a?Nt
z?9}tHVa>j^-a2}WuIBqqx^H&TDcy8%cVxZX?fW*F?@KD<%w`09^5Z&3+I3O{2I9iZ
z%_^YfhPjauk8&h=wXU4UbeO*hJFvYEJ$B^GfW#mo;sFaA4blrnN$`)4)(42gts2Xd
z+=%DWKo#d@6$|8(UlbF-Qq(LwDJf~&({774nc~fHU{v>kJp&8~nQSwFkRjfBqBrz$
zx$TZkN}jTEufU4w<@dy4885G;%8xIAR*`|<<bu1gfbOR34Y<+xDaUjx-gNjDL&VEY
zUVO|mK{Thf4h!k+_tTI+=qbeg%0;!88_KYI!MS~E7wLT9Y4yot-hvUu@Y)*;<ewk!
zx(=B_0dl<cqS!!W<bHDnhh+SOU_*Dhu9W8jLryfA<duZ!#pRp}h02%fC2;qV%s|Sb
zw!B7I&wF=%sv!_)bQ#D~t1<|D%Eo$_%V;@qn&-uOAvw=&xZl0ZbC$05b8oQ8ecn^$
z4yLptv3EU?7}RGLav4tKl=6F$fFB={7pvl;YDmR}RlIt(24lxlpsg^;0NNa$FVoO$
z3jH)bzzJQ+^j$CEZ$MA4BLHrv>m|8u7qHwKz=>2LznB3IvWky5)o!cX&^YDmuJA6}
z6Oar>Qm%8Lgh%u8sVzEw&Ky5<f&a2l`%>)GABvnsbgn_c8Yt7ro$}Nl&W2eU%<eAe
zzms|VG+)%92jt*w!_bmgbHl#cMF;3QuBYo@u>NgSUnaPB{ZV&_!4S|_ovX271svd3
zK+#KPU{B|+2wjs@rvnm`3h}$C#hGc29E4-==A_s|po6NtkShDex-E9y#%FEQoMkm4
z*@~R5%=%D|9IS6S2&I~6<V;bjFPo$e$p`pzWTaQ7Q->Ol;iBVdB$lrfGtj_<N_><x
zFIRy@(<aBm$&D_W-Ot{Z(Y#5$A1t*);<_RFgQ5n7A1BM)B<1z203xc>tNqc^liXym
zC7?O?F13&SUBVf~<E?YqRgDu}Qh4)CAT+f@l^j*%UfJliTt(?>`=c#Y_wu#th4@Tt
z@LYt+z}jkCe`iS%_A)P8RC$N=Hg<eZ$*!Ebd3r(1i0)49Y_}0IZ_P6-eAA89z`!Su
zdfo`+8{6*Sv;)|dtv!tGF!Zyy1iPkiZq2=?+iHMJ-w8w})C!hcwjBb<5xjTE36&~t
zZ}Ya|<`r6aJRSEcdxf<z<fvYb-OA#dW6EqeQt6_7xQk?sOThV4!2rLlM^Y$pSaE90
zHWeCVxwrNbI}K_;h$<gW<Jd&ynGBu}Nj&6j>CtE9tmQzbzXjk44-Ls%UP>H*go@=)
zIZoZLXaBs8@W%6yV%WQl@VeHSF6`S_)NZIwT}YHCq=`6oV}9eb$DLS2x)x?i$>aQX
zF!y1`F6-;1Q4ki9_ul%PlSbboEA0>WAu6~B$S?{^|EG(Ul-70H0y&C~!z69s14wW9
z%S?wfe8l-*GTiZsae7a6hcElIcdLQn|2y42v?RZ~sEO=bx{qEy7i|pvt2Pw;7WbcM
z6ej*5+5c!FK-C43eJlhDWV{Nz%70B4%GJFkx&tg~;o1XHUN9R4Qsnsl_?_qgM!p)h
zQXoOpI&f%BKpwg$u%3e<>T!ypfp~W#=N}`@UvOLBe0*2c*03CAhA)uz1FDJ_O8<?h
zyu!!tXbHIJ2jGK6Kq(JHHn8Qk8U}f&;FN{6UJH#vb9^RH!`~%bwZWqC1x@LEg-t+%
z1A@x8olt}L2(47%bLbcH<w2lL{O<Mmla~Ye)7RrqUXJ(W_K#kVKY2Oeuj%{;Z4Jk6
zlnNEe_Mijkz_SSh1~KNPtpF4MfKYn7Vl1=A#={Y{EzKS&kimm{TT%QM0Cs6|DThZP
zFe{*{Q4fRdzsPa>X7;EyKjaFR@tCNOo8bX;Oa!LTuLsJWF(z9lZ4|=N#=w?SL5IZ(
zftw~pr!{uxT+!&f;Gl@dFkm&N&dD6vC#WaN{&MxoN1%?w2HEMf5u`XqTATK!99r!>
z+;71|taB?<iK?J-UW0Q%*-C2csQ?r<S7+=j)fPBCly}`V+_NWa#{wfefhbDjhVD;W
z+>f7qLqWdAt$htcLrvZ$n3Iy=9c^phgTSu+^sWSPNyNFZljW#%kL-FNJeWvq3lbu!
zA0UtcuVbW}AppJvi2@8I3Nfc1uK9eR9b=O{C(UyN?RlqpG&r@gTjnj4On8Y{5`c;z
z{Q*4GciDir4@5M=@kG^^85URD=z7as)uO<WU<ugE0Bb?120-Eq&Wg=ka-ITsB#<=j
zsd4uX%#PY-8+kfytt*(R6au+TXmD@^^JZbqBGv&Ub`0So>$M!xt84bCcGl*6d@Xnq
zCqi|rOv2A})VjXH4bYkTn%my1qk;MP5opVW=IwHo-_2sxoo})Rq~D%|0QpyM8C=hA
zi<K|U0}ouFaXM~`2q({noCN^4*${jczBZWa3Vl-s;SwXvjP~JizMZpR*~=kx&c;aK
z2jRT2XZol%>L#V<0cSfi#@Ih<Ykn;%`e6VVn{Ep{`qR^MlX!$L5c}k?<tvv4Ys61&
z%{N0LEnWAKYl35e@lA+y@yGZIvkKW8-C>b|t;$m3p_Itm<<5?QdUcZ}=NKgeF3Set
zfS?!?=JML)LOCw?L*b*ZwowOHl~q!l-TbnlY5}ZegJNS-?A9u~b5Mf2a=q<<BvhIs
z*^+D#IvgZH6|&NW?uFJN`s~Kz8Q{&16Rl-tC?%*n!V5;u^^p6ajgrmvv@Mtt*l1>J
zIwhoyhAB8<cmQXjn*~WWgd0#dATwPbPogV#n&ps{(fNaUCZ8Mt>+6TuGMkeq+9`J6
z)WXqB%XrieQw$$R1zIME8WRU`a9iX+954(L)qHY&3J(Zg;r)Jd8LYjeV0$zJTFc9t
z<%xIo>Zwnf+iS08NLr9V6>BL}*fq2skXF8zUS)(N;nQEJvsQC?&SkL(<3o|xdH2~7
zg2#=DLJAbOH}5T08#T7wa&?dZsj!E}ofAPkc(fj|N%fmy*h45<S^`(HeXlsC%C~We
zEoESksQ7C!+&;NQYsWuToP-$B3%6{_zS<Px3)0y$&>o?JXlrSo&LzV>-4Qt*Nwz!y
zn9F)9yYwl`u1enHc~403R#i?qaYfBT#g2y8d9Nf;z;U`Yz`&EYN7wc><qlWgnd_gB
zd>kStc^{Omq@?>oBp!<1*kr=bn8kxgxktIX;Q1_dfsBfRPOpS01U!4MPO<iI0fUzl
z5;Qnoa8<Bp13uR~>I}w&Esx00+sE$k-)d_<moEQ5TbuX5<#+83JHOkT>f4~@-QN6|
zBnhRw|4nQ2F6EG5`rm17-lZJmTEA{>OpakcTbs)tv^MWjj{K^%F&&Nj$e=PmhA?l$
z$%lCaSGgEkoVV5}etjfm0ePGA-&~cqDCx_Ca~5_#iI9>X%p?39FhCrsp^pL$^5?Ve
zfH?ejLk6sdUw35xt^@fprD4hJdlZ=J7_;#?A?Mdoh8ZZBM5w#JFboW-AQcRMe+(I2
zU@&alHMoLpl*lscO0Cf8VF@@^z_hu3zn4PZxm<6kiuOCK$w|qKG(TrhGRP+wP=iJU
z{^99lZ4$#eD38+Bk>qq;fXVPEug^x`vhIO)d2<&m7l9#a1DJ+A*|pDVwD0Zn%;!&=
zb?!S2E+5B3GJDQ;)NZ?UduLm*QxLblA3|e--`J{(q&;KyU$G{?W;FkDZ}pkHfWRvR
zzzCyYNhLVZ{G~lVEAC048!FB!ssiK7Pf?NIuqL;Q(8y%hzX~%<@3jTtLSm7|YWaj-
z_`2)U8_Dn>sz@nIPId_>HaY!Hfm;CeAo}Yf!+wfM1)NT(Zf|ur=q4d;Y&%$BeGnG#
zgC{X1K)zcp9EQR$b}CP0WK)}Cb9fZD3kb_u#r&u9=B<CZ*VD%x={=7Xe~DH4-T5Gd
z@ACowcKrdHJ>zB@W7KQ&np*^??PDNk_~y^9A`aA^u;l;t9)a!~HbGTjsV4uy(}w*i
zN9SK|FtPZ;!GHj^=I%o}IM4RSM3Nbj_r&sj5qU3rKt2eUj1>7$uh2jx@;~xYxsP2G
zc)s9P;(UAT`)Ba)p1>D>z<>4(;3fGtz5xVgGCIHsSvmvg)O$VyBDQ=Vg_BZk+N-^D
z%S-5%*p)Ia?ob+W)8GY@*nYW5Pr+%kBM9ppTmnh~l`HZfk6{rA_2ia^k=AaNeHeMh
z%WlKwfJ#R$*cGE!`XzBHR{N%LZCM|bC)hu=tGsX7V{ZE6u9{A|qCr6SlZ??sjn5`a
z=C$1$`!afUx3E`s^eH~>PC3z65hF*t9lpKqeULTKXdLPokT`sB^9}y8pbKqIrCTD5
zA4op0u9Y50U8d<Z#ERPoso&j~0xgYc`?x9MHmt@SaodHZlX6EAg(Mb0%_msyv&&`f
zLj}dlyZuDYrnZwn`!1Qz>ygbcW_#r^s=nOLe`R3X=C9<OOc&=M9C#4XOL5ouuVAaX
za}gft^>Fh#;qtso8P^cVN*Z=T9NOJ;1A#!X906(sFEm3v({2J<Jjy~*4t`}>?4BGn
zNPflYX1C-0n@~-<RZzYso#Fnyn=X$SAzS9jiZFK|Vr&my<ape!i+w23)ypCC<}Sm<
zpJG58bgAkEPWpsxHXs6b4|}-@hNMHyW(Di54Xc?1A%&epVNrJqW<{}po#g_7rJ}2$
z9rT8QJXt+7BK{_G(csJMK0t5F`B1cdzO`JEKN+cP@Wxe~k1Ssz<+431th9v9iiv$B
zz$vy3oQ;e6oV=0F_38TL{j;!-d`c-GkNOJBpR`p){@B_H<lGuYXGYX5rZ(mV=_1TM
z0LGbH@vg{y>2sI&zwr&c7t#p_lMIHjx*2>PRhV-EbY@JEVt}IZ>v;mc!Oy;dUw<z0
z^^^Z~&+|3dt`&kO?d>7V8}PM5G@o*`BDu_8xihEO6V#vz^TFS7pxGSR?aZa54DSj0
zpCmFpNP{@A9^~M}V2p-M@<vJ@?Z(>>422b)BbA^nO1+0tLnIXHpnKF(wMzFeZ`4H?
znh*fArrCW~@q}@M(PY2nWVM<|&8VJKmYDp8I1w}a0CDm}M6h(RtUYThqp=Z6y@nYe
z5SFRxG8-sD%!}^fa}t`H=V4RjrLI3~6hcg24UtP$+~F1;^HfPd&VjRtk#Rb4n=#nB
zwx$P~x5wC5Up82MpEd%rqtCpv?x?m)4=W@WZhProG?zCM2eRA7GNEGUdGDOY6qot3
zn{706WE+@N)aKC*Q^)PPJyUxkU(_DLYQ`2P<M!dy!JQHxrgTqBVBgdSN_N0mTA|UL
z5gfJNq+Gq>(M{rW|7AFRj|$tnwh7PDAoogNN!Jt^9c|e)R+;n9Bd26d2n2zB9hAY8
zv##DsQi2*K{=6+5^DJ~ydvR*)?1~ZvAj4rY;0MaPa5zi>#GoRsDaE_K)&9TUuC3Wo
z<i5L-H(v6Xr&M9rmS;9P9zX~ot-X!`AwWWa09`9pQ`{tRN8BOh<P+sR5BVzj5_!%;
z{^*>uduDdLYsZdHZPl!x>2A8AoBsFjVnjx<J^*bH&zKbK@(y~@kG$E`pM(8l*>X4V
z47^V+3}JBGcxJ&ox(9XP`9d9ME5(H%0(-!aQ7o_vCcFxYyj?cEY%uJ0dbGGeFpuh;
z5%*5xZJb4|=6W<5sUR#<c?&4Qq5@~EfG@+;bO|ECZEzq&?uDx39L2Gq%wX;{WhE~`
z3YgO83YLX+`jfcTPv`9~ZlwOyyv>Qj9AvFu!<H~_FV3Me1v*Wzq%bUF?wID#?f$bb
zoW<ueCh`}VvtQsN2t*41B6Ic-_5@#oKjEt(_;GnYgx%oUfv`9dW0yi{mEdmCh)!HY
zrRAAfb16YWs!J`*zEq%Wb&Q04f24t4W*sgkPhH2IW7_gD1m0<-AWS@Kr4Sb;W46=J
zIA9MvuP?CZGCA0SYJvzW)#0sWn#KKmBDx!EWCeSN5UK#*M#JW+jzipSa1h%rKrQu;
z)8|_%ci1IfQ9+pj=Bo*oimx>tm#+md|6pgshn0%}xBRqWE#^lPfBMl2ThcH~3b1h~
zE=C1YpXU?xNNa!=0`F6&Xil&_qJq4CL50AIk-RZDb&v;Bac<ej!=&vj2UTm(cHfD~
zR%#jPv|Cn3rx1=1VVdj`dKNC@AVO*_qAa%qCAFF?gJlmYOwo=~bs<wLN63<@^gQao
zTgRV3B&gY|f7A=f_78md{^ZhlX+J)Fy*IBA*N4j;bve$)_QB*bYM!4aRPy<!??Zcm
zXTJ35D-8sKL0kYneIw0*CG(3X)LU*X0I22-tl)Ofdeq<3Y5^4u@%86bDh{CS4%7@L
z%3?=6D;$7f;tO@=LEZ&F=j|IL#2_OyaBKO1**-6!(+$Gy-juZ9-3|K!*lqw|MzY_z
zSv@b6!MO^Sg&;cj#mG18bM7E)Eq%75znFMCFl{S#%`wedrX#>CwrW62$mzWkxbJWw
zZ;-GG0y9{1tmCO$fKHZ$oU*Ci_~tRdL_53N7oZ$L@Cg)`Z$*0^%Ot-q@YQ@znQ3Pq
zWZdT&!dW%FyHO7*PJq&jxIJRk=2C_;8lV=WeF0;kK!ioUBj=2M+Q+P+;%62pwixU_
z#}m)Y8^bXD=qx%X(*Sg@j??sV9x$S>iX?j+XP1?{R*}=qD?QM=17SdNA{z&TpU2yl
z5tjZI?v-<b;w+exi4}VSaLp$@$Od_b_4vBDyWyT)X1Z|JOm>}s1<o9472R&PK(pS6
z=lC+|`m4v6Jgso`Y-}FnHXAEPN-6Kzbc(I%WJTpt;(Pm1d}|eJW3z%X>DL56;@VC{
z<~_a{ok03*W73;`7tEZ#T^Amh$8dgMQvgvTSYdohi_D&8)5+S*Lu!K?j=K>|Q46k#
zzm2w-U`vzr9U;5sS}4ufysG_$k#|&5SohR1>wXGyZI!*KL?)!LV<y<^sx9bAzrX?M
zyoMHCz$Qa)T{o>Aj}F2}1@+?T$Y@6-vCW;xTofj|nK{vY=15>2cW*%2n3aNE+0S7s
zEJ@f2xQ7s&G%tcHt2Ol!CjFHWyFI;im%<zbXtt7SUSzR1vdxMxKzVNuoWdzd>=;<{
z9AtaBGLCEJevprib~^D}9R_Qi?#Ii<yXdqG2GL>NM}RcT5fh9fQ{z%)Hmx~^!c|#0
zmR~{d&$@kI3?Y2U66yy`q6mWc`;I3-`g-4DTB|_Vl+(bs)IK%hbG#y8!4YejGP}>y
zB5xDF#>qizPyx8x_hE7nuLnK**<^7@sO&~hfuwrIS#6dnfR)%%zF`$Oj`N5nSE6Lj
z$hpWz<FJ~-jz0%p^JCft%`f3_oW%1G;=u<xUaZAVqjfvSg3fHZd)Qt80dbW#<tER<
z^O&m<Kolw@$Uoy#pDQl$DU$LH7<OM-wmQR_Lx9Mzl;$fT2@Poq-+=qpHkQZLMGmW+
zSP?;`DiU>>@}Lv7MiPAEV!3k))APb;Km`(l;V12w8R~sc9~rQ*+|`(QfCUO>gKO})
zazI2&!W0jn4reX^=f=gto~fh5UcMp6iJ@IJ8DE%I>E>!sddtuavJ)v1+A%s8)P8^7
zPmh<`DT_B!zuQ3!$EA(YS^?C2Vc6^iu>Wpt&%7CqBswuM(%a!YrhpuWN&ZMs>)krr
zG}(CzRJ?V4Z<C#>vghhNjselqTfJzX6weqRDz)xFb#0TKO^}UnVni@V02Q9M8%>-a
z7Guh2UGP`d@3*CatNe^guj7NY*`$$1f-%>8ZQ<%c2K5SFD<)aW%eXuz@VP<tHIE^7
zGjTv%5k4t>(bYo?P*5JYz~H&PSX0%`bg#1(7$mBEHV?#UuQ)A)ud#KfRYXx=a5Xuc
z`(iQ1=QO~eKs_;R*AxLkw}9cVSAYn+Tm$!`F5(`fg*EWeof@^5U2iL<F}8jcYxT$<
zm)t#&Cb%QY5=6>;jfq+YQIAJ3vEn&*PAd~$6^Y820?M@D6Ku4uP9Qnhwm%lO`oiq}
z#qZz$)9ih-!1EIn*yg*;(xlN%s=GfBxcYyZy}Qq6@8>|(vogkCYxe#!9Pba!-e5K7
zMH^}hz{QnGj+dhad7^rv=XT`IHIDa#s*HVSoVWIz^3v!|IfgJolrp|ya$ni&+e~4z
zR;aF^Ryb32#xAqbqQ_%io-Ed22Fp*t3fN+F@F`ih+Eksc#00jb=$Xgar?C~;Ju(5Q
zGZ5<x*~(P5LAEA6&I11_PPesDXO+%nVAZ!yy@01JzaO-2-9%iBO^My%ga^CT9OOYF
z4TPsVZjX2;?&O#8O0i2_m~P3-r8}$B8mp%OVZhZkK7#IrIB{+K!Y@xVP{zLsaPO0w
za&qgH5O0CtPaN)&MmHt0a|JK(STLQwfI-~;G^P!0B?X~Bhrk(O8dsz0`fMeO>|7Y#
z$WvI9<1@Wm#oHtT_xj!bxDd28d*EP>MMyz9lQW;#h!~-9XM(rMBWU1?$<?dRM=?q_
zGb?Mtt$0>f=crCKLF!6-H?^xl9HR2JlQwa>+xal61BB%IO=g_C(+xo6S+~%&v!1Ck
z<N5lryG|pmzY_p02T>8Kp)%Q9MooI4ob!s3fw#2U_1T8vZk7m60X;IehgC3JmGi}Y
zbdWlqH|e@_(lX%oM_o!QC^H~7C&~hU-D-8nWU1h_ovTd27XY)jV0F5h&w$r()^P;(
z8%6?cH&hAe`?)mhra>^*gI?0-6^#RO2>?FF1GVrin;BUgcD>OWI0fR<uD6eu^Zl}8
zLiV^`sf=JE9Vv3^9vAq_8JsJw<id>k)$7LfpF4YlweNK6%OsVCjJc=gn+eRK0!7~L
zvmeafpQlCMCjM;p{vNxx3gkP^%qY;7iO5-bn_3#Nq}Dlo^(eY*M-w(&L4Xet0xAWI
zrZ}Thh-5$jxB*C8JMmq%&cY?u<Qoh#$G54v-fnGX+_qDlyt4inf_c1qcYp{d@bWA>
z<Dn-Z@s9UFqT+NqOy>9Xb!of&r2)FxJ^^^gR(6lsb-k^Mrk;#%F(KU5(cRQzbIFK_
zCxN!#e#!|%0{{Y%6%b=m_9S4vT$6T02mC<sc`r)ffSSzxZndw=QlSGmGn;s1>)?Zn
zvqV$Oy$e$t9XNT46d=u|W*UPdX@>BLD~*l0fSGlJiFy;l=%O-!R$Fjm@Cl`rd$SGh
ztg#uZF4ON1?IE7GzGAH?bDXg<I2Uea>E4Qt^V1obdIS)R39>ibuwR*$8N#GJT2Rei
zWyHKHN4N_(Q6PubiMR0Aq-H8-TQR-)c_(bHgf8rsjuy^@IUVN6mAi~n_-4K^_H8@c
znCd~Yl$->`&FPM*wwWBW`#fakg=iQ@phBEiC&20KU0)u){MT3beA$t~6{Kdk6fxqx
zu&ld#P3Wq*uSP1blIzV~-b_<#Wv9`R3+n@jlm(y}l5qK$J=#~7@)%zs5`H{$EzXcl
z+QqNl5l2p|`4n&!BV`U)o1P-`_2lHNU>R<N3%{mWmEK^&brh<^<Z@UKf~7_6;Q3ZQ
zOs&HfY~aLesVza;b5h4hMX#EB*yv<EpRN1L&MTLsB4No0B!D!D-x!4Sqok-W%--bZ
z%%}$q;3vaZ(!d?GzVm?^!dLN#?`JES8?igC7^fRs5g6<T;j1?@_pssf62AJg*WLZF
z>_!$Q|GMt}vu5sL*^PX+!vCY!-5bD24U2aOSBC-c;Uri)?DFe&%OD_u<uhQ9PDg%a
zf{i_SJx$YAoK4uEv2WcJ_~IMiO%8Nc_8_7YEKf~+IAx&eX5N#1m~UxEpcm_LRZ%Qb
zT7;XPj+^NX-<|IYXvEomq-jYRS%5Q-Z)WcuD?{>XUCie)NMRS7U0fZ{5fulYZyCIm
zDgOa9dWdm-ij3ApceVjtaft$H?Rv6j6&&_x$<;x!e<+~-qlrI7F*HiXEZCR{cqW(z
zPGUpwgJmz%r5DcSE@UHcN?ryE6OQ}Y$QyhRNB3Y}@l9G2ijxYYpw|{_-vUaYhD^qQ
zV((`1mIKFkfZkaMxg#zsrfL|DkLbxLVYzcVN!pCskF62UU(`M}LyFiB`e3t5&HmU&
zuw>}MvsTRU(@oo-dTl=6t<@Px+V=Fq*s97|7qn|NRyiLC7oW{Y4@RN+#`OjO@E|%3
z+)>c;NnWBw<^Vt)?9c{RmO;Y!M8aqS7I>e6RNW4wRhqZf6d-KSu6D#qe2pR<07&s2
zmJ}E+z<g;y!EN$vTn6~kx5oljd&_niR1)@JkD6^b=b627+YC?!974i_aoeDY39|Dm
zfXRlRz#VjQRl7%~U7C#f3L!xjd}QuI=XHbmv>Aq-Kd`m?gPq+qnXW|G9$j4en409s
z9?0c4ns&<h47l}er2!Epg-s;aS^HSMzuWbRF}uJfV_&d&!z}Y<fazPqqGuPpO53<w
zj&PX5sxTlB@|b{pTY&L3e~ZJK_0w}}?XvhBciJUrn|g|8($(%km=EY9v~NsCWPF6#
zhywpH%AJ9Rb+8*<CwU0=3if1EfsW@fkXXDd(;8p+iU%e-l3o(CfSHZbQAN600PwAY
zIa{mzt{nAEonrJJJWxT34bQI-qKOtF2f^(whAos$NV6Jnl4FdQm36QftpF&dllM+!
zDHDeUcOh4+`at>Y<P<rI7!_^>4raoNqRJ|a{aM_)%odq!X>Y7^>awoL`7|Kg(GGT0
zeGUBC_tZ!tL1Jy}o@siKtXuJ1%@{0Nv98aO+XKEX&+OD{SU20P7xUnP5H@az{jOfK
zZh>b3duO819aEx0=W4!;JZnW@PHC;$Ey$~w)LAR>80^YdyyaRc-ZCO?-D3NYsL`@Q
zZb9W;j_ZWK?k5KiAMK9UJ=dsO0Bm8_mPQ_)ItB@`+PW8}6}#ohb`g?B-z8=e+-5AN
zFyMMrw!jAvk*WYV(Xs`jXo-&3;6oza{S?@Ly?mcvo2{LuAWCbA2lm2)9)aE5#ZFLh
z3sgYw5-D1io6lgQeh0$aSdOTg5|YI=Uu@G9!!GvZw&w^=-mVDp2Cx(0q;OIW;`(Ed
z$!|+M*&S%C7EEG;E86SM?lVghoN*qzwGG(CH^8E7U`F}lyw|95nZzCq#+l5rzrrii
zvDpckSFvpepR@8_v`0bZZLiNycf0HE%f)sfi{RT4URPZx<zUZrlwk8xk|WE&_Veg0
zBb9yY`DJ2CXLm&5Z9Af6;{>ajJ5%46-STj|EVm0_`ScZgp4$0p>5YWsG~%zC9Zt`}
zmSjD^<(12sVC@k+Y3rs+X&S<C-Q22Hn2;44zrv2hD6UhVRcmZ~#O|R{_=WDM0k&Q%
z%kne^L-%fPP!*sT{q)-VJ6QJJ79B^9riXn4`nYmUw(><ZgNmyIQIZm+4@(i0p0==g
z!SvE(wb521p#W<Z5qu(UY2(Kf?OT0$RMS<1m;j@MnL_5Su41b<Hz<<2zt>_#?{YId
z9*hzMgVcT3U2k`w?*&UsBr}zSJw-48$414Pk*d0M>1qTR51_ps@W%D@NSP@Os52=V
zA7%YMws-f*YG;A`Z)Gk*u`gONTdwzO`aT&=Nn=AOb4n)^+dbS0$K`CWazIx7-bp*3
zWG7@2nCh15)XRwA9PiKyguUGt;Bb6_R*V3@wexX|i0a!27^XLE65iM9YJ6|+6Oaur
z004lT_O&U_8#*}Jo$UFF>&}jwaWmgxCEel63pVe3SjSB!XKRu#$Feeq{Zvz9V3QgY
z+(p(&4t#xGvD;{#AS>SA;b&;5tAoK~qFs-=J~`e=`iut?SUnNOmFN4WWx%4)g8{R>
zI;S?lJbg03)~gA41&<=I&LwNFTOsi%JZ|T65JAO8!Hqhm1;Hdo%9qOA6FBF`c58q0
zPuVR)5dVU{*F^23!Dtft^KZXJ-#L;0?rZd45ETAA=a!=H4pyG{$@LqQOXDmpyc{iV
zX%)Gs?V)~Bry;OsYu`f4n}>3;gw|eO49Z2IR|Hu<p$T}f4(Hx|crEF}J9H@U`2erV
zzeawKAh~z1{M<u7eq2cU$B!T0VIMwV1@Mbuh$)LQx3UjV<#)eEU`T?3Uh4ACOKWgx
z?|%L1(IZG2S631EH<6crc8&x7O&WRs%w!qZygyWyAX6{T)BI!ZW$7Sxdwu^a;%hjH
zVJMe-R_UQ>Rc2L*UceFllgu9+pEp?{U%jOK7JaGOH{UpqEYwF1?}u!^Y9IAOzeN+T
z`82=xFV{n|msG#{E%*mJ;c#J<6khrM<@k`%(0`XIaRyS+N3ZpZvUvZ_PyCON?%mg~
z{rJ0=Bk$k2evyG|$j8Us7Vn<W$6XG6{Z;J6(2i2VcBNN9&0e!b-+xtGQRTg6_DEbM
zehF!yTgnt{^t}9iQQ7e8CG8XXD`LnSsyHMc{+=^|Yu@`w`Sr`s?@t$vUplw<Ukw?*
zEor#av7p1>UEh84U1r14;qN8WfBY6D@bTpP+@8;(D0E+vqBb;th`x8KxQd{=YY&D0
zDShaDFaP?JJfBO7|7s<Le?m#{ioD)d?v+)ZeA0%$hlk<OBG26@`|+cL%G{5k!Tpa{
zcy8W3pZgdMC*kRMZZBvfDH)znXymsEP<XZ=4-kERbTGWaR)P`&`zU63?p~gl=c0c8
zEp$2h%U^!>9DS1LYxFmdenS88?LTjO`}+s=ZB5`YeMboIUh4b%M+IKqgy&>05AD-y
z;g2doa%eKRySz#s;}2l(;(?H7egEDgU%mP9@;1JCmxovGZHiz07V>(ZR8a&)J-_CV
z0s;O)m;2XS3D4Nq1Ep_@Dt=U;jO*o4V)BHp(%kh?HoabbD@hSVgA;!77$u>7L5171
zJM@}cN$I&y=q2}Hyt$yx4Y)Vn6I!M)$aH81<LWT1f@Uf7XmUj@)Pa%_NLk(>n<53E
zoCPE3<0gwDb$l4OT+~hB$@daKh76<f#gE|iem@MzzX0^^<BDH={V2rstO)U+8|(EH
z3hyP1_9dEo@Ot}?f#wN@eFqj)LmggEKc1=Q5{9fF%~2pwa~Q0j9`$@e<J9#p9sKJ(
zZY!(o$S-bBC=5!FwGA+dClrPe7?eF2=bnc9z+kwAZD1hwr<o7+e%z~vzSi!MW@sO1
zo8f_fZlYp1(%jPc(e1BkhbQ?0uMHIGxkJs|6q@W&V&FM2MWv7Amo$RseYiCTWUT!M
zTk{(PG2x~xJB6`hX!GaZhKUWT`e^P)2RthA+^a9AiklTh&-3KaV9*)Rym08d+kv|q
z#!cvW--^)3+%vF4KfBEjP`=2chPpHljLxHyfoi(Cmlpg2{RLyd?_WHDV5=#rI}?Qz
zI-{5;R9&KsazfWYynRCXqo(tQfvP$pZ8TZp;h3aw@@^q10?NX5;Ju3iMOrc;Q&UBU
z0$;L{c(`tfZg`y&A!jfIl_m4^ge)af8S-4JI?A9L142XGPG)pe+v%EW@KF2$vQ;Ey
zsY5A19aYQ^P%0b)+e-L>8X_ahLs1B`gWBsu{ZLNTPP!zBCMv2jI3F1(%R|i=R_33J
zf}7%GM%p}~3ueOzL+x}_g<N!GNcLPi^dR!Xp`knm{^LwZRfgN(fU>Q_&nLJ`-Tdy-
z16YYYp^PpWLxU`Jb@PM_%?VdP?vM#wr|{3Y3{CxLpznGFk_-*DgPVS>3%cML8S;Ye
uQ9kKD=*@rsx4$3#E%QJB@Zmq=fBo-oe*MjV{@3sR?r(nfv!4m<hW%e+{BZRE

literal 0
HcmV?d00001

diff --git a/openapi/generate_json_api.bat b/openapi/generate_json_api.bat
new file mode 100644
index 000000000..87426a95c
--- /dev/null
+++ b/openapi/generate_json_api.bat
@@ -0,0 +1 @@
+.\vendor\bin\openapi -o swagger.json --format json -e vendor -e libraries ..\openml_OS
\ No newline at end of file
diff --git a/openapi/generate_json_api.sh b/openapi/generate_json_api.sh
new file mode 100644
index 000000000..0b651d7ed
--- /dev/null
+++ b/openapi/generate_json_api.sh
@@ -0,0 +1,3 @@
+#!/bin/bash
+
+./vendor/bin/openapi -o swagger.json --format json -e vendor -e libraries ../openml_OS
\ No newline at end of file
diff --git a/openapi/generate_yaml_api.bat b/openapi/generate_yaml_api.bat
new file mode 100644
index 000000000..82fb2f145
--- /dev/null
+++ b/openapi/generate_yaml_api.bat
@@ -0,0 +1 @@
+.\vendor\bin\openapi -o swagger.yaml --format yaml -e vendor -e libraries ..\openml_OS
\ No newline at end of file
diff --git a/openapi/generate_yaml_api.sh b/openapi/generate_yaml_api.sh
new file mode 100644
index 000000000..bd779dd20
--- /dev/null
+++ b/openapi/generate_yaml_api.sh
@@ -0,0 +1,3 @@
+#!/bin/bash
+
+./vendor/bin/openapi -o swagger.yaml --format yaml -e vendor -e libraries ../openml_OS
\ No newline at end of file
diff --git a/openapi/swagger.json b/openapi/swagger.json
new file mode 100644
index 000000000..bfb522c48
--- /dev/null
+++ b/openapi/swagger.json
@@ -0,0 +1,5647 @@
+{
+    "openapi": "3.0.0",
+    "info": {
+        "title": "OpenML API",
+        "description": "REST API for sharing, organizing and reusing machine learning datasets, code, and experiments. Follows a predictive URL scheme from endpoint https://www.openml.org/api/v1/json (or /xml). You need to add your api_key for every call except GET calls. The API key can be found in your profile on openml.org.",
+        "version": "1.0.0"
+    },
+    "servers": [
+        {
+            "url": "https://openml.org/api/v1/json",
+            "description": "The current OpenML API"
+        },
+        {
+            "url": "https://test.openml.org/api/v1/json",
+            "description": "The test OpenML API"
+        }
+    ],
+    "paths": {
+        "/data/tag": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Tag a dataset",
+                "description": "Tags a dataset.",
+                "operationId": "Api_data::data_tag",
+                "parameters": [
+                    {
+                        "name": "data_id",
+                        "in": "query",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the tagged dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "data_tag": {
+                                            "$ref": "#/components/schemas/inline_response_200_2_data_tag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "data_tag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\\n474 - Internal error tagging the entity. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/untag": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Untag a dataset",
+                "description": "Untags a dataset.",
+                "operationId": "Api_data::data_untag",
+                "parameters": [
+                    {
+                        "name": "data_id",
+                        "in": "query",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The ID of the untagged dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "data_untag": {
+                                            "$ref": "#/components/schemas/inline_response_200_3_data_untag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "data_untag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\\n479 - Internal error removing the tag. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/list/{filters}": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "List and filter datasets",
+                "description": "List datasets, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/data/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all datasets that match the constraints.",
+                "operationId": "Api_data::data_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  /status/{status} - returns only datasets with a given status, either 'active', 'deactivated', or 'in_preparation'.\r\n  /tag/{tag} - returns only datasets tagged with the given tag.\r\n  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, data_version, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of datasets with the given task",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataList"
+                                },
+                                "example": {
+                                    "data": {
+                                        "dataset": [
+                                            {
+                                                "did": "1",
+                                                "name": "anneal",
+                                                "status": "active",
+                                                "format": "ARFF",
+                                                "quality": [
+                                                    {
+                                                        "name": "MajorityClassSize",
+                                                        "value": "684"
+                                                    },
+                                                    {
+                                                        "name": "MaxNominalAttDistinctValues",
+                                                        "value": "10.0"
+                                                    },
+                                                    {
+                                                        "name": "MinorityClassSize",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumBinaryAtts",
+                                                        "value": "14.0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfClasses",
+                                                        "value": "6"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfFeatures",
+                                                        "value": "39"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfInstances",
+                                                        "value": "898"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfInstancesWithMissingValues",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfMissingValues",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfNumericFeatures",
+                                                        "value": "6"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfSymbolicFeatures",
+                                                        "value": "32"
+                                                    }
+                                                ]
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n370 - Illegal filter specified.\\n371 - Filter values/ranges not properly specified.\\n372 - No results. There where no matches for the given constraints.\\n373 - Can not specify an offset without a limit.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/{id}": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Get dataset description",
+                "description": "Returns information about a dataset. The information includes the name, information about the creator, URL to download it and more.",
+                "operationId": "Api_data::data",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A dataset description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Data"
+                                },
+                                "example": {
+                                    "data_set_description": {
+                                        "id": "1",
+                                        "name": "anneal",
+                                        "version": "2",
+                                        "description": "...",
+                                        "format": "ARFF",
+                                        "upload_date": "2014-04-06 23:19:20",
+                                        "licence": "Public",
+                                        "url": "https://www.openml.org/data/download/1/dataset_1_anneal.arff",
+                                        "file_id": "1",
+                                        "default_target_attribute": "class",
+                                        "version_label": "2",
+                                        "tag": [
+                                            "study_1",
+                                            "uci"
+                                        ],
+                                        "visibility": "public",
+                                        "original_data_url": "https://www.openml.org/d/2",
+                                        "status": "active",
+                                        "md5_checksum": "d01f6ccd68c88b749b20bbe897de3713"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned\\n110 - Please provide data_id.\\n111 - Unknown dataset. Data set description with data_id was not found in the database.\\n112 - No access granted. This dataset is not shared with you.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Delete dataset",
+                "description": "Deletes a dataset. Upon success, it returns the ID of the deleted dataset.",
+                "operationId": "Api_data::data_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "data_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_data_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "data_delete": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned\\n- 350 - Please provide API key. In order to remove your content, please authenticate.\\n- 351 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\\n- 352 - Dataset does not exists. The data ID could not be linked to an existing dataset.\\n- 353 - Dataset is not owned by you. The dataset is owned by another user. Hence you cannot delete it.\\n- 354 - Dataset is in use by other content. Can not be deleted. The data is used in tasks or runs. Delete other content before deleting this dataset.\\n- 355 - Deleting dataset failed. Deleting the dataset failed. Please contact support team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Upload dataset",
+                "description": "Uploads a dataset. Upon success, it returns the data id.",
+                "operationId": "Api_data::data_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/data).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "dataset",
+                        "in": "query",
+                        "description": "The actual dataset, being an ARFF file.",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the uploaded dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_data_set": {
+                                            "$ref": "#/components/schemas/inline_response_200_1_upload_data_set"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "upload_data_set": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n130 - Problem with file uploading. There was a problem with the file upload.\\n131 - Problem validating uploaded description file. The XML description format does not meet the standards.\\n132 - Failed to move the files. Internal server error, please contact API administrators.\\n133 - Failed to make checksum of datafile. Internal server error, please contact API administrators.\\n134 - Failed to insert record in database. Internal server error, please contact API administrators.\\n135 - Please provide description xml.\\n136 - File failed format verification. The uploaded file is not valid according to the selected file format. Please check the file format specification and try again.\\n137 - Please provide API key. In order to share content, please log in or provide your API key.\\n138 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\\n139 - Combination name / version already exists. Leave version out for auto increment\\n140 - Both dataset file and dataset url provided. The system is confused since both a dataset file (post) and a dataset url (xml) are provided. Please remove one.\\n141 - Neither dataset file or dataset url are provided. Please provide either a dataset file as POST variable, or a dataset url in the description XML.\\n142 - Error in processing arff file. Can be a syntax error, or the specified target feature does not exists. For now, we only check on arff files. If a dataset is claimed to be in such a format, and it can not be parsed, this error is returned.\\n143 - Suggested target feature not legal. It is possible to suggest a default target feature (for predictive tasks). However, it should be provided in the data.\\n144 - Unable to update dataset. The dataset with id could not be found in the database. If you upload a new dataset, unset the id.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/status/update/": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Change the status of a dataset",
+                "description": "Change the status of a dataset, either 'active' or 'deactivated'",
+                "operationId": "Api_data::status_update",
+                "parameters": [
+                    {
+                        "name": "data_id",
+                        "in": "query",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "status",
+                        "in": "query",
+                        "description": "The status on which to filter the results, either 'active' or 'deactivated'.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n691 - Illegal status\\n692 - Dataset does not exists\\n693 - Dataset is not owned by you\\n694 - Illegal status transition\\n695 - Status update failed\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/features/{id}": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Get data features",
+                "description": "Returns the features of a dataset.",
+                "operationId": "Api_data::data_features",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "All the features of the dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataFeatures"
+                                },
+                                "example": {
+                                    "data_features": {
+                                        "feature": [
+                                            {
+                                                "index": "0",
+                                                "name": "sepallength",
+                                                "data_type": "numeric",
+                                                "is_target": "false",
+                                                "is_ignore": "false",
+                                                "is_row_identifier": "false"
+                                            },
+                                            {
+                                                "index": "1",
+                                                "name": "sepalwidth",
+                                                "data_type": "numeric",
+                                                "is_target": "false",
+                                                "is_ignore": "false",
+                                                "is_row_identifier": "false"
+                                            },
+                                            {
+                                                "index": "2",
+                                                "name": "petallength",
+                                                "data_type": "numeric",
+                                                "is_target": "false",
+                                                "is_ignore": "false",
+                                                "is_row_identifier": "false"
+                                            },
+                                            {
+                                                "index": "3",
+                                                "name": "petalwidth",
+                                                "data_type": "numeric",
+                                                "is_target": "false",
+                                                "is_ignore": "false",
+                                                "is_row_identifier": "false"
+                                            },
+                                            {
+                                                "index": "4",
+                                                "name": "class",
+                                                "data_type": "nominal",
+                                                "is_target": "true",
+                                                "is_ignore": "false",
+                                                "is_row_identifier": "false"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n270 - Please provide dataset ID.\\n271 - Unknown dataset. Data set with the given data ID was not found (or is not shared with you).\\n272 - No features found. The dataset did not contain any features, or we could not extract them.\\n273 - Dataset not processed yet. The dataset was not processed yet, features are not yet available. Please wait for a few minutes.\\n274 - Dataset processed with error. The feature extractor has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, please contact the API admins.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/features": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Upload dataset feature description",
+                "description": "Uploads dataset feature description. Upon success, it returns the data id.",
+                "operationId": "Api_data::data_features_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.features) and an [XML example](https://www.openml.org/api/v1/xml_example/data.features).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n431 - Dataset already processed\\n432 - Please provide description xml\\n433 - Problem validating uploaded description file\\n434 - Could not find dataset\\n436 - Something wrong with XML, check data id and evaluation engine id\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/qualities/list": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "List all data qualities",
+                "description": "Returns a list of all data qualities in the system.",
+                "operationId": "Api_data::data_qualities_list",
+                "responses": {
+                    "200": {
+                        "description": "A list of data qualities",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataQualityList"
+                                },
+                                "example": {
+                                    "data_qualities_list": {
+                                        "quality": [
+                                            "NumberOfClasses",
+                                            "NumberOfFeatures",
+                                            "NumberOfInstances",
+                                            "NumberOfInstancesWithMissingValues",
+                                            "NumberOfMissingValues",
+                                            "NumberOfNumericFeatures",
+                                            "NumberOfSymbolicFeatures"
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned\\n370 - No data qualities available. There are no data qualities in the system.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/qualities/{id}": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Get data qualities",
+                "description": "Returns the qualities of a dataset.",
+                "operationId": "Api_data::data_qualities",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the dataset.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "All the qualities of the dataset",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataQualities"
+                                },
+                                "example": {
+                                    "data_qualities": {
+                                        "quality": [
+                                            {
+                                                "name": "ClassCount",
+                                                "value": "3.0"
+                                            },
+                                            {
+                                                "name": "ClassEntropy",
+                                                "value": "1.584962500721156"
+                                            },
+                                            {
+                                                "name": "NumberOfClasses",
+                                                "value": "3"
+                                            },
+                                            {
+                                                "name": "NumberOfFeatures",
+                                                "value": "5"
+                                            },
+                                            {
+                                                "name": "NumberOfInstances",
+                                                "value": "150"
+                                            },
+                                            {
+                                                "name": "NumberOfInstancesWithMissingValues",
+                                                "value": "0"
+                                            },
+                                            {
+                                                "name": "NumberOfMissingValues",
+                                                "value": "0"
+                                            },
+                                            {
+                                                "name": "NumberOfNumericFeatures",
+                                                "value": "4"
+                                            },
+                                            {
+                                                "name": "NumberOfSymbolicFeatures",
+                                                "value": "0"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n360 - Please provide data set ID\\n361 - Unknown dataset. The data set with the given ID was not found in the database, or is not shared with you.\\n362 - No qualities found. The registered dataset did not contain any calculated qualities.\\n363 - Dataset not processed yet. The dataset was not processed yet, no qualities are available. Please wait for a few minutes.\\n364 - Dataset processed with error. The quality calculator has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, contact the support team.\\n365 - Interval start or end illegal. There was a problem with the interval\\nstart or end.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/qualities": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Upload dataset qualities",
+                "description": "Uploads dataset qualities. Upon success, it returns the data id.",
+                "operationId": "Api_data::data_qualities_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.qualities) and an [XML example](https://www.openml.org/api/v1/xml_example/data.qualities).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n381 - Something wrong with XML, please check did and evaluation_engine_id\\n382 - Please provide description xml\\n383 - Problem validating uploaded description file\\n384 - Dataset not processed yet\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/unprocessed/{data_engine_id}/{order}": {
+            "get": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Get a list of unprocessed datasets",
+                "description": "This call is for people running their own dataset processing engines. It returns the details of datasets that are not yet processed by the given processing engine. It doesn't process the datasets, it just returns the dataset info.",
+                "operationId": "Api_data::data_unprocessed",
+                "parameters": [
+                    {
+                        "name": "data_engine_id",
+                        "in": "path",
+                        "description": "The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "order",
+                        "in": "path",
+                        "description": "When there are multiple datasets still to process, this defines which ones to return. Options are 'normal' - the oldest datasets, or 'random'.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of unprocessed datasets",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataUnprocessed"
+                                },
+                                "example": {
+                                    "data_unprocessed": {
+                                        "run": [
+                                            {
+                                                "did": "1",
+                                                "status": "deactivated",
+                                                "version": "2",
+                                                "name": "anneal",
+                                                "format": "ARFF"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n681 - No unprocessed datasets.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/data/qualities/unprocessed/{data_engine_id}/{order}": {
+            "post": {
+                "tags": [
+                    "data"
+                ],
+                "summary": "Get a list of datasets with unprocessed qualities",
+                "description": "This call is for people running their own dataset processing engines. It returns the details of datasets for which certain qualities are not yet processed by the given processing engine. It doesn't process the datasets, it just returns the dataset info.",
+                "operationId": "Api_data::dataqualities_unprocessed",
+                "parameters": [
+                    {
+                        "name": "data_engine_id",
+                        "in": "path",
+                        "description": "The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "order",
+                        "in": "path",
+                        "description": "When there are multiple datasets still to process, this defines which ones to return. Options are 'normal' - the oldest datasets, or 'random'.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "API key to authenticate the user",
+                        "required": false,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "qualities",
+                        "in": "query",
+                        "description": "Comma-separated list of (at least two) quality names, e.g. 'NumberOfInstances,NumberOfFeatures'.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of unprocessed datasets",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/DataUnprocessed"
+                                },
+                                "example": {
+                                    "data_unprocessed": {
+                                        "run": [
+                                            {
+                                                "did": "1",
+                                                "status": "deactivated",
+                                                "version": "2",
+                                                "name": "anneal",
+                                                "format": "ARFF"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n686 - Please specify the features the evaluation engine wants to calculate (at least 2).\\n687 - No unprocessed datasets according to the given set of meta-features.\\n688 - Illegal qualities.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/estimationprocedure/list": {
+            "get": {
+                "tags": [
+                    "estimationprocedure"
+                ],
+                "summary": "List all estimation procedures",
+                "description": "Returns an array with all model performance estimation procedures in the system.",
+                "operationId": "Api_estimationprocedure::estimationprocedure_list",
+                "responses": {
+                    "200": {
+                        "description": "A list of estimation procedures",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/EstimationProcedureList"
+                                },
+                                "example": {
+                                    "estimationprocedures": {
+                                        "estimationprocedure": [
+                                            {
+                                                "id": "1",
+                                                "ttid": "1",
+                                                "name": "10-fold Crossvalidation",
+                                                "type": "crossvalidation",
+                                                "repeats": "1",
+                                                "folds": "10",
+                                                "stratified_sampling": "true"
+                                            },
+                                            {
+                                                "id": "2",
+                                                "ttid": "1",
+                                                "name": "5 times 2-fold Crossvalidation",
+                                                "type": "crossvalidation",
+                                                "repeats": "5",
+                                                "folds": "2",
+                                                "stratified_sampling": "true"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n500 - No model performance estimation procedures available.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/evaluation/request/{evaluation_engine_id}/{order}": {
+            "get": {
+                "tags": [
+                    "evaluation"
+                ],
+                "summary": "Get an unevaluated run",
+                "description": "This call is for people running their own evaluation engines. It returns the details of a run that is not yet evaluated by the given evaluation engine. It doesn't evaluate the run, it just returns the run info.",
+                "operationId": "Api_evaluation::evaluation_request",
+                "parameters": [
+                    {
+                        "name": "evaluation_engine_id",
+                        "in": "path",
+                        "description": "The ID of the evaluation engine. You get this ID when you register a new evaluation engine with OpenML. The ID of the main evaluation engine is 1.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "order",
+                        "in": "path",
+                        "description": "When there are multiple runs still to evaluate, this defines which one to return. Options are 'normal' - the oldest run, 'reverse' - the newest run, or 'random' - a random run.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of evaluations descriptions",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/EvaluationRequest"
+                                },
+                                "example": {
+                                    "evaluation_request": {
+                                        "run": [
+                                            {
+                                                "setup_id": "68799271",
+                                                "upload_time": "2018-04-03 21:05:38",
+                                                "uploader": "1935",
+                                                "task_id": "3021",
+                                                "run_id": "8943712"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n100 - Function not valid.\\n545 - No unevaluated runs according to the criteria.\\n546 - Illegal filter.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/evaluation/list/{filters}": {
+            "get": {
+                "tags": [
+                    "evaluation"
+                ],
+                "summary": "List and filter evaluations",
+                "description": "List evaluations, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/evaluation/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all evaluations that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, uploaders or runs.",
+                "operationId": "Api_evaluation::evaluation_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /function/{name} - name of the evaluation measure, e.g. area_under_auc or predictive_accuracy. See the OpenML website for the complete list of measures.\r\n  /tag/{tag} - returns only evaluations of runs tagged with the given tag.\r\n  /run/{ids} - return only evaluations for specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''\r\n  /task/{ids} - return only evaluations for specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''\r\n  /flow/{ids} - return only evaluations for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only evaluations for specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  /per_fold/{true,false} - whether or not to return crossvalidation scores per fold. Defaults to 'false'. Setting it to 'true' leads to large numbers of results, use only for very specific sets of runs.\r\n  /sort_order/{asc,desc} - sorts the results by the evaluation value, according to the selected evaluation measure (function)\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of evaluations descriptions",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/EvaluationList"
+                                },
+                                "example": {
+                                    "evaluations": {
+                                        "evaluation": [
+                                            {
+                                                "function": "area_under_roc_curve",
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "68",
+                                                "run_id": "1",
+                                                "array_data": "[0,0.99113,0.898048,0.874862,0.791282,0.807343,0.820674]",
+                                                "value": "0.839359",
+                                                "uploader": "1",
+                                                "flow_id": "61"
+                                            },
+                                            {
+                                                "function": "f_measure",
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "68",
+                                                "run_id": "1",
+                                                "array_data": "[0,0,0.711934,0.735714,0.601363,0.435678,0.430913]",
+                                                "value": "0.600026",
+                                                "uploader": "1",
+                                                "flow_id": "61"
+                                            },
+                                            {
+                                                "function": "predictive_accuracy",
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "68",
+                                                "run_id": "1",
+                                                "array_data": [],
+                                                "value": "0.614634",
+                                                "uploader": "1",
+                                                "flow_id": "61"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n540 - Please provide at least task, flow or setup, uploader or run, to\\nfilter results, or limit the number of responses.\\n541 - The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\\n542 - There where no results. Check whether there are runs under the given constraint.\\n543 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\\n544 - Illegal filter specified.\\n545 - Offset specified without limit.\\n546 - Requested result limit too high.\\n547 - Per fold can only be set to value 'true' or 'false'.\\n548 - Per fold queries are experimental and require a fair amount of filters on resulting run records to keep the query fast (use, e.g., flow, setup, task and uploader filter)\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/evaluationmeasure/list": {
+            "get": {
+                "tags": [
+                    "evaluationmeasure"
+                ],
+                "summary": "List all evaluation measures",
+                "description": "Returns an array with all model evaluation measures in the system.",
+                "operationId": "Api_evaluationmeasure::evaluationmeasure_list",
+                "responses": {
+                    "200": {
+                        "description": "A list of evaluation measures",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/EvaluationMeasureList"
+                                },
+                                "example": {
+                                    "evaluation_measures": {
+                                        "measures": {
+                                            "measure": [
+                                                "area_under_roc_curve",
+                                                "average_cost",
+                                                "binominal_test",
+                                                "build_cpu_time"
+                                            ]
+                                        }
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow/list/{filters}": {
+            "get": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "List and filter flows",
+                "description": "List flows, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/task/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all flows that match the constraints.",
+                "operationId": "Api_flow::flow_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.\r\n  /tag/{tag} - returns only tasks tagged with the given tag.\r\n  /uploader/{id} - return only evaluations uploaded by a specific user, specified by user ID.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of flows",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/FlowList"
+                                },
+                                "example": {
+                                    "flows": {
+                                        "flow": [
+                                            {
+                                                "id": "65",
+                                                "full_name": "weka.RandomForest(1)",
+                                                "name": "weka.RandomForest",
+                                                "version": "1",
+                                                "external_version": "Weka_3.7.10_9186",
+                                                "uploader": "1"
+                                            },
+                                            {
+                                                "id": "66",
+                                                "full_name": "weka.IBk(1)",
+                                                "name": "weka.IBk",
+                                                "version": "1",
+                                                "external_version": "Weka_3.7.10_8034",
+                                                "uploader": "1"
+                                            },
+                                            {
+                                                "id": "67",
+                                                "full_name": "weka.BayesNet_K2(1)",
+                                                "name": "weka.BayesNet_K2",
+                                                "version": "1",
+                                                "external_version": "Weka_3.7.10_8034",
+                                                "uploader": "1"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n500 - No results. There where no matches for the given constraints.\\n501 - Illegal filter specified.\\n502 - Filter values/ranges not properly specified.\\n503 - Can not specify an offset without a limit.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow/tag": {
+            "post": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Tag a flow",
+                "description": "Tags a flow.",
+                "operationId": "Api_flow::flow_tag",
+                "parameters": [
+                    {
+                        "name": "flow_id",
+                        "in": "query",
+                        "description": "Id of the flow.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the tagged flow",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_tag": {
+                                            "$ref": "#/components/schemas/inline_response_200_12_flow_tag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "flow_tag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\\n474 - Internal error tagging the entity. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow/untag": {
+            "post": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Untag a flow",
+                "description": "Untags a flow.",
+                "operationId": "Api_flow::flow_untag",
+                "parameters": [
+                    {
+                        "name": "flow_id",
+                        "in": "query",
+                        "description": "Id of the flow.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the untagged flow",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_untag": {
+                                            "$ref": "#/components/schemas/inline_response_200_13_flow_untag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "flow_untag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\\nby another user. Hence you cannot delete it.\\n479 - Internal error removing the tag. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow/exists/{name}/{version}": {
+            "get": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Check whether flow exists",
+                "description": "Checks whether a flow with the given name and (external) version exists.",
+                "operationId": "Api_flow::flow_exists",
+                "parameters": [
+                    {
+                        "name": "name",
+                        "in": "path",
+                        "description": "The name of the flow.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "version",
+                        "in": "path",
+                        "description": "The external version of the flow",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of flows",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_exists": {
+                                            "$ref": "#/components/schemas/inline_response_200_10_flow_exists"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "flow_exists": {
+                                        "exists": "true",
+                                        "id": "65"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n330 - Mandatory fields not present. Please provide name and external_version.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow/{id}": {
+            "get": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Get flow description",
+                "description": "Returns information about a flow. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.",
+                "operationId": "Api_flow::flow",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the flow.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A flow description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Flow"
+                                },
+                                "example": {
+                                    "flow": {
+                                        "id": "100",
+                                        "uploader": "1",
+                                        "name": "weka.J48",
+                                        "version": "2",
+                                        "external_version": "Weka_3.7.5_9117",
+                                        "description": "...",
+                                        "upload_date": "2014-04-23 18:00:36",
+                                        "language": "Java",
+                                        "dependencies": "Weka_3.7.5",
+                                        "parameter": [
+                                            {
+                                                "name": "A",
+                                                "data_type": "flag",
+                                                "default_value": [],
+                                                "description": "Laplace smoothing..."
+                                            },
+                                            {
+                                                "name": "C",
+                                                "data_type": "option",
+                                                "default_value": "0.25",
+                                                "description": "Set confidence threshold..."
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n180 - Please provide flow id.\\n181 - Unknown flow. The flow with the given ID was not found in the database.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Delete a flow",
+                "description": "Deletes a flow. Upon success, it returns the ID of the deleted flow.",
+                "operationId": "Api_flow::flow_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the flow.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "API key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted flow",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_8_flow_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "flow_delete": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n320 - Please provide API key. In order to remove your content, please authenticate.\\n321 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\\n322 - Flow does not exists. The flow ID could not be linked to an existing flow.\\n323 - Flow is not owned by you. The flow is owned by another user. Hence you cannot delete it.\\n324 - Flow is in use by other content. Can not be deleted. The flow is used in runs, evaluations or as a component of another flow. Delete other content before deleting this flow.\\n325 - Deleting flow failed. Deleting the flow failed. Please contact\\nsupport team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/flow": {
+            "post": {
+                "tags": [
+                    "flow"
+                ],
+                "summary": "Upload a flow",
+                "description": "Uploads a flow. Upon success, it returns the flow id.",
+                "operationId": "Api_flow::flow_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the flow. Only name and description are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.implementation.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/flow).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "flow",
+                        "in": "query",
+                        "description": "The actual flow, being a source (or binary) file.",
+                        "required": false,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the uploaded flow",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_flow": {
+                                            "$ref": "#/components/schemas/inline_response_200_9_upload_flow"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "upload_flow": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n160 - Error in file uploading. There was a problem with the file upload.\\n161 - Please provide description xml.\\n163 - Problem validating uploaded description file. The XML description format does not meet the standards.\\n164 - Flow already stored in database. Please change name or version number\\n165 - Failed to insert flow. There can be many causes for this error. If you included the implements field, it should be an existing entry in the algorithm or math_function table. Otherwise it could be an internal server error. Please contact API support team.\\n166 - Failed to add flow to database. Internal server error, please contact API administrators\\n167 - Illegal files uploaded. An non required file was uploaded.\\n168 - The provided md5 hash equals not the server generated md5 hash of the file.\\n169 - Please provide API key. In order to share content, please authenticate and provide API key.\\n170 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\\n171 - Flow already exists. This flow is already in the database\\n172 - XSD not found. Please contact API support team\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/tag": {
+            "post": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Tag a run",
+                "description": "Tags a run.",
+                "operationId": "Api_run::run_tag",
+                "parameters": [
+                    {
+                        "name": "run_id",
+                        "in": "query",
+                        "description": "Id of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the tagged run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "run_tag": {
+                                            "$ref": "#/components/schemas/inline_response_200_19_run_tag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_tag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\\n474 - Internal error tagging the entity. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/untag": {
+            "post": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Untag a run",
+                "description": "Untags a run.",
+                "operationId": "Api_run::run_untag",
+                "parameters": [
+                    {
+                        "name": "run_id",
+                        "in": "query",
+                        "description": "Id of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the untagged run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "run_untag": {
+                                            "$ref": "#/components/schemas/inline_response_200_20_run_untag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_untag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\\n479 - Internal error removing the tag. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/list/{filters}": {
+            "get": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "List and filter runs",
+                "description": "List runs, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/run/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all runs that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, or uploaders.",
+                "operationId": "Api_run::run_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /tag/{tag} - return only runs tagged with the given tag.\r\n  /run/{ids} - return only specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''\r\n  /task/{ids} - return only runs on specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''\r\n  /flow/{ids} - return only runs on specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only runs with specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /uploader/{ids} - return only runs uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of runs descriptions",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/RunList"
+                                },
+                                "example": {
+                                    "runs": {
+                                        "run": [
+                                            {
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "28",
+                                                "run_id": "100",
+                                                "error_message": [],
+                                                "setup_id": "12",
+                                                "uploader": "1",
+                                                "flow_id": "67"
+                                            },
+                                            {
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "48",
+                                                "run_id": "101",
+                                                "error_message": [],
+                                                "setup_id": "6",
+                                                "uploader": "1",
+                                                "flow_id": "61"
+                                            },
+                                            {
+                                                "upload_time": "2014-04-06 23:30:40",
+                                                "task_id": "41",
+                                                "run_id": "102",
+                                                "error_message": [],
+                                                "setup_id": "3",
+                                                "uploader": "1",
+                                                "flow_id": "58"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n510 - Please provide at least task, flow or setup, uploader or run, to filter results, or limit the number of responses. The number of runs is huge. Please limit the result space.\\n511 - Input not safe. The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\\n512 - There where no results. Check whether there are runs under the given constraint.\\n513 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\\n514 - Illegal filter specified.\\n515 - Offset specified without limit.\\n516 - Requested result limit too high.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/{id}": {
+            "get": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Get run description",
+                "description": "Returns information about a run. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.",
+                "operationId": "Api_run::run",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A run description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Run"
+                                },
+                                "example": {
+                                    "run": {
+                                        "run_id": "100",
+                                        "uploader": "1",
+                                        "uploader_name": "Jan van Rijn",
+                                        "task_id": "28",
+                                        "task_type": "Supervised Classification",
+                                        "task_evaluation_measure": "predictive_accuracy",
+                                        "flow_id": "67",
+                                        "flow_name": "weka.BayesNet_K2(1)",
+                                        "setup_string": "weka.classifiers.bayes.BayesNet -- -D -Q weka.classifiers.bayes.net.search.local.K2 -- -P 1 -S BAYES -E weka.classifiers.bayes.net.estimate.SimpleEstimator -- -A 0.5",
+                                        "parameter_setting": [
+                                            {
+                                                "name": "D",
+                                                "value": "true"
+                                            },
+                                            {
+                                                "name": "Q",
+                                                "value": "weka.classifiers.bayes.net.search.local.K2"
+                                            },
+                                            {
+                                                "name": "P",
+                                                "value": "1"
+                                            },
+                                            {
+                                                "name": "S",
+                                                "value": "BAYES"
+                                            }
+                                        ],
+                                        "input_data": {
+                                            "dataset": {
+                                                "did": "28",
+                                                "name": "optdigits",
+                                                "url": "https:\\\\/\\\\/www.openml.org\\\\/data\\\\/download\\\\/28\\\\/dataset_28_optdigits.arff"
+                                            }
+                                        },
+                                        "output_data": {
+                                            "file": [
+                                                {
+                                                    "did": "48838",
+                                                    "file_id": "261",
+                                                    "name": "description",
+                                                    "url": "https:\\\\/\\\\/www.openml.org\\\\/data\\\\/download\\\\/261\\\\/weka_generated_run935374685998857626.xml"
+                                                },
+                                                {
+                                                    "did": "48839",
+                                                    "file_id": "262",
+                                                    "name": "predictions",
+                                                    "url": "https:\\\\/\\\\/www.openml.org\\\\/data\\\\/download\\\\/262\\\\/weka_generated_predictions576954524972002741.arff"
+                                                }
+                                            ],
+                                            "evaluation": [
+                                                {
+                                                    "name": "area_under_roc_curve",
+                                                    "flow_id": "4",
+                                                    "value": "0.990288",
+                                                    "array_data": "[0.99724,0.989212,0.992776,0.994279,0.980578,0.98649,0.99422,0.99727,0.994858,0.976143]"
+                                                },
+                                                {
+                                                    "name": "confusion_matrix",
+                                                    "flow_id": "10",
+                                                    "array_data": "[[544,1,0,0,7,0,1,0,0,1],[0,511,21,1,0,1,3,1,5,28],[0,7,511,1,0,1,0,3,23,11],[0,2,2,519,0,3,0,12,16,18],[0,3,0,0,528,0,4,21,6,6],[0,1,0,7,5,488,2,0,4,51],[1,7,0,0,2,0,548,0,0,0],[0,2,0,1,9,1,0,545,4,4],[1,25,2,2,3,6,2,1,503,9],[0,7,0,20,16,5,0,19,9,486]]"
+                                                },
+                                                {
+                                                    "name": "f_measure",
+                                                    "flow_id": "12",
+                                                    "value": "0.922723",
+                                                    "array_data": "[0.989091,0.898857,0.935041,0.92431,0.927944,0.918156,0.980322,0.933219,0.895018,0.826531]"
+                                                },
+                                                {
+                                                    "name": "kappa",
+                                                    "flow_id": "13",
+                                                    "value": "0.913601"
+                                                }
+                                            ]
+                                        }
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n220 - Please provide run ID. In order to view run details, please provide the run ID.\\n221 - Run not found. The run ID was invalid, run does not exist (anymore).\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Delete run",
+                "description": "Deletes a run. Upon success, it returns the ID of the deleted run.",
+                "operationId": "Api_run::run_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "API key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "data_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_17_data_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_delete": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n390 - Please provide API key. In order to remove your content, please authenticate.\\n391 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators\\n392 - Run does not exists. The run ID could not be linked to an existing run.\\n393 - Run is not owned by you. The run was owned by another user. Hence you cannot delete it.\\n394 - Deleting run failed. Deleting the run failed. Please contact support team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/reset/{id}": {
+            "get": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Resets a run.",
+                "description": "Removes all run evaluations. When a run is reset, the runs will automatically be evaluated as soon as they are picked up by the evaluation engine again.",
+                "operationId": "Api_run::run_reset",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Run ID.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the evaluated run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "run_reset": {
+                                            "$ref": "#/components/schemas/inline_response_200_21_upload_flow"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_reset": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n412 - Run does not exist\\n413 - Run is not owned by you\\n394 - Resetting run failed\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run": {
+            "post": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Upload run",
+                "description": "Uploads a run. Upon success, it returns the run id.",
+                "operationId": "Api_run::run_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/run).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "predictions",
+                        "in": "query",
+                        "description": "The predictions generated by the run",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "model_readable",
+                        "in": "query",
+                        "description": "The human-readable model generated by the run",
+                        "required": false,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "model_serialized",
+                        "in": "query",
+                        "description": "The serialized model generated by the run",
+                        "required": false,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the uploaded run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_flow": {
+                                            "$ref": "#/components/schemas/inline_response_200_18_upload_flow"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "upload_run": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n201 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\\n202 - Please provide run XML.\\n203 - Could not validate run xml by XSD. Please double check that the xml is valid.\\n204 - Unknown task. The task with the given ID was not found in the database.\\n205 - Unknown flow. The flow with the given ID was not found in the database.\\n206 - Invalid number of files. The number of uploaded files did not match the number of files expected for the task type\\n207 - File upload failed. One of the files uploaded has a problem.\\n208 - Error inserting setup record. Please contact api administrators\\n210 - Unable to store run. Please contact api administrators.\\n211 - Dataset not in database. One of the datasets of the task was not included in database, please contact api administrators.\\n212 - Unable to store file. Please contact api administrators.\\n213 - Parameter in run xml unknown. One of the parameters provided in the run xml is not registered as parameter for the flow nor its components.\\n214 - Unable to store input setting. Please contact API support team.\\n215 - Unable to evaluate predictions. Please contact API support team.\\n216 - Error thrown by Java Application. Additional information field is provided.\\n217 - Error processing output data. Unknown or inconsistent evaluation measure. One of the provided evaluation measures could not be matched with a record in the math_function or flow table.\\n218 - Wrong flow associated with run. The flow implements a math_function, which is unable to generate predictions. Please select another flow.\\n219 - Error reading the XML document. The XML description file could not be verified.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/trace/{id}": {
+            "get": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Get run trace",
+                "description": "Returns the optimization trace of run. The trace contains every setup tried, its evaluation, and whether it was selected.",
+                "operationId": "Api_run::run_trace",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A run trace description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/RunTrace"
+                                },
+                                "example": {
+                                    "trace": {
+                                        "run_id": "573055",
+                                        "trace_iteration": {
+                                            "repeat": "0",
+                                            "fold": "0",
+                                            "iteration": "1",
+                                            "setup_string": {
+                                                "parameter_minNumObj": "1",
+                                                "parameter_confidenceFactor": "0.25"
+                                            },
+                                            "evaluation": "94.074074",
+                                            "selected": "true"
+                                        }
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n570 - No successful trace associated with this run\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "post": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Upload run trace",
+                "description": "Uploads a run trace. Upon success, it returns the run id.",
+                "operationId": "Api_run::run_trace_upload",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the run.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the trace. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.trace) and an [XML example](https://www.openml.org/api/v1/xml_example/run.trace).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the run with the trace",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_flow": {
+                                            "$ref": "#/components/schemas/inline_response_200_23_upload_flow"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_trace": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n561 - Problem with uploaded trace file.\\n562 - Problem validating xml trace file.\\n563 - Problem loading xml trace file.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/run/evaluate": {
+            "post": {
+                "tags": [
+                    "run"
+                ],
+                "summary": "Uploads a run evaluation",
+                "description": "Uploads a run evaluation. When successful, it returns the run id.",
+                "operationId": "Api_run::run_evaluate",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the run evaluation.Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.evaluate) and an [XML example](https://www.openml.org/api/v1/xml_example/run.evaluate).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the evaluated run",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_flow": {
+                                            "$ref": "#/components/schemas/inline_response_200_21_upload_flow"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "run_evaluate": {
+                                        "id": "2520"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n422 - Upload problem description XML\\n423 - Problem validating uploaded description file\\n424 - Problem opening description xml\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/setup/tag": {
+            "post": {
+                "tags": [
+                    "setup"
+                ],
+                "summary": "Tag a setup",
+                "description": "Tags a setup.",
+                "operationId": "Api_setup::setup_tag",
+                "parameters": [
+                    {
+                        "name": "setup_id",
+                        "in": "query",
+                        "description": "Id of the setup.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the tagged setup",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_tag": {
+                                            "$ref": "#/components/schemas/inline_response_200_15_flow_tag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "setup_tag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\\n474 - Internal error tagging the entity. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/setup/untag": {
+            "post": {
+                "tags": [
+                    "setup"
+                ],
+                "summary": "Untag a setup",
+                "description": "Untags a setup.",
+                "operationId": "Api_setup::setup_untag",
+                "parameters": [
+                    {
+                        "name": "setup_id",
+                        "in": "query",
+                        "description": "Id of the setup.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the untagged setup",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "flow_untag": {
+                                            "$ref": "#/components/schemas/inline_response_200_16_flow_untag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "setup_untag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\\nby another user. Hence you cannot delete it.\\n479 - Internal error removing the tag. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/setup/{id}": {
+            "get": {
+                "tags": [
+                    "setup"
+                ],
+                "summary": "Get a hyperparameter setup",
+                "description": "Returns information about a setup. The information includes the list of hyperparameters, with name, value, and default value.",
+                "operationId": "Api_setup::setup",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the hyperparameter setup (configuration). These IDs are stated in run descriptions.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A setup description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Setup"
+                                },
+                                "example": {
+                                    "setup_parameters": {
+                                        "flow_id": "59",
+                                        "parameter": [
+                                            {
+                                                "full_name": "weka.JRip(1)_F",
+                                                "parameter_name": "F",
+                                                "data_type": "option",
+                                                "default_value": "3",
+                                                "value": "3"
+                                            },
+                                            {
+                                                "full_name": "weka.JRip(1)_N",
+                                                "parameter_name": "N",
+                                                "data_type": "option",
+                                                "default_value": "2.0",
+                                                "value": "2.0"
+                                            },
+                                            {
+                                                "full_name": "weka.JRip(1)_O",
+                                                "parameter_name": "O",
+                                                "data_type": "option",
+                                                "default_value": "2",
+                                                "value": "2"
+                                            },
+                                            {
+                                                "full_name": "weka.JRip(1)_S",
+                                                "parameter_name": "S",
+                                                "data_type": "option",
+                                                "default_value": "1",
+                                                "value": "1"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n280 - Please provide setup ID. In order to view setup details, please provide the run ID\\n281 - Setup not found. The setup ID was invalid, or setup does not exist (anymore).\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "setup"
+                ],
+                "summary": "Delete setup",
+                "description": "Deletes a setup. Upon success, it returns the ID of the deleted setup.",
+                "operationId": "Api_setup::setup_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the setup.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted setup",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "study_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_14_study_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "setup_delete": {
+                                        "id": "1"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n401 - Authentication failed. Please provide API key. In order to remove your content, please authenticate.\\n402 - Setup does not exists. The setup ID could not be linked to an existing setup.\\n404 - Setup deletion failed. Setup is in use by other content (runs, schedules, etc). Can not be deleted.\\n405 - Setup deletion failed. Please try again later.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/setup/list/{filters}": {
+            "get": {
+                "tags": [
+                    "setup"
+                ],
+                "summary": "List and filter setups",
+                "description": "List setups, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/setup/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all evaluations that match the constraints. A maximum of 1,000 results are returned at a time, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain flows, setups, or tags.",
+                "operationId": "Api_setup::setup_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /tag/{tag} - returns only setups tagged with the given tag.\r\n  /flow/{ids} - return only setups for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of setup descriptions",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/SetupList"
+                                },
+                                "example": {
+                                    "setups": {
+                                        "setup": [
+                                            {
+                                                "setup_id": "10",
+                                                "flow_id": "65",
+                                                "parameter": [
+                                                    {
+                                                        "id": "4144",
+                                                        "flow_id": "65",
+                                                        "flow_name": "weka.RandomForest",
+                                                        "full_name": "weka.RandomForest(1)_I",
+                                                        "parameter_name": "I",
+                                                        "data_type": "option",
+                                                        "default_value": "10",
+                                                        "value": "10"
+                                                    },
+                                                    {
+                                                        "id": "4145",
+                                                        "flow_id": "65",
+                                                        "flow_name": "weka.RandomForest",
+                                                        "full_name": "weka.RandomForest(1)_K",
+                                                        "parameter_name": "K",
+                                                        "data_type": "option",
+                                                        "default_value": "0",
+                                                        "value": "0"
+                                                    }
+                                                ]
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n670 - Please specify at least one filter.\\n671 - Illegal filter.\\n672 - Illegal filter input.\\n673 - Result set too big. Please use one of the filters or the limit option.\\n674 - No results, please check the filter.\\n675 - Cannot specify offset without limit.\\n676 - Requested result limit too high.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/study": {
+            "post": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "Create new study",
+                "description": "Creates a new study. Upon success, it returns the study id.",
+                "operationId": "Api_study::study_create",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the study. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.study.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/study).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the uploaded study",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_study": {
+                                            "$ref": "#/components/schemas/inline_response_200_25_upload_study"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "upload_study": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n1031 - Description file not present. Please upload the study description.\\n1032 - Problem validating uploaded description file. The XML description format does not meet the standards. See the XSD schema.\\n1033 - Illegal main entity type. Currently only collections of tasks and can be created.\\n1034 - Linked entities are not of the correct type fot this study.\\n1035 - Benchmark suites can only be linked to run studies.\\n1036 - Referred benchmark suite cannot be found.\\n1037 - Referred benchmark suite should be a task collection.\\n1038 - Study alias is not unique.\\n1039 - Dataset insertion problem. Please contact the administrators.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/study/{id}/attach": {
+            "post": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "Attach a new entity to a study",
+                "description": "Attach a new entity to an exising study. Upon success, it returns the study id, type, and linked entities.",
+                "operationId": "Api_study::study_attach",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the study. Supplied in the URL path.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "ids",
+                        "in": "query",
+                        "description": "Comma-separated list of entity IDs to be attached to the study. For instance, if this is a run study, the list of run IDs that need to be added (attached) to the study. Must be supplied as a POST variable.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Properties of the updated study",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "study_attach": {
+                                            "$ref": "#/components/schemas/inline_response_200_26_study_attach"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "study_attach": {
+                                        "id": "1",
+                                        "main_entity_type": "task",
+                                        "linked_entities": "5"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n1041 - Could not find study. Check the study ID in your request.\\n1042 - Cannnot attach entities to legacy studies.\\n1043 - Please provide POST field 'ids'.\\n1044 - Please ensure that the 'ids' in the POST field is a list of natural numbers.\\n1045 - Could not attach entities to the study. It appears as if the entity does not exist.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/study/{id}/detach": {
+            "post": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "Detach an entity from a study",
+                "description": "Detach an entity from an exising study. Upon success, it returns the study id, type, and linked entities.",
+                "operationId": "Api_study::study_detach",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the study.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "ids",
+                        "in": "query",
+                        "description": "Comma-separated list of entity IDs to be detached from the study. For instance, if this is a run study, the list of run IDs that need to be removed (detached) from the study. Must be supplied as a POST variable.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Properties of the updated study",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_study": {
+                                            "$ref": "#/components/schemas/inline_response_200_26_study_attach"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "study_detach": {
+                                        "id": "1",
+                                        "main_entity_type": "task",
+                                        "linked_entities": "5"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n1041 - Could not find study. Check the study ID in your request.\\n1042 - Cannot attach entities to legacy studies.\\n1043 - Please provide POST field 'ids'.\\n1044 - Please ensure that the 'ids' in the POST field is a list of natural numbers.\\n1046 - Could not detach entities from the study. It appears as if the entity does not exist.       \\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/study/{id}": {
+            "get": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "Get study description by study id or alias",
+                "description": "Returns information about the study with the given id or alias.",
+                "operationId": "Api_study::study_get",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID or alias of the study.",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A study description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Study"
+                                },
+                                "example": {
+                                    "study": {
+                                        "id": "99",
+                                        "main_entity_type": "task",
+                                        "name": "CC18 benchmark suite",
+                                        "description": "CC18 benchmark suite",
+                                        "creation_date": "2019-02-16T17:35:58",
+                                        "creator": "1159",
+                                        "data": {
+                                            "data_id": [
+                                                "1",
+                                                "2",
+                                                "3"
+                                            ]
+                                        },
+                                        "tasks": {
+                                            "task_id": [
+                                                "1",
+                                                "2",
+                                                "3"
+                                            ]
+                                        }
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n601 - Unknown study. The study with the given id or alias was not found in the database\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "Delete study",
+                "description": "Deletes a study. Upon success, it returns the ID of the deleted study.",
+                "operationId": "Api_study::study_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the study.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted study",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "study_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_24_study_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "study_delete": {
+                                        "id": "1"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n591 - Please provide API key. In order to remove your content, please authenticate.\\n592 - Study does not exists. The study ID could not be linked to an existing study.\\n593 - Study deletion failed. Please try again later.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/study/list/{filters}": {
+            "get": {
+                "tags": [
+                    "study"
+                ],
+                "summary": "List all studies (collections of items)",
+                "description": "List studies, optionally filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/study/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all studies that match the constraints.",
+                "operationId": "Api_study::study_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /main_entity_type/{type} - only return studies collecting entities of a given type (e.g. 'task' or 'run').\r\n  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of studies",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/StudyList"
+                                },
+                                "example": {
+                                    "study_list": {
+                                        "study": [
+                                            {
+                                                "id": "1",
+                                                "alias": "Study_1",
+                                                "name": "A large-scale comparison of classification algorithms",
+                                                "creation_date": "2017-07-20 15:51:20",
+                                                "creator": "2"
+                                            },
+                                            {
+                                                "id": "2",
+                                                "alias": "Study_2",
+                                                "name": "Fast Algorithm Selection using Learning Curves",
+                                                "creation_date": "2017-07-20 15:51:20",
+                                                "creator": "2"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/task/tag": {
+            "post": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "Tag a task",
+                "description": "Tags a task.",
+                "operationId": "Api_task::task_tag",
+                "parameters": [
+                    {
+                        "name": "task_id",
+                        "in": "query",
+                        "description": "Id of the task.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "The id of the tagged task",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "task_tag": {
+                                            "$ref": "#/components/schemas/inline_response_200_6_task_tag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "task_tag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n470 - In order to add a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\\n471 - Entity not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\\n472 - Entity already tagged by this tag. The entity {dataset, task, flow, run} already had this tag.\\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\\n474 - Internal error tagging the entity. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/task/untag": {
+            "post": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "Untag a task",
+                "description": "Untags a task.",
+                "operationId": "Api_task::task_untag",
+                "parameters": [
+                    {
+                        "name": "task_id",
+                        "in": "query",
+                        "description": "Id of the task.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "tag",
+                        "in": "query",
+                        "description": "Tag name",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A the features of the task",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "task_untag": {
+                                            "$ref": "#/components/schemas/inline_response_200_7_task_untag"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "task_untag": {
+                                        "id": "2"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\\n476 - Entity {dataset, task, flow, run} not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\\n477 - Tag not found. The provided tag is not associated with the entity {dataset, task, flow, run}.\\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\\n479 - Internal error removing the tag. Please contact OpenML Team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/task/list/{filters}": {
+            "get": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "List and filter tasks",
+                "description": "List tasks, possibly filtered by a range of properties from the task itself or from the underlying dataset. Any number of properties can be combined by listing them one after the other in the form '/task/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all tasks that match the constraints.",
+                "operationId": "Api_task::task_list",
+                "parameters": [
+                    {
+                        "name": "filters",
+                        "in": "path",
+                        "description": "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.\r\n  /status/{status} - returns only tasks with a given status, either 'active', 'deactivated', or 'in_preparation'.\r\n  /type/{type_id} - returns only tasks with a given task type id. See the list of task types of the ID's (e.g. 1 = Supervised Classification).\r\n  /tag/{tag} - returns only tasks tagged with the given tag.\r\n  /data_tag/{tag} - returns only tasks for which the underlying dataset is tagged with the given tag.\r\n  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.\r\n  ",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A list of tasks with the given tag",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/TaskList"
+                                },
+                                "example": {
+                                    "task": {
+                                        "task": [
+                                            {
+                                                "task_id": "1",
+                                                "task_type": "Supervised Classification",
+                                                "did": "1",
+                                                "name": "anneal",
+                                                "status": "active",
+                                                "format": "ARFF",
+                                                "input": [
+                                                    {
+                                                        "name": "estimation_procedure",
+                                                        "value": "1"
+                                                    },
+                                                    {
+                                                        "name": "evaluation_measures",
+                                                        "value": "predictive_accuracy"
+                                                    },
+                                                    {
+                                                        "name": "source_data",
+                                                        "value": "1"
+                                                    },
+                                                    {
+                                                        "name": "target_feature",
+                                                        "value": "class"
+                                                    }
+                                                ],
+                                                "quality": [
+                                                    {
+                                                        "name": "MajorityClassSize",
+                                                        "value": "684"
+                                                    },
+                                                    {
+                                                        "name": "MaxNominalAttDistinctValues",
+                                                        "value": "10.0"
+                                                    },
+                                                    {
+                                                        "name": "MinorityClassSize",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumBinaryAtts",
+                                                        "value": "14.0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfClasses",
+                                                        "value": "6"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfFeatures",
+                                                        "value": "39"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfInstances",
+                                                        "value": "898"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfInstancesWithMissingValues",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfMissingValues",
+                                                        "value": "0"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfNumericFeatures",
+                                                        "value": "6"
+                                                    },
+                                                    {
+                                                        "name": "NumberOfSymbolicFeatures",
+                                                        "value": "32"
+                                                    }
+                                                ],
+                                                "tag": [
+                                                    "basic",
+                                                    "study_1",
+                                                    "study_7",
+                                                    "under100k",
+                                                    "under1m"
+                                                ]
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n480 - Illegal filter specified.\\n481 - Filter values/ranges not properly specified.\\n482 - No results. There where no matches for the given constraints.\\n483 - Can not specify an offset without a limit.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/task/{id}": {
+            "get": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "Get task description",
+                "description": "Returns information about a task. The information includes the task type, input data, train/test sets, and more.",
+                "operationId": "Api_task::task",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the task.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A task description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Task"
+                                },
+                                "example": {
+                                    "task": {
+                                        "task_id": "1",
+                                        "task_type": "Supervised Classification",
+                                        "input": [
+                                            {
+                                                "name": "source_data",
+                                                "data_set": {
+                                                    "data_set_id": "1",
+                                                    "target_feature": "class"
+                                                }
+                                            },
+                                            {
+                                                "name": "estimation_procedure",
+                                                "estimation_procedure": {
+                                                    "type": "crossvalidation",
+                                                    "data_splits_url": "https://www.openml.org/api_splits/get/1/Task_1_splits.arff",
+                                                    "parameter": [
+                                                        {
+                                                            "name": "number_repeats",
+                                                            "value": "1"
+                                                        },
+                                                        {
+                                                            "name": "number_folds",
+                                                            "value": "10"
+                                                        },
+                                                        {
+                                                            "name": "percentage"
+                                                        },
+                                                        {
+                                                            "name": "stratified_sampling",
+                                                            "value": "true"
+                                                        }
+                                                    ]
+                                                }
+                                            },
+                                            {
+                                                "name": "cost_matrix",
+                                                "cost_matrix": []
+                                            },
+                                            {
+                                                "name": "evaluation_measures",
+                                                "evaluation_measures": {
+                                                    "evaluation_measure": "predictive_accuracy"
+                                                }
+                                            }
+                                        ],
+                                        "output": {
+                                            "name": "predictions",
+                                            "predictions": {
+                                                "format": "ARFF",
+                                                "feature": [
+                                                    {
+                                                        "name": "repeat",
+                                                        "type": "integer"
+                                                    },
+                                                    {
+                                                        "name": "fold",
+                                                        "type": "integer"
+                                                    },
+                                                    {
+                                                        "name": "row_id",
+                                                        "type": "integer"
+                                                    },
+                                                    {
+                                                        "name": "confidence.classname",
+                                                        "type": "numeric"
+                                                    },
+                                                    {
+                                                        "name": "prediction",
+                                                        "type": "string"
+                                                    }
+                                                ]
+                                            }
+                                        },
+                                        "tag": [
+                                            "basic",
+                                            "study_1",
+                                            "under100k",
+                                            "under1m"
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n150 - Please provide task_id.\\n151 - Unknown task. The task with the given id was not found in the database\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            },
+            "delete": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "Delete task",
+                "description": "Deletes a task. Upon success, it returns the ID of the deleted task.",
+                "operationId": "Api_task::task_delete",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "Id of the task.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "ID of the deleted task",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "task_delete": {
+                                            "$ref": "#/components/schemas/inline_response_200_4_task_delete"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "task_delete": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n450 - Please provide API key. In order to remove your content, please authenticate.\\n451 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\\n452 - Task does not exists. The task ID could not be linked to an existing task.\\n454 - Task is executed in some runs. Delete these first.\\n455 - Deleting the task failed. Please contact support team.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/task": {
+            "post": {
+                "tags": [
+                    "task"
+                ],
+                "summary": "Upload task",
+                "description": "Uploads a task. Upon success, it returns the task id.",
+                "operationId": "Api_task::task_upload",
+                "parameters": [
+                    {
+                        "name": "description",
+                        "in": "query",
+                        "description": "An XML file describing the task. Only name, description, and task format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.task.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/task).",
+                        "required": true,
+                        "schema": {
+                            "type": "file"
+                        }
+                    },
+                    {
+                        "name": "api_key",
+                        "in": "query",
+                        "description": "Api key to authenticate the user",
+                        "required": true,
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "Id of the uploaded task",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "properties": {
+                                        "upload_task": {
+                                            "$ref": "#/components/schemas/inline_response_200_5_upload_task"
+                                        }
+                                    },
+                                    "type": "object"
+                                },
+                                "example": {
+                                    "upload_task": {
+                                        "id": "4328"
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n530 - Description file not present. Please upload the task description.\\n531 - Internal error. Please contact api support team\\n532 - Problem validating uploaded description file. The XML description format does not meet the standards\\n533 - Task already exists.\\n534 - Error creating the task.\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/tasktype/list": {
+            "get": {
+                "tags": [
+                    "tasktype"
+                ],
+                "summary": "List all task types",
+                "description": "Returns an array with all task types in the system.",
+                "responses": {
+                    "default": {
+                        "description": "Unexpected error",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    },
+                    "200": {
+                        "description": "A task description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/TaskTypeList"
+                                },
+                                "example": {
+                                    "task_types": {
+                                        "task_type": [
+                                            {
+                                                "id": "1",
+                                                "name": "Supervised Classification",
+                                                "description": "In supervised classification, you are given ...",
+                                                "creator": "Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl"
+                                            },
+                                            {
+                                                "id": "2",
+                                                "name": "Supervised Regression",
+                                                "description": "Given a dataset with a numeric target ...",
+                                                "creator": "Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/tasktype/{id}": {
+            "get": {
+                "tags": [
+                    "tasktype"
+                ],
+                "summary": "Get task type description",
+                "description": "Returns information about a task type. The information includes a description, the given inputs and the expected outputs.",
+                "parameters": [
+                    {
+                        "name": "id",
+                        "in": "path",
+                        "description": "ID of the task.",
+                        "required": true,
+                        "schema": {
+                            "type": "integer"
+                        }
+                    }
+                ],
+                "responses": {
+                    "200": {
+                        "description": "A task type description",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/TaskType"
+                                },
+                                "example": {
+                                    "task_type": {
+                                        "id": "1",
+                                        "name": "Supervised Classification",
+                                        "description": "In supervised classification, you are given an input dataset in which instances are labeled with a certain class. The goal is to build a model that predicts the class for future unlabeled instances. The model is evaluated using a train-test procedure, e.g. cross-validation.<br><br>\\\r\n     *   \\\r\n     *   To make results by different users comparable, you are given the exact train-test folds to be used, and you need to return at least the predictions generated by your model for each of the test instances. OpenML will use these predictions to calculate a range of evaluation measures on the server.<br><br>\\\r\n     *   \\\r\n     *   You can also upload your own evaluation measures, provided that the code for doing so is available from the implementation used. For extremely large datasets, it may be infeasible to upload all predictions. In those cases, you need to compute and provide the evaluations yourself.<br><br>\\\r\n     *   \\\r\n     *   Optionally, you can upload the model trained on all the input data. There is no restriction on the file format, but please use a well-known format or PMML.",
+                                        "creator": [
+                                            "Joaquin Vanschoren",
+                                            "Jan van Rijn",
+                                            "Luis Torgo",
+                                            "Bernd Bischl"
+                                        ],
+                                        "contributor": [
+                                            "Bo Gao",
+                                            "Simon Fischer",
+                                            "Venkatesh Umaashankar",
+                                            "Michael Berthold",
+                                            "Bernd Wiswedel",
+                                            "Patrick Winter"
+                                        ],
+                                        "creation_date": "2013-01-24 00:00:00",
+                                        "input": [
+                                            {
+                                                "name": "source_data",
+                                                "requirement": "required",
+                                                "data_type": "numeric"
+                                            },
+                                            {
+                                                "name": "target_feature",
+                                                "requirement": "required",
+                                                "data_type": "string"
+                                            },
+                                            {
+                                                "name": "estimation_procedure",
+                                                "requirement": "required",
+                                                "data_type": "numeric"
+                                            },
+                                            {
+                                                "name": "cost_matrix",
+                                                "data_type": "json"
+                                            },
+                                            {
+                                                "name": "custom_testset",
+                                                "data_type": "json"
+                                            },
+                                            {
+                                                "name": "evaluation_measures",
+                                                "data_type": "string"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    },
+                    "412": {
+                        "description": "Precondition failed. An error code and message are returned.\\n240 - Please provide task type ID.\\n241 - Unknown task type. The task type with the given id was not found in the database\\n",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/Error"
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        },
+        "/user/list": {
+            "get": {
+                "tags": [
+                    "user"
+                ],
+                "summary": "List all users by user id",
+                "description": "Returns an array with all user ids and names.",
+                "operationId": "Api_user::username_list",
+                "responses": {
+                    "200": {
+                        "description": "A list of users",
+                        "content": {
+                            "application/json": {
+                                "schema": {
+                                    "$ref": "#/components/schemas/UserList"
+                                },
+                                "example": {
+                                    "users": {
+                                        "user": [
+                                            {
+                                                "id": "1",
+                                                "username": "janvanrijn@gmail.com"
+                                            },
+                                            {
+                                                "id": "2",
+                                                "username": "joaquin.vanschoren@gmail.com"
+                                            }
+                                        ]
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        }
+    },
+    "components": {
+        "schemas": {
+            "inline_response_200_16_flow_untag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the untagged setup",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationMeasureList_evaluation_measures_measures": {
+                "properties": {
+                    "measure": {
+                        "description": "The evaluation measure names",
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_10_flow_exists": {
+                "properties": {
+                    "id": {
+                        "description": "The id of the flow with the given name and (external) version",
+                        "type": "string"
+                    },
+                    "exists": {
+                        "description": "true or false",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Data_data_set_description": {
+                "properties": {
+                    "default_target_attribute": {
+                        "description": "For tabular data, the name of the column that is typically used as the target attribute for that data set",
+                        "type": "string"
+                    },
+                    "upload_date": {
+                        "description": "The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+                        "type": "string"
+                    },
+                    "version_label": {
+                        "description": "The version of the dataset, as defined by the uploader, for reference. Can be any format as long as it is unique.",
+                        "type": "string"
+                    },
+                    "description": {
+                        "description": "Wiki description of the dataset, in (Git flavoured) markdown format",
+                        "type": "string"
+                    },
+                    "format": {
+                        "description": "Data format, for instance ARFF",
+                        "type": "string"
+                    },
+                    "url": {
+                        "description": "The URL where the data can be downloaded",
+                        "type": "string"
+                    },
+                    "tag": {
+                        "description": "Tags added by OpenML users. Includes study tags in the form `study_1`",
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "visibility": {
+                        "description": "Who can see the dataset. For instance `public`.",
+                        "type": "string"
+                    },
+                    "md5_checksum": {
+                        "description": "Checksum to verify downloads of the dataset",
+                        "type": "string"
+                    },
+                    "version": {
+                        "description": "The version of the dataset, set by OpenML. A positive integer",
+                        "type": "string"
+                    },
+                    "status": {
+                        "description": "active, in_preparation, or deactivated",
+                        "type": "string"
+                    },
+                    "file_id": {
+                        "description": "The ID of the dataset file stored on the OpenML server",
+                        "type": "string"
+                    },
+                    "licence": {
+                        "description": "The licence granted for using the dataset, for instance Public or CC-BY",
+                        "type": "string"
+                    },
+                    "original_data_url": {
+                        "description": "The URL where the original data is hosted.",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "ID of the dataset, a positive integer",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the dataset",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataList_data_dataset": {
+                "properties": {
+                    "did": {
+                        "description": "The dataset ID",
+                        "type": "string"
+                    },
+                    "status": {
+                        "description": "The dataset status, either in_preparation, active, or deactivated",
+                        "type": "string"
+                    },
+                    "quality": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/DataList_data_quality"
+                        }
+                    },
+                    "name": {
+                        "description": "The dataset name",
+                        "type": "string"
+                    },
+                    "format": {
+                        "description": "The data format of the dataset, e.g. ARFF",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_estimation_procedure": {
+                "properties": {
+                    "parameter": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Task_task_description_estimation_procedure_parameter"
+                        }
+                    },
+                    "type": {
+                        "description": "The type of procedure, e.g. crossvalidation",
+                        "type": "string"
+                    },
+                    "data_splits_url": {
+                        "description": "The url where the data splits can be downloaded",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataList_data_quality": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the property",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the property",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "FlowList_flows_flow": {
+                "properties": {
+                    "full_name": {
+                        "description": "The full flow name (name + internal version number)",
+                        "type": "string"
+                    },
+                    "external_version": {
+                        "description": "The external flow version",
+                        "type": "string"
+                    },
+                    "version": {
+                        "description": "The internal flow version",
+                        "type": "string"
+                    },
+                    "uploader": {
+                        "description": "The ID of the person who uploaded the flow",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The flow ID",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The flow name",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_input_data_dataset": {
+                "properties": {
+                    "did": {
+                        "description": "The id of the dataset",
+                        "type": "string"
+                    },
+                    "url": {
+                        "description": "The download url of the dataset",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the dataset",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationList_evaluations": {
+                "properties": {
+                    "evaluation": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/EvaluationList_evaluations_evaluation"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_14_study_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted setup, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataUnprocessed": {
+                "properties": {
+                    "data_unprocessed": {
+                        "$ref": "#/components/schemas/DataUnprocessed_data_unprocessed"
+                    }
+                },
+                "type": "object"
+            },
+            "DataUnprocessed_data_unprocessed_dataset": {
+                "properties": {
+                    "did": {
+                        "description": "ID of the dataset a positive integer",
+                        "type": "string"
+                    },
+                    "status": {
+                        "description": "Status of the dataset",
+                        "type": "string"
+                    },
+                    "version": {
+                        "description": "Version of the dataset, a positive integer",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the dataset",
+                        "type": "string"
+                    },
+                    "format": {
+                        "description": "The dataset format, e.g. ARFF",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EstimationProcedureList_estimationprocedures": {
+                "properties": {
+                    "estimationprocedure": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/EstimationProcedureList_estimationprocedures_estimationprocedure"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "TaskTypeList": {
+                "properties": {
+                    "task_types": {
+                        "$ref": "#/components/schemas/TaskTypeList_task_types"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_output_data_evaluation": {
+                "properties": {
+                    "array_data": {
+                        "description": "For composite evaluation measures (e.g. per-class measures, confusion matrix), a string (JSON) representation of the   evaluation.",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the evaluation measure",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The result of the evaluation",
+                        "type": "string"
+                    },
+                    "flow_id": {
+                        "description": "The id of the code used to compute this evaluation method",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskList_task_quality": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the quality",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the quality",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description": {
+                "properties": {
+                    "input": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Task_task_description_input"
+                        }
+                    },
+                    "task_type": {
+                        "description": "The type of the task, e.g. Supervised Classification",
+                        "type": "string"
+                    },
+                    "tag": {
+                        "description": "Tags added by OpenML uers. Includes study tags in the form 'study_1'",
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "task_id": {
+                        "description": "ID of the task, a positive integer",
+                        "type": "string"
+                    },
+                    "output": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Task_task_description_output"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_9_upload_flow": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the uploaded flow, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Flow_flow_description": {
+                "properties": {
+                    "upload_date": {
+                        "description": "The datetime that the flow was uploaded, format yyyy-MM-dd HH:mm:ss",
+                        "type": "string"
+                    },
+                    "description": {
+                        "description": "Wiki description of the flow, in (Git flavoured) markdown format",
+                        "type": "string"
+                    },
+                    "language": {
+                        "description": "The programming language the flow is written in.",
+                        "type": "string"
+                    },
+                    "parameter": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Flow_flow_description_parameter"
+                        }
+                    },
+                    "tag": {
+                        "description": "Tags added by OpenML users. Includes study tags in the form `study_1`",
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "version": {
+                        "description": "The version of the flow, set by OpenML. A positive integer",
+                        "type": "string"
+                    },
+                    "version_label": {
+                        "description": "The version of the flow, as defined by the uploader, for reference. Can be any format as long as it is unique.",
+                        "type": "string"
+                    },
+                    "dependencies": {
+                        "description": "The libraries that this flow depends on, and their version numbers.",
+                        "type": "string"
+                    },
+                    "uploader": {
+                        "description": "The uploader of the flow",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "ID of the flow, a positive integer",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the flow",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task": {
+                "properties": {
+                    "task_description": {
+                        "$ref": "#/components/schemas/Task_task_description"
+                    }
+                },
+                "type": "object"
+            },
+            "Setup": {
+                "properties": {
+                    "setup_parameters": {
+                        "$ref": "#/components/schemas/Setup_setup_parameters"
+                    }
+                },
+                "type": "object"
+            },
+            "FlowList_flows": {
+                "properties": {
+                    "flow": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/FlowList_flows_flow"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "RunTrace": {
+                "properties": {
+                    "trace": {
+                        "$ref": "#/components/schemas/RunTrace_trace"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_predictions_feature": {
+                "properties": {
+                    "type": {
+                        "description": "The type of the prediction feature, e.g. integer",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the prediction feature, e.g. row_id",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_24_study_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted study, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_5_upload_task": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the uploaded task, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_13_flow_untag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the untagged flow",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "UserList_users": {
+                "properties": {
+                    "user": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/UserList_users_user"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "DataFeatures": {
+                "properties": {
+                    "data_features": {
+                        "$ref": "#/components/schemas/DataFeatures_data_features"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_predictions": {
+                "properties": {
+                    "feature": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Task_task_description_predictions_feature"
+                        }
+                    },
+                    "format": {
+                        "description": "The fromat of the predictions, e.g. ARFF",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_15_flow_tag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the tagged setup",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Setup_setup_parameters": {
+                "properties": {
+                    "parameter_setting": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Setup_setup_parameters_parameter_setting"
+                        }
+                    },
+                    "flow_id": {
+                        "description": "ID of the flow, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataList": {
+                "properties": {
+                    "data": {
+                        "$ref": "#/components/schemas/DataList_data"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_26_study_attach": {
+                "properties": {
+                    "linked_entities": {
+                        "description": "The number of linked entities",
+                        "type": "string"
+                    },
+                    "main_entity_type": {
+                        "description": "Main entity type of the of the study",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "ID of the study, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_2_data_tag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the tagged dataset",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_23_upload_flow": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the run with the trace, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationRequest_evaluation_request": {
+                "properties": {
+                    "run": {
+                        "$ref": "#/components/schemas/EvaluationRequest_evaluation_request_run"
+                    }
+                },
+                "type": "object"
+            },
+            "Run": {
+                "properties": {
+                    "run_description": {
+                        "$ref": "#/components/schemas/Run_run_description"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_6_task_tag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the tagged task",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "RunList": {
+                "properties": {
+                    "runs": {
+                        "$ref": "#/components/schemas/RunList_runs"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_17_data_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted run, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_input": {
+                "properties": {
+                    "data_set": {
+                        "$ref": "#/components/schemas/Task_task_description_data_set"
+                    },
+                    "cost_matrix": {
+                        "description": "The cost matrix, indicating the costs for each type of misclassification",
+                        "type": "array",
+                        "items": {
+                            "type": "array",
+                            "items": {
+                                "type": "integer",
+                                "format": "int64"
+                            }
+                        }
+                    },
+                    "name": {
+                        "description": "The name of the input, e.g. source_data",
+                        "type": "string"
+                    },
+                    "evaluation_measures": {
+                        "$ref": "#/components/schemas/Task_task_description_evaluation_measures"
+                    },
+                    "estimation_procedure": {
+                        "$ref": "#/components/schemas/Task_task_description_estimation_procedure"
+                    }
+                },
+                "type": "object"
+            },
+            "UserList": {
+                "properties": {
+                    "users": {
+                        "$ref": "#/components/schemas/UserList_users"
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study_runs": {
+                "properties": {
+                    "run_id": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "DataFeatures_data_features_feature": {
+                "properties": {
+                    "index": {
+                        "description": "Feature index",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "Feature name",
+                        "type": "string"
+                    },
+                    "data_type": {
+                        "description": "Feature data type",
+                        "type": "string"
+                    },
+                    "is_target": {
+                        "description": "Whether this feature is seen as a target feature",
+                        "type": "string"
+                    },
+                    "is_ignore": {
+                        "description": "Whether this feature should be ignored in modelling (e.g. every value is unique)",
+                        "type": "string"
+                    },
+                    "is_row_identifier": {
+                        "description": "Whether this feature is a row identifier",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_21_upload_flow": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the evaluated run, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskType": {
+                "properties": {
+                    "description": {
+                        "description": "A description of the task type",
+                        "type": "string"
+                    },
+                    "date": {
+                        "description": "The date when the task type was created",
+                        "type": "string"
+                    },
+                    "output": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskType_output"
+                        }
+                    },
+                    "contributor": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "input": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskType_input"
+                        }
+                    },
+                    "id": {
+                        "description": "ID of the task type, a positive integer",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the task type, e.g. Supervised Classification",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_output_data_file": {
+                "properties": {
+                    "did": {
+                        "description": "The id of the uploaded file",
+                        "type": "string"
+                    },
+                    "file_id": {
+                        "description": "The reference id of the uploaded file, for downloading afterward",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the uploaded file (e.g., description, predictions, model,...)",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskType_input": {
+                "properties": {
+                    "data_set": {
+                        "$ref": "#/components/schemas/Task_task_description_data_set"
+                    },
+                    "cost_matrix": {
+                        "type": "array",
+                        "items": {
+                            "type": "array",
+                            "items": {
+                                "type": "integer",
+                                "format": "int64"
+                            }
+                        }
+                    },
+                    "name": {
+                        "description": "The name of the input, e.g. source_data",
+                        "type": "string"
+                    },
+                    "evaluation_measures": {
+                        "$ref": "#/components/schemas/Task_task_description_evaluation_measures"
+                    },
+                    "estimation_procedure": {
+                        "$ref": "#/components/schemas/Task_task_description_estimation_procedure"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_output_data": {
+                "properties": {
+                    "evaluation": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Run_run_description_output_data_evaluation"
+                        }
+                    },
+                    "file": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Run_run_description_output_data_file"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_3_data_untag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the untagged dataset",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_output": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the output, e.g. predictions",
+                        "type": "string"
+                    },
+                    "predictions": {
+                        "$ref": "#/components/schemas/Task_task_description_predictions"
+                    }
+                },
+                "type": "object"
+            },
+            "DataQualityList": {
+                "properties": {
+                    "data_qualities_list": {
+                        "$ref": "#/components/schemas/DataQualityList_data_qualities_list"
+                    }
+                },
+                "type": "object"
+            },
+            "DataUnprocessed_data_unprocessed": {
+                "properties": {
+                    "dataset": {
+                        "$ref": "#/components/schemas/DataUnprocessed_data_unprocessed_dataset"
+                    }
+                },
+                "type": "object"
+            },
+            "Study": {
+                "properties": {
+                    "study": {
+                        "$ref": "#/components/schemas/Study_study"
+                    }
+                },
+                "type": "object"
+            },
+            "DataList_data": {
+                "properties": {
+                    "dataset": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/DataList_data_dataset"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "UserList_users_user": {
+                "properties": {
+                    "username": {
+                        "description": "The full user name",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The user ID",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskType_output": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the output, e.g. predictions",
+                        "type": "string"
+                    },
+                    "predictions": {
+                        "$ref": "#/components/schemas/TaskType_predictions"
+                    }
+                },
+                "type": "object"
+            },
+            "RunList_runs_run": {
+                "properties": {
+                    "task_id": {
+                        "description": "The ID of the task solved by this run",
+                        "type": "string"
+                    },
+                    "run_id": {
+                        "description": "The run ID",
+                        "type": "string"
+                    },
+                    "error_message": {
+                        "description": "Error message generated by the run (if any)",
+                        "type": "string"
+                    },
+                    "setup_id": {
+                        "description": "Ignore (internal representation of the parameter setting)",
+                        "type": "string"
+                    },
+                    "uploader": {
+                        "description": "The ID of the person uploading this run",
+                        "type": "string"
+                    },
+                    "flow_id": {
+                        "description": "The ID of the flow used in this run",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataFeatures_data_features": {
+                "properties": {
+                    "feature": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/DataFeatures_data_features_feature"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "DataQualities_data_qualities_quality": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the dataset quality measures",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value for this dataset",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_12_flow_tag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the tagged flow",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_8_flow_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted flow, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataQualities_data_qualities": {
+                "properties": {
+                    "quality": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/DataQualities_data_qualities_quality"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationRequest": {
+                "properties": {
+                    "evaluation_request": {
+                        "$ref": "#/components/schemas/EvaluationRequest_evaluation_request"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_18_upload_flow": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the uploaded run, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "SetupList": {
+                "properties": {
+                    "setups": {
+                        "$ref": "#/components/schemas/SetupList_setups"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskList": {
+                "properties": {
+                    "task": {
+                        "$ref": "#/components/schemas/TaskList_task"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_4_task_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted task, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_data_set": {
+                "properties": {
+                    "data_set_id": {
+                        "description": "The id of the dataset",
+                        "type": "string"
+                    },
+                    "target_feature": {
+                        "description": "The name of the target feature for this task",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationList": {
+                "properties": {
+                    "evaluations": {
+                        "$ref": "#/components/schemas/EvaluationList_evaluations"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description": {
+                "properties": {
+                    "setup_string": {
+                        "description": "Configuration of the flow as a string, to be interpreted by the flow, its library, or command line interface.",
+                        "type": "string"
+                    },
+                    "task_type": {
+                        "description": "The type of task solved by this run (e.g., classification)",
+                        "type": "string"
+                    },
+                    "task_id": {
+                        "description": "The id of the task solved by this run",
+                        "type": "string"
+                    },
+                    "task_evaluation_measure": {
+                        "description": "The evaluation measure that is supposed to be optimized in the task, if any",
+                        "type": "string"
+                    },
+                    "uploader_name": {
+                        "description": "The name of the uploader of the run",
+                        "type": "string"
+                    },
+                    "input_data": {
+                        "$ref": "#/components/schemas/Run_run_description_input_data"
+                    },
+                    "tag": {
+                        "description": "Tags added by OpenML users. Includes study tags in the form `study_1`",
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "output_data": {
+                        "$ref": "#/components/schemas/Run_run_description_output_data"
+                    },
+                    "uploader": {
+                        "description": "The uploader of the run",
+                        "type": "string"
+                    },
+                    "flow_id": {
+                        "description": "The id of the flow used in this run",
+                        "type": "string"
+                    },
+                    "flow_name": {
+                        "description": "The name of the flow used in this run",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "ID of the run, a positive integer",
+                        "type": "string"
+                    },
+                    "parameter_setting": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Run_run_description_parameter_setting"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "FlowList": {
+                "properties": {
+                    "flows": {
+                        "$ref": "#/components/schemas/FlowList_flows"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_19_run_tag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the tagged run",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study_data": {
+                "properties": {
+                    "data_id": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_data_delete": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the deleted dataset, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "RunTrace_trace": {
+                "properties": {
+                    "trace_iteration": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/RunTrace_trace_trace_iteration"
+                        }
+                    },
+                    "run_id": {
+                        "description": "run ID",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_7_task_untag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the untagged task",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Data": {
+                "properties": {
+                    "data_set_description": {
+                        "$ref": "#/components/schemas/Data_data_set_description"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_evaluation_measures": {
+                "properties": {
+                    "evaluation_measure": {
+                        "description": "The evaluation measure to optimize in this task",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "SetupList_setups_parameter": {
+                "properties": {
+                    "default_value": {
+                        "description": "The parameter's default value",
+                        "type": "string"
+                    },
+                    "data_type": {
+                        "description": "The parameter's data type",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The parameter value in this setup",
+                        "type": "string"
+                    },
+                    "parameter_name": {
+                        "description": "The parameter's short name",
+                        "type": "string"
+                    },
+                    "full_name": {
+                        "description": "The parameter's full name",
+                        "type": "string"
+                    },
+                    "flow_id": {
+                        "description": "The (sub)flow ID",
+                        "type": "string"
+                    },
+                    "flow_name": {
+                        "description": "The (sub)flow name",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The parameter ID",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskList_task_input": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the input",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the input",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "StudyList_study_list": {
+                "properties": {
+                    "study": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/StudyList_study_list_study"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study_tag": {
+                "properties": {
+                    "write_access": {
+                        "description": "The write access level of the study (e.g. public)",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the study (e.g. study_1)",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Flow": {
+                "properties": {
+                    "flow_description": {
+                        "$ref": "#/components/schemas/Flow_flow_description"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_parameter_setting": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the parameter",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the parameter used",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationRequest_evaluation_request_run": {
+                "properties": {
+                    "setup_id": {
+                        "description": "ID of the setup, a positive integer",
+                        "type": "string"
+                    },
+                    "upload_time": {
+                        "description": "The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+                        "type": "string"
+                    },
+                    "uploader": {
+                        "description": "ID of the uploader, a positive integer",
+                        "type": "string"
+                    },
+                    "task_id": {
+                        "description": "ID of the task, a positive integer",
+                        "type": "string"
+                    },
+                    "run_id": {
+                        "description": "ID of the run, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Error": {
+                "properties": {
+                    "message": {
+                        "type": "string"
+                    },
+                    "code": {
+                        "type": "integer"
+                    },
+                    "additional_message": {
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EstimationProcedureList_estimationprocedures_estimationprocedure": {
+                "properties": {
+                    "name": {
+                        "description": "The estimation procedure name, e.g. '10 fold Crossvalidation'",
+                        "type": "string"
+                    },
+                    "folds": {
+                        "description": "The number of cross-validation folds, e.g. '10'",
+                        "type": "string"
+                    },
+                    "stratified_sampling": {
+                        "description": "Whether or not the sampling is stratified, 'true' or 'false'",
+                        "type": "string"
+                    },
+                    "ttid": {
+                        "description": "The task type ID",
+                        "type": "string"
+                    },
+                    "repeats": {
+                        "description": "The number of repeats, e.g. '10'",
+                        "type": "string"
+                    },
+                    "type": {
+                        "description": "The estimation procedure type, e.g. 'crossvalidation'",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The estimation procedure ID",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskTypeList_task_types": {
+                "properties": {
+                    "task_type": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskTypeList_task_types_task_type"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "TaskList_task_task": {
+                "properties": {
+                    "status": {
+                        "description": "The status of the source dataset, active, in_preparation, or deactivated",
+                        "type": "string"
+                    },
+                    "task_type": {
+                        "description": "The type of task (e.g. Supervised Classificationr)",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the source dataset",
+                        "type": "string"
+                    },
+                    "task_id": {
+                        "description": "The ID of the task",
+                        "type": "string"
+                    },
+                    "format": {
+                        "description": "The format of the source dataset",
+                        "type": "string"
+                    },
+                    "did": {
+                        "description": "The id of the source dataset",
+                        "type": "string"
+                    },
+                    "tag": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    },
+                    "input": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskList_task_input"
+                        }
+                    },
+                    "quality": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskList_task_quality"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_20_run_untag": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the untagged run",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "StudyList": {
+                "properties": {
+                    "study_list": {
+                        "$ref": "#/components/schemas/StudyList_study_list"
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study": {
+                "properties": {
+                    "runs": {
+                        "$ref": "#/components/schemas/Study_study_runs"
+                    },
+                    "tasks": {
+                        "$ref": "#/components/schemas/Study_study_tasks"
+                    },
+                    "name": {
+                        "description": "The name of the study",
+                        "type": "string"
+                    },
+                    "creator": {
+                        "description": "A comma-separated list of the study creators",
+                        "type": "string"
+                    },
+                    "flows": {
+                        "$ref": "#/components/schemas/Study_study_flows"
+                    },
+                    "creation_date": {
+                        "description": "The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+                        "type": "string"
+                    },
+                    "alias": {
+                        "description": "The alias of the study",
+                        "type": "string"
+                    },
+                    "tag": {
+                        "$ref": "#/components/schemas/Study_study_tag"
+                    },
+                    "main_entity_type": {
+                        "description": "The type of entity collected in the study (e.g. task or run)",
+                        "type": "string"
+                    },
+                    "data": {
+                        "$ref": "#/components/schemas/Study_study_data"
+                    },
+                    "id": {
+                        "description": "The ID of the study",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_25_upload_study": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the uploaded study, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationList_evaluations_evaluation": {
+                "properties": {
+                    "function": {
+                        "description": "The name of the evaluation function",
+                        "type": "string"
+                    },
+                    "task_id": {
+                        "description": "The ID of the tasks solved by this run",
+                        "type": "string"
+                    },
+                    "run_id": {
+                        "description": "The run ID",
+                        "type": "string"
+                    },
+                    "array_data": {
+                        "description": "For structured evaluation measures, an array of evaluation values (e.g. per-class predictions, evaluation matrices,...)",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The outcome of the evaluation",
+                        "type": "string"
+                    },
+                    "flow_id": {
+                        "description": "The ID of the flow used by this run",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EstimationProcedureList": {
+                "properties": {
+                    "estimationprocedures": {
+                        "$ref": "#/components/schemas/EstimationProcedureList_estimationprocedures"
+                    }
+                },
+                "type": "object"
+            },
+            "DataQualityList_data_qualities_list": {
+                "properties": {
+                    "quality": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study_flows": {
+                "properties": {
+                    "flow_id": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "StudyList_study_list_study": {
+                "properties": {
+                    "alias": {
+                        "description": "The alias of the study",
+                        "type": "string"
+                    },
+                    "creation_date": {
+                        "description": "The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+                        "type": "string"
+                    },
+                    "creator": {
+                        "description": "A comma-separated list of the study creators",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The ID of the study",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the study",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Run_run_description_input_data": {
+                "properties": {
+                    "dataset": {
+                        "$ref": "#/components/schemas/Run_run_description_input_data_dataset"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskTypeList_task_types_task_type": {
+                "properties": {
+                    "description": {
+                        "description": "A description of the task type",
+                        "type": "string"
+                    },
+                    "creator": {
+                        "description": "A comma-separated list of the task type creators",
+                        "type": "string"
+                    },
+                    "id": {
+                        "description": "The ID of the task type",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the task type",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationMeasureList_evaluation_measures": {
+                "properties": {
+                    "measures": {
+                        "$ref": "#/components/schemas/EvaluationMeasureList_evaluation_measures_measures"
+                    }
+                },
+                "type": "object"
+            },
+            "TaskType_predictions": {
+                "properties": {
+                    "feature": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/Task_task_description_predictions_feature"
+                        }
+                    },
+                    "format": {
+                        "description": "The format of the predictions, e.g. ARFF",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Flow_flow_description_parameter": {
+                "properties": {
+                    "default_value": {
+                        "description": "The default value of the parameter",
+                        "type": "string"
+                    },
+                    "name": {
+                        "description": "The name of the parameter",
+                        "type": "string"
+                    },
+                    "data_type": {
+                        "description": "The data type of the parameter",
+                        "type": "string"
+                    },
+                    "description": {
+                        "description": "A description of the parameter",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "RunTrace_trace_trace_iteration": {
+                "properties": {
+                    "setup_string": {
+                        "description": "A JSON representation of the setup (configuration)",
+                        "type": "string"
+                    },
+                    "repeat": {
+                        "description": "The number of the repeat in the outer cross-valudation",
+                        "type": "string"
+                    },
+                    "selected": {
+                        "description": "Whether this setup was selected as the best one (true or false)",
+                        "type": "string"
+                    },
+                    "iteration": {
+                        "description": "A number of the optimization iteration",
+                        "type": "string"
+                    },
+                    "fold": {
+                        "description": "The number of the fold in the inner cross-validation",
+                        "type": "string"
+                    },
+                    "evaluation": {
+                        "description": "The evaluation score of the setup",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "EvaluationMeasureList": {
+                "properties": {
+                    "evaluation_measures": {
+                        "$ref": "#/components/schemas/EvaluationMeasureList_evaluation_measures"
+                    }
+                },
+                "type": "object"
+            },
+            "SetupList_setups": {
+                "properties": {
+                    "setup": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/SetupList_setups_setup"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "TaskList_task": {
+                "properties": {
+                    "task": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/TaskList_task_task"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "SetupList_setups_setup": {
+                "properties": {
+                    "setup_id": {
+                        "description": "The setup ID",
+                        "type": "string"
+                    },
+                    "parameter": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/SetupList_setups_parameter"
+                        }
+                    },
+                    "flow_id": {
+                        "description": "The ID of the flow used by this run",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Task_task_description_estimation_procedure_parameter": {
+                "properties": {
+                    "name": {
+                        "description": "The name of the parameter",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the parameter",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "RunList_runs": {
+                "properties": {
+                    "run": {
+                        "type": "array",
+                        "items": {
+                            "$ref": "#/components/schemas/RunList_runs_run"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "Study_study_tasks": {
+                "properties": {
+                    "task_id": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        }
+                    }
+                },
+                "type": "object"
+            },
+            "inline_response_200_1_upload_data_set": {
+                "properties": {
+                    "id": {
+                        "description": "ID of the uploaded dataset, a positive integer",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "Setup_setup_parameters_parameter_setting": {
+                "properties": {
+                    "default_value": {
+                        "description": "The default value of the parameter used",
+                        "type": "string"
+                    },
+                    "value": {
+                        "description": "The value of the parameter used",
+                        "type": "string"
+                    },
+                    "data_type": {
+                        "description": "The data type of the hyperparameter value",
+                        "type": "string"
+                    },
+                    "full_name": {
+                        "description": "The full name of the hyperparameter",
+                        "type": "string"
+                    },
+                    "parameter_name": {
+                        "description": "The short name of the hyperparameter",
+                        "type": "string"
+                    }
+                },
+                "type": "object"
+            },
+            "DataQualities": {
+                "properties": {
+                    "data_qualities": {
+                        "$ref": "#/components/schemas/DataQualities_data_qualities"
+                    }
+                },
+                "type": "object"
+            }
+        }
+    }
+}
\ No newline at end of file
diff --git a/openapi/swagger.yaml b/openapi/swagger.yaml
new file mode 100644
index 000000000..3c86cfc16
--- /dev/null
+++ b/openapi/swagger.yaml
@@ -0,0 +1,3338 @@
+openapi: 3.0.0
+info:
+  title: 'OpenML API'
+  description: 'REST API for sharing, organizing and reusing machine learning datasets, code, and experiments. Follows a predictive URL scheme from endpoint https://www.openml.org/api/v1/json (or /xml). You need to add your api_key for every call except GET calls. The API key can be found in your profile on openml.org.'
+  version: 1.0.0
+servers:
+  -
+    url: 'https://openml.org/api/v1/json'
+    description: 'The current OpenML API'
+  -
+    url: 'https://test.openml.org/api/v1/json'
+    description: 'The test OpenML API'
+paths:
+  /data/tag:
+    post:
+      tags:
+        - data
+      summary: 'Tag a dataset'
+      description: 'Tags a dataset.'
+      operationId: 'Api_data::data_tag'
+      parameters:
+        -
+          name: data_id
+          in: query
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the tagged dataset'
+          content:
+            application/json:
+              schema:
+                properties:
+                  data_tag: { $ref: '#/components/schemas/inline_response_200_2_data_tag' }
+                type: object
+              example:
+                data_tag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data/untag:
+    post:
+      tags:
+        - data
+      summary: 'Untag a dataset'
+      description: 'Untags a dataset.'
+      operationId: 'Api_data::data_untag'
+      parameters:
+        -
+          name: data_id
+          in: query
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The ID of the untagged dataset'
+          content:
+            application/json:
+              schema:
+                properties:
+                  data_untag: { $ref: '#/components/schemas/inline_response_200_3_data_untag' }
+                type: object
+              example:
+                data_untag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/list/{filters}':
+    get:
+      tags:
+        - data
+      summary: 'List and filter datasets'
+      description: 'List datasets, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/data/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all datasets that match the constraints.'
+      operationId: 'Api_data::data_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  /status/{status} - returns only datasets with a given status, either 'active', 'deactivated', or 'in_preparation'.\r\n  /tag/{tag} - returns only datasets tagged with the given tag.\r\n  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, data_version, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of datasets with the given task'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataList'
+              example:
+                data:
+                  dataset: [{ did: '1', name: anneal, status: active, format: ARFF, quality: [{ name: MajorityClassSize, value: '684' }, { name: MaxNominalAttDistinctValues, value: '10.0' }, { name: MinorityClassSize, value: '0' }, { name: NumBinaryAtts, value: '14.0' }, { name: NumberOfClasses, value: '6' }, { name: NumberOfFeatures, value: '39' }, { name: NumberOfInstances, value: '898' }, { name: NumberOfInstancesWithMissingValues, value: '0' }, { name: NumberOfMissingValues, value: '0' }, { name: NumberOfNumericFeatures, value: '6' }, { name: NumberOfSymbolicFeatures, value: '32' }] }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n370 - Illegal filter specified.\n371 - Filter values/ranges not properly specified.\n372 - No results. There where no matches for the given constraints.\n373 - Can not specify an offset without a limit.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/{id}':
+    get:
+      tags:
+        - data
+      summary: 'Get dataset description'
+      description: 'Returns information about a dataset. The information includes the name, information about the creator, URL to download it and more.'
+      operationId: 'Api_data::data'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A dataset description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Data'
+              example:
+                data_set_description:
+                  id: '1'
+                  name: anneal
+                  version: '2'
+                  description: ...
+                  format: ARFF
+                  upload_date: '2014-04-06 23:19:20'
+                  licence: Public
+                  url: 'https://www.openml.org/data/download/1/dataset_1_anneal.arff'
+                  file_id: '1'
+                  default_target_attribute: class
+                  version_label: '2'
+                  tag: [study_1, uci]
+                  visibility: public
+                  original_data_url: 'https://www.openml.org/d/2'
+                  status: active
+                  md5_checksum: d01f6ccd68c88b749b20bbe897de3713
+        '412':
+          description: 'Precondition failed. An error code and message are returned\n110 - Please provide data_id.\n111 - Unknown dataset. Data set description with data_id was not found in the database.\n112 - No access granted. This dataset is not shared with you.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - data
+      summary: 'Delete dataset'
+      description: 'Deletes a dataset. Upon success, it returns the ID of the deleted dataset.'
+      operationId: 'Api_data::data_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted dataset'
+          content:
+            application/json:
+              schema:
+                properties:
+                  data_delete: { $ref: '#/components/schemas/inline_response_200_data_delete' }
+                type: object
+              example:
+                data_delete:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned\n- 350 - Please provide API key. In order to remove your content, please authenticate.\n- 351 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n- 352 - Dataset does not exists. The data ID could not be linked to an existing dataset.\n- 353 - Dataset is not owned by you. The dataset is owned by another user. Hence you cannot delete it.\n- 354 - Dataset is in use by other content. Can not be deleted. The data is used in tasks or runs. Delete other content before deleting this dataset.\n- 355 - Deleting dataset failed. Deleting the dataset failed. Please contact support team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data:
+    post:
+      tags:
+        - data
+      summary: 'Upload dataset'
+      description: 'Uploads a dataset. Upon success, it returns the data id.'
+      operationId: 'Api_data::data_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/data).'
+          required: true
+          schema:
+            type: file
+        -
+          name: dataset
+          in: query
+          description: 'The actual dataset, being an ARFF file.'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the uploaded dataset'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_data_set: { $ref: '#/components/schemas/inline_response_200_1_upload_data_set' }
+                type: object
+              example:
+                upload_data_set:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n130 - Problem with file uploading. There was a problem with the file upload.\n131 - Problem validating uploaded description file. The XML description format does not meet the standards.\n132 - Failed to move the files. Internal server error, please contact API administrators.\n133 - Failed to make checksum of datafile. Internal server error, please contact API administrators.\n134 - Failed to insert record in database. Internal server error, please contact API administrators.\n135 - Please provide description xml.\n136 - File failed format verification. The uploaded file is not valid according to the selected file format. Please check the file format specification and try again.\n137 - Please provide API key. In order to share content, please log in or provide your API key.\n138 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\n139 - Combination name / version already exists. Leave version out for auto increment\n140 - Both dataset file and dataset url provided. The system is confused since both a dataset file (post) and a dataset url (xml) are provided. Please remove one.\n141 - Neither dataset file or dataset url are provided. Please provide either a dataset file as POST variable, or a dataset url in the description XML.\n142 - Error in processing arff file. Can be a syntax error, or the specified target feature does not exists. For now, we only check on arff files. If a dataset is claimed to be in such a format, and it can not be parsed, this error is returned.\n143 - Suggested target feature not legal. It is possible to suggest a default target feature (for predictive tasks). However, it should be provided in the data.\n144 - Unable to update dataset. The dataset with id could not be found in the database. If you upload a new dataset, unset the id.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data/status/update/:
+    post:
+      tags:
+        - data
+      summary: 'Change the status of a dataset'
+      description: 'Change the status of a dataset, either ''active'' or ''deactivated'''
+      operationId: 'Api_data::status_update'
+      parameters:
+        -
+          name: data_id
+          in: query
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: status
+          in: query
+          description: 'The status on which to filter the results, either ''active'' or ''deactivated''.'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n691 - Illegal status\n692 - Dataset does not exists\n693 - Dataset is not owned by you\n694 - Illegal status transition\n695 - Status update failed\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/features/{id}':
+    get:
+      tags:
+        - data
+      summary: 'Get data features'
+      description: 'Returns the features of a dataset.'
+      operationId: 'Api_data::data_features'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'All the features of the dataset'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataFeatures'
+              example:
+                data_features:
+                  feature: [{ index: '0', name: sepallength, data_type: numeric, is_target: 'false', is_ignore: 'false', is_row_identifier: 'false' }, { index: '1', name: sepalwidth, data_type: numeric, is_target: 'false', is_ignore: 'false', is_row_identifier: 'false' }, { index: '2', name: petallength, data_type: numeric, is_target: 'false', is_ignore: 'false', is_row_identifier: 'false' }, { index: '3', name: petalwidth, data_type: numeric, is_target: 'false', is_ignore: 'false', is_row_identifier: 'false' }, { index: '4', name: class, data_type: nominal, is_target: 'true', is_ignore: 'false', is_row_identifier: 'false' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n270 - Please provide dataset ID.\n271 - Unknown dataset. Data set with the given data ID was not found (or is not shared with you).\n272 - No features found. The dataset did not contain any features, or we could not extract them.\n273 - Dataset not processed yet. The dataset was not processed yet, features are not yet available. Please wait for a few minutes.\n274 - Dataset processed with error. The feature extractor has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, please contact the API admins.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data/features:
+    post:
+      tags:
+        - data
+      summary: 'Upload dataset feature description'
+      description: 'Uploads dataset feature description. Upon success, it returns the data id.'
+      operationId: 'Api_data::data_features_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.features) and an [XML example](https://www.openml.org/api/v1/xml_example/data.features).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n431 - Dataset already processed\n432 - Please provide description xml\n433 - Problem validating uploaded description file\n434 - Could not find dataset\n436 - Something wrong with XML, check data id and evaluation engine id\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data/qualities/list:
+    get:
+      tags:
+        - data
+      summary: 'List all data qualities'
+      description: 'Returns a list of all data qualities in the system.'
+      operationId: 'Api_data::data_qualities_list'
+      responses:
+        '200':
+          description: 'A list of data qualities'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataQualityList'
+              example:
+                data_qualities_list:
+                  quality: [NumberOfClasses, NumberOfFeatures, NumberOfInstances, NumberOfInstancesWithMissingValues, NumberOfMissingValues, NumberOfNumericFeatures, NumberOfSymbolicFeatures]
+        '412':
+          description: 'Precondition failed. An error code and message are returned\n370 - No data qualities available. There are no data qualities in the system.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/qualities/{id}':
+    get:
+      tags:
+        - data
+      summary: 'Get data qualities'
+      description: 'Returns the qualities of a dataset.'
+      operationId: 'Api_data::data_qualities'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the dataset.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'All the qualities of the dataset'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataQualities'
+              example:
+                data_qualities:
+                  quality: [{ name: ClassCount, value: '3.0' }, { name: ClassEntropy, value: '1.584962500721156' }, { name: NumberOfClasses, value: '3' }, { name: NumberOfFeatures, value: '5' }, { name: NumberOfInstances, value: '150' }, { name: NumberOfInstancesWithMissingValues, value: '0' }, { name: NumberOfMissingValues, value: '0' }, { name: NumberOfNumericFeatures, value: '4' }, { name: NumberOfSymbolicFeatures, value: '0' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n360 - Please provide data set ID\n361 - Unknown dataset. The data set with the given ID was not found in the database, or is not shared with you.\n362 - No qualities found. The registered dataset did not contain any calculated qualities.\n363 - Dataset not processed yet. The dataset was not processed yet, no qualities are available. Please wait for a few minutes.\n364 - Dataset processed with error. The quality calculator has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, contact the support team.\n365 - Interval start or end illegal. There was a problem with the interval\nstart or end.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /data/qualities:
+    post:
+      tags:
+        - data
+      summary: 'Upload dataset qualities'
+      description: 'Uploads dataset qualities. Upon success, it returns the data id.'
+      operationId: 'Api_data::data_qualities_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.qualities) and an [XML example](https://www.openml.org/api/v1/xml_example/data.qualities).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n381 - Something wrong with XML, please check did and evaluation_engine_id\n382 - Please provide description xml\n383 - Problem validating uploaded description file\n384 - Dataset not processed yet\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/unprocessed/{data_engine_id}/{order}':
+    get:
+      tags:
+        - data
+      summary: 'Get a list of unprocessed datasets'
+      description: 'This call is for people running their own dataset processing engines. It returns the details of datasets that are not yet processed by the given processing engine. It doesn''t process the datasets, it just returns the dataset info.'
+      operationId: 'Api_data::data_unprocessed'
+      parameters:
+        -
+          name: data_engine_id
+          in: path
+          description: 'The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: order
+          in: path
+          description: 'When there are multiple datasets still to process, this defines which ones to return. Options are ''normal'' - the oldest datasets, or ''random''.'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of unprocessed datasets'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataUnprocessed'
+              example:
+                data_unprocessed:
+                  run: [{ did: '1', status: deactivated, version: '2', name: anneal, format: ARFF }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n681 - No unprocessed datasets.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/data/qualities/unprocessed/{data_engine_id}/{order}':
+    post:
+      tags:
+        - data
+      summary: 'Get a list of datasets with unprocessed qualities'
+      description: 'This call is for people running their own dataset processing engines. It returns the details of datasets for which certain qualities are not yet processed by the given processing engine. It doesn''t process the datasets, it just returns the dataset info.'
+      operationId: 'Api_data::dataqualities_unprocessed'
+      parameters:
+        -
+          name: data_engine_id
+          in: path
+          description: 'The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.'
+          required: true
+          schema:
+            type: string
+        -
+          name: order
+          in: path
+          description: 'When there are multiple datasets still to process, this defines which ones to return. Options are ''normal'' - the oldest datasets, or ''random''.'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'API key to authenticate the user'
+          required: false
+          schema:
+            type: string
+        -
+          name: qualities
+          in: query
+          description: 'Comma-separated list of (at least two) quality names, e.g. ''NumberOfInstances,NumberOfFeatures''.'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of unprocessed datasets'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DataUnprocessed'
+              example:
+                data_unprocessed:
+                  run: [{ did: '1', status: deactivated, version: '2', name: anneal, format: ARFF }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n686 - Please specify the features the evaluation engine wants to calculate (at least 2).\n687 - No unprocessed datasets according to the given set of meta-features.\n688 - Illegal qualities.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /estimationprocedure/list:
+    get:
+      tags:
+        - estimationprocedure
+      summary: 'List all estimation procedures'
+      description: 'Returns an array with all model performance estimation procedures in the system.'
+      operationId: 'Api_estimationprocedure::estimationprocedure_list'
+      responses:
+        '200':
+          description: 'A list of estimation procedures'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/EstimationProcedureList'
+              example:
+                estimationprocedures:
+                  estimationprocedure: [{ id: '1', ttid: '1', name: '10-fold Crossvalidation', type: crossvalidation, repeats: '1', folds: '10', stratified_sampling: 'true' }, { id: '2', ttid: '1', name: '5 times 2-fold Crossvalidation', type: crossvalidation, repeats: '5', folds: '2', stratified_sampling: 'true' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n500 - No model performance estimation procedures available.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/evaluation/request/{evaluation_engine_id}/{order}':
+    get:
+      tags:
+        - evaluation
+      summary: 'Get an unevaluated run'
+      description: 'This call is for people running their own evaluation engines. It returns the details of a run that is not yet evaluated by the given evaluation engine. It doesn''t evaluate the run, it just returns the run info.'
+      operationId: 'Api_evaluation::evaluation_request'
+      parameters:
+        -
+          name: evaluation_engine_id
+          in: path
+          description: 'The ID of the evaluation engine. You get this ID when you register a new evaluation engine with OpenML. The ID of the main evaluation engine is 1.'
+          required: true
+          schema:
+            type: string
+        -
+          name: order
+          in: path
+          description: 'When there are multiple runs still to evaluate, this defines which one to return. Options are ''normal'' - the oldest run, ''reverse'' - the newest run, or ''random'' - a random run.'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of evaluations descriptions'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/EvaluationRequest'
+              example:
+                evaluation_request:
+                  run: [{ setup_id: '68799271', upload_time: '2018-04-03 21:05:38', uploader: '1935', task_id: '3021', run_id: '8943712' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n100 - Function not valid.\n545 - No unevaluated runs according to the criteria.\n546 - Illegal filter.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/evaluation/list/{filters}':
+    get:
+      tags:
+        - evaluation
+      summary: 'List and filter evaluations'
+      description: 'List evaluations, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/evaluation/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all evaluations that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, uploaders or runs.'
+      operationId: 'Api_evaluation::evaluation_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /function/{name} - name of the evaluation measure, e.g. area_under_auc or predictive_accuracy. See the OpenML website for the complete list of measures.\r\n  /tag/{tag} - returns only evaluations of runs tagged with the given tag.\r\n  /run/{ids} - return only evaluations for specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''\r\n  /task/{ids} - return only evaluations for specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''\r\n  /flow/{ids} - return only evaluations for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only evaluations for specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  /per_fold/{true,false} - whether or not to return crossvalidation scores per fold. Defaults to 'false'. Setting it to 'true' leads to large numbers of results, use only for very specific sets of runs.\r\n  /sort_order/{asc,desc} - sorts the results by the evaluation value, according to the selected evaluation measure (function)\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of evaluations descriptions'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/EvaluationList'
+              example:
+                evaluations:
+                  evaluation: [{ function: area_under_roc_curve, upload_time: '2014-04-06 23:30:40', task_id: '68', run_id: '1', array_data: '[0,0.99113,0.898048,0.874862,0.791282,0.807343,0.820674]', value: '0.839359', uploader: '1', flow_id: '61' }, { function: f_measure, upload_time: '2014-04-06 23:30:40', task_id: '68', run_id: '1', array_data: '[0,0,0.711934,0.735714,0.601363,0.435678,0.430913]', value: '0.600026', uploader: '1', flow_id: '61' }, { function: predictive_accuracy, upload_time: '2014-04-06 23:30:40', task_id: '68', run_id: '1', array_data: [], value: '0.614634', uploader: '1', flow_id: '61' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n540 - Please provide at least task, flow or setup, uploader or run, to\nfilter results, or limit the number of responses.\n541 - The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\n542 - There where no results. Check whether there are runs under the given constraint.\n543 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\n544 - Illegal filter specified.\n545 - Offset specified without limit.\n546 - Requested result limit too high.\n547 - Per fold can only be set to value ''true'' or ''false''.\n548 - Per fold queries are experimental and require a fair amount of filters on resulting run records to keep the query fast (use, e.g., flow, setup, task and uploader filter)\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /evaluationmeasure/list:
+    get:
+      tags:
+        - evaluationmeasure
+      summary: 'List all evaluation measures'
+      description: 'Returns an array with all model evaluation measures in the system.'
+      operationId: 'Api_evaluationmeasure::evaluationmeasure_list'
+      responses:
+        '200':
+          description: 'A list of evaluation measures'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/EvaluationMeasureList'
+              example:
+                evaluation_measures:
+                  measures: { measure: [area_under_roc_curve, average_cost, binominal_test, build_cpu_time] }
+  '/flow/list/{filters}':
+    get:
+      tags:
+        - flow
+      summary: 'List and filter flows'
+      description: 'List flows, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/task/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all flows that match the constraints.'
+      operationId: 'Api_flow::flow_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.\r\n  /tag/{tag} - returns only tasks tagged with the given tag.\r\n  /uploader/{id} - return only evaluations uploaded by a specific user, specified by user ID.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of flows'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/FlowList'
+              example:
+                flows:
+                  flow: [{ id: '65', full_name: weka.RandomForest(1), name: weka.RandomForest, version: '1', external_version: Weka_3.7.10_9186, uploader: '1' }, { id: '66', full_name: weka.IBk(1), name: weka.IBk, version: '1', external_version: Weka_3.7.10_8034, uploader: '1' }, { id: '67', full_name: weka.BayesNet_K2(1), name: weka.BayesNet_K2, version: '1', external_version: Weka_3.7.10_8034, uploader: '1' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n500 - No results. There where no matches for the given constraints.\n501 - Illegal filter specified.\n502 - Filter values/ranges not properly specified.\n503 - Can not specify an offset without a limit.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /flow/tag:
+    post:
+      tags:
+        - flow
+      summary: 'Tag a flow'
+      description: 'Tags a flow.'
+      operationId: 'Api_flow::flow_tag'
+      parameters:
+        -
+          name: flow_id
+          in: query
+          description: 'Id of the flow.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the tagged flow'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_tag: { $ref: '#/components/schemas/inline_response_200_12_flow_tag' }
+                type: object
+              example:
+                flow_tag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /flow/untag:
+    post:
+      tags:
+        - flow
+      summary: 'Untag a flow'
+      description: 'Untags a flow.'
+      operationId: 'Api_flow::flow_untag'
+      parameters:
+        -
+          name: flow_id
+          in: query
+          description: 'Id of the flow.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the untagged flow'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_untag: { $ref: '#/components/schemas/inline_response_200_13_flow_untag' }
+                type: object
+              example:
+                flow_untag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\nby another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/flow/exists/{name}/{version}':
+    get:
+      tags:
+        - flow
+      summary: 'Check whether flow exists'
+      description: 'Checks whether a flow with the given name and (external) version exists.'
+      operationId: 'Api_flow::flow_exists'
+      parameters:
+        -
+          name: name
+          in: path
+          description: 'The name of the flow.'
+          required: true
+          schema:
+            type: string
+        -
+          name: version
+          in: path
+          description: 'The external version of the flow'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of flows'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_exists: { $ref: '#/components/schemas/inline_response_200_10_flow_exists' }
+                type: object
+              example:
+                flow_exists:
+                  exists: 'true'
+                  id: '65'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n330 - Mandatory fields not present. Please provide name and external_version.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/flow/{id}':
+    get:
+      tags:
+        - flow
+      summary: 'Get flow description'
+      description: 'Returns information about a flow. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.'
+      operationId: 'Api_flow::flow'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the flow.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A flow description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Flow'
+              example:
+                flow:
+                  id: '100'
+                  uploader: '1'
+                  name: weka.J48
+                  version: '2'
+                  external_version: Weka_3.7.5_9117
+                  description: ...
+                  upload_date: '2014-04-23 18:00:36'
+                  language: Java
+                  dependencies: Weka_3.7.5
+                  parameter: [{ name: A, data_type: flag, default_value: [], description: 'Laplace smoothing...' }, { name: C, data_type: option, default_value: '0.25', description: 'Set confidence threshold...' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n180 - Please provide flow id.\n181 - Unknown flow. The flow with the given ID was not found in the database.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - flow
+      summary: 'Delete a flow'
+      description: 'Deletes a flow. Upon success, it returns the ID of the deleted flow.'
+      operationId: 'Api_flow::flow_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the flow.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'API key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted flow'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_delete: { $ref: '#/components/schemas/inline_response_200_8_flow_delete' }
+                type: object
+              example:
+                flow_delete:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n320 - Please provide API key. In order to remove your content, please authenticate.\n321 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n322 - Flow does not exists. The flow ID could not be linked to an existing flow.\n323 - Flow is not owned by you. The flow is owned by another user. Hence you cannot delete it.\n324 - Flow is in use by other content. Can not be deleted. The flow is used in runs, evaluations or as a component of another flow. Delete other content before deleting this flow.\n325 - Deleting flow failed. Deleting the flow failed. Please contact\nsupport team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /flow:
+    post:
+      tags:
+        - flow
+      summary: 'Upload a flow'
+      description: 'Uploads a flow. Upon success, it returns the flow id.'
+      operationId: 'Api_flow::flow_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the flow. Only name and description are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.implementation.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/flow).'
+          required: true
+          schema:
+            type: file
+        -
+          name: flow
+          in: query
+          description: 'The actual flow, being a source (or binary) file.'
+          required: false
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the uploaded flow'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_flow: { $ref: '#/components/schemas/inline_response_200_9_upload_flow' }
+                type: object
+              example:
+                upload_flow:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n160 - Error in file uploading. There was a problem with the file upload.\n161 - Please provide description xml.\n163 - Problem validating uploaded description file. The XML description format does not meet the standards.\n164 - Flow already stored in database. Please change name or version number\n165 - Failed to insert flow. There can be many causes for this error. If you included the implements field, it should be an existing entry in the algorithm or math_function table. Otherwise it could be an internal server error. Please contact API support team.\n166 - Failed to add flow to database. Internal server error, please contact API administrators\n167 - Illegal files uploaded. An non required file was uploaded.\n168 - The provided md5 hash equals not the server generated md5 hash of the file.\n169 - Please provide API key. In order to share content, please authenticate and provide API key.\n170 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\n171 - Flow already exists. This flow is already in the database\n172 - XSD not found. Please contact API support team\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /run/tag:
+    post:
+      tags:
+        - run
+      summary: 'Tag a run'
+      description: 'Tags a run.'
+      operationId: 'Api_run::run_tag'
+      parameters:
+        -
+          name: run_id
+          in: query
+          description: 'Id of the run.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the tagged run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  run_tag: { $ref: '#/components/schemas/inline_response_200_19_run_tag' }
+                type: object
+              example:
+                run_tag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /run/untag:
+    post:
+      tags:
+        - run
+      summary: 'Untag a run'
+      description: 'Untags a run.'
+      operationId: 'Api_run::run_untag'
+      parameters:
+        -
+          name: run_id
+          in: query
+          description: 'Id of the run.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the untagged run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  run_untag: { $ref: '#/components/schemas/inline_response_200_20_run_untag' }
+                type: object
+              example:
+                run_untag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/run/list/{filters}':
+    get:
+      tags:
+        - run
+      summary: 'List and filter runs'
+      description: 'List runs, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/run/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all runs that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, or uploaders.'
+      operationId: 'Api_run::run_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /tag/{tag} - return only runs tagged with the given tag.\r\n  /run/{ids} - return only specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''\r\n  /task/{ids} - return only runs on specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''\r\n  /flow/{ids} - return only runs on specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only runs with specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /uploader/{ids} - return only runs uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of runs descriptions'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RunList'
+              example:
+                runs:
+                  run: [{ upload_time: '2014-04-06 23:30:40', task_id: '28', run_id: '100', error_message: [], setup_id: '12', uploader: '1', flow_id: '67' }, { upload_time: '2014-04-06 23:30:40', task_id: '48', run_id: '101', error_message: [], setup_id: '6', uploader: '1', flow_id: '61' }, { upload_time: '2014-04-06 23:30:40', task_id: '41', run_id: '102', error_message: [], setup_id: '3', uploader: '1', flow_id: '58' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n510 - Please provide at least task, flow or setup, uploader or run, to filter results, or limit the number of responses. The number of runs is huge. Please limit the result space.\n511 - Input not safe. The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\n512 - There where no results. Check whether there are runs under the given constraint.\n513 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\n514 - Illegal filter specified.\n515 - Offset specified without limit.\n516 - Requested result limit too high.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/run/{id}':
+    get:
+      tags:
+        - run
+      summary: 'Get run description'
+      description: 'Returns information about a run. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.'
+      operationId: 'Api_run::run'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the run.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A run description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Run'
+              example:
+                run:
+                  run_id: '100'
+                  uploader: '1'
+                  uploader_name: 'Jan van Rijn'
+                  task_id: '28'
+                  task_type: 'Supervised Classification'
+                  task_evaluation_measure: predictive_accuracy
+                  flow_id: '67'
+                  flow_name: weka.BayesNet_K2(1)
+                  setup_string: 'weka.classifiers.bayes.BayesNet -- -D -Q weka.classifiers.bayes.net.search.local.K2 -- -P 1 -S BAYES -E weka.classifiers.bayes.net.estimate.SimpleEstimator -- -A 0.5'
+                  parameter_setting: [{ name: D, value: 'true' }, { name: Q, value: weka.classifiers.bayes.net.search.local.K2 }, { name: P, value: '1' }, { name: S, value: BAYES }]
+                  input_data: { dataset: { did: '28', name: optdigits, url: 'https:\\/\\/www.openml.org\\/data\\/download\\/28\\/dataset_28_optdigits.arff' } }
+                  output_data: { file: [{ did: '48838', file_id: '261', name: description, url: 'https:\\/\\/www.openml.org\\/data\\/download\\/261\\/weka_generated_run935374685998857626.xml' }, { did: '48839', file_id: '262', name: predictions, url: 'https:\\/\\/www.openml.org\\/data\\/download\\/262\\/weka_generated_predictions576954524972002741.arff' }], evaluation: [{ name: area_under_roc_curve, flow_id: '4', value: '0.990288', array_data: '[0.99724,0.989212,0.992776,0.994279,0.980578,0.98649,0.99422,0.99727,0.994858,0.976143]' }, { name: confusion_matrix, flow_id: '10', array_data: '[[544,1,0,0,7,0,1,0,0,1],[0,511,21,1,0,1,3,1,5,28],[0,7,511,1,0,1,0,3,23,11],[0,2,2,519,0,3,0,12,16,18],[0,3,0,0,528,0,4,21,6,6],[0,1,0,7,5,488,2,0,4,51],[1,7,0,0,2,0,548,0,0,0],[0,2,0,1,9,1,0,545,4,4],[1,25,2,2,3,6,2,1,503,9],[0,7,0,20,16,5,0,19,9,486]]' }, { name: f_measure, flow_id: '12', value: '0.922723', array_data: '[0.989091,0.898857,0.935041,0.92431,0.927944,0.918156,0.980322,0.933219,0.895018,0.826531]' }, { name: kappa, flow_id: '13', value: '0.913601' }] }
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n220 - Please provide run ID. In order to view run details, please provide the run ID.\n221 - Run not found. The run ID was invalid, run does not exist (anymore).\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - run
+      summary: 'Delete run'
+      description: 'Deletes a run. Upon success, it returns the ID of the deleted run.'
+      operationId: 'Api_run::run_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the run.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'API key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  data_delete: { $ref: '#/components/schemas/inline_response_200_17_data_delete' }
+                type: object
+              example:
+                run_delete:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n390 - Please provide API key. In order to remove your content, please authenticate.\n391 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators\n392 - Run does not exists. The run ID could not be linked to an existing run.\n393 - Run is not owned by you. The run was owned by another user. Hence you cannot delete it.\n394 - Deleting run failed. Deleting the run failed. Please contact support team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/run/reset/{id}':
+    get:
+      tags:
+        - run
+      summary: 'Resets a run.'
+      description: 'Removes all run evaluations. When a run is reset, the runs will automatically be evaluated as soon as they are picked up by the evaluation engine again.'
+      operationId: 'Api_run::run_reset'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Run ID.'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the evaluated run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  run_reset: { $ref: '#/components/schemas/inline_response_200_21_upload_flow' }
+                type: object
+              example:
+                run_reset:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n412 - Run does not exist\n413 - Run is not owned by you\n394 - Resetting run failed\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /run:
+    post:
+      tags:
+        - run
+      summary: 'Upload run'
+      description: 'Uploads a run. Upon success, it returns the run id.'
+      operationId: 'Api_run::run_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/run).'
+          required: true
+          schema:
+            type: file
+        -
+          name: predictions
+          in: query
+          description: 'The predictions generated by the run'
+          required: true
+          schema:
+            type: file
+        -
+          name: model_readable
+          in: query
+          description: 'The human-readable model generated by the run'
+          required: false
+          schema:
+            type: file
+        -
+          name: model_serialized
+          in: query
+          description: 'The serialized model generated by the run'
+          required: false
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the uploaded run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_flow: { $ref: '#/components/schemas/inline_response_200_18_upload_flow' }
+                type: object
+              example:
+                upload_run:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n201 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n202 - Please provide run XML.\n203 - Could not validate run xml by XSD. Please double check that the xml is valid.\n204 - Unknown task. The task with the given ID was not found in the database.\n205 - Unknown flow. The flow with the given ID was not found in the database.\n206 - Invalid number of files. The number of uploaded files did not match the number of files expected for the task type\n207 - File upload failed. One of the files uploaded has a problem.\n208 - Error inserting setup record. Please contact api administrators\n210 - Unable to store run. Please contact api administrators.\n211 - Dataset not in database. One of the datasets of the task was not included in database, please contact api administrators.\n212 - Unable to store file. Please contact api administrators.\n213 - Parameter in run xml unknown. One of the parameters provided in the run xml is not registered as parameter for the flow nor its components.\n214 - Unable to store input setting. Please contact API support team.\n215 - Unable to evaluate predictions. Please contact API support team.\n216 - Error thrown by Java Application. Additional information field is provided.\n217 - Error processing output data. Unknown or inconsistent evaluation measure. One of the provided evaluation measures could not be matched with a record in the math_function or flow table.\n218 - Wrong flow associated with run. The flow implements a math_function, which is unable to generate predictions. Please select another flow.\n219 - Error reading the XML document. The XML description file could not be verified.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/run/trace/{id}':
+    get:
+      tags:
+        - run
+      summary: 'Get run trace'
+      description: 'Returns the optimization trace of run. The trace contains every setup tried, its evaluation, and whether it was selected.'
+      operationId: 'Api_run::run_trace'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the run.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A run trace description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RunTrace'
+              example:
+                trace:
+                  run_id: '573055'
+                  trace_iteration: { repeat: '0', fold: '0', iteration: '1', setup_string: { parameter_minNumObj: '1', parameter_confidenceFactor: '0.25' }, evaluation: '94.074074', selected: 'true' }
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n570 - No successful trace associated with this run\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    post:
+      tags:
+        - run
+      summary: 'Upload run trace'
+      description: 'Uploads a run trace. Upon success, it returns the run id.'
+      operationId: 'Api_run::run_trace_upload'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the run.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the trace. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.trace) and an [XML example](https://www.openml.org/api/v1/xml_example/run.trace).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the run with the trace'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_flow: { $ref: '#/components/schemas/inline_response_200_23_upload_flow' }
+                type: object
+              example:
+                run_trace:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n561 - Problem with uploaded trace file.\n562 - Problem validating xml trace file.\n563 - Problem loading xml trace file.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /run/evaluate:
+    post:
+      tags:
+        - run
+      summary: 'Uploads a run evaluation'
+      description: 'Uploads a run evaluation. When successful, it returns the run id.'
+      operationId: 'Api_run::run_evaluate'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the run evaluation.Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.evaluate) and an [XML example](https://www.openml.org/api/v1/xml_example/run.evaluate).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the evaluated run'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_flow: { $ref: '#/components/schemas/inline_response_200_21_upload_flow' }
+                type: object
+              example:
+                run_evaluate:
+                  id: '2520'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n422 - Upload problem description XML\n423 - Problem validating uploaded description file\n424 - Problem opening description xml\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /setup/tag:
+    post:
+      tags:
+        - setup
+      summary: 'Tag a setup'
+      description: 'Tags a setup.'
+      operationId: 'Api_setup::setup_tag'
+      parameters:
+        -
+          name: setup_id
+          in: query
+          description: 'Id of the setup.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the tagged setup'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_tag: { $ref: '#/components/schemas/inline_response_200_15_flow_tag' }
+                type: object
+              example:
+                setup_tag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /setup/untag:
+    post:
+      tags:
+        - setup
+      summary: 'Untag a setup'
+      description: 'Untags a setup.'
+      operationId: 'Api_setup::setup_untag'
+      parameters:
+        -
+          name: setup_id
+          in: query
+          description: 'Id of the setup.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the untagged setup'
+          content:
+            application/json:
+              schema:
+                properties:
+                  flow_untag: { $ref: '#/components/schemas/inline_response_200_16_flow_untag' }
+                type: object
+              example:
+                setup_untag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\nby another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/setup/{id}':
+    get:
+      tags:
+        - setup
+      summary: 'Get a hyperparameter setup'
+      description: 'Returns information about a setup. The information includes the list of hyperparameters, with name, value, and default value.'
+      operationId: 'Api_setup::setup'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the hyperparameter setup (configuration). These IDs are stated in run descriptions.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A setup description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Setup'
+              example:
+                setup_parameters:
+                  flow_id: '59'
+                  parameter: [{ full_name: weka.JRip(1)_F, parameter_name: F, data_type: option, default_value: '3', value: '3' }, { full_name: weka.JRip(1)_N, parameter_name: 'N', data_type: option, default_value: '2.0', value: '2.0' }, { full_name: weka.JRip(1)_O, parameter_name: O, data_type: option, default_value: '2', value: '2' }, { full_name: weka.JRip(1)_S, parameter_name: S, data_type: option, default_value: '1', value: '1' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n280 - Please provide setup ID. In order to view setup details, please provide the run ID\n281 - Setup not found. The setup ID was invalid, or setup does not exist (anymore).\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - setup
+      summary: 'Delete setup'
+      description: 'Deletes a setup. Upon success, it returns the ID of the deleted setup.'
+      operationId: 'Api_setup::setup_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the setup.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted setup'
+          content:
+            application/json:
+              schema:
+                properties:
+                  study_delete: { $ref: '#/components/schemas/inline_response_200_14_study_delete' }
+                type: object
+              example:
+                setup_delete:
+                  id: '1'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n401 - Authentication failed. Please provide API key. In order to remove your content, please authenticate.\n402 - Setup does not exists. The setup ID could not be linked to an existing setup.\n404 - Setup deletion failed. Setup is in use by other content (runs, schedules, etc). Can not be deleted.\n405 - Setup deletion failed. Please try again later.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/setup/list/{filters}':
+    get:
+      tags:
+        - setup
+      summary: 'List and filter setups'
+      description: 'List setups, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/setup/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all evaluations that match the constraints. A maximum of 1,000 results are returned at a time, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain flows, setups, or tags.'
+      operationId: 'Api_setup::setup_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /tag/{tag} - returns only setups tagged with the given tag.\r\n  /flow/{ids} - return only setups for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''\r\n  /setup/{ids} - return only specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of setup descriptions'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/SetupList'
+              example:
+                setups:
+                  setup: [{ setup_id: '10', flow_id: '65', parameter: [{ id: '4144', flow_id: '65', flow_name: weka.RandomForest, full_name: weka.RandomForest(1)_I, parameter_name: I, data_type: option, default_value: '10', value: '10' }, { id: '4145', flow_id: '65', flow_name: weka.RandomForest, full_name: weka.RandomForest(1)_K, parameter_name: K, data_type: option, default_value: '0', value: '0' }] }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n670 - Please specify at least one filter.\n671 - Illegal filter.\n672 - Illegal filter input.\n673 - Result set too big. Please use one of the filters or the limit option.\n674 - No results, please check the filter.\n675 - Cannot specify offset without limit.\n676 - Requested result limit too high.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /study:
+    post:
+      tags:
+        - study
+      summary: 'Create new study'
+      description: 'Creates a new study. Upon success, it returns the study id.'
+      operationId: 'Api_study::study_create'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the study. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.study.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/study).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the uploaded study'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_study: { $ref: '#/components/schemas/inline_response_200_25_upload_study' }
+                type: object
+              example:
+                upload_study:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n1031 - Description file not present. Please upload the study description.\n1032 - Problem validating uploaded description file. The XML description format does not meet the standards. See the XSD schema.\n1033 - Illegal main entity type. Currently only collections of tasks and can be created.\n1034 - Linked entities are not of the correct type fot this study.\n1035 - Benchmark suites can only be linked to run studies.\n1036 - Referred benchmark suite cannot be found.\n1037 - Referred benchmark suite should be a task collection.\n1038 - Study alias is not unique.\n1039 - Dataset insertion problem. Please contact the administrators.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/study/{id}/attach':
+    post:
+      tags:
+        - study
+      summary: 'Attach a new entity to a study'
+      description: 'Attach a new entity to an exising study. Upon success, it returns the study id, type, and linked entities.'
+      operationId: 'Api_study::study_attach'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the study. Supplied in the URL path.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: ids
+          in: query
+          description: 'Comma-separated list of entity IDs to be attached to the study. For instance, if this is a run study, the list of run IDs that need to be added (attached) to the study. Must be supplied as a POST variable.'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Properties of the updated study'
+          content:
+            application/json:
+              schema:
+                properties:
+                  study_attach: { $ref: '#/components/schemas/inline_response_200_26_study_attach' }
+                type: object
+              example:
+                study_attach:
+                  id: '1'
+                  main_entity_type: task
+                  linked_entities: '5'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n1041 - Could not find study. Check the study ID in your request.\n1042 - Cannnot attach entities to legacy studies.\n1043 - Please provide POST field ''ids''.\n1044 - Please ensure that the ''ids'' in the POST field is a list of natural numbers.\n1045 - Could not attach entities to the study. It appears as if the entity does not exist.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/study/{id}/detach':
+    post:
+      tags:
+        - study
+      summary: 'Detach an entity from a study'
+      description: 'Detach an entity from an exising study. Upon success, it returns the study id, type, and linked entities.'
+      operationId: 'Api_study::study_detach'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the study.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: ids
+          in: query
+          description: 'Comma-separated list of entity IDs to be detached from the study. For instance, if this is a run study, the list of run IDs that need to be removed (detached) from the study. Must be supplied as a POST variable.'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Properties of the updated study'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_study: { $ref: '#/components/schemas/inline_response_200_26_study_attach' }
+                type: object
+              example:
+                study_detach:
+                  id: '1'
+                  main_entity_type: task
+                  linked_entities: '5'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n1041 - Could not find study. Check the study ID in your request.\n1042 - Cannot attach entities to legacy studies.\n1043 - Please provide POST field ''ids''.\n1044 - Please ensure that the ''ids'' in the POST field is a list of natural numbers.\n1046 - Could not detach entities from the study. It appears as if the entity does not exist.       \n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/study/{id}':
+    get:
+      tags:
+        - study
+      summary: 'Get study description by study id or alias'
+      description: 'Returns information about the study with the given id or alias.'
+      operationId: 'Api_study::study_get'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID or alias of the study.'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A study description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Study'
+              example:
+                study:
+                  id: '99'
+                  main_entity_type: task
+                  name: 'CC18 benchmark suite'
+                  description: 'CC18 benchmark suite'
+                  creation_date: '2019-02-16T17:35:58'
+                  creator: '1159'
+                  data: { data_id: ['1', '2', '3'] }
+                  tasks: { task_id: ['1', '2', '3'] }
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n601 - Unknown study. The study with the given id or alias was not found in the database\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - study
+      summary: 'Delete study'
+      description: 'Deletes a study. Upon success, it returns the ID of the deleted study.'
+      operationId: 'Api_study::study_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the study.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted study'
+          content:
+            application/json:
+              schema:
+                properties:
+                  study_delete: { $ref: '#/components/schemas/inline_response_200_24_study_delete' }
+                type: object
+              example:
+                study_delete:
+                  id: '1'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n591 - Please provide API key. In order to remove your content, please authenticate.\n592 - Study does not exists. The study ID could not be linked to an existing study.\n593 - Study deletion failed. Please try again later.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/study/list/{filters}':
+    get:
+      tags:
+        - study
+      summary: 'List all studies (collections of items)'
+      description: 'List studies, optionally filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form ''/study/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all studies that match the constraints.'
+      operationId: 'Api_study::study_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /main_entity_type/{type} - only return studies collecting entities of a given type (e.g. 'task' or 'run').\r\n  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of studies'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/StudyList'
+              example:
+                study_list:
+                  study: [{ id: '1', alias: Study_1, name: 'A large-scale comparison of classification algorithms', creation_date: '2017-07-20 15:51:20', creator: '2' }, { id: '2', alias: Study_2, name: 'Fast Algorithm Selection using Learning Curves', creation_date: '2017-07-20 15:51:20', creator: '2' }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /task/tag:
+    post:
+      tags:
+        - task
+      summary: 'Tag a task'
+      description: 'Tags a task.'
+      operationId: 'Api_task::task_tag'
+      parameters:
+        -
+          name: task_id
+          in: query
+          description: 'Id of the task.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'The id of the tagged task'
+          content:
+            application/json:
+              schema:
+                properties:
+                  task_tag: { $ref: '#/components/schemas/inline_response_200_6_task_tag' }
+                type: object
+              example:
+                task_tag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, task, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /task/untag:
+    post:
+      tags:
+        - task
+      summary: 'Untag a task'
+      description: 'Untags a task.'
+      operationId: 'Api_task::task_untag'
+      parameters:
+        -
+          name: task_id
+          in: query
+          description: 'Id of the task.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: tag
+          in: query
+          description: 'Tag name'
+          required: true
+          schema:
+            type: string
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A the features of the task'
+          content:
+            application/json:
+              schema:
+                properties:
+                  task_untag: { $ref: '#/components/schemas/inline_response_200_7_task_untag' }
+                type: object
+              example:
+                task_untag:
+                  id: '2'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, task, flow, run} not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, task, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/task/list/{filters}':
+    get:
+      tags:
+        - task
+      summary: 'List and filter tasks'
+      description: 'List tasks, possibly filtered by a range of properties from the task itself or from the underlying dataset. Any number of properties can be combined by listing them one after the other in the form ''/task/list/{filter}/{value}/{filter}/{value}/...'' Returns an array with all tasks that match the constraints.'
+      operationId: 'Api_task::task_list'
+      parameters:
+        -
+          name: filters
+          in: path
+          description: "Any combination of these filters\r\n  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.\r\n  /status/{status} - returns only tasks with a given status, either 'active', 'deactivated', or 'in_preparation'.\r\n  /type/{type_id} - returns only tasks with a given task type id. See the list of task types of the ID's (e.g. 1 = Supervised Classification).\r\n  /tag/{tag} - returns only tasks tagged with the given tag.\r\n  /data_tag/{tag} - returns only tasks for which the underlying dataset is tagged with the given tag.\r\n  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.\r\n  "
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'A list of tasks with the given tag'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/TaskList'
+              example:
+                task:
+                  task: [{ task_id: '1', task_type: 'Supervised Classification', did: '1', name: anneal, status: active, format: ARFF, input: [{ name: estimation_procedure, value: '1' }, { name: evaluation_measures, value: predictive_accuracy }, { name: source_data, value: '1' }, { name: target_feature, value: class }], quality: [{ name: MajorityClassSize, value: '684' }, { name: MaxNominalAttDistinctValues, value: '10.0' }, { name: MinorityClassSize, value: '0' }, { name: NumBinaryAtts, value: '14.0' }, { name: NumberOfClasses, value: '6' }, { name: NumberOfFeatures, value: '39' }, { name: NumberOfInstances, value: '898' }, { name: NumberOfInstancesWithMissingValues, value: '0' }, { name: NumberOfMissingValues, value: '0' }, { name: NumberOfNumericFeatures, value: '6' }, { name: NumberOfSymbolicFeatures, value: '32' }], tag: [basic, study_1, study_7, under100k, under1m] }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n480 - Illegal filter specified.\n481 - Filter values/ranges not properly specified.\n482 - No results. There where no matches for the given constraints.\n483 - Can not specify an offset without a limit.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  '/task/{id}':
+    get:
+      tags:
+        - task
+      summary: 'Get task description'
+      description: 'Returns information about a task. The information includes the task type, input data, train/test sets, and more.'
+      operationId: 'Api_task::task'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the task.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A task description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Task'
+              example:
+                task:
+                  task_id: '1'
+                  task_type: 'Supervised Classification'
+                  input: [{ name: source_data, data_set: { data_set_id: '1', target_feature: class } }, { name: estimation_procedure, estimation_procedure: { type: crossvalidation, data_splits_url: 'https://www.openml.org/api_splits/get/1/Task_1_splits.arff', parameter: [{ name: number_repeats, value: '1' }, { name: number_folds, value: '10' }, { name: percentage }, { name: stratified_sampling, value: 'true' }] } }, { name: cost_matrix, cost_matrix: [] }, { name: evaluation_measures, evaluation_measures: { evaluation_measure: predictive_accuracy } }]
+                  output: { name: predictions, predictions: { format: ARFF, feature: [{ name: repeat, type: integer }, { name: fold, type: integer }, { name: row_id, type: integer }, { name: confidence.classname, type: numeric }, { name: prediction, type: string }] } }
+                  tag: [basic, study_1, under100k, under1m]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n150 - Please provide task_id.\n151 - Unknown task. The task with the given id was not found in the database\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+    delete:
+      tags:
+        - task
+      summary: 'Delete task'
+      description: 'Deletes a task. Upon success, it returns the ID of the deleted task.'
+      operationId: 'Api_task::task_delete'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'Id of the task.'
+          required: true
+          schema:
+            type: integer
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'ID of the deleted task'
+          content:
+            application/json:
+              schema:
+                properties:
+                  task_delete: { $ref: '#/components/schemas/inline_response_200_4_task_delete' }
+                type: object
+              example:
+                task_delete:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n450 - Please provide API key. In order to remove your content, please authenticate.\n451 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n452 - Task does not exists. The task ID could not be linked to an existing task.\n454 - Task is executed in some runs. Delete these first.\n455 - Deleting the task failed. Please contact support team.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /task:
+    post:
+      tags:
+        - task
+      summary: 'Upload task'
+      description: 'Uploads a task. Upon success, it returns the task id.'
+      operationId: 'Api_task::task_upload'
+      parameters:
+        -
+          name: description
+          in: query
+          description: 'An XML file describing the task. Only name, description, and task format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.task.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/task).'
+          required: true
+          schema:
+            type: file
+        -
+          name: api_key
+          in: query
+          description: 'Api key to authenticate the user'
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: 'Id of the uploaded task'
+          content:
+            application/json:
+              schema:
+                properties:
+                  upload_task: { $ref: '#/components/schemas/inline_response_200_5_upload_task' }
+                type: object
+              example:
+                upload_task:
+                  id: '4328'
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n530 - Description file not present. Please upload the task description.\n531 - Internal error. Please contact api support team\n532 - Problem validating uploaded description file. The XML description format does not meet the standards\n533 - Task already exists.\n534 - Error creating the task.\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /tasktype/list:
+    get:
+      tags:
+        - tasktype
+      summary: 'List all task types'
+      description: 'Returns an array with all task types in the system.'
+      responses:
+        default:
+          description: 'Unexpected error'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+        '200':
+          description: 'A task description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/TaskTypeList'
+              example:
+                task_types:
+                  task_type: [{ id: '1', name: 'Supervised Classification', description: 'In supervised classification, you are given ...', creator: 'Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl' }, { id: '2', name: 'Supervised Regression', description: 'Given a dataset with a numeric target ...', creator: 'Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl' }]
+  '/tasktype/{id}':
+    get:
+      tags:
+        - tasktype
+      summary: 'Get task type description'
+      description: 'Returns information about a task type. The information includes a description, the given inputs and the expected outputs.'
+      parameters:
+        -
+          name: id
+          in: path
+          description: 'ID of the task.'
+          required: true
+          schema:
+            type: integer
+      responses:
+        '200':
+          description: 'A task type description'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/TaskType'
+              example:
+                task_type:
+                  id: '1'
+                  name: 'Supervised Classification'
+                  description: "In supervised classification, you are given an input dataset in which instances are labeled with a certain class. The goal is to build a model that predicts the class for future unlabeled instances. The model is evaluated using a train-test procedure, e.g. cross-validation.<br><br>\\\r\n     *   \\\r\n     *   To make results by different users comparable, you are given the exact train-test folds to be used, and you need to return at least the predictions generated by your model for each of the test instances. OpenML will use these predictions to calculate a range of evaluation measures on the server.<br><br>\\\r\n     *   \\\r\n     *   You can also upload your own evaluation measures, provided that the code for doing so is available from the implementation used. For extremely large datasets, it may be infeasible to upload all predictions. In those cases, you need to compute and provide the evaluations yourself.<br><br>\\\r\n     *   \\\r\n     *   Optionally, you can upload the model trained on all the input data. There is no restriction on the file format, but please use a well-known format or PMML."
+                  creator: ['Joaquin Vanschoren', 'Jan van Rijn', 'Luis Torgo', 'Bernd Bischl']
+                  contributor: ['Bo Gao', 'Simon Fischer', 'Venkatesh Umaashankar', 'Michael Berthold', 'Bernd Wiswedel', 'Patrick Winter']
+                  creation_date: '2013-01-24 00:00:00'
+                  input: [{ name: source_data, requirement: required, data_type: numeric }, { name: target_feature, requirement: required, data_type: string }, { name: estimation_procedure, requirement: required, data_type: numeric }, { name: cost_matrix, data_type: json }, { name: custom_testset, data_type: json }, { name: evaluation_measures, data_type: string }]
+        '412':
+          description: 'Precondition failed. An error code and message are returned.\n240 - Please provide task type ID.\n241 - Unknown task type. The task type with the given id was not found in the database\n'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/Error'
+  /user/list:
+    get:
+      tags:
+        - user
+      summary: 'List all users by user id'
+      description: 'Returns an array with all user ids and names.'
+      operationId: 'Api_user::username_list'
+      responses:
+        '200':
+          description: 'A list of users'
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/UserList'
+              example:
+                users:
+                  user: [{ id: '1', username: janvanrijn@gmail.com }, { id: '2', username: joaquin.vanschoren@gmail.com }]
+components:
+  schemas:
+    inline_response_200_16_flow_untag:
+      properties:
+        id:
+          description: 'ID of the untagged setup'
+          type: string
+      type: object
+    EvaluationMeasureList_evaluation_measures_measures:
+      properties:
+        measure:
+          description: 'The evaluation measure names'
+          type: array
+          items:
+            type: string
+      type: object
+    inline_response_200_10_flow_exists:
+      properties:
+        id:
+          description: 'The id of the flow with the given name and (external) version'
+          type: string
+        exists:
+          description: 'true or false'
+          type: string
+      type: object
+    Data_data_set_description:
+      properties:
+        default_target_attribute:
+          description: 'For tabular data, the name of the column that is typically used as the target attribute for that data set'
+          type: string
+        upload_date:
+          description: 'The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss'
+          type: string
+        version_label:
+          description: 'The version of the dataset, as defined by the uploader, for reference. Can be any format as long as it is unique.'
+          type: string
+        description:
+          description: 'Wiki description of the dataset, in (Git flavoured) markdown format'
+          type: string
+        format:
+          description: 'Data format, for instance ARFF'
+          type: string
+        url:
+          description: 'The URL where the data can be downloaded'
+          type: string
+        tag:
+          description: 'Tags added by OpenML users. Includes study tags in the form `study_1`'
+          type: array
+          items:
+            type: string
+        visibility:
+          description: 'Who can see the dataset. For instance `public`.'
+          type: string
+        md5_checksum:
+          description: 'Checksum to verify downloads of the dataset'
+          type: string
+        version:
+          description: 'The version of the dataset, set by OpenML. A positive integer'
+          type: string
+        status:
+          description: 'active, in_preparation, or deactivated'
+          type: string
+        file_id:
+          description: 'The ID of the dataset file stored on the OpenML server'
+          type: string
+        licence:
+          description: 'The licence granted for using the dataset, for instance Public or CC-BY'
+          type: string
+        original_data_url:
+          description: 'The URL where the original data is hosted.'
+          type: string
+        id:
+          description: 'ID of the dataset, a positive integer'
+          type: string
+        name:
+          description: 'The name of the dataset'
+          type: string
+      type: object
+    DataList_data_dataset:
+      properties:
+        did:
+          description: 'The dataset ID'
+          type: string
+        status:
+          description: 'The dataset status, either in_preparation, active, or deactivated'
+          type: string
+        quality:
+          type: array
+          items:
+            $ref: '#/components/schemas/DataList_data_quality'
+        name:
+          description: 'The dataset name'
+          type: string
+        format:
+          description: 'The data format of the dataset, e.g. ARFF'
+          type: string
+      type: object
+    Task_task_description_estimation_procedure:
+      properties:
+        parameter:
+          type: array
+          items:
+            $ref: '#/components/schemas/Task_task_description_estimation_procedure_parameter'
+        type:
+          description: 'The type of procedure, e.g. crossvalidation'
+          type: string
+        data_splits_url:
+          description: 'The url where the data splits can be downloaded'
+          type: string
+      type: object
+    DataList_data_quality:
+      properties:
+        name:
+          description: 'The name of the property'
+          type: string
+        value:
+          description: 'The value of the property'
+          type: string
+      type: object
+    FlowList_flows_flow:
+      properties:
+        full_name:
+          description: 'The full flow name (name + internal version number)'
+          type: string
+        external_version:
+          description: 'The external flow version'
+          type: string
+        version:
+          description: 'The internal flow version'
+          type: string
+        uploader:
+          description: 'The ID of the person who uploaded the flow'
+          type: string
+        id:
+          description: 'The flow ID'
+          type: string
+        name:
+          description: 'The flow name'
+          type: string
+      type: object
+    Run_run_description_input_data_dataset:
+      properties:
+        did:
+          description: 'The id of the dataset'
+          type: string
+        url:
+          description: 'The download url of the dataset'
+          type: string
+        name:
+          description: 'The name of the dataset'
+          type: string
+      type: object
+    EvaluationList_evaluations:
+      properties:
+        evaluation:
+          type: array
+          items:
+            $ref: '#/components/schemas/EvaluationList_evaluations_evaluation'
+      type: object
+    inline_response_200_14_study_delete:
+      properties:
+        id:
+          description: 'ID of the deleted setup, a positive integer'
+          type: string
+      type: object
+    DataUnprocessed:
+      properties:
+        data_unprocessed:
+          $ref: '#/components/schemas/DataUnprocessed_data_unprocessed'
+      type: object
+    DataUnprocessed_data_unprocessed_dataset:
+      properties:
+        did:
+          description: 'ID of the dataset a positive integer'
+          type: string
+        status:
+          description: 'Status of the dataset'
+          type: string
+        version:
+          description: 'Version of the dataset, a positive integer'
+          type: string
+        name:
+          description: 'The name of the dataset'
+          type: string
+        format:
+          description: 'The dataset format, e.g. ARFF'
+          type: string
+      type: object
+    EstimationProcedureList_estimationprocedures:
+      properties:
+        estimationprocedure:
+          type: array
+          items:
+            $ref: '#/components/schemas/EstimationProcedureList_estimationprocedures_estimationprocedure'
+      type: object
+    TaskTypeList:
+      properties:
+        task_types:
+          $ref: '#/components/schemas/TaskTypeList_task_types'
+      type: object
+    Run_run_description_output_data_evaluation:
+      properties:
+        array_data:
+          description: 'For composite evaluation measures (e.g. per-class measures, confusion matrix), a string (JSON) representation of the   evaluation.'
+          type: string
+        name:
+          description: 'The name of the evaluation measure'
+          type: string
+        value:
+          description: 'The result of the evaluation'
+          type: string
+        flow_id:
+          description: 'The id of the code used to compute this evaluation method'
+          type: string
+      type: object
+    TaskList_task_quality:
+      properties:
+        name:
+          description: 'The name of the quality'
+          type: string
+        value:
+          description: 'The value of the quality'
+          type: string
+      type: object
+    Task_task_description:
+      properties:
+        input:
+          type: array
+          items:
+            $ref: '#/components/schemas/Task_task_description_input'
+        task_type:
+          description: 'The type of the task, e.g. Supervised Classification'
+          type: string
+        tag:
+          description: 'Tags added by OpenML uers. Includes study tags in the form ''study_1'''
+          type: array
+          items:
+            type: string
+        task_id:
+          description: 'ID of the task, a positive integer'
+          type: string
+        output:
+          type: array
+          items:
+            $ref: '#/components/schemas/Task_task_description_output'
+      type: object
+    inline_response_200_9_upload_flow:
+      properties:
+        id:
+          description: 'ID of the uploaded flow, a positive integer'
+          type: string
+      type: object
+    Flow_flow_description:
+      properties:
+        upload_date:
+          description: 'The datetime that the flow was uploaded, format yyyy-MM-dd HH:mm:ss'
+          type: string
+        description:
+          description: 'Wiki description of the flow, in (Git flavoured) markdown format'
+          type: string
+        language:
+          description: 'The programming language the flow is written in.'
+          type: string
+        parameter:
+          type: array
+          items:
+            $ref: '#/components/schemas/Flow_flow_description_parameter'
+        tag:
+          description: 'Tags added by OpenML users. Includes study tags in the form `study_1`'
+          type: array
+          items:
+            type: string
+        version:
+          description: 'The version of the flow, set by OpenML. A positive integer'
+          type: string
+        version_label:
+          description: 'The version of the flow, as defined by the uploader, for reference. Can be any format as long as it is unique.'
+          type: string
+        dependencies:
+          description: 'The libraries that this flow depends on, and their version numbers.'
+          type: string
+        uploader:
+          description: 'The uploader of the flow'
+          type: string
+        id:
+          description: 'ID of the flow, a positive integer'
+          type: string
+        name:
+          description: 'The name of the flow'
+          type: string
+      type: object
+    Task:
+      properties:
+        task_description:
+          $ref: '#/components/schemas/Task_task_description'
+      type: object
+    Setup:
+      properties:
+        setup_parameters:
+          $ref: '#/components/schemas/Setup_setup_parameters'
+      type: object
+    FlowList_flows:
+      properties:
+        flow:
+          type: array
+          items:
+            $ref: '#/components/schemas/FlowList_flows_flow'
+      type: object
+    RunTrace:
+      properties:
+        trace:
+          $ref: '#/components/schemas/RunTrace_trace'
+      type: object
+    Task_task_description_predictions_feature:
+      properties:
+        type:
+          description: 'The type of the prediction feature, e.g. integer'
+          type: string
+        name:
+          description: 'The name of the prediction feature, e.g. row_id'
+          type: string
+      type: object
+    inline_response_200_24_study_delete:
+      properties:
+        id:
+          description: 'ID of the deleted study, a positive integer'
+          type: string
+      type: object
+    inline_response_200_5_upload_task:
+      properties:
+        id:
+          description: 'ID of the uploaded task, a positive integer'
+          type: string
+      type: object
+    inline_response_200_13_flow_untag:
+      properties:
+        id:
+          description: 'ID of the untagged flow'
+          type: string
+      type: object
+    UserList_users:
+      properties:
+        user:
+          type: array
+          items:
+            $ref: '#/components/schemas/UserList_users_user'
+      type: object
+    DataFeatures:
+      properties:
+        data_features:
+          $ref: '#/components/schemas/DataFeatures_data_features'
+      type: object
+    Task_task_description_predictions:
+      properties:
+        feature:
+          type: array
+          items:
+            $ref: '#/components/schemas/Task_task_description_predictions_feature'
+        format:
+          description: 'The fromat of the predictions, e.g. ARFF'
+          type: string
+      type: object
+    inline_response_200_15_flow_tag:
+      properties:
+        id:
+          description: 'ID of the tagged setup'
+          type: string
+      type: object
+    Setup_setup_parameters:
+      properties:
+        parameter_setting:
+          type: array
+          items:
+            $ref: '#/components/schemas/Setup_setup_parameters_parameter_setting'
+        flow_id:
+          description: 'ID of the flow, a positive integer'
+          type: string
+      type: object
+    DataList:
+      properties:
+        data:
+          $ref: '#/components/schemas/DataList_data'
+      type: object
+    inline_response_200_26_study_attach:
+      properties:
+        linked_entities:
+          description: 'The number of linked entities'
+          type: string
+        main_entity_type:
+          description: 'Main entity type of the of the study'
+          type: string
+        id:
+          description: 'ID of the study, a positive integer'
+          type: string
+      type: object
+    inline_response_200_2_data_tag:
+      properties:
+        id:
+          description: 'ID of the tagged dataset'
+          type: string
+      type: object
+    inline_response_200_23_upload_flow:
+      properties:
+        id:
+          description: 'ID of the run with the trace, a positive integer'
+          type: string
+      type: object
+    EvaluationRequest_evaluation_request:
+      properties:
+        run:
+          $ref: '#/components/schemas/EvaluationRequest_evaluation_request_run'
+      type: object
+    Run:
+      properties:
+        run_description:
+          $ref: '#/components/schemas/Run_run_description'
+      type: object
+    inline_response_200_6_task_tag:
+      properties:
+        id:
+          description: 'ID of the tagged task'
+          type: string
+      type: object
+    RunList:
+      properties:
+        runs:
+          $ref: '#/components/schemas/RunList_runs'
+      type: object
+    inline_response_200_17_data_delete:
+      properties:
+        id:
+          description: 'ID of the deleted run, a positive integer'
+          type: string
+      type: object
+    Task_task_description_input:
+      properties:
+        data_set:
+          $ref: '#/components/schemas/Task_task_description_data_set'
+        cost_matrix:
+          description: 'The cost matrix, indicating the costs for each type of misclassification'
+          type: array
+          items:
+            type: array
+            items:
+              type: integer
+              format: int64
+        name:
+          description: 'The name of the input, e.g. source_data'
+          type: string
+        evaluation_measures:
+          $ref: '#/components/schemas/Task_task_description_evaluation_measures'
+        estimation_procedure:
+          $ref: '#/components/schemas/Task_task_description_estimation_procedure'
+      type: object
+    UserList:
+      properties:
+        users:
+          $ref: '#/components/schemas/UserList_users'
+      type: object
+    Study_study_runs:
+      properties:
+        run_id:
+          type: array
+          items:
+            type: string
+      type: object
+    DataFeatures_data_features_feature:
+      properties:
+        index:
+          description: 'Feature index'
+          type: string
+        name:
+          description: 'Feature name'
+          type: string
+        data_type:
+          description: 'Feature data type'
+          type: string
+        is_target:
+          description: 'Whether this feature is seen as a target feature'
+          type: string
+        is_ignore:
+          description: 'Whether this feature should be ignored in modelling (e.g. every value is unique)'
+          type: string
+        is_row_identifier:
+          description: 'Whether this feature is a row identifier'
+          type: string
+      type: object
+    inline_response_200_21_upload_flow:
+      properties:
+        id:
+          description: 'ID of the evaluated run, a positive integer'
+          type: string
+      type: object
+    TaskType:
+      properties:
+        description:
+          description: 'A description of the task type'
+          type: string
+        date:
+          description: 'The date when the task type was created'
+          type: string
+        output:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskType_output'
+        contributor:
+          type: array
+          items:
+            type: string
+        input:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskType_input'
+        id:
+          description: 'ID of the task type, a positive integer'
+          type: string
+        name:
+          description: 'The name of the task type, e.g. Supervised Classification'
+          type: string
+      type: object
+    Run_run_description_output_data_file:
+      properties:
+        did:
+          description: 'The id of the uploaded file'
+          type: string
+        file_id:
+          description: 'The reference id of the uploaded file, for downloading afterward'
+          type: string
+        name:
+          description: 'The name of the uploaded file (e.g., description, predictions, model,...)'
+          type: string
+      type: object
+    TaskType_input:
+      properties:
+        data_set:
+          $ref: '#/components/schemas/Task_task_description_data_set'
+        cost_matrix:
+          type: array
+          items:
+            type: array
+            items:
+              type: integer
+              format: int64
+        name:
+          description: 'The name of the input, e.g. source_data'
+          type: string
+        evaluation_measures:
+          $ref: '#/components/schemas/Task_task_description_evaluation_measures'
+        estimation_procedure:
+          $ref: '#/components/schemas/Task_task_description_estimation_procedure'
+      type: object
+    Run_run_description_output_data:
+      properties:
+        evaluation:
+          type: array
+          items:
+            $ref: '#/components/schemas/Run_run_description_output_data_evaluation'
+        file:
+          type: array
+          items:
+            $ref: '#/components/schemas/Run_run_description_output_data_file'
+      type: object
+    inline_response_200_3_data_untag:
+      properties:
+        id:
+          description: 'ID of the untagged dataset'
+          type: string
+      type: object
+    Task_task_description_output:
+      properties:
+        name:
+          description: 'The name of the output, e.g. predictions'
+          type: string
+        predictions:
+          $ref: '#/components/schemas/Task_task_description_predictions'
+      type: object
+    DataQualityList:
+      properties:
+        data_qualities_list:
+          $ref: '#/components/schemas/DataQualityList_data_qualities_list'
+      type: object
+    DataUnprocessed_data_unprocessed:
+      properties:
+        dataset:
+          $ref: '#/components/schemas/DataUnprocessed_data_unprocessed_dataset'
+      type: object
+    Study:
+      properties:
+        study:
+          $ref: '#/components/schemas/Study_study'
+      type: object
+    DataList_data:
+      properties:
+        dataset:
+          type: array
+          items:
+            $ref: '#/components/schemas/DataList_data_dataset'
+      type: object
+    UserList_users_user:
+      properties:
+        username:
+          description: 'The full user name'
+          type: string
+        id:
+          description: 'The user ID'
+          type: string
+      type: object
+    TaskType_output:
+      properties:
+        name:
+          description: 'The name of the output, e.g. predictions'
+          type: string
+        predictions:
+          $ref: '#/components/schemas/TaskType_predictions'
+      type: object
+    RunList_runs_run:
+      properties:
+        task_id:
+          description: 'The ID of the task solved by this run'
+          type: string
+        run_id:
+          description: 'The run ID'
+          type: string
+        error_message:
+          description: 'Error message generated by the run (if any)'
+          type: string
+        setup_id:
+          description: 'Ignore (internal representation of the parameter setting)'
+          type: string
+        uploader:
+          description: 'The ID of the person uploading this run'
+          type: string
+        flow_id:
+          description: 'The ID of the flow used in this run'
+          type: string
+      type: object
+    DataFeatures_data_features:
+      properties:
+        feature:
+          type: array
+          items:
+            $ref: '#/components/schemas/DataFeatures_data_features_feature'
+      type: object
+    DataQualities_data_qualities_quality:
+      properties:
+        name:
+          description: 'The name of the dataset quality measures'
+          type: string
+        value:
+          description: 'The value for this dataset'
+          type: string
+      type: object
+    inline_response_200_12_flow_tag:
+      properties:
+        id:
+          description: 'ID of the tagged flow'
+          type: string
+      type: object
+    inline_response_200_8_flow_delete:
+      properties:
+        id:
+          description: 'ID of the deleted flow, a positive integer'
+          type: string
+      type: object
+    DataQualities_data_qualities:
+      properties:
+        quality:
+          type: array
+          items:
+            $ref: '#/components/schemas/DataQualities_data_qualities_quality'
+      type: object
+    EvaluationRequest:
+      properties:
+        evaluation_request:
+          $ref: '#/components/schemas/EvaluationRequest_evaluation_request'
+      type: object
+    inline_response_200_18_upload_flow:
+      properties:
+        id:
+          description: 'ID of the uploaded run, a positive integer'
+          type: string
+      type: object
+    SetupList:
+      properties:
+        setups:
+          $ref: '#/components/schemas/SetupList_setups'
+      type: object
+    TaskList:
+      properties:
+        task:
+          $ref: '#/components/schemas/TaskList_task'
+      type: object
+    inline_response_200_4_task_delete:
+      properties:
+        id:
+          description: 'ID of the deleted task, a positive integer'
+          type: string
+      type: object
+    Task_task_description_data_set:
+      properties:
+        data_set_id:
+          description: 'The id of the dataset'
+          type: string
+        target_feature:
+          description: 'The name of the target feature for this task'
+          type: string
+      type: object
+    EvaluationList:
+      properties:
+        evaluations:
+          $ref: '#/components/schemas/EvaluationList_evaluations'
+      type: object
+    Run_run_description:
+      properties:
+        setup_string:
+          description: 'Configuration of the flow as a string, to be interpreted by the flow, its library, or command line interface.'
+          type: string
+        task_type:
+          description: 'The type of task solved by this run (e.g., classification)'
+          type: string
+        task_id:
+          description: 'The id of the task solved by this run'
+          type: string
+        task_evaluation_measure:
+          description: 'The evaluation measure that is supposed to be optimized in the task, if any'
+          type: string
+        uploader_name:
+          description: 'The name of the uploader of the run'
+          type: string
+        input_data:
+          $ref: '#/components/schemas/Run_run_description_input_data'
+        tag:
+          description: 'Tags added by OpenML users. Includes study tags in the form `study_1`'
+          type: array
+          items:
+            type: string
+        output_data:
+          $ref: '#/components/schemas/Run_run_description_output_data'
+        uploader:
+          description: 'The uploader of the run'
+          type: string
+        flow_id:
+          description: 'The id of the flow used in this run'
+          type: string
+        flow_name:
+          description: 'The name of the flow used in this run'
+          type: string
+        id:
+          description: 'ID of the run, a positive integer'
+          type: string
+        parameter_setting:
+          type: array
+          items:
+            $ref: '#/components/schemas/Run_run_description_parameter_setting'
+      type: object
+    FlowList:
+      properties:
+        flows:
+          $ref: '#/components/schemas/FlowList_flows'
+      type: object
+    inline_response_200_19_run_tag:
+      properties:
+        id:
+          description: 'ID of the tagged run'
+          type: string
+      type: object
+    Study_study_data:
+      properties:
+        data_id:
+          type: array
+          items:
+            type: string
+      type: object
+    inline_response_200_data_delete:
+      properties:
+        id:
+          description: 'ID of the deleted dataset, a positive integer'
+          type: string
+      type: object
+    RunTrace_trace:
+      properties:
+        trace_iteration:
+          type: array
+          items:
+            $ref: '#/components/schemas/RunTrace_trace_trace_iteration'
+        run_id:
+          description: 'run ID'
+          type: string
+      type: object
+    inline_response_200_7_task_untag:
+      properties:
+        id:
+          description: 'ID of the untagged task'
+          type: string
+      type: object
+    Data:
+      properties:
+        data_set_description:
+          $ref: '#/components/schemas/Data_data_set_description'
+      type: object
+    Task_task_description_evaluation_measures:
+      properties:
+        evaluation_measure:
+          description: 'The evaluation measure to optimize in this task'
+          type: string
+      type: object
+    SetupList_setups_parameter:
+      properties:
+        default_value:
+          description: 'The parameter''s default value'
+          type: string
+        data_type:
+          description: 'The parameter''s data type'
+          type: string
+        value:
+          description: 'The parameter value in this setup'
+          type: string
+        parameter_name:
+          description: 'The parameter''s short name'
+          type: string
+        full_name:
+          description: 'The parameter''s full name'
+          type: string
+        flow_id:
+          description: 'The (sub)flow ID'
+          type: string
+        flow_name:
+          description: 'The (sub)flow name'
+          type: string
+        id:
+          description: 'The parameter ID'
+          type: string
+      type: object
+    TaskList_task_input:
+      properties:
+        name:
+          description: 'The name of the input'
+          type: string
+        value:
+          description: 'The value of the input'
+          type: string
+      type: object
+    StudyList_study_list:
+      properties:
+        study:
+          type: array
+          items:
+            $ref: '#/components/schemas/StudyList_study_list_study'
+      type: object
+    Study_study_tag:
+      properties:
+        write_access:
+          description: 'The write access level of the study (e.g. public)'
+          type: string
+        name:
+          description: 'The name of the study (e.g. study_1)'
+          type: string
+      type: object
+    Flow:
+      properties:
+        flow_description:
+          $ref: '#/components/schemas/Flow_flow_description'
+      type: object
+    Run_run_description_parameter_setting:
+      properties:
+        name:
+          description: 'The name of the parameter'
+          type: string
+        value:
+          description: 'The value of the parameter used'
+          type: string
+      type: object
+    EvaluationRequest_evaluation_request_run:
+      properties:
+        setup_id:
+          description: 'ID of the setup, a positive integer'
+          type: string
+        upload_time:
+          description: 'The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss'
+          type: string
+        uploader:
+          description: 'ID of the uploader, a positive integer'
+          type: string
+        task_id:
+          description: 'ID of the task, a positive integer'
+          type: string
+        run_id:
+          description: 'ID of the run, a positive integer'
+          type: string
+      type: object
+    Error:
+      properties:
+        message:
+          type: string
+        code:
+          type: integer
+        additional_message:
+          type: string
+      type: object
+    EstimationProcedureList_estimationprocedures_estimationprocedure:
+      properties:
+        name:
+          description: 'The estimation procedure name, e.g. ''10 fold Crossvalidation'''
+          type: string
+        folds:
+          description: 'The number of cross-validation folds, e.g. ''10'''
+          type: string
+        stratified_sampling:
+          description: 'Whether or not the sampling is stratified, ''true'' or ''false'''
+          type: string
+        ttid:
+          description: 'The task type ID'
+          type: string
+        repeats:
+          description: 'The number of repeats, e.g. ''10'''
+          type: string
+        type:
+          description: 'The estimation procedure type, e.g. ''crossvalidation'''
+          type: string
+        id:
+          description: 'The estimation procedure ID'
+          type: string
+      type: object
+    TaskTypeList_task_types:
+      properties:
+        task_type:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskTypeList_task_types_task_type'
+      type: object
+    TaskList_task_task:
+      properties:
+        status:
+          description: 'The status of the source dataset, active, in_preparation, or deactivated'
+          type: string
+        task_type:
+          description: 'The type of task (e.g. Supervised Classificationr)'
+          type: string
+        name:
+          description: 'The name of the source dataset'
+          type: string
+        task_id:
+          description: 'The ID of the task'
+          type: string
+        format:
+          description: 'The format of the source dataset'
+          type: string
+        did:
+          description: 'The id of the source dataset'
+          type: string
+        tag:
+          type: array
+          items:
+            type: string
+        input:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskList_task_input'
+        quality:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskList_task_quality'
+      type: object
+    inline_response_200_20_run_untag:
+      properties:
+        id:
+          description: 'ID of the untagged run'
+          type: string
+      type: object
+    StudyList:
+      properties:
+        study_list:
+          $ref: '#/components/schemas/StudyList_study_list'
+      type: object
+    Study_study:
+      properties:
+        runs:
+          $ref: '#/components/schemas/Study_study_runs'
+        tasks:
+          $ref: '#/components/schemas/Study_study_tasks'
+        name:
+          description: 'The name of the study'
+          type: string
+        creator:
+          description: 'A comma-separated list of the study creators'
+          type: string
+        flows:
+          $ref: '#/components/schemas/Study_study_flows'
+        creation_date:
+          description: 'The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss'
+          type: string
+        alias:
+          description: 'The alias of the study'
+          type: string
+        tag:
+          $ref: '#/components/schemas/Study_study_tag'
+        main_entity_type:
+          description: 'The type of entity collected in the study (e.g. task or run)'
+          type: string
+        data:
+          $ref: '#/components/schemas/Study_study_data'
+        id:
+          description: 'The ID of the study'
+          type: string
+      type: object
+    inline_response_200_25_upload_study:
+      properties:
+        id:
+          description: 'ID of the uploaded study, a positive integer'
+          type: string
+      type: object
+    EvaluationList_evaluations_evaluation:
+      properties:
+        function:
+          description: 'The name of the evaluation function'
+          type: string
+        task_id:
+          description: 'The ID of the tasks solved by this run'
+          type: string
+        run_id:
+          description: 'The run ID'
+          type: string
+        array_data:
+          description: 'For structured evaluation measures, an array of evaluation values (e.g. per-class predictions, evaluation matrices,...)'
+          type: string
+        value:
+          description: 'The outcome of the evaluation'
+          type: string
+        flow_id:
+          description: 'The ID of the flow used by this run'
+          type: string
+      type: object
+    EstimationProcedureList:
+      properties:
+        estimationprocedures:
+          $ref: '#/components/schemas/EstimationProcedureList_estimationprocedures'
+      type: object
+    DataQualityList_data_qualities_list:
+      properties:
+        quality:
+          type: array
+          items:
+            type: string
+      type: object
+    Study_study_flows:
+      properties:
+        flow_id:
+          type: array
+          items:
+            type: string
+      type: object
+    StudyList_study_list_study:
+      properties:
+        alias:
+          description: 'The alias of the study'
+          type: string
+        creation_date:
+          description: 'The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss'
+          type: string
+        creator:
+          description: 'A comma-separated list of the study creators'
+          type: string
+        id:
+          description: 'The ID of the study'
+          type: string
+        name:
+          description: 'The name of the study'
+          type: string
+      type: object
+    Run_run_description_input_data:
+      properties:
+        dataset:
+          $ref: '#/components/schemas/Run_run_description_input_data_dataset'
+      type: object
+    TaskTypeList_task_types_task_type:
+      properties:
+        description:
+          description: 'A description of the task type'
+          type: string
+        creator:
+          description: 'A comma-separated list of the task type creators'
+          type: string
+        id:
+          description: 'The ID of the task type'
+          type: string
+        name:
+          description: 'The name of the task type'
+          type: string
+      type: object
+    EvaluationMeasureList_evaluation_measures:
+      properties:
+        measures:
+          $ref: '#/components/schemas/EvaluationMeasureList_evaluation_measures_measures'
+      type: object
+    TaskType_predictions:
+      properties:
+        feature:
+          type: array
+          items:
+            $ref: '#/components/schemas/Task_task_description_predictions_feature'
+        format:
+          description: 'The format of the predictions, e.g. ARFF'
+          type: string
+      type: object
+    Flow_flow_description_parameter:
+      properties:
+        default_value:
+          description: 'The default value of the parameter'
+          type: string
+        name:
+          description: 'The name of the parameter'
+          type: string
+        data_type:
+          description: 'The data type of the parameter'
+          type: string
+        description:
+          description: 'A description of the parameter'
+          type: string
+      type: object
+    RunTrace_trace_trace_iteration:
+      properties:
+        setup_string:
+          description: 'A JSON representation of the setup (configuration)'
+          type: string
+        repeat:
+          description: 'The number of the repeat in the outer cross-valudation'
+          type: string
+        selected:
+          description: 'Whether this setup was selected as the best one (true or false)'
+          type: string
+        iteration:
+          description: 'A number of the optimization iteration'
+          type: string
+        fold:
+          description: 'The number of the fold in the inner cross-validation'
+          type: string
+        evaluation:
+          description: 'The evaluation score of the setup'
+          type: string
+      type: object
+    EvaluationMeasureList:
+      properties:
+        evaluation_measures:
+          $ref: '#/components/schemas/EvaluationMeasureList_evaluation_measures'
+      type: object
+    SetupList_setups:
+      properties:
+        setup:
+          type: array
+          items:
+            $ref: '#/components/schemas/SetupList_setups_setup'
+      type: object
+    TaskList_task:
+      properties:
+        task:
+          type: array
+          items:
+            $ref: '#/components/schemas/TaskList_task_task'
+      type: object
+    SetupList_setups_setup:
+      properties:
+        setup_id:
+          description: 'The setup ID'
+          type: string
+        parameter:
+          type: array
+          items:
+            $ref: '#/components/schemas/SetupList_setups_parameter'
+        flow_id:
+          description: 'The ID of the flow used by this run'
+          type: string
+      type: object
+    Task_task_description_estimation_procedure_parameter:
+      properties:
+        name:
+          description: 'The name of the parameter'
+          type: string
+        value:
+          description: 'The value of the parameter'
+          type: string
+      type: object
+    RunList_runs:
+      properties:
+        run:
+          type: array
+          items:
+            $ref: '#/components/schemas/RunList_runs_run'
+      type: object
+    Study_study_tasks:
+      properties:
+        task_id:
+          type: array
+          items:
+            type: string
+      type: object
+    inline_response_200_1_upload_data_set:
+      properties:
+        id:
+          description: 'ID of the uploaded dataset, a positive integer'
+          type: string
+      type: object
+    Setup_setup_parameters_parameter_setting:
+      properties:
+        default_value:
+          description: 'The default value of the parameter used'
+          type: string
+        value:
+          description: 'The value of the parameter used'
+          type: string
+        data_type:
+          description: 'The data type of the hyperparameter value'
+          type: string
+        full_name:
+          description: 'The full name of the hyperparameter'
+          type: string
+        parameter_name:
+          description: 'The short name of the hyperparameter'
+          type: string
+      type: object
+    DataQualities:
+      properties:
+        data_qualities:
+          $ref: '#/components/schemas/DataQualities_data_qualities'
+      type: object
diff --git a/openapi/unusedSchemas.py b/openapi/unusedSchemas.py
new file mode 100644
index 000000000..5cf8cc3e1
--- /dev/null
+++ b/openapi/unusedSchemas.py
@@ -0,0 +1,55 @@
+import json
+
+with open('swagger.json',) as f:
+	data = json.load(f)
+
+schemasAvailable = list()
+schemasReferenced = set()
+for k,v in data.items():
+	if k=="components":
+		for k2,v2 in v["schemas"].items():
+			schemasAvailable.append(k2)
+	elif k=="paths":
+		for k2,v2 in v.items():
+			for k3,v3 in v2.items():
+				if "parameters" in v3:
+					for x in v3["parameters"]:
+						if "$ref" in x["schema"]:
+							schemasReferenced.add(x["schema"]["$ref"][21:])
+				for k4,v4 in v3["responses"].items():
+					cur = v4["content"]["application/json"]["schema"]
+					if "$ref" in cur:
+						schemasReferenced.add(cur["$ref"][21:])
+					if "properties" in cur:
+						for x,y in cur["properties"].items():
+							schemasReferenced.add(y["$ref"][21:])
+
+schemasIndirectlyReferenced = set()
+schemasIndirectlyReferenced.update(schemasReferenced)
+
+schemasReferencedQueue = list(schemasReferenced)
+
+while len(schemasReferencedQueue) > 0:
+	ele = schemasReferencedQueue[0]
+	schemasReferencedQueue.remove(ele)
+	schemasIndirectlyReferenced.add(ele)
+	for k,v in data["components"]["schemas"][ele]["properties"].items():
+		if "$ref" in v:
+			schemaName = v["$ref"][21:]
+			if not (schemaName in schemasIndirectlyReferenced):
+				schemasReferencedQueue.append(schemaName)
+		if "items" in v and "$ref" in v["items"]:
+			schemaName = v["items"]["$ref"][21:]
+			if not (schemaName in schemasIndirectlyReferenced):
+				schemasReferencedQueue.append(schemaName)
+
+print("Schemas listed: ", len(schemasAvailable))
+print("Schemas directly referenced: ", len(schemasReferenced))
+print("Schemas (in)directly referenced in total: ", len(schemasIndirectlyReferenced))
+
+for e in schemasIndirectlyReferenced:
+	schemasAvailable.remove(e)
+
+print("Found ", len(schemasAvailable), " unused schemas: ")
+for s in schemasAvailable:
+	print(s)
\ No newline at end of file
diff --git a/openml_OS/controllers/Api.php b/openml_OS/controllers/Api.php
index 43eb01b8b..12a529d4e 100644
--- a/openml_OS/controllers/Api.php
+++ b/openml_OS/controllers/Api.php
@@ -1,12 +1,2132 @@
 <?php
 require_once('Api_new.php');
 
-class Api extends Api_new {
+/**
+ * @OA\Info(title="OpenML API", version="1.0.0", description="REST API for sharing, organizing and reusing machine learning datasets, code, and experiments. Follows a predictive URL scheme from endpoint https://www.openml.org/api/v1/json (or /xml). You need to add your api_key for every call except GET calls. The API key can be found in your profile on openml.org."
+ * )
+ */
+ 
+ /**
+ * @OA\Server(
+ *	url="https://openml.org/api/v1/json",
+ *  description="The current OpenML API"
+ * )
+ * @OA\Server(
+ *	url="https://test.openml.org/api/v1/json",
+ *  description="The test OpenML API"
+ * )
+ */
+class Api extends Api_new
+{
 
-  private $version;
+    private $version;
 
-  function __construct() {
-    parent::__construct();
-  }
+    function __construct()
+    {
+        parent::__construct();
+    }
 }
+
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_16_flow_untag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the untagged setup",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationMeasureList_evaluation_measures_measures",
+ *	  @OA\Property(
+ *        property="measure",
+ *        type="array",
+ *        description="The evaluation measure names",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_10_flow_exists",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The id of the flow with the given name and (external) version",
+ *    ),
+ *	@OA\Property(
+ *        property="exists",
+ *        type="string",
+ *        description="true or false",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Data_data_set_description",
+ *	@OA\Property(
+ *        property="default_target_attribute",
+ *        type="string",
+ *        description="For tabular data, the name of the column that is typically used as the target attribute for that data set",
+ *    ),
+ *	@OA\Property(
+ *        property="upload_date",
+ *        type="string",
+ *        description="The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+ *    ),
+ *	@OA\Property(
+ *        property="version_label",
+ *        type="string",
+ *        description="The version of the dataset, as defined by the uploader, for reference. Can be any format as long as it is unique.",
+ *    ),
+ *	@OA\Property(
+ *        property="description",
+ *        type="string",
+ *        description="Wiki description of the dataset, in (Git flavoured) markdown format",
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="Data format, for instance ARFF",
+ *    ),
+ *	@OA\Property(
+ *        property="url",
+ *        type="string",
+ *        description="The URL where the data can be downloaded",
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        type="array",
+ *        description="Tags added by OpenML users. Includes study tags in the form `study_1`",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="visibility",
+ *        type="string",
+ *        description="Who can see the dataset. For instance `public`.",
+ *    ),
+ *	@OA\Property(
+ *        property="md5_checksum",
+ *        type="string",
+ *        description="Checksum to verify downloads of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="version",
+ *        type="string",
+ *        description="The version of the dataset, set by OpenML. A positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="status",
+ *        type="string",
+ *        description="active, in_preparation, or deactivated",
+ *    ),
+ *	@OA\Property(
+ *        property="file_id",
+ *        type="string",
+ *        description="The ID of the dataset file stored on the OpenML server",
+ *    ),
+ *	@OA\Property(
+ *        property="licence",
+ *        type="string",
+ *        description="The licence granted for using the dataset, for instance Public or CC-BY",
+ *    ),
+ *	@OA\Property(
+ *        property="original_data_url",
+ *        type="string",
+ *        description="The URL where the original data is hosted.",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the dataset, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataList_data_dataset",
+ *	@OA\Property(
+ *        property="did",
+ *        type="string",
+ *        description="The dataset ID",
+ *    ),
+ *	@OA\Property(
+ *        property="status",
+ *        type="string",
+ *        description="The dataset status, either in_preparation, active, or deactivated",
+ *    ),
+ *	@OA\Property(
+ *        property="quality",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/DataList_data_quality"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The dataset name",
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="The data format of the dataset, e.g. ARFF",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_estimation_procedure",
+ *	@OA\Property(
+ *        property="parameter",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Task_task_description_estimation_procedure_parameter"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="type",
+ *        type="string",
+ *        description="The type of procedure, e.g. crossvalidation",
+ *    ),
+ *	@OA\Property(
+ *        property="data_splits_url",
+ *        type="string",
+ *        description="The url where the data splits can be downloaded",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataList_data_quality",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the property",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the property",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="FlowList_flows_flow",
+ *	@OA\Property(
+ *        property="full_name",
+ *        type="string",
+ *        description="The full flow name (name + internal version number)",
+ *    ),
+ *	@OA\Property(
+ *        property="external_version",
+ *        type="string",
+ *        description="The external flow version",
+ *    ),
+ *	@OA\Property(
+ *        property="version",
+ *        type="string",
+ *        description="The internal flow version",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader",
+ *        type="string",
+ *        description="The ID of the person who uploaded the flow",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The flow ID",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The flow name",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_input_data_dataset",
+ *	@OA\Property(
+ *        property="did",
+ *        type="string",
+ *        description="The id of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="url",
+ *        type="string",
+ *        description="The download url of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationList_evaluations",
+ *	  @OA\Property(
+ *        property="evaluation",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/EvaluationList_evaluations_evaluation"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_14_study_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted setup, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataUnprocessed",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data_unprocessed",
+ *        ref="#/components/schemas/DataUnprocessed_data_unprocessed",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataUnprocessed_data_unprocessed_dataset",
+ *	@OA\Property(
+ *        property="did",
+ *        type="string",
+ *        description="ID of the dataset a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="status",
+ *        type="string",
+ *        description="Status of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="version",
+ *        type="string",
+ *        description="Version of the dataset, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="The dataset format, e.g. ARFF",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EstimationProcedureList_estimationprocedures",
+ *	@OA\Property(
+ *        property="estimationprocedure",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/EstimationProcedureList_estimationprocedures_estimationprocedure"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskTypeList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="task_types",
+ *        ref="#/components/schemas/TaskTypeList_task_types",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_output_data_evaluation",
+ *	@OA\Property(
+ *        property="array_data",
+ *        type="string",
+ *        description="For composite evaluation measures (e.g. per-class measures, confusion matrix), a string (JSON) representation of the   evaluation.",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the evaluation measure",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The result of the evaluation",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The id of the code used to compute this evaluation method",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskList_task_quality",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the quality",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the quality",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description",
+ *	@OA\Property(
+ *        property="input",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Task_task_description_input"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="task_type",
+ *        type="string",
+ *        description="The type of the task, e.g. Supervised Classification",
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        type="array",
+ *        description="Tags added by OpenML uers. Includes study tags in the form 'study_1'",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="ID of the task, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="output",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Task_task_description_output"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_9_upload_flow",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the uploaded flow, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Flow_flow_description",
+ *	@OA\Property(
+ *        property="upload_date",
+ *        type="string",
+ *        description="The datetime that the flow was uploaded, format yyyy-MM-dd HH:mm:ss",
+ *    ),
+ *	@OA\Property(
+ *        property="description",
+ *        type="string",
+ *        description="Wiki description of the flow, in (Git flavoured) markdown format",
+ *    ),
+ *	@OA\Property(
+ *        property="language",
+ *        type="string",
+ *        description="The programming language the flow is written in.",
+ *    ),
+ *	@OA\Property(
+ *        property="parameter",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Flow_flow_description_parameter"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        type="array",
+ *        description="Tags added by OpenML users. Includes study tags in the form `study_1`",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="version",
+ *        type="string",
+ *        description="The version of the flow, set by OpenML. A positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="version_label",
+ *        type="string",
+ *        description="The version of the flow, as defined by the uploader, for reference. Can be any format as long as it is unique.",
+ *    ),
+ *	@OA\Property(
+ *        property="dependencies",
+ *        type="string",
+ *        description="The libraries that this flow depends on, and their version numbers.",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader",
+ *        type="string",
+ *        description="The uploader of the flow",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the flow, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the flow",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task",
+ *    type="object",
+ *	@OA\Property(
+ *        property="task_description",
+ *        ref="#/components/schemas/Task_task_description",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Setup",
+ *    type="object",
+ *	@OA\Property(
+ *        property="setup_parameters",
+ *        ref="#/components/schemas/Setup_setup_parameters",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="FlowList_flows",
+ *	@OA\Property(
+ *        property="flow",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/FlowList_flows_flow"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunTrace",
+ *    type="object",
+ *	@OA\Property(
+ *        property="trace",
+ *        ref="#/components/schemas/RunTrace_trace",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_predictions_feature",
+ *	@OA\Property(
+ *        property="type",
+ *        type="string",
+ *        description="The type of the prediction feature, e.g. integer",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the prediction feature, e.g. row_id",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_24_study_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted study, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_5_upload_task",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the uploaded task, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_13_flow_untag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the untagged flow",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="UserList_users",
+ *	@OA\Property(
+ *        property="user",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/UserList_users_user"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataFeatures",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data_features",
+ *        ref="#/components/schemas/DataFeatures_data_features",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_predictions",
+ *	@OA\Property(
+ *        property="feature",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Task_task_description_predictions_feature"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="The fromat of the predictions, e.g. ARFF",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_15_flow_tag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the tagged setup",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Setup_setup_parameters",
+ *	@OA\Property(
+ *        property="parameter_setting",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Setup_setup_parameters_parameter_setting"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="ID of the flow, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data",
+ *        ref="#/components/schemas/DataList_data",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_26_study_attach",
+ *	@OA\Property(
+ *        property="linked_entities",
+ *        type="string",
+ *        description="The number of linked entities",
+ *    ),
+ *	@OA\Property(
+ *        property="main_entity_type",
+ *        type="string",
+ *        description="Main entity type of the of the study",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the study, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_2_data_tag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the tagged dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_23_upload_flow",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the run with the trace, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationRequest_evaluation_request",
+ *	@OA\Property(
+ *        property="run",
+ *        ref="#/components/schemas/EvaluationRequest_evaluation_request_run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run",
+ *    type="object",
+ *	@OA\Property(
+ *        property="run_description",
+ *        ref="#/components/schemas/Run_run_description",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_6_task_tag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the tagged task",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="runs",
+ *        ref="#/components/schemas/RunList_runs",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_17_data_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted run, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_input",
+ *	@OA\Property(
+ *        property="data_set",
+ *        ref="#/components/schemas/Task_task_description_data_set",
+ *    ),
+ *	@OA\Property(
+ *        property="cost_matrix",
+ *        type="array",
+ *        description="The cost matrix, indicating the costs for each type of misclassification",
+ *		  @OA\Items(
+ *			type="array",
+ *          @OA\Items(
+ *              type="integer",
+ *              format="int64"
+ *          )
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the input, e.g. source_data",
+ *    ),
+ *	@OA\Property(
+ *        property="evaluation_measures",
+ *        ref="#/components/schemas/Task_task_description_evaluation_measures",
+ *    ),
+ *	@OA\Property(
+ *        property="estimation_procedure",
+ *        ref="#/components/schemas/Task_task_description_estimation_procedure",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="UserList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="users",
+ *        ref="#/components/schemas/UserList_users",
+ *    )
+ *
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study_runs",
+ *	@OA\Property(
+ *        property="run_id",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataFeatures_data_features_feature",
+ *	@OA\Property(
+ *        property="index",
+ *        type="string",
+ *        description="Feature index",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="Feature name",
+ *    ),
+ *	@OA\Property(
+ *        property="data_type",
+ *        type="string",
+ *        description="Feature data type",
+ *    ),
+ *	@OA\Property(
+ *        property="is_target",
+ *        type="string",
+ *        description="Whether this feature is seen as a target feature",
+ *    ),
+ *	@OA\Property(
+ *        property="is_ignore",
+ *        type="string",
+ *        description="Whether this feature should be ignored in modelling (e.g. every value is unique)",
+ *    ),
+ *	@OA\Property(
+ *        property="is_row_identifier",
+ *        type="string",
+ *        description="Whether this feature is a row identifier",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_21_upload_flow",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the evaluated run, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskType",
+ *    type="object",
+ *	@OA\Property(
+ *        property="description",
+ *        type="string",
+ *        description="A description of the task type",
+ *    ),
+ *	@OA\Property(
+ *        property="date",
+ *        type="string",
+ *        description="The date when the task type was created",
+ *    ),
+ *	@OA\Property(
+ *        property="output",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskType_output"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="contributor",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="input",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskType_input"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the task type, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the task type, e.g. Supervised Classification",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_output_data_file",
+ *	@OA\Property(
+ *        property="did",
+ *        type="string",
+ *        description="The id of the uploaded file",
+ *    ),
+ *	@OA\Property(
+ *        property="file_id",
+ *        type="string",
+ *        description="The reference id of the uploaded file, for downloading afterward",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the uploaded file (e.g., description, predictions, model,...)",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskType_input",
+ *	@OA\Property(
+ *        property="data_set",
+ *        ref="#/components/schemas/Task_task_description_data_set",
+ *    ),
+ *	@OA\Property(
+ *        property="cost_matrix",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="array",
+ *          @OA\Items(
+ *              type="integer",
+ *              format="int64"
+ *          )
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the input, e.g. source_data",
+ *    ),
+ *	@OA\Property(
+ *        property="evaluation_measures",
+ *        ref="#/components/schemas/Task_task_description_evaluation_measures",
+ *    ),
+ *	@OA\Property(
+ *        property="estimation_procedure",
+ *        ref="#/components/schemas/Task_task_description_estimation_procedure",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_output_data",
+ *	@OA\Property(
+ *        property="evaluation",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Run_run_description_output_data_evaluation"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="file",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Run_run_description_output_data_file"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_3_data_untag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the untagged dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_output",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the output, e.g. predictions",
+ *    ),
+ *	@OA\Property(
+ *        property="predictions",
+ *        ref="#/components/schemas/Task_task_description_predictions",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataQualityList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data_qualities_list",
+ *        ref="#/components/schemas/DataQualityList_data_qualities_list",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataUnprocessed_data_unprocessed",
+ *	@OA\Property(
+ *        property="dataset",
+ *        ref="#/components/schemas/DataUnprocessed_data_unprocessed_dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study",
+ *    type="object",
+ *	@OA\Property(
+ *        property="study",
+ *        ref="#/components/schemas/Study_study",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataList_data",
+ *	@OA\Property(
+ *        property="dataset",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/DataList_data_dataset"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="UserList_users_user",
+ *	@OA\Property(
+ *        property="username",
+ *        type="string",
+ *        description="The full user name",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The user ID",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskType_output",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the output, e.g. predictions",
+ *    ),
+ *	@OA\Property(
+ *        property="predictions",
+ *        ref="#/components/schemas/TaskType_predictions",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunList_runs_run",
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="The ID of the task solved by this run",
+ *    ),
+ *	@OA\Property(
+ *        property="run_id",
+ *        type="string",
+ *        description="The run ID",
+ *    ),
+ *	@OA\Property(
+ *        property="error_message",
+ *        type="string",
+ *        description="Error message generated by the run (if any)",
+ *    ),
+ *	@OA\Property(
+ *        property="setup_id",
+ *        type="string",
+ *        description="Ignore (internal representation of the parameter setting)",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader",
+ *        type="string",
+ *        description="The ID of the person uploading this run",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The ID of the flow used in this run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataFeatures_data_features",
+ *	@OA\Property(
+ *        property="feature",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/DataFeatures_data_features_feature"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataQualities_data_qualities_quality",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the dataset quality measures",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value for this dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_12_flow_tag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the tagged flow",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_8_flow_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted flow, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataQualities_data_qualities",
+ *	@OA\Property(
+ *        property="quality",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/DataQualities_data_qualities_quality"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationRequest",
+ *    type="object",
+ *	@OA\Property(
+ *        property="evaluation_request",
+ *        ref="#/components/schemas/EvaluationRequest_evaluation_request",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_18_upload_flow",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the uploaded run, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="SetupList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="setups",
+ *        ref="#/components/schemas/SetupList_setups",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="task",
+ *        ref="#/components/schemas/TaskList_task",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_4_task_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted task, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_data_set",
+ *	@OA\Property(
+ *        property="data_set_id",
+ *        type="string",
+ *        description="The id of the dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="target_feature",
+ *        type="string",
+ *        description="The name of the target feature for this task",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="evaluations",
+ *        ref="#/components/schemas/EvaluationList_evaluations",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description",
+ *	@OA\Property(
+ *        property="setup_string",
+ *        type="string",
+ *        description="Configuration of the flow as a string, to be interpreted by the flow, its library, or command line interface.",
+ *    ),
+ *	@OA\Property(
+ *        property="task_type",
+ *        type="string",
+ *        description="The type of task solved by this run (e.g., classification)",
+ *    ),
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="The id of the task solved by this run",
+ *    ),
+ *	@OA\Property(
+ *        property="task_evaluation_measure",
+ *        type="string",
+ *        description="The evaluation measure that is supposed to be optimized in the task, if any",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader_name",
+ *        type="string",
+ *        description="The name of the uploader of the run",
+ *    ),
+ *	@OA\Property(
+ *        property="input_data",
+ *        ref="#/components/schemas/Run_run_description_input_data",
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  ),
+ *        description="Tags added by OpenML users. Includes study tags in the form `study_1`",
+ *    ),
+ *	@OA\Property(
+ *        property="output_data",
+ *        ref="#/components/schemas/Run_run_description_output_data",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader",
+ *        type="string",
+ *        description="The uploader of the run",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The id of the flow used in this run",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_name",
+ *        type="string",
+ *        description="The name of the flow used in this run",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the run, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="parameter_setting",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Run_run_description_parameter_setting"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="FlowList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="flows",
+ *        ref="#/components/schemas/FlowList_flows",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_19_run_tag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the tagged run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study_data",
+ *	@OA\Property(
+ *        property="data_id",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_data_delete",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the deleted dataset, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunTrace_trace",
+ *	@OA\Property(
+ *        property="trace_iteration",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/RunTrace_trace_trace_iteration"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="run_id",
+ *        type="string",
+ *        description="run ID",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_7_task_untag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the untagged task",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Data",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data_set_description",
+ *        ref="#/components/schemas/Data_data_set_description",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_evaluation_measures",
+ *	@OA\Property(
+ *        property="evaluation_measure",
+ *        type="string",
+ *        description="The evaluation measure to optimize in this task",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="SetupList_setups_parameter",
+ *	@OA\Property(
+ *        property="default_value",
+ *        type="string",
+ *        description="The parameter's default value",
+ *    ),
+ *	@OA\Property(
+ *        property="data_type",
+ *        type="string",
+ *        description="The parameter's data type",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The parameter value in this setup",
+ *    ),
+ *	@OA\Property(
+ *        property="parameter_name",
+ *        type="string",
+ *        description="The parameter's short name",
+ *    ),
+ *	@OA\Property(
+ *        property="full_name",
+ *        type="string",
+ *        description="The parameter's full name",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The (sub)flow ID",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_name",
+ *        type="string",
+ *        description="The (sub)flow name",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The parameter ID",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskList_task_input",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the input",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the input",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="StudyList_study_list",
+ *	@OA\Property(
+ *        property="study",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/StudyList_study_list_study"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study_tag",
+ *	@OA\Property(
+ *        property="write_access",
+ *        type="string",
+ *        description="The write access level of the study (e.g. public)",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the study (e.g. study_1)",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Flow",
+ *    type="object",
+ *	@OA\Property(
+ *        property="flow_description",
+ *        ref="#/components/schemas/Flow_flow_description",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_parameter_setting",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the parameter",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the parameter used",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationRequest_evaluation_request_run",
+ *	@OA\Property(
+ *        property="setup_id",
+ *        type="string",
+ *        description="ID of the setup, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="upload_time",
+ *        type="string",
+ *        description="The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+ *    ),
+ *	@OA\Property(
+ *        property="uploader",
+ *        type="string",
+ *        description="ID of the uploader, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="ID of the task, a positive integer",
+ *    ),
+ *	@OA\Property(
+ *        property="run_id",
+ *        type="string",
+ *        description="ID of the run, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Error",
+ *    type="object",
+ *	@OA\Property(
+ *        property="message",
+ *        type="string",
+ *    ),
+ *	@OA\Property(
+ *        property="code",
+ *        type="integer",
+ *    ),
+ *	@OA\Property(
+ *        property="additional_message",
+ *        type="string",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EstimationProcedureList_estimationprocedures_estimationprocedure",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The estimation procedure name, e.g. '10 fold Crossvalidation'",
+ *    ),
+ *	@OA\Property(
+ *        property="folds",
+ *        type="string",
+ *        description="The number of cross-validation folds, e.g. '10'",
+ *    ),
+ *	@OA\Property(
+ *        property="stratified_sampling",
+ *        type="string",
+ *        description="Whether or not the sampling is stratified, 'true' or 'false'",
+ *    ),
+ *	@OA\Property(
+ *        property="ttid",
+ *        type="string",
+ *        description="The task type ID",
+ *    ),
+ *	@OA\Property(
+ *        property="repeats",
+ *        type="string",
+ *        description="The number of repeats, e.g. '10'",
+ *    ),
+ *	@OA\Property(
+ *        property="type",
+ *        type="string",
+ *        description="The estimation procedure type, e.g. 'crossvalidation'",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The estimation procedure ID",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskTypeList_task_types",
+ *	@OA\Property(
+ *        property="task_type",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskTypeList_task_types_task_type"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskList_task_task",
+ *	@OA\Property(
+ *        property="status",
+ *        type="string",
+ *        description="The status of the source dataset, active, in_preparation, or deactivated",
+ *    ),
+ *	@OA\Property(
+ *        property="task_type",
+ *        type="string",
+ *        description="The type of task (e.g. Supervised Classificationr)",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the source dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="The ID of the task",
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="The format of the source dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="did",
+ *        type="string",
+ *        description="The id of the source dataset",
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="input",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskList_task_input"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="quality",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskList_task_quality"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_20_run_untag",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the untagged run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="StudyList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="study_list",
+ *        ref="#/components/schemas/StudyList_study_list",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study",
+ *	@OA\Property(
+ *        property="runs",
+ *        ref="#/components/schemas/Study_study_runs",
+ *    ),
+ *	@OA\Property(
+ *        property="tasks",
+ *        ref="#/components/schemas/Study_study_tasks",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the study",
+ *    ),
+ *	@OA\Property(
+ *        property="creator",
+ *        type="string",
+ *        description="A comma-separated list of the study creators",
+ *    ),
+ *	@OA\Property(
+ *        property="flows",
+ *        ref="#/components/schemas/Study_study_flows",
+ *    ),
+ *	@OA\Property(
+ *        property="creation_date",
+ *        type="string",
+ *        description="The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+ *    ),
+ *	@OA\Property(
+ *        property="alias",
+ *        type="string",
+ *        description="The alias of the study",
+ *    ),
+ *	@OA\Property(
+ *        property="tag",
+ *        ref="#/components/schemas/Study_study_tag",
+ *    ),
+ *	@OA\Property(
+ *        property="main_entity_type",
+ *        type="string",
+ *        description="The type of entity collected in the study (e.g. task or run)",
+ *    ),
+ *	@OA\Property(
+ *        property="data",
+ *        ref="#/components/schemas/Study_study_data",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The ID of the study",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_25_upload_study",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the uploaded study, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationList_evaluations_evaluation",
+ *	@OA\Property(
+ *        property="function",
+ *        type="string",
+ *        description="The name of the evaluation function",
+ *    ),
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="string",
+ *        description="The ID of the tasks solved by this run",
+ *    ),
+ *	@OA\Property(
+ *        property="run_id",
+ *        type="string",
+ *        description="The run ID",
+ *    ),
+ *	@OA\Property(
+ *        property="array_data",
+ *        type="string",
+ *        description="For structured evaluation measures, an array of evaluation values (e.g. per-class predictions, evaluation matrices,...)",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The outcome of the evaluation",
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The ID of the flow used by this run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EstimationProcedureList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="estimationprocedures",
+ *        ref="#/components/schemas/EstimationProcedureList_estimationprocedures",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataQualityList_data_qualities_list",
+ *	@OA\Property(
+ *        property="quality",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study_flows",
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="StudyList_study_list_study",
+ *	@OA\Property(
+ *        property="alias",
+ *        type="string",
+ *        description="The alias of the study",
+ *    ),
+ *	@OA\Property(
+ *        property="creation_date",
+ *        type="string",
+ *        description="The datetime that the dataset was uploaded, format yyyy-MM-dd HH:mm:ss",
+ *    ),
+ *	@OA\Property(
+ *        property="creator",
+ *        type="string",
+ *        description="A comma-separated list of the study creators",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The ID of the study",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the study",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Run_run_description_input_data",
+ *	@OA\Property(
+ *        property="dataset",
+ *        ref="#/components/schemas/Run_run_description_input_data_dataset",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskTypeList_task_types_task_type",
+ *	@OA\Property(
+ *        property="description",
+ *        type="string",
+ *        description="A description of the task type",
+ *    ),
+ *	@OA\Property(
+ *        property="creator",
+ *        type="string",
+ *        description="A comma-separated list of the task type creators",
+ *    ),
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="The ID of the task type",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the task type",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationMeasureList_evaluation_measures",
+ *	@OA\Property(
+ *        property="measures",
+ *        ref="#/components/schemas/EvaluationMeasureList_evaluation_measures_measures",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskType_predictions",
+ *	@OA\Property(
+ *        property="feature",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/Task_task_description_predictions_feature"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="format",
+ *        type="string",
+ *        description="The format of the predictions, e.g. ARFF",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Flow_flow_description_parameter",
+ *	@OA\Property(
+ *        property="default_value",
+ *        type="string",
+ *        description="The default value of the parameter",
+ *    ),
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the parameter",
+ *    ),
+ *	@OA\Property(
+ *        property="data_type",
+ *        type="string",
+ *        description="The data type of the parameter",
+ *    ),
+ *	@OA\Property(
+ *        property="description",
+ *        type="string",
+ *        description="A description of the parameter",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunTrace_trace_trace_iteration",
+ *	@OA\Property(
+ *        property="setup_string",
+ *        type="string",
+ *        description="A JSON representation of the setup (configuration)",
+ *    ),
+ *	@OA\Property(
+ *        property="repeat",
+ *        type="string",
+ *        description="The number of the repeat in the outer cross-valudation",
+ *    ),
+ *	@OA\Property(
+ *        property="selected",
+ *        type="string",
+ *        description="Whether this setup was selected as the best one (true or false)",
+ *    ),
+ *	@OA\Property(
+ *        property="iteration",
+ *        type="string",
+ *        description="A number of the optimization iteration",
+ *    ),
+ *	@OA\Property(
+ *        property="fold",
+ *        type="string",
+ *        description="The number of the fold in the inner cross-validation",
+ *    ),
+ *	@OA\Property(
+ *        property="evaluation",
+ *        type="string",
+ *        description="The evaluation score of the setup",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="EvaluationMeasureList",
+ *    type="object",
+ *	@OA\Property(
+ *        property="evaluation_measures",
+ *        ref="#/components/schemas/EvaluationMeasureList_evaluation_measures",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="SetupList_setups",
+ *	@OA\Property(
+ *        property="setup",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/SetupList_setups_setup"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="TaskList_task",
+ *	@OA\Property(
+ *        property="task",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/TaskList_task_task"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="SetupList_setups_setup",
+ *	@OA\Property(
+ *        property="setup_id",
+ *        type="string",
+ *        description="The setup ID",
+ *    ),
+ *	@OA\Property(
+ *        property="parameter",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/SetupList_setups_parameter"
+ *		  )
+ *    ),
+ *	@OA\Property(
+ *        property="flow_id",
+ *        type="string",
+ *        description="The ID of the flow used by this run",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Task_task_description_estimation_procedure_parameter",
+ *	@OA\Property(
+ *        property="name",
+ *        type="string",
+ *        description="The name of the parameter",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the parameter",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="RunList_runs",
+ *	@OA\Property(
+ *        property="run",
+ *        type="array",
+ *		  @OA\Items(
+ *			ref="#/components/schemas/RunList_runs_run"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Study_study_tasks",
+ *	@OA\Property(
+ *        property="task_id",
+ *        type="array",
+ *		  @OA\Items(
+ *			type="string"
+ *		  )
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="inline_response_200_1_upload_data_set",
+ *	@OA\Property(
+ *        property="id",
+ *        type="string",
+ *        description="ID of the uploaded dataset, a positive integer",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="Setup_setup_parameters_parameter_setting",
+ *	@OA\Property(
+ *        property="default_value",
+ *        type="string",
+ *        description="The default value of the parameter used",
+ *    ),
+ *	@OA\Property(
+ *        property="value",
+ *        type="string",
+ *        description="The value of the parameter used",
+ *    ),
+ *	@OA\Property(
+ *        property="data_type",
+ *        type="string",
+ *        description="The data type of the hyperparameter value",
+ *    ),
+ *	@OA\Property(
+ *        property="full_name",
+ *        type="string",
+ *        description="The full name of the hyperparameter",
+ *    ),
+ *	@OA\Property(
+ *        property="parameter_name",
+ *        type="string",
+ *        description="The short name of the hyperparameter",
+ *    ),
+ *)
+ */
+/**
+ * @OA\Schema(
+ *    schema="DataQualities",
+ *    type="object",
+ *	@OA\Property(
+ *        property="data_qualities",
+ *        ref="#/components/schemas/DataQualities_data_qualities",
+ *    ),
+ *)
+ */
+
 ?>
diff --git a/openml_OS/models/api/v1/Api_data.php b/openml_OS/models/api/v1/Api_data.php
index c2c52de77..3f7779f22 100644
--- a/openml_OS/models/api/v1/Api_data.php
+++ b/openml_OS/models/api/v1/Api_data.php
@@ -1,4 +1,5 @@
 <?php
+
 class Api_data extends MY_Api_Model {
 
   protected $version = 'v1';
@@ -30,7 +31,6 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->outputFormat = $format;
 
     $getpost = array('get','post');
-
     if (count($segments) >= 1 && $segments[0] == 'list') {
       array_shift($segments);
       $this->data_list($segments);
@@ -43,6 +43,7 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     $order_values = array('random', 'normal');
+
     if (count($segments) == 3 && $segments[0] == 'unprocessed' && is_numeric($segments[1]) && in_array($segments[2], $order_values)) {
       $this->data_unprocessed($segments[1], $segments[2]);
       return;
@@ -79,6 +80,16 @@ function bootstrap($format, $segments, $request_type, $user_id) {
       return;
     }
 
+    if ( $segments[0] == 'fork' && $request_type == 'post') {
+      $this->data_fork();
+      return;
+    }
+
+    if ( $segments[0] == 'edit' && $request_type == 'post') {
+      $this->data_edit();
+      return;
+    }
+
     if (count($segments) == 2 && $segments[0] == 'features' && is_numeric($segments[1]) && in_array($request_type, $getpost)) {
       $this->data_features($segments[1]);
       return;
@@ -121,12 +132,12 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     if (count($segments) == 1 && $segments[0] == 'tag' && $request_type == 'post') {
-      $this->entity_tag_untag('dataset', $this->input->post('data_id'), $this->input->post('tag'), false, 'data');
+      $this->data_tag($this->input->post('data_id'), $this->input->post('tag'));
       return;
     }
 
     if (count($segments) == 1 && $segments[0] == 'untag' && $request_type == 'post') {
-      $this->entity_tag_untag('dataset', $this->input->post('data_id'), $this->input->post('tag'), true, 'data');
+      $this->data_untag($this->input->post('data_id'), $this->input->post('tag'));
       return;
     }
 
@@ -143,21 +154,244 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError(100, $this->version);
   }
 
+  /**
+   *@OA\Post(
+   *	path="/data/tag",
+   *	tags={"data"},
+   *	summary="Tag a dataset",
+   *	description="Tags a dataset.",
+   *	@OA\Parameter(
+   *		name="data_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *        @OA\Schema(
+   *          type="string"
+   *        )
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the tagged dataset",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="data_tag",
+   *				ref="#/components/schemas/inline_response_200_2_data_tag",
+   *			),
+   *			example={
+   *			  "data_tag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function data_tag($data_id, $tag) {
+    $this->data_tag_untag($data_id, $tag, false);
+  }
+
+  /**
+   *@OA\Post(
+   *	path="/data/untag",
+   *	tags={"data"},
+   *	summary="Untag a dataset",
+   *	description="Untags a dataset.",
+   *	@OA\Parameter(
+   *		name="data_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The ID of the untagged dataset",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="data_untag",
+   *				ref="#/components/schemas/inline_response_200_3_data_untag",
+   *			),
+   *			example={
+   *			  "data_untag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function data_untag($data_id, $tag) {
+    $this->data_tag_untag($data_id, $tag, true);
+  }
+
+  private function data_tag_untag($data_id,$tag, $do_untag) {
+    // forward action to superclass
+    $this->entity_tag_untag('dataset', $data_id, $tag, $do_untag, 'data');
+  }
+
+  /**
+   *@OA\Get(
+   *	path="/data/list/{filters}",
+   *	tags={"data"},
+   *	summary="List and filter datasets",
+   *	description="List datasets, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/data/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all datasets that match the constraints.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.
+  /status/{status} - returns only datasets with a given status, either 'active', 'deactivated', or 'in_preparation'.
+  /tag/{tag} - returns only datasets tagged with the given tag.
+  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, data_version, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of datasets with the given task",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataList",
+   *			example={
+   *			  "data": {
+   *			    "dataset": {
+   *			      {
+   *			        "did":"1",
+   *			        "name":"anneal",
+   *			        "status":"active",
+   *			        "format":"ARFF",
+   *			        "quality":{
+   *			          {
+   *			            "name":"MajorityClassSize",
+   *			            "value":"684"
+   *			          },
+   *			          {
+   *			            "name":"MaxNominalAttDistinctValues",
+   *			            "value":"10.0"
+   *			          },
+   *			          {
+   *			            "name":"MinorityClassSize"
+   *			            ,"value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumBinaryAtts",
+   *			            "value":"14.0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfClasses",
+   *			            "value":"6"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfFeatures",
+   *			            "value":"39"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfInstances",
+   *			            "value":"898"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfInstancesWithMissingValues",
+   *			            "value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfMissingValues",
+   *			            "value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfNumericFeatures",
+   *			            "value":"6"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfSymbolicFeatures",
+   *			            "value":"32"
+   *			          }
+   *			        }
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n370 - Illegal filter specified.\n371 - Filter values/ranges not properly specified.\n372 - No results. There where no matches for the given constraints.\n373 - Can not specify an offset without a limit.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_list($segs) {
     $legal_filters = array('tag', 'status', 'limit', 'offset', 'data_id', 'data_name', 'data_version', 'uploader', 'number_instances', 'number_features', 'number_classes', 'number_missing_values');
-    
+
     list($query_string, $illegal_filters) = $this->parse_filters($segs, $legal_filters);
     if (count($illegal_filters) > 0) {
       $this->returnError(370, $this->version, $this->openmlGeneralErrorCode, 'Legal filter operators: ' . implode(',', $legal_filters) .'. Found illegal filter(s): ' . implode(', ', $illegal_filters));
       return;
     }
-    
+
     $illegal_filter_inputs = $this->check_filter_inputs($query_string, $legal_filters, array('tag', 'status', 'data_name', 'number_instances', 'number_features', 'number_classes', 'number_missing_values'));
     if (count($illegal_filter_inputs) > 0) {
       $this->returnError(371, $this->version, $this->openmlGeneralErrorCode, 'Filters with illegal values: ' . implode(',', $illegal_filter_inputs));
       return;
     }
-    
+
     $tag = element('tag', $query_string, null);
     $name = element('data_name', $query_string, null);
     $data_id = element('data_id', $query_string, null);
@@ -170,7 +404,7 @@ private function data_list($segs) {
     $nr_feats = element('number_features', $query_string, null);
     $nr_class = element('number_classes', $query_string, null);
     $nr_miss = element('number_missing_values', $query_string, null);
-    
+
     if ($offset && !$limit) {
       $this->returnError(373, $this->version);
       return;
@@ -189,17 +423,17 @@ private function data_list($segs) {
     $status_sql_variable = 'IFNULL(`s`.`status`, \'' . $this->config->item('default_dataset_status') . '\')';
     $where_status = $status === null ? ' AND ' . $status_sql_variable . ' = "active" ' : ($status != "all" ? ' AND ' . $status_sql_variable . ' = "'. $status . '" ' : '');
     $where_total = $where_tag . $where_did . $where_name . $where_version . $where_uploader . $where_insts . $where_feats . $where_class . $where_miss . $where_status;
-    
+
     $where_limit = $limit === null ? '' : ' LIMIT ' . $limit;
     if($limit && $offset){
       $where_limit =  ' LIMIT ' . $offset . ',' . $limit;
     }
-    
+
     $sql = 'SELECT d.*, ' . $status_sql_variable . ' AS `status` '.
-           'FROM dataset d ' . 
+           'FROM dataset d ' .
            'LEFT JOIN (SELECT `did`, MAX(`status`) AS `status` FROM `dataset_status` GROUP BY `did`) s ON d.did = s.did ' .
            'WHERE (visibility = "public" or uploader='.$this->user_id.') '. $where_total . $where_limit;
-    
+
     $datasets_res = $this->Dataset->query($sql);
     if( is_array( $datasets_res ) == false || count( $datasets_res ) == 0 ) {
       $this->returnError( 372, $this->version );
@@ -214,12 +448,12 @@ private function data_list($segs) {
     }
 
     # JvR: This is a BAD idea and this will break in the future, when OpenML grows.
-    $sql = 
+    $sql =
       'SELECT data, quality, value FROM data_quality ' .
       'WHERE `data` IN (' . implode(',', array_keys($datasets)) . ') ' .
       'AND evaluation_engine_id = ' . $this->config->item('default_evaluation_engine_id') . ' ' .
-      'AND quality IN ("' . implode('","', $this->config->item('basic_qualities')) . '") ' . 
-      'AND value IS NOT NULL ' . 
+      'AND quality IN ("' . implode('","', $this->config->item('basic_qualities')) . '") ' .
+      'AND value IS NOT NULL ' .
       'ORDER BY `data`;';
     $dq = $this->Data_quality->query($sql);
 
@@ -231,7 +465,174 @@ private function data_list($segs) {
 
     $this->xmlContents('data', $this->version, array('datasets' => $datasets));
   }
+  
+  private function data_fork() {
+    // get data id
+    $data_id = $this->input->post('data_id');
+    // If data id is not given
+    if( $data_id == false ) {
+      $this->returnError( 1070, $this->version );
+      return;
+    }
+    // If dataset does not exist
+    $dataset = $this->Dataset->getById( $data_id );
+    if( $dataset == false ) {
+      $this->returnError( 1071, $this->version );
+      return;
+    }
+    // create a copy
+    $dataset->uploader = $this->user_id;   
+    $latest_version = $this->Dataset-> getWhereSingle('`name` = "' . $dataset->name . '"', 'CAST(`version` AS DECIMAL) DESC');
+    $dataset->version = $latest_version->version + 1;
+    unset($dataset->did);
+    $data_id = $this->Dataset->insert($dataset);
+    if (!$data_id) {
+      $this->returnError(1072, $this->version);
+      return;
+    }
+    // Return data id, for user to verify changes
+    $this->xmlContents( 'data-fork', $this->version, array( 'id' => $data_id) );
+  }
+
+  private function data_edit() {
+    // get data id
+    $data_id = $this->input->post('data_id');
+    // get edit parameters as xml
+    $xsdFile = xsd('openml.data.edit', $this->controller, $this->version);
+
+    if($this->input->post('edit_parameters')) {
+      // get fields from string upload
+      $edit_parameters = $this->input->post('edit_parameters', false);
+      if(validateXml($edit_parameters, $xsdFile, $xmlErrors, false ) == false) {
+        $this->returnError(1060, $this->version, $this->openmlGeneralErrorCode, $xmlErrors);
+        return;
+      }
+      $xml = simplexml_load_string( $edit_parameters );
+    } elseif (isset($_FILES['edit_parameters'])) {
+      $uploadError = '';
+      $xmlErrors = '';
+      if (check_uploaded_file($_FILES['edit_parameters'], false, $uploadError) == false) {
+        $this->returnError(1061, $this->version, $this->openmlGeneralErrorCode, $uploadError);
+      }
+      // get fields from file upload
+      $edit_parameters = $_FILES['edit_parameters'];
+
+      if (validateXml($edit_parameters['tmp_name'], $xsdFile, $xmlErrors) == false) {
+        $this->returnError(1060, $this->version, $this->openmlGeneralErrorCode, $xmlErrors);
+        return;
+      }
+      $xml = simplexml_load_file($edit_parameters['tmp_name']);
+    } else {
+      $this->returnError(1061, $this->version);
+      return;
+    }
+
+    // create an array of update fields for the update
+    $update_fields = array();
+    foreach($xml->children('oml', true) as $input) {
+      // iterate over all fields, does not check for legal fields, as it wont match the xsd.  
+      $name = $input->getName() . '';
+      $update_fields[$name] = $input . '';
+    }
+
+    // If data id is not given
+    if( $data_id == false ) {
+      $this->returnError( 1062, $this->version );
+      return;
+    }
+
+    // If dataset does not exist
+    $dataset = $this->Dataset->getById( $data_id );
+    if( $dataset == false ) {
+      $this->returnError( 1063, $this->version );
+      return;
+    }
+
+    // If all the fields are false, there is nothing to update, return error 
+    if(!$update_fields) {
+      $this->returnError( 1064, $this->version );
+      return;
+    }
+
+    // If critical fields need to be edited
+    if (isset($update_fields['default_target_attribute']) || isset($update_fields['row_id_attribute']) || isset($update_fields['ignore_attribute'])) {
+      # Only owner can edit critical features
+      if($dataset->uploader != $this->user_id and !$this->user_has_admin_rights) {
+        $this->returnError(1065, $this->version);
+        return;
+      }
+      # Only datasets without tasks can allow critical feature edits
+      $tasks = $this->Task->getTasksWithValue( array( 'source_data' => $dataset->did ) );
+      if( $tasks !== false ) {
+        $this->returnError( 1066, $this->version );
+        return;
+      }
+    }
 
+    $update_result = $this->Dataset->update($data_id, $update_fields);
+    // If result returns error    
+    if( $update_result == false ) {
+      $this->returnError( 1067, $this->version );
+      return;
+    }
+    // Return data id, for user to verify changes    
+    $this->xmlContents( 'data-edit', $this->version, array( 'id' => $data_id) );
+  }
+
+  /**
+   *@OA\Get(
+   *	path="/data/{id}",
+   *	tags={"data"},
+   *	summary="Get dataset description",
+   *	description="Returns information about a dataset. The information includes the name, information about the creator, URL to download it and more.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A dataset description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Data",
+   *			example={
+   *			  "data_set_description": {
+   *			    "id": "1",
+   *			    "name": "anneal",
+   *			    "version": "2",
+   *			    "description": "...",
+   *			    "format": "ARFF",
+   *			    "upload_date": "2014-04-06 23:19:20",
+   *			    "licence": "Public",
+   *			    "url": "https://www.openml.org/data/download/1/dataset_1_anneal.arff",
+   *			    "file_id": "1",
+   *			    "default_target_attribute": "class",
+   *			    "version_label": "2",
+   *			    "tag": {
+   *			      "study_1",
+   *			      "uci"
+   *			    },
+   *			    "visibility": "public",
+   *			    "original_data_url": "https://www.openml.org/d/2",
+   *			    "status": "active",
+   *			    "md5_checksum": "d01f6ccd68c88b749b20bbe897de3713"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned\n110 - Please provide data_id.\n111 - Unknown dataset. Data set description with data_id was not found in the database.\n112 - No access granted. This dataset is not shared with you.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data($data_id) {
     if( $data_id == false ) {
       $this->returnError( 110, $this->version );
@@ -270,7 +671,7 @@ private function data($data_id) {
     foreach( $this->xml_fields_dataset['csv'] as $field ) {
       $dataset->{$field} = getcsv( $dataset->{$field} );
     }
-    
+
     $data_processed = $this->Data_processed->getById(array($data_id, $this->config->item('default_evaluation_engine_id')));
     $relevant_fields = array('processing_date', 'error', 'warning');
     foreach ($relevant_fields as $field) {
@@ -280,7 +681,7 @@ private function data($data_id) {
         $dataset->{$field} = null;
       }
     }
-    
+
     $dataset->status = $this->config->item('default_dataset_status');
     $data_status = $this->Dataset_status->getWhereSingle('did =' . $data_id, 'status_date DESC');
     if ($data_status != false) {
@@ -289,7 +690,7 @@ private function data($data_id) {
 
     $this->xmlContents( 'data-get', $this->version, $dataset );
   }
-  
+
   private function data_reset($data_id) {
     $dataset = $this->Dataset->getById($data_id);
     if ($dataset == false) {
@@ -301,7 +702,7 @@ private function data_reset($data_id) {
       $this->returnError(1022, $this->version);
       return;
     }
-    
+
     $result = $this->Data_processed->deleteWhere('`did` = "' . $dataset->did . '" ');
 
     if ($result == false) {
@@ -311,6 +712,55 @@ private function data_reset($data_id) {
     $this->xmlContents('data-reset', $this->version, array('dataset' => $dataset));
   }
 
+  /**
+   *@OA\Delete(
+   *	path="/data/{id}",
+   *	tags={"data"},
+   *	summary="Delete dataset",
+   *	description="Deletes a dataset. Upon success, it returns the ID of the deleted dataset.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted dataset",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="data_delete",
+   *				ref="#/components/schemas/inline_response_200_data_delete",
+   *			),
+   *			example={
+   *			  "data_delete": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned\n- 350 - Please provide API key. In order to remove your content, please authenticate.\n- 351 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n- 352 - Dataset does not exists. The data ID could not be linked to an existing dataset.\n- 353 - Dataset is not owned by you. The dataset is owned by another user. Hence you cannot delete it.\n- 354 - Dataset is in use by other content. Can not be deleted. The data is used in tasks or runs. Delete other content before deleting this dataset.\n- 355 - Deleting dataset failed. Deleting the dataset failed. Please contact support team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_delete($data_id) {
 
     $dataset = $this->Dataset->getById( $data_id );
@@ -359,6 +809,64 @@ private function data_delete($data_id) {
   }
 
 
+  /**
+   *@OA\Post(
+   *	path="/data",
+   *	tags={"data"},
+   *	summary="Upload dataset",
+   *	description="Uploads a dataset. Upon success, it returns the data id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/data).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="dataset",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="The actual dataset, being an ARFF file.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the uploaded dataset",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_data_set",
+   *				ref="#/components/schemas/inline_response_200_1_upload_data_set",
+   *			),
+   *			example={
+   *			  "upload_data_set": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n130 - Problem with file uploading. There was a problem with the file upload.\n131 - Problem validating uploaded description file. The XML description format does not meet the standards.\n132 - Failed to move the files. Internal server error, please contact API administrators.\n133 - Failed to make checksum of datafile. Internal server error, please contact API administrators.\n134 - Failed to insert record in database. Internal server error, please contact API administrators.\n135 - Please provide description xml.\n136 - File failed format verification. The uploaded file is not valid according to the selected file format. Please check the file format specification and try again.\n137 - Please provide API key. In order to share content, please log in or provide your API key.\n138 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\n139 - Combination name / version already exists. Leave version out for auto increment\n140 - Both dataset file and dataset url provided. The system is confused since both a dataset file (post) and a dataset url (xml) are provided. Please remove one.\n141 - Neither dataset file or dataset url are provided. Please provide either a dataset file as POST variable, or a dataset url in the description XML.\n142 - Error in processing arff file. Can be a syntax error, or the specified target feature does not exists. For now, we only check on arff files. If a dataset is claimed to be in such a format, and it can not be parsed, this error is returned.\n143 - Suggested target feature not legal. It is possible to suggest a default target feature (for predictive tasks). However, it should be provided in the data.\n144 - Unable to update dataset. The dataset with id could not be found in the database. If you upload a new dataset, unset the id.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_upload() {
     // get correct description
     $xsdFile = xsd('openml.data.upload', $this->controller, $this->version);
@@ -440,7 +948,7 @@ private function data_upload() {
         $this->returnError(145, $this->version, $this->openmlGeneralErrorCode, 'Arff error in dataset file: ' . $uploadedFileCheck);
         return;
       }
-      
+
       $to_folder = $this->data_folders['dataset'];
       $file_id = $this->File->register_uploaded_file($_FILES['dataset'], $to_folder, $this->user_id, 'dataset', $access_control);
       if ($file_id === false) {
@@ -484,7 +992,7 @@ private function data_upload() {
       'isOriginal' => 'true',
       'file_id' => $file_id
     );
-    
+
     // extract all other necessary info from the XML description
     $dataset = all_tags_from_xml(
       $xml->children('oml', true),
@@ -505,15 +1013,15 @@ private function data_upload() {
       $this->returnError(134, $this->version);
       return;
     }
-    
-    // try to move the file to a new directory. If it fails, the dataset is 
+
+    // try to move the file to a new directory. If it fails, the dataset is
     // still valid, but we probably want to make some mechanism to inform administrators
     if ($file_record->type != 'url') {
       $subdirectory = floor($id / $this->content_folder_modulo) * $this->content_folder_modulo;
       $to_folder = $this->data_folders['dataset'] . '/' . $subdirectory . '/' . $id . '/';
       $this->File->move_file($file_id, $to_folder);
     }
-    
+
     // try making the ES stuff
     try {
       // update elastic search index.
@@ -538,7 +1046,49 @@ private function data_upload() {
     // create
     $this->xmlContents('data-upload', $this->version, array('id' => $id));
   }
-  
+
+  /**
+   *@OA\Post(
+   *	path="/data/status/update/",
+   *	tags={"data"},
+   *	summary="Change the status of a dataset",
+   *	description="Change the status of a dataset, either 'active' or 'deactivated'",
+   *	@OA\Parameter(
+   *		name="data_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="status",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="The status on which to filter the results, either 'active' or 'deactivated'.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n691 - Illegal status\n692 - Dataset does not exists\n693 - Dataset is not owned by you\n694 - Illegal status transition\n695 - Status update failed\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function status_update($data_id, $status) {
     // in_preparation is not a legal status to change to
     $legal_status = array('active', 'deactivated');
@@ -546,7 +1096,7 @@ private function status_update($data_id, $status) {
       $this->returnError(691, $this->version);
       return;
     }
-    
+
     $dataset = $this->Dataset->getById($data_id);
     if ($dataset == false) {
       $this->returnError(692, $this->version);
@@ -557,12 +1107,12 @@ private function status_update($data_id, $status) {
       $this->returnError(693, $this->version);
       return;
     }
-    
+
     if ($status == 'active' && !$this->user_has_admin_rights) {
       $this->returnError(696, $this->version);
       return;
     }
-    
+
     $status_record = $this->Dataset_status->getWhereSingle('did = ' . $data_id, 'status DESC');
     $in_preparation = $this->config->item('default_dataset_status');
     if ($status_record == false) {
@@ -570,14 +1120,14 @@ private function status_update($data_id, $status) {
     } else {
       $old_status = $status_record->status;
     }
-    
+
     $record = array(
       'did' => $data_id,
-      'status' => $status, 
+      'status' => $status,
       'status_date' => now(),
       'user_id' => $this->user_id
     );
-    
+
     if (
         ($old_status == $in_preparation && $status == 'active') ||
         ($old_status == $in_preparation && $status == 'deactivated') ||
@@ -586,10 +1136,10 @@ private function status_update($data_id, $status) {
       $this->Dataset_status->insert($record);
     } elseif ($old_status == 'deactivated' && $status == 'active') {
       $this->Dataset_status->delete(array($data_id, 'deactivated'));
-      
+
       // see if the dataset is still active
       $status_record = $this->Dataset_status->getWhereSingle('did = ' . $data_id, 'status DESC');
-      
+
       $result = true;
       if (!$status_record || $status_record->status != 'active') {
         $result = $this->Dataset_status->insert($record);
@@ -602,10 +1152,87 @@ private function status_update($data_id, $status) {
       $this->returnError(694, $this->version);
       return;
     }
-    
+
     $this->xmlContents('data-status-update', $this->version, array('did' => $data_id, 'status' => $status));
   }
 
+  /**
+   *@OA\Get(
+   *	path="/data/features/{id}",
+   *	tags={"data"},
+   *	summary="Get data features",
+   *	description="Returns the features of a dataset.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="All the features of the dataset",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataFeatures",
+   *			example={
+   *			  "data_features": {
+   *			    "feature": {
+   *			      {
+   *			        "index": "0",
+   *			        "name": "sepallength",
+   *			        "data_type": "numeric",
+   *			        "is_target": "false",
+   *			        "is_ignore": "false",
+   *			        "is_row_identifier": "false"
+   *			      },
+   *			      {
+   *			        "index": "1",
+   *			        "name": "sepalwidth",
+   *			        "data_type": "numeric",
+   *			        "is_target": "false",
+   *			        "is_ignore": "false",
+   *			        "is_row_identifier": "false"
+   *			      },
+   *			      {
+   *			        "index": "2",
+   *			        "name": "petallength",
+   *			        "data_type": "numeric",
+   *			        "is_target": "false",
+   *			        "is_ignore": "false",
+   *			        "is_row_identifier": "false"
+   *			      },
+   *			      {
+   *			        "index": "3",
+   *			        "name": "petalwidth",
+   *			        "data_type": "numeric",
+   *			        "is_target": "false",
+   *			        "is_ignore": "false",
+   *			        "is_row_identifier": "false"
+   *			      },
+   *			      {
+   *			        "index": "4",
+   *			        "name": "class",
+   *			        "data_type": "nominal",
+   *			        "is_target": "true",
+   *			        "is_ignore": "false",
+   *			        "is_row_identifier": "false"
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n270 - Please provide dataset ID.\n271 - Unknown dataset. Data set with the given data ID was not found (or is not shared with you).\n272 - No features found. The dataset did not contain any features, or we could not extract them.\n273 - Dataset not processed yet. The dataset was not processed yet, features are not yet available. Please wait for a few minutes.\n274 - Dataset processed with error. The feature extractor has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, please contact the API admins.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_features($data_id) {
     $dataset = $this->Dataset->getById($data_id);
     if ($dataset === false) {
@@ -637,12 +1264,12 @@ private function data_features($data_id) {
       }
     }
     $dataset->index_values = $index_values;
-    
+
     if ($data_processed->error && $dataset->features === false) {
       $this->returnError(274, $this->version);
       return;
     }
-    
+
     if ($dataset->features === false) {
       $this->returnError(272, $this->version);
       return;
@@ -659,6 +1286,39 @@ private function data_features($data_id) {
     $this->xmlContents('data-features', $this->version, $dataset);
   }
 
+  /**
+   *@OA\Post(
+   *	path="/data/features",
+   *	tags={"data"},
+   *	summary="Upload dataset feature description",
+   *	description="Uploads dataset feature description. Upon success, it returns the data id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.features) and an [XML example](https://www.openml.org/api/v1/xml_example/data.features).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n431 - Dataset already processed\n432 - Please provide description xml\n433 - Problem validating uploaded description file\n434 - Could not find dataset\n436 - Something wrong with XML, check data id and evaluation engine id\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_features_upload() {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
@@ -670,11 +1330,11 @@ private function data_features_upload() {
       $this->returnError(442, $this->version);
       return;
     }
-    
+
     // get description from file upload. Note that we will check the XSD later on (after we have assembled fields for error handling)
     $description = $_FILES['description'];
     $xml = simplexml_load_file($description['tmp_name']);
-    
+
     // precheck XSD (if this pre-check succeeds, we can do database error logging later)
     if (!($xml->children('oml', true)->{'did'} && $xml->children('oml', true)->{'evaluation_engine_id'})) {
       $this->returnError(443, $this->version, $this->openmlGeneralErrorCode, 'XML misses basic fields did or evaluation_engine_id');
@@ -693,13 +1353,13 @@ private function data_features_upload() {
       $this->returnError(444, $this->version);
       return;
     }
-    
+
     $data_processed_record = $this->Data_processed->getById(array($did, $eval_id));
     if ($data_processed_record && $data_processed_record->error == null) {
       $this->returnError(441, $this->version);
       return;
     }
-    
+
     $num_tries = 0;
     if ($data_processed_record) {
       $num_tries = $data_processed_record->num_tries;
@@ -709,12 +1369,12 @@ private function data_features_upload() {
     $data = array('did' => $did,
                   'evaluation_engine_id' => $eval_id,
                   'user_id' => $this->user_id,
-                  'processing_date' => now(), 
+                  'processing_date' => now(),
                   'num_tries' => $num_tries + 1);
     if ($xml->children('oml', true)->{'error'}) {
       $data['error'] = htmlentities($xml->children('oml', true)->{'error'});
     }
-    
+
     if (validateXml($description['tmp_name'], xsd('openml.data.features', $this->controller, $this->version), $xmlErrors) == false) {
       $data['error'] = 'XSD does not comply. XSD errors: ' . $xmlErrors;
       $success = $this->Data_processed->replace($data);
@@ -729,10 +1389,10 @@ private function data_features_upload() {
     }
 
     $this->db->trans_start();
-    
+
     # replace is delete then insert again
     $success = $this->Data_processed->replace($data);
-    if (!$success) {  
+    if (!$success) {
       $this->returnError(445, $this->version, $this->openmlGeneralErrorCode, 'Failed to create data processed record. ');
       return;
     }
@@ -765,7 +1425,7 @@ private function data_features_upload() {
       if (in_array($feature['name'], $ignores)) {
         $feature['is_ignore'] = 'true';
       }
-      
+
       if (in_array('ClassDistribution', $feature)) {
         // check class distributions field
         json_decode($feature['ClassDistribution']);
@@ -775,7 +1435,7 @@ private function data_features_upload() {
           return;
         }
       }
-      
+
       //actual insert of the feature
       if (array_key_exists('nominal_value', $feature)) {
         $nominal_values = $feature['nominal_value'];
@@ -783,19 +1443,19 @@ private function data_features_upload() {
       } else {
         $nominal_values = false;
       }
-      
+
       $result = $this->Data_feature->insert($feature);
       if (!$result) {
         $this->db->trans_rollback();
         $this->returnError(450, $this->version, $this->openmlGeneralErrorCode, 'feature: ' . $feature['name']);
         return;
       }
-      
+
       if ($nominal_values) {
         // check the nominal value property
         foreach ($nominal_values as $value) {
           $data = array(
-            'did' => $did, 
+            'did' => $did,
             'index' => $feature['index'],
             'value' => $value
           );
@@ -806,7 +1466,7 @@ private function data_features_upload() {
             return;
           }
         }
-        
+
         if ($feature['data_type'] != 'nominal') {
           // only allowed for nominal values
           $this->db->trans_rollback();
@@ -837,6 +1497,41 @@ private function data_features_upload() {
     }
   }
 
+  /**
+   *@OA\Get(
+   *	path="/data/qualities/list",
+   *	tags={"data"},
+   *	summary="List all data qualities",
+   *	description="Returns a list of all data qualities in the system.",
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of data qualities",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataQualityList",
+   *			example={
+   *			  "data_qualities_list":{
+   *			    "quality":{
+   *			      "NumberOfClasses",
+   *			      "NumberOfFeatures",
+   *			      "NumberOfInstances",
+   *			      "NumberOfInstancesWithMissingValues",
+   *			      "NumberOfMissingValues",
+   *			      "NumberOfNumericFeatures",
+   *			      "NumberOfSymbolicFeatures"
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned\n370 - No data qualities available. There are no data qualities in the system.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_qualities_list() {
     $result = $this->Quality->allUsed( );
     $qualities = array();
@@ -867,6 +1562,79 @@ private function feature_qualities_list() {
   }
 
 
+  /**
+   *@OA\Get(
+   *	path="/data/qualities/{id}",
+   *	tags={"data"},
+   *	summary="Get data qualities",
+   *	description="Returns the qualities of a dataset.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the dataset.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="All the qualities of the dataset",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataQualities",
+   *			example={
+   *			  "data_qualities": {
+   *			    "quality": {
+   *			      {
+   *			        "name": "ClassCount",
+   *			        "value": "3.0"
+   *			      },
+   *			      {
+   *			        "name": "ClassEntropy",
+   *			        "value": "1.584962500721156"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfClasses",
+   *			        "value": "3"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfFeatures",
+   *			        "value": "5"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfInstances",
+   *			        "value": "150"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfInstancesWithMissingValues",
+   *			        "value": "0"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfMissingValues",
+   *			        "value": "0"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfNumericFeatures",
+   *			        "value": "4"
+   *			      },
+   *			      {
+   *			        "name": "NumberOfSymbolicFeatures",
+   *			        "value": "0"
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n360 - Please provide data set ID\n361 - Unknown dataset. The data set with the given ID was not found in the database, or is not shared with you.\n362 - No qualities found. The registered dataset did not contain any calculated qualities.\n363 - Dataset not processed yet. The dataset was not processed yet, no qualities are available. Please wait for a few minutes.\n364 - Dataset processed with error. The quality calculator has run into an error while processing the dataset. Please check whether it is a valid supported file. If so, contact the support team.\n365 - Interval start or end illegal. There was a problem with the interval\nstart or end.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_qualities($data_id, $evaluation_engine_id) {
     if( $data_id == false ) {
       $this->returnError( 360, $this->version );
@@ -893,7 +1661,7 @@ private function data_qualities($data_id, $evaluation_engine_id) {
     $interval_start = false; // $this->input->get( 'interval_start' );
     $interval_end   = false; // $this->input->get( 'interval_end' );
     $interval_size  = false; // $this->input->get( 'interval_size' );
-    
+
     if($interval_start !== false || $interval_end !== false || $interval_size !== false) {
       $interval_constraints = '';
       if( $interval_start !== false && is_numeric( $interval_start ) ) {
@@ -909,7 +1677,7 @@ private function data_qualities($data_id, $evaluation_engine_id) {
     } else {
       $dataset->qualities = $this->Data_quality->getWhere('data = "' . $dataset->did . '" AND evaluation_engine_id = ' . $evaluation_engine_id);
     }
-    
+
     if($data_processed->error && $dataset->qualities === false) {
       $this->returnError(364, $this->version, $this->openmlGeneralErrorCode, $data_processed->error);
       return;
@@ -978,6 +1746,39 @@ private function feature_qualities($data_id, $evaluation_engine_id) {
     $this->xmlContents( 'feature-qualities', $this->version, $dataset );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/data/qualities",
+   *	tags={"data"},
+   *	summary="Upload dataset qualities",
+   *	description="Uploads dataset qualities. Upon success, it returns the data id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.data.qualities) and an [XML example](https://www.openml.org/api/v1/xml_example/data.qualities).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n381 - Something wrong with XML, please check did and evaluation_engine_id\n382 - Please provide description xml\n383 - Problem validating uploaded description file\n384 - Dataset not processed yet\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_qualities_upload() {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
@@ -1084,7 +1885,7 @@ private function data_qualities_upload() {
         $result = $this->Data_quality->insert_ignore($data);
       }
     }
-    
+
     if ($this->db->trans_status() === FALSE) {
       $this->db->trans_rollback();
       $this->returnError(389, $this->version);
@@ -1104,12 +1905,53 @@ private function data_qualities_upload() {
     }
   }
 
+  /**
+   *@OA\Get(
+   *	path="/data/unprocessed/{data_engine_id}/{order}",
+   *	tags={"data"},
+   *	summary="Get a list of unprocessed datasets",
+   *	description="This call is for people running their own dataset processing engines. It returns the details of datasets that are not yet processed by the given processing engine. It doesn't process the datasets, it just returns the dataset info.",
+   *	@OA\Parameter(
+   *		name="data_engine_id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="order",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="When there are multiple datasets still to process, this defines which ones to return. Options are 'normal' - the oldest datasets, or 'random'.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of unprocessed datasets",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataUnprocessed",
+   *			example={"data_unprocessed": {"run": {{"did": "1", "status": "deactivated", "version": "2", "name": "anneal", "format": "ARFF"}}}}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n681 - No unprocessed datasets.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function data_unprocessed($evaluation_engine_id, $order) {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
       return;
     }
-    
+
     $this->db->select('d.*')->from('dataset d');
     $this->db->join('data_processed p', 'd.did = p.did AND evaluation_engine_id = ' . $evaluation_engine_id, 'left');
     $this->db->where('(p.did IS NULL OR (p.error IS NOT NULL AND p.num_tries < ' . $this->config->item('process_data_tries') . ' AND p.processing_date < "' . now_offset('-' . $this->config->item('process_data_offset')) . '"))');
@@ -1144,6 +1986,65 @@ private function data_unprocessed($evaluation_engine_id, $order) {
     $this->xmlContents('data-unprocessed', $this->version, array('res' => $result));
   }
 
+  /**
+   *@OA\Post(
+   *	path="/data/qualities/unprocessed/{data_engine_id}/{order}",
+   *	tags={"data"},
+   *	summary="Get a list of datasets with unprocessed qualities",
+   *	description="This call is for people running their own dataset processing engines. It returns the details of datasets for which certain qualities are not yet processed by the given processing engine. It doesn't process the datasets, it just returns the dataset info.",
+   *	@OA\Parameter(
+   *		name="data_engine_id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="The ID of the data processing engine. You get this ID when you register a new data processing engine with OpenML. The ID of the main data processing engine is 1.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="order",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="When there are multiple datasets still to process, this defines which ones to return. Options are 'normal' - the oldest datasets, or 'random'.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="API key to authenticate the user",
+   *		required=false,
+   *	),
+   *	@OA\Parameter(
+   *		name="qualities",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Comma-separated list of (at least two) quality names, e.g. 'NumberOfInstances,NumberOfFeatures'.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of unprocessed datasets",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/DataUnprocessed",
+   *			example={"data_unprocessed": {"run": {{"did": "1", "status": "deactivated", "version": "2", "name": "anneal", "format": "ARFF"}}}}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n686 - Please specify the features the evaluation engine wants to calculate (at least 2).\n687 - No unprocessed datasets according to the given set of meta-features.\n688 - Illegal qualities.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function dataqualities_unprocessed($evaluation_engine_id, $order, $feature_attributes = false, $priorityTag = null) {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
diff --git a/openml_OS/models/api/v1/Api_estimationprocedure.php b/openml_OS/models/api/v1/Api_estimationprocedure.php
index 3ad1709bc..beccb51ba 100644
--- a/openml_OS/models/api/v1/Api_estimationprocedure.php
+++ b/openml_OS/models/api/v1/Api_estimationprocedure.php
@@ -12,7 +12,7 @@ function __construct() {
   
   function bootstrap($format, $segments, $request_type, $user_id) {
     $this->outputFormat = $format;
-    
+
     if (count($segments) == 1 && $segments[0] == 'list') {
       $this->estimationprocedure_list();
       return;
@@ -41,8 +41,54 @@ private function estimationprocedure($id) {
     }
     $this->xmlContents( 'estimationprocedure-get', $this->version, array( 'ep' => $ep ) );
   }
-  private function estimationprocedure_list() {
 
+  /**
+   *@OA\Get(
+   *	path="/estimationprocedure/list",
+   *	tags={"estimationprocedure"},
+   *	summary="List all estimation procedures",
+   *	description="Returns an array with all model performance estimation procedures in the system.",
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of estimation procedures",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/EstimationProcedureList",
+   *			example={
+   *			  "estimationprocedures": {
+   *			    "estimationprocedure": {
+   *			      {
+   *			         "id":"1",
+   *			         "ttid":"1",
+   *			         "name":"10-fold Crossvalidation",
+   *			         "type":"crossvalidation",
+   *			         "repeats":"1",
+   *			         "folds":"10",
+   *			         "stratified_sampling":"true"
+   *			      },
+   *			      {
+   *			        "id":"2",
+   *			        "ttid":"1",
+   *			        "name":"5 times 2-fold Crossvalidation",
+   *			        "type":"crossvalidation",
+   *			        "repeats":"5",
+   *			        "folds":"2",
+   *			        "stratified_sampling":"true"
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n500 - No model performance estimation procedures available.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function estimationprocedure_list() {
     $estimationprocedures = $this->Estimation_procedure->get();
     if( $estimationprocedures == false ) {
       $this->returnError( 500, $this->version );
diff --git a/openml_OS/models/api/v1/Api_evaluation.php b/openml_OS/models/api/v1/Api_evaluation.php
index 42e460722..84541a6d0 100644
--- a/openml_OS/models/api/v1/Api_evaluation.php
+++ b/openml_OS/models/api/v1/Api_evaluation.php
@@ -40,6 +40,47 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError(100, $this->version);
   }
 
+  /**
+   *@OA\Get(
+   *	path="/evaluation/request/{evaluation_engine_id}/{order}",
+   *	tags={"evaluation"},
+   *	summary="Get an unevaluated run",
+   *	description="This call is for people running their own evaluation engines. It returns the details of a run that is not yet evaluated by the given evaluation engine. It doesn't evaluate the run, it just returns the run info.",
+   *	@OA\Parameter(
+   *		name="evaluation_engine_id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="The ID of the evaluation engine. You get this ID when you register a new evaluation engine with OpenML. The ID of the main evaluation engine is 1.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="order",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="When there are multiple runs still to evaluate, this defines which one to return. Options are 'normal' - the oldest run, 'reverse' - the newest run, or 'random' - a random run.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of evaluations descriptions",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/EvaluationRequest",
+   *			example={"evaluation_request": {"run": {{"setup_id": "68799271", "upload_time": "2018-04-03 21:05:38", "uploader": "1935", "task_id": "3021", "run_id": "8943712"}}}}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n100 - Function not valid.\n545 - No unevaluated runs according to the criteria.\n546 - Illegal filter.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function evaluation_request($evaluation_engine_id, $order,  $num_requests, $segs) {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
@@ -94,6 +135,49 @@ private function evaluation_request($evaluation_engine_id, $order,  $num_request
   }
 
 
+  /**
+   *@OA\Get(
+   *	path="/evaluation/list/{filters}",
+   *	tags={"evaluation"},
+   *	summary="List and filter evaluations",
+   *	description="List evaluations, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/evaluation/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all evaluations that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, uploaders or runs.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /function/{name} - name of the evaluation measure, e.g. area_under_auc or predictive_accuracy. See the OpenML website for the complete list of measures.
+  /tag/{tag} - returns only evaluations of runs tagged with the given tag.
+  /run/{ids} - return only evaluations for specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''
+  /task/{ids} - return only evaluations for specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''
+  /flow/{ids} - return only evaluations for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''
+  /setup/{ids} - return only evaluations for specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''
+  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.
+  /per_fold/{true,false} - whether or not to return crossvalidation scores per fold. Defaults to 'false'. Setting it to 'true' leads to large numbers of results, use only for very specific sets of runs.
+  /sort_order/{asc,desc} - sorts the results by the evaluation value, according to the selected evaluation measure (function)
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of evaluations descriptions",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/EvaluationList",
+   *			example={"evaluations": {"evaluation": {{"function": "area_under_roc_curve", "upload_time": "2014-04-06 23:30:40", "task_id": "68", "run_id": "1", "array_data": "[0,0.99113,0.898048,0.874862,0.791282,0.807343,0.820674]", "value": "0.839359", "uploader": "1", "flow_id": "61"}, {"function": "f_measure", "upload_time": "2014-04-06 23:30:40", "task_id": "68", "run_id": "1", "array_data": "[0,0,0.711934,0.735714,0.601363,0.435678,0.430913]", "value": "0.600026", "uploader": "1", "flow_id": "61"}, {"function": "predictive_accuracy", "upload_time": "2014-04-06 23:30:40", "task_id": "68", "run_id": "1", "array_data": {}, "value": "0.614634", "uploader": "1", "flow_id": "61"}}}}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n540 - Please provide at least task, flow or setup, uploader or run, to\nfilter results, or limit the number of responses.\n541 - The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\n542 - There where no results. Check whether there are runs under the given constraint.\n543 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\n544 - Illegal filter specified.\n545 - Offset specified without limit.\n546 - Requested result limit too high.\n547 - Per fold can only be set to value 'true' or 'false'.\n548 - Per fold queries are experimental and require a fair amount of filters on resulting run records to keep the query fast (use, e.g., flow, setup, task and uploader filter)\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function evaluation_list($segs, $user_id, $show_params) {
     $result_limit = 10000;
     $legal_filters = array('task', 'setup', 'flow', 'uploader', 'run', 'tag', 'limit', 'offset', 'function', 'per_fold', 'sort_order', 'study');
@@ -102,7 +186,7 @@ private function evaluation_list($segs, $user_id, $show_params) {
       $this->returnError(544, $this->version, $this->openmlGeneralErrorCode, 'Legal filter operators: ' . implode(',', $legal_filters) .'. Found illegal filter(s): ' . implode(', ', $illegal_filters));
       return;
     }
-    
+
     $illegal_filter_inputs = $this->check_filter_inputs($query_string, $legal_filters, array('tag', 'function', 'per_fold', 'sort_order'));
     if (count($illegal_filter_inputs) > 0) {
       $this->returnError(541, $this->version, $this->openmlGeneralErrorCode, 'Filters with illegal values: ' . implode(',', $illegal_filter_inputs));
@@ -126,7 +210,7 @@ private function evaluation_list($segs, $user_id, $show_params) {
       return;
     }
     $per_fold = ($per_fold === 'true') ? true : false; // cast to boolean. we only accept lowercase entrees
-    
+
     if ($offset && !$limit) {
       $this->returnError(545, $this->version);
       return;
@@ -139,7 +223,7 @@ private function evaluation_list($segs, $user_id, $show_params) {
       $this->returnError(549, $this->version);
       return;
     }
-    
+
     if ($study_id) {
       $study = $this->Study->getById($study_id);
       if ($study === false || $study->legacy != 'n' || $study->main_entity_type != 'run') {
@@ -166,7 +250,7 @@ private function evaluation_list($segs, $user_id, $show_params) {
       $where_limit =  ' LIMIT ' . $offset . ',' . $limit;
     }
     $where_task_closed = ' AND (`t`.`embargo_end_date` is NULL OR `t`.`embargo_end_date` < NOW() OR `r`.`uploader` = '.$user_id.')';
-    
+
     $where_runs = $where_task . $where_setup . $where_uploader . $where_impl . $where_run . $where_tag . $where_study . $where_task_closed;
     $where_total = $where_runs . $where_function;
 
@@ -176,17 +260,17 @@ private function evaluation_list($segs, $user_id, $show_params) {
       //shortcuts
       if (!$implementation_id) {
         $sql_test =
-          'SELECT count(distinct r.rid) as count ' .
-          'FROM run r, task t '.
-          'WHERE r.task_id = t.task_id ' . 
-          $where_runs;
+            'SELECT count(distinct r.rid) as count ' .
+            'FROM run r, task t '.
+            'WHERE r.task_id = t.task_id ' .
+            $where_runs;
         $count = $this->Evaluation->query($sql_test)[0]->count;
       } else {
         $sql_test =
-          'SELECT count(distinct r.rid) as count ' .
-          'FROM run r, task t, algorithm_setup s ' .
-          'WHERE r.setup = s.sid AND r.task_id = t.task_id ' .
-          $where_runs;
+            'SELECT count(distinct r.rid) as count ' .
+            'FROM run r, task t, algorithm_setup s ' .
+            'WHERE r.setup = s.sid AND r.task_id = t.task_id ' .
+            $where_runs;
         $count = $this->Evaluation->query($sql_test)[0]->count;
       }
       if ($count > $result_limit) {
@@ -200,12 +284,12 @@ private function evaluation_list($segs, $user_id, $show_params) {
         }
       }
     }
-    
+
     $order_by = null;
     if ($sort_order) {
       $order_by = 'ORDER BY `e`.`value` ' . $sort_order;
     }
-    
+
     if ($per_fold) {
       $eval_table = 'evaluation_fold';
       $columns = 'NULL as value, NULL as array_data, CONCAT("[", GROUP_CONCAT(e.value), "]") AS `values`';
@@ -222,17 +306,17 @@ private function evaluation_list($segs, $user_id, $show_params) {
     // TODO: remove dependency on task_inputs and dataset table
     // TODO (2): transform into subquery where all columns except evaluation_fold are obtained in subquery (along with limit requirements, as MYSQL query optimizer does not seem to understand this query has an upper limit to the number of obtained runs that need to be inspected)
     $sql =
-      'SELECT r.rid, r.task_id, r.start_time, r.uploader, s.implementation_id, s.sid, f.name AS `function`, i.fullName, d.did, d.name, e.evaluation_engine_id, ' . $columns . ' ' .
-      'FROM run r, ' . $eval_table . ' e, algorithm_setup s, implementation i, dataset d, task t, task_inputs ti, math_function f ' .
-      'WHERE r.setup = s.sid ' .
-      'AND e.source = r.rid ' .
-      'AND e.function_id = f.id ' . 
-      'AND s.implementation_id = i.id ' .
-      'AND r.task_id = t.task_id ' .
-      'AND r.task_id = ti.task_id ' .
-      'AND ti.input = "source_data" ' .
-      'AND ti.value = d.did ' . $where_total;
-    
+        'SELECT r.rid, r.task_id, r.start_time, r.uploader, s.implementation_id, s.sid, f.name AS `function`, i.fullName, d.did, d.name, e.evaluation_engine_id, ' . $columns . ' ' .
+        'FROM run r, ' . $eval_table . ' e, algorithm_setup s, implementation i, dataset d, task t, task_inputs ti, math_function f ' .
+        'WHERE r.setup = s.sid ' .
+        'AND e.source = r.rid ' .
+        'AND e.function_id = f.id ' .
+        'AND s.implementation_id = i.id ' .
+        'AND r.task_id = t.task_id ' .
+        'AND r.task_id = ti.task_id ' .
+        'AND ti.input = "source_data" ' .
+        'AND ti.value = d.did ' . $where_total;
+
     if ($group_by) {
       $sql .= $group_by;
     }
@@ -242,14 +326,14 @@ private function evaluation_list($segs, $user_id, $show_params) {
     if ($where_limit) {
       $sql .= $where_limit;
     }
-    
+
     $res = $this->Evaluation->query($sql);
 
     if ($res == false) {
       $this->returnError(542, $this->version);
       return;
     }
-    
+
     if ($show_params) {
       # 2 stage query .. unfortunately. Can break when too much results. let's take the damage for now
       $setup_ids = array();
diff --git a/openml_OS/models/api/v1/Api_evaluationmeasure.php b/openml_OS/models/api/v1/Api_evaluationmeasure.php
index 7cde44a37..e3c5aadfe 100644
--- a/openml_OS/models/api/v1/Api_evaluationmeasure.php
+++ b/openml_OS/models/api/v1/Api_evaluationmeasure.php
@@ -1,4 +1,5 @@
 <?php
+
 class Api_evaluationmeasure extends MY_Api_Model {
   
   protected $version = 'v1';
@@ -14,7 +15,7 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $getpost = array('get','post');
     
     $this->outputFormat = $format;
-    
+
     if (count($segments) == 1 && $segments[0] == 'list') {
       $this->evaluationmeasure_list();
       return;
@@ -23,8 +24,34 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     
     $this->returnError( 100, $this->version );
   }
-  
-  
+
+  /**
+   *@OA\Get(
+   *	path="/evaluationmeasure/list",
+   *	tags={"evaluationmeasure"},
+   *	summary="List all evaluation measures",
+   *	description="Returns an array with all model evaluation measures in the system.",
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of evaluation measures",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/EvaluationMeasureList",
+   *			example={
+   *			  "evaluation_measures":{
+   *			    "measures":{
+   *			      "measure":{
+   *			        "area_under_roc_curve",
+   *			        "average_cost",
+   *			        "binominal_test",
+   *			        "build_cpu_time"
+   *			        }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *)
+   */
   private function evaluationmeasure_list() {
     $data = new stdClass();
     $data->measures = $this->Math_function->getWhere( 'functionType = "EvaluationFunction"' );
diff --git a/openml_OS/models/api/v1/Api_flow.php b/openml_OS/models/api/v1/Api_flow.php
index 7c1f61f55..142f80e8a 100644
--- a/openml_OS/models/api/v1/Api_flow.php
+++ b/openml_OS/models/api/v1/Api_flow.php
@@ -1,4 +1,5 @@
 <?php
+
 class Api_flow extends MY_Api_Model {
 
   protected $version = 'v1';
@@ -65,12 +66,12 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     if (count($segments) == 1 && $segments[0] == 'tag' && $request_type == 'post') {
-      $this->entity_tag_untag('implementation', $this->input->post('flow_id'), $this->input->post('tag'), false, 'flow');
+      $this->flow_tag($this->input->post('flow_id'), $this->input->post('tag'));
       return;
     }
 
     if (count($segments) == 1 && $segments[0] == 'untag' && $request_type == 'post') {
-      $this->entity_tag_untag('implementation', $this->input->post('flow_id'), $this->input->post('tag'), true, 'flow');
+      $this->flow_untag($this->input->post('flow_id'), $this->input->post('tag'));
       return;
     }
 
@@ -82,7 +83,72 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError( 100, $this->version );
   }
 
-
+  /**
+   *@OA\Get(
+   *	path="/flow/list/{filters}",
+   *	tags={"flow"},
+   *	summary="List and filter flows",
+   *	description="List flows, possibly filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/task/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all flows that match the constraints.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.
+  /tag/{tag} - returns only tasks tagged with the given tag.
+  /uploader/{id} - return only evaluations uploaded by a specific user, specified by user ID.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of flows",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/FlowList",
+   *			example={
+   *			  "flows":
+   *			    {
+   *			      "flow":{
+   *			        {
+   *			          "id":"65",
+   *			          "full_name":"weka.RandomForest(1)",
+   *			          "name":"weka.RandomForest",
+   *			          "version":"1",
+   *			          "external_version":"Weka_3.7.10_9186",
+   *			          "uploader":"1"
+   *			        },
+   *			        {
+   *			          "id":"66",
+   *			          "full_name":"weka.IBk(1)",
+   *			          "name":"weka.IBk",
+   *			          "version":"1",
+   *			          "external_version":"Weka_3.7.10_8034",
+   *			          "uploader":"1"
+   *			        },
+   *			        {
+   *			          "id":"67",
+   *			          "full_name":"weka.BayesNet_K2(1)",
+   *			          "name":"weka.BayesNet_K2",
+   *			          "version":"1",
+   *			          "external_version":"Weka_3.7.10_8034",
+   *			          "uploader":"1"
+   *			        }
+   *			      }
+   *			    }
+   *			  }
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n500 - No results. There where no matches for the given constraints.\n501 - Illegal filter specified.\n502 - Filter values/ranges not properly specified.\n503 - Can not specify an offset without a limit.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function flow_list($segs) {
     $legal_filters = array('uploader', 'tag', 'limit', 'offset');
     
@@ -136,7 +202,186 @@ private function flow_list($segs) {
     $this->xmlContents('implementations', $this->version, array('implementations' => $implementations_res));
   }
 
+  /**
+   *@OA\Post(
+   *	path="/flow/tag",
+   *	tags={"flow"},
+   *	summary="Tag a flow",
+   *	description="Tags a flow.",
+   *	@OA\Parameter(
+   *		name="flow_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the flow.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the tagged flow",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_tag",
+   *				ref="#/components/schemas/inline_response_200_12_flow_tag",
+   *			),
+   *			example={
+   *			  "flow_tag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function flow_tag($flow_id, $tag) {
+    $this->flow_tag_untag($flow_id,$tag, false);
+  }
+
+  /**
+   *@OA\Post(
+   *	path="/flow/untag",
+   *	tags={"flow"},
+   *	summary="Untag a flow",
+   *	description="Untags a flow.",
+   *	@OA\Parameter(
+   *		name="flow_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the flow.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the untagged flow",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_untag",
+   *				ref="#/components/schemas/inline_response_200_13_flow_untag",
+   *			),
+   *			example={
+   *			  "flow_untag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\nby another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function flow_untag($flow_id, $tag) {
+    $this->flow_tag_untag($flow_id, $tag, true);
+  }
+
+  private function flow_tag_untag($flow_id, $tag, $do_untag) {
+    //forward execution of logic to superclass.
+    $this->entity_tag_untag('implementation', $flow_id, $tag, $do_untag, 'flow');
+  }
+
 
+  /**
+   *@OA\Get(
+   *	path="/flow/exists/{name}/{version}",
+   *	tags={"flow"},
+   *	summary="Check whether flow exists",
+   *	description="Checks whether a flow with the given name and (external) version exists.",
+   *	@OA\Parameter(
+   *		name="name",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="The name of the flow.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="version",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="The external version of the flow",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of flows",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_exists",
+   *				ref="#/components/schemas/inline_response_200_10_flow_exists",
+   *			),
+   *			example={
+   *			  "flow_exists": {
+   *			    "exists": "true",
+   *			    "id": "65"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n330 - Mandatory fields not present. Please provide name and external_version.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function flow_exists($name, $external_version) {
 
     $similar = false;
@@ -154,7 +399,65 @@ private function flow_exists($name, $external_version) {
     $this->xmlContents( 'implementation-exists', $this->version, $result );
   }
 
-  // TODO: check what is going wrong with implementation id 1
+  /**
+   *@OA\Get(
+   *	path="/flow/{id}",
+   *	tags={"flow"},
+   *	summary="Get flow description",
+   *	description="Returns information about a flow. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the flow.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A flow description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Flow",
+   *			example={
+   *			  "flow": {
+   *			    "id":"100",
+   *			    "uploader":"1",
+   *			    "name":"weka.J48",
+   *			    "version":"2",
+   *			    "external_version":"Weka_3.7.5_9117",
+   *			    "description":"...",
+   *			    "upload_date":"2014-04-23 18:00:36",
+   *			    "language":"Java",
+   *			    "dependencies":"Weka_3.7.5",
+   *			    "parameter": {
+   *			      {
+   *			        "name":"A",
+   *			        "data_type":"flag",
+   *			        "default_value":{},
+   *			        "description":"Laplace smoothing..."
+   *			      },
+   *			      {
+   *			        "name":"C",
+   *			        "data_type":"option",
+   *			        "default_value":"0.25",
+   *			        "description":"Set confidence threshold..."
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n180 - Please provide flow id.\n181 - Unknown flow. The flow with the given ID was not found in the database.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  //  TODO: check what is going wrong with implementation id 1
   private function flow($id) {
     if( $id == false ) {
       $this->returnError( 180, $this->version );
@@ -171,6 +474,64 @@ private function flow($id) {
     $this->xmlContents( 'implementation-get', $this->version, array( 'source' => $implementation ) );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/flow",
+   *	tags={"flow"},
+   *	summary="Upload a flow",
+   *	description="Uploads a flow. Upon success, it returns the flow id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the flow. Only name and description are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.implementation.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/flow).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="flow",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="The actual flow, being a source (or binary) file.",
+   *		required=false,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the uploaded flow",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_flow",
+   *				ref="#/components/schemas/inline_response_200_9_upload_flow",
+   *			),
+   *			example={
+   *			  "upload_flow": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n160 - Error in file uploading. There was a problem with the file upload.\n161 - Please provide description xml.\n163 - Problem validating uploaded description file. The XML description format does not meet the standards.\n164 - Flow already stored in database. Please change name or version number\n165 - Failed to insert flow. There can be many causes for this error. If you included the implements field, it should be an existing entry in the algorithm or math_function table. Otherwise it could be an internal server error. Please contact API support team.\n166 - Failed to add flow to database. Internal server error, please contact API administrators\n167 - Illegal files uploaded. An non required file was uploaded.\n168 - The provided md5 hash equals not the server generated md5 hash of the file.\n169 - Please provide API key. In order to share content, please authenticate and provide API key.\n170 - Authentication failed. The API key was not valid. Please try to login again, or contact API administrators\n171 - Flow already exists. This flow is already in the database\n172 - XSD not found. Please contact API support team\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function flow_upload() {
 
     if(isset($_FILES['source']) && $_FILES['source']['error'] == 0) {
@@ -297,6 +658,55 @@ private function flow_upload() {
     $this->xmlContents( 'implementation-upload', $this->version, $implementation );
   }
 
+  /**
+   *@OA\Delete(
+   *	path="/flow/{id}",
+   *	tags={"flow"},
+   *	summary="Delete a flow",
+   *	description="Deletes a flow. Upon success, it returns the ID of the deleted flow.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the flow.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="API key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted flow",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_delete",
+   *				ref="#/components/schemas/inline_response_200_8_flow_delete",
+   *			),
+   *			example={
+   *			  "flow_delete": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n320 - Please provide API key. In order to remove your content, please authenticate.\n321 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n322 - Flow does not exists. The flow ID could not be linked to an existing flow.\n323 - Flow is not owned by you. The flow is owned by another user. Hence you cannot delete it.\n324 - Flow is in use by other content. Can not be deleted. The flow is used in runs, evaluations or as a component of another flow. Delete other content before deleting this flow.\n325 - Deleting flow failed. Deleting the flow failed. Please contact\nsupport team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function flow_delete($flow_id) {
 
     $implementation = $this->Implementation->getById($flow_id);
diff --git a/openml_OS/models/api/v1/Api_run.php b/openml_OS/models/api/v1/Api_run.php
index bf014e4f8..be5add8a5 100644
--- a/openml_OS/models/api/v1/Api_run.php
+++ b/openml_OS/models/api/v1/Api_run.php
@@ -1,6 +1,11 @@
 <?php
+
+
 class Api_run extends MY_Api_Model {
 
+  const ENTITY_TYPE = "run";
+  const ENTITY_SPECIAL_NAME = "run";
+
   protected $version = 'v1';
 
   function __construct() {
@@ -84,12 +89,12 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     if (count($segments) == 1 && $segments[0] == 'tag' && $request_type == 'post') {
-      $this->entity_tag_untag('run', $this->input->post('run_id'), $this->input->post('tag'), false, 'run');
+      $this->run_tag($this->input->post('run_id'), $this->input->post('tag'));
       return;
     }
 
     if (count($segments) == 1 && $segments[0] == 'untag' && $request_type == 'post') {
-      $this->entity_tag_untag('run', $this->input->post('run_id'), $this->input->post('tag'), true, 'run');
+      $this->run_untag($this->input->post('run_id'), $this->input->post('tag'));
       return;
     }
     
@@ -101,7 +106,170 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError( 100, $this->version );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/run/tag",
+   *	tags={"run"},
+   *	summary="Tag a run",
+   *	description="Tags a run.",
+   *	@OA\Parameter(
+   *		name="run_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the tagged run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="run_tag",
+   *				ref="#/components/schemas/inline_response_200_19_run_tag",
+   *			),
+   *			example={
+   *			  "run_tag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function run_tag($run_id, $tag) {
+    $this->entity_tag_untag(self::ENTITY_TYPE, $run_id, $tag, false, self::ENTITY_SPECIAL_NAME);
+  }
+
+  /**
+   *@OA\Post(
+   *	path="/run/untag",
+   *	tags={"run"},
+   *	summary="Untag a run",
+   *	description="Untags a run.",
+   *	@OA\Parameter(
+   *		name="run_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the untagged run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="run_untag",
+   *				ref="#/components/schemas/inline_response_200_20_run_untag",
+   *			),
+   *			example={
+   *			  "run_untag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function run_untag($run_id, $tag) {
+    $this->entity_tag_untag(self::ENTITY_TYPE, $run_id, $tag, true, self::ENTITY_SPECIAL_NAME);
+  }
 
+  /**
+   *@OA\Get(
+   *	path="/run/list/{filters}",
+   *	tags={"run"},
+   *	summary="List and filter runs",
+   *	description="List runs, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/run/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all runs that match the constraints. A maximum of 10,000 results are returned, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain tasks, flows, setups, or uploaders.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /tag/{tag} - return only runs tagged with the given tag.
+  /run/{ids} - return only specific runs, specified as a comma-separated list of run IDs, e.g. ''1,2,3''
+  /task/{ids} - return only runs on specific tasks, specified as a comma-separated list of task IDs, e.g. ''1,2,3''
+  /flow/{ids} - return only runs on specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''
+  /setup/{ids} - return only runs with specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''
+  /uploader/{ids} - return only runs uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of runs descriptions",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/RunList",
+   *			example={"runs": {"run": {{"upload_time": "2014-04-06 23:30:40", "task_id": "28", "run_id": "100", "error_message": {}, "setup_id": "12", "uploader": "1", "flow_id": "67"}, {"upload_time": "2014-04-06 23:30:40", "task_id": "48", "run_id": "101", "error_message": {}, "setup_id": "6", "uploader": "1", "flow_id": "61"}, {"upload_time": "2014-04-06 23:30:40", "task_id": "41", "run_id": "102", "error_message": {}, "setup_id": "3", "uploader": "1", "flow_id": "58"}}}}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n510 - Please provide at least task, flow or setup, uploader or run, to filter results, or limit the number of responses. The number of runs is huge. Please limit the result space.\n511 - Input not safe. The input parameters (task_id, setup_id, flow_id, run_id, uploader_id) did not meet the constraints (comma separated list of integers).\n512 - There where no results. Check whether there are runs under the given constraint.\n513 - Too many results. Given the constraints, there were still too many results. Please add filters to narrow down the list.\n514 - Illegal filter specified.\n515 - Offset specified without limit.\n516 - Requested result limit too high.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_list($segs, $user_id) {
     $result_limit = 10000;
     $legal_filters = array('task', 'setup', 'flow', 'uploader', 'run', 'tag', 'limit', 'offset', 'task_type', 'study', 'show_errors');
@@ -203,7 +371,118 @@ private function run_list($segs, $user_id) {
     $this->xmlContents('runs', $this->version, array('runs' => $res));
   }
 
-
+  /**
+   *@OA\Get(
+   *	path="/run/{id}",
+   *	tags={"run"},
+   *	summary="Get run description",
+   *	description="Returns information about a run. The information includes the name, information about the creator, dependencies, parameters, run instructions and more.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A run description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Run",
+   *			example={
+   *			  "run": {
+   *			    "run_id":"100",
+   *			    "uploader":"1",
+   *			    "uploader_name":"Jan van Rijn",
+   *			    "task_id":"28",
+   *			    "task_type":"Supervised Classification",
+   *			    "task_evaluation_measure":"predictive_accuracy",
+   *			    "flow_id":"67",
+   *			    "flow_name":"weka.BayesNet_K2(1)",
+   *			    "setup_string":"weka.classifiers.bayes.BayesNet -- -D -Q weka.classifiers.bayes.net.search.local.K2 -- -P 1 -S BAYES -E weka.classifiers.bayes.net.estimate.SimpleEstimator -- -A 0.5",
+   *			    "parameter_setting": {
+   *			      {
+   *			        "name":"D",
+   *			        "value":"true"
+   *			      },
+   *			      {
+   *			        "name":"Q",
+   *			        "value":"weka.classifiers.bayes.net.search.local.K2"
+   *			      },
+   *			      {
+   *			        "name":"P",
+   *			        "value":"1"
+   *			      },
+   *			      {
+   *			        "name":"S",
+   *			        "value":"BAYES"
+   *			      }
+   *			    },
+   *			    "input_data":
+   *			      {
+   *			        "dataset":
+   *			          {
+   *			            "did":"28",
+   *			            "name":"optdigits",
+   *			            "url":"https:\\/\\/www.openml.org\\/data\\/download\\/28\\/dataset_28_optdigits.arff"
+   *			          }
+   *			      },
+   *			    "output_data":
+   *			      {
+   *			        "file": {
+   *			          {
+   *			            "did":"48838",
+   *			            "file_id":"261",
+   *			            "name":"description",
+   *			            "url":"https:\\/\\/www.openml.org\\/data\\/download\\/261\\/weka_generated_run935374685998857626.xml"
+   *			          },
+   *			          {
+   *			            "did":"48839",
+   *			            "file_id":"262",
+   *			            "name":"predictions",
+   *			            "url":"https:\\/\\/www.openml.org\\/data\\/download\\/262\\/weka_generated_predictions576954524972002741.arff"
+   *			          }
+   *			        },
+   *			        "evaluation": {
+   *			          {
+   *			            "name":"area_under_roc_curve",
+   *			            "flow_id":"4",
+   *			            "value":"0.990288",
+   *			            "array_data":"[0.99724,0.989212,0.992776,0.994279,0.980578,0.98649,0.99422,0.99727,0.994858,0.976143]"
+   *			          },
+   *			          {
+   *			            "name":"confusion_matrix",
+   *			            "flow_id":"10",
+   *			            "array_data":"[[544,1,0,0,7,0,1,0,0,1],[0,511,21,1,0,1,3,1,5,28],[0,7,511,1,0,1,0,3,23,11],[0,2,2,519,0,3,0,12,16,18],[0,3,0,0,528,0,4,21,6,6],[0,1,0,7,5,488,2,0,4,51],[1,7,0,0,2,0,548,0,0,0],[0,2,0,1,9,1,0,545,4,4],[1,25,2,2,3,6,2,1,503,9],[0,7,0,20,16,5,0,19,9,486]]"
+   *			          },
+   *			          {
+   *			            "name":"f_measure",
+   *			            "flow_id":"12",
+   *			            "value":"0.922723",
+   *			            "array_data":"[0.989091,0.898857,0.935041,0.92431,0.927944,0.918156,0.980322,0.933219,0.895018,0.826531]"
+   *			          },
+   *			          {
+   *			            "name":"kappa",
+   *			            "flow_id":"13",
+   *			            "value":"0.913601"
+   *			          }
+   *			        }
+   *			      }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n220 - Please provide run ID. In order to view run details, please provide the run ID.\n221 - Run not found. The run ID was invalid, run does not exist (anymore).\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run($run_id) {
     if( $run_id == false ) {
       $this->returnError( 235, $this->version );
@@ -230,6 +509,55 @@ private function run($run_id) {
     $this->xmlContents( 'run-get', $this->version, array( 'source' => $run ) );
   }
 
+  /**
+   *@OA\Delete(
+   *	path="/run/{id}",
+   *	tags={"run"},
+   *	summary="Delete run",
+   *	description="Deletes a run. Upon success, it returns the ID of the deleted run.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="API key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="data_delete",
+   *				ref="#/components/schemas/inline_response_200_17_data_delete",
+   *			),
+   *			example={
+   *			  "run_delete": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n390 - Please provide API key. In order to remove your content, please authenticate.\n391 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators\n392 - Run does not exists. The run ID could not be linked to an existing run.\n393 - Run is not owned by you. The run was owned by another user. Hence you cannot delete it.\n394 - Deleting run failed. Deleting the run failed. Please contact support team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_delete($run_id) {
 
     $run = $this->Run->getById( $run_id );
@@ -275,7 +603,46 @@ private function run_delete($run_id) {
     $this->xmlContents( 'run-delete', $this->version, array( 'run' => $run ) );
   }
 
-
+  /**
+   *@OA\Get(
+   *	path="/run/reset/{id}",
+   *	tags={"run"},
+   *	summary="Resets a run.",
+   *	description="Removes all run evaluations. When a run is reset, the runs will automatically be evaluated as soon as they are picked up by the evaluation engine again.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Run ID.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the evaluated run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="run_reset",
+   *				ref="#/components/schemas/inline_response_200_21_upload_flow",
+   *			),
+   *			example={
+   *			  "run_reset": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n412 - Run does not exist\n413 - Run is not owned by you\n394 - Resetting run failed\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_reset($run_id) {
 
     $run = $this->Run->getById( $run_id );
@@ -298,6 +665,82 @@ private function run_reset($run_id) {
     $this->xmlContents( 'run-reset', $this->version, array( 'run' => $run ) );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/run",
+   *	tags={"run"},
+   *	summary="Upload run",
+   *	description="Uploads a run. Upon success, it returns the run id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the dataset. Only name, description, and data format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/run).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="predictions",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="The predictions generated by the run",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="model_readable",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="The human-readable model generated by the run",
+   *		required=false,
+   *	),
+   *	@OA\Parameter(
+   *		name="model_serialized",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="The serialized model generated by the run",
+   *		required=false,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the uploaded run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_flow",
+   *				ref="#/components/schemas/inline_response_200_18_upload_flow",
+   *			),
+   *			example={
+   *			  "upload_run": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n201 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n202 - Please provide run XML.\n203 - Could not validate run xml by XSD. Please double check that the xml is valid.\n204 - Unknown task. The task with the given ID was not found in the database.\n205 - Unknown flow. The flow with the given ID was not found in the database.\n206 - Invalid number of files. The number of uploaded files did not match the number of files expected for the task type\n207 - File upload failed. One of the files uploaded has a problem.\n208 - Error inserting setup record. Please contact api administrators\n210 - Unable to store run. Please contact api administrators.\n211 - Dataset not in database. One of the datasets of the task was not included in database, please contact api administrators.\n212 - Unable to store file. Please contact api administrators.\n213 - Parameter in run xml unknown. One of the parameters provided in the run xml is not registered as parameter for the flow nor its components.\n214 - Unable to store input setting. Please contact API support team.\n215 - Unable to evaluate predictions. Please contact API support team.\n216 - Error thrown by Java Application. Additional information field is provided.\n217 - Error processing output data. Unknown or inconsistent evaluation measure. One of the provided evaluation measures could not be matched with a record in the math_function or flow table.\n218 - Wrong flow associated with run. The flow implements a math_function, which is unable to generate predictions. Please select another flow.\n219 - Error reading the XML document. The XML description file could not be verified.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_upload() {
 
     /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
@@ -600,6 +1043,62 @@ private function run_upload() {
     $this->xmlContents( 'run-upload', $this->version, $result );
   }
 
+  /**
+   *@OA\Get(
+   *	path="/run/trace/{id}",
+   *	tags={"run"},
+   *	summary="Get run trace",
+   *	description="Returns the optimization trace of run. The trace contains every setup tried, its evaluation, and whether it was selected.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A run trace description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/RunTrace",
+   *			example={
+   *			  "trace": {
+   *			    "run_id":"573055",
+   *			    "trace_iteration": {
+   *			        "repeat":"0",
+   *			        "fold":"0",
+   *			        "repeat":"0",
+   *			        "iteration":"0",
+   *			        "setup_string":{"parameter_minNumObj": "1",
+   *			                        "parameter_confidenceFactor": "0.1"},
+   *			        "evaluation":"94.814815",
+   *			        "selected": "true"
+   *			      },
+   *			    "trace_iteration": {
+   *			        "repeat":"0",
+   *			        "fold":"0",
+   *			        "repeat":"0",
+   *			        "iteration":"1",
+   *			        "setup_string":{"parameter_minNumObj": "1",
+   *			                        "parameter_confidenceFactor": "0.25"},
+   *			        "evaluation": "94.074074",
+   *			        "selected": "true"
+   *			      }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n570 - No successful trace associated with this run\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_trace($run_id) {
     $run = $this->Run->getById($run_id);
     if ($run === false) {
@@ -616,6 +1115,64 @@ private function run_trace($run_id) {
     $this->xmlContents('run-trace-get', $this->version, array('run_id' => $run_id, 'trace' => $trace));
   }
 
+  /**
+   *@OA\Post(
+   *	path="/run/trace/{id}",
+   *	tags={"run"},
+   *	summary="Upload run trace",
+   *	description="Uploads a run trace. Upon success, it returns the run id.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the run.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the trace. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.trace) and an [XML example](https://www.openml.org/api/v1/xml_example/run.trace).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the run with the trace",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_flow",
+   *				ref="#/components/schemas/inline_response_200_23_upload_flow",
+   *			),
+   *			example={
+   *			  "run_trace": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n561 - Problem with uploaded trace file.\n562 - Problem validating xml trace file.\n563 - Problem loading xml trace file.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_trace_upload() {
     if (!$this->user_has_admin_rights) {
       $this->returnError(106, $this->version);
@@ -666,6 +1223,55 @@ private function run_trace_upload() {
     $this->xmlContents('run-trace', $this->version, array('run_id' => $run_id));
   }
 
+  /**
+   *@OA\Post(
+   *	path="/run/evaluate",
+   *	tags={"run"},
+   *	summary="Uploads a run evaluation",
+   *	description="Uploads a run evaluation. When successful, it returns the run id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the run evaluation.Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.run.evaluate) and an [XML example](https://www.openml.org/api/v1/xml_example/run.evaluate).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the evaluated run",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_flow",
+   *				ref="#/components/schemas/inline_response_200_21_upload_flow",
+   *			),
+   *			example={
+   *			  "run_evaluate": {
+   *			    "id": "2520"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n422 - Upload problem description XML\n423 - Problem validating uploaded description file\n424 - Problem opening description xml\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function run_evaluate() {
     $timestamps = array(microtime(true)); // profiling 0
     if (!$this->user_has_admin_rights) {
diff --git a/openml_OS/models/api/v1/Api_setup.php b/openml_OS/models/api/v1/Api_setup.php
index ca66360d1..fff2dd4bd 100644
--- a/openml_OS/models/api/v1/Api_setup.php
+++ b/openml_OS/models/api/v1/Api_setup.php
@@ -41,12 +41,12 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     if (count($segments) == 1 && $segments[0] == 'tag' && $request_type == 'post') {
-      $this->entity_tag_untag('algorithm_setup', $this->input->post('setup_id'), $this->input->post('tag'), false, 'setup');
+      $this->setup_tag( $this->input->post('setup_id'), $this->input->post('tag'));
       return;
     }
 
     if (count($segments) == 1 && $segments[0] == 'untag' && $request_type == 'post') {
-      $this->entity_tag_untag('algorithm_setup', $this->input->post('setup_id'), $this->input->post('tag'), true, 'setup');
+      $this->setup_untag($this->input->post('setup_id'), $this->input->post('tag'));
       return;
     }
     
@@ -84,6 +84,196 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError( 100, $this->version );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/setup/tag",
+   *	tags={"setup"},
+   *	summary="Tag a setup",
+   *	description="Tags a setup.",
+   *	@OA\Parameter(
+   *		name="setup_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the setup.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the tagged setup",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_tag",
+   *				ref="#/components/schemas/inline_response_200_15_flow_tag",
+   *			),
+   *			example={
+   *			  "setup_tag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function setup_tag($setup_id, $tag) {
+    $this->setup_tag_untag($setup_id, $tag, false);
+  }
+
+  /**
+   *@OA\Post(
+   *	path="/setup/untag",
+   *	tags={"setup"},
+   *	summary="Untag a setup",
+   *	description="Untags a setup.",
+   *	@OA\Parameter(
+   *		name="setup_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the setup.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the untagged setup",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="flow_untag",
+   *				ref="#/components/schemas/inline_response_200_16_flow_untag",
+   *			),
+   *			example={
+   *			  "setup_untag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, flow, run} not found. The provided entity_id {data_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged\nby another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function setup_untag($setup_id, $tag) {
+    $this->setup_tag_untag($setup_id, $tag, true);
+  }
+
+  private function setup_tag_untag($setup_id, $tag, $do_untag) {
+    $this->entity_tag_untag('algorithm_setup', $setup_id, $tag, $do_untag, 'setup');
+  }
+
+  /**
+   *@OA\Get(
+   *	path="/setup/{id}",
+   *	tags={"setup"},
+   *	summary="Get a hyperparameter setup",
+   *	description="Returns information about a setup. The information includes the list of hyperparameters, with name, value, and default value.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the hyperparameter setup (configuration). These IDs are stated in run descriptions.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A setup description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Setup",
+   *			example={
+   *			  "setup_parameters":{
+   *			    "flow_id":"59",
+   *			    "parameter":{
+   *			      {
+   *			        "full_name":"weka.JRip(1)_F",
+   *			        "parameter_name":"F",
+   *			        "data_type":"option",
+   *			        "default_value":"3",
+   *			        "value":"3"
+   *			      },{
+   *			        "full_name":"weka.JRip(1)_N",
+   *			        "parameter_name":"N",
+   *			        "data_type":"option",
+   *			        "default_value":"2.0",
+   *			        "value":"2.0"
+   *			      },{
+   *			        "full_name":"weka.JRip(1)_O",
+   *			        "parameter_name":"O",
+   *			        "data_type":"option",
+   *			        "default_value":"2",
+   *			        "value":"2"
+   *			      },{
+   *			        "full_name":"weka.JRip(1)_S",
+   *			        "parameter_name":"S",
+   *			        "data_type":"option",
+   *			        "default_value":"1",
+   *			        "value":"1"
+   *			      }}
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n280 - Please provide setup ID. In order to view setup details, please provide the run ID\n281 - Setup not found. The setup ID was invalid, or setup does not exist (anymore).\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function setup($setup_id) {
     if($setup_id == false) {
       $this->returnError(280, $this->version);
@@ -105,7 +295,75 @@ private function setup($setup_id) {
       $this->xmlContents('setup-parameters', $this->version, array('parameters' => $this->parameters, 'setup' => $setup));
     }
   }
-  
+
+  /**
+   *@OA\Get(
+   *	path="/setup/list/{filters}",
+   *	tags={"setup"},
+   *	summary="List and filter setups",
+   *	description="List setups, filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/setup/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all evaluations that match the constraints. A maximum of 1,000 results are returned at a time, an error is returned if the result set is bigger. Use pagination (via limit and offset filters), or limit the results to certain flows, setups, or tags.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /tag/{tag} - returns only setups tagged with the given tag.
+  /flow/{ids} - return only setups for specific flows, specified as a comma-separated list of flow IDs, e.g. ''1,2,3''
+  /setup/{ids} - return only specific setups, specified as a comma-separated list of setup IDs, e.g. ''1,2,3''
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of setup descriptions",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/SetupList",
+   *			example={
+   *			  "setups": {
+   *			    "setup": {
+   *			      {
+   *			        "setup_id":"10",
+   *			        "flow_id":"65",
+   *			        "parameter": {
+   *			          {
+   *			            "id":"4144",
+   *			            "flow_id":"65",
+   *			            "flow_name":"weka.RandomForest",
+   *			            "full_name":"weka.RandomForest(1)_I",
+   *			            "parameter_name":"I",
+   *			            "data_type":"option",
+   *			            "default_value":"10",
+   *			            "value":"10"
+   *			          },
+   *			          {
+   *			            "id":"4145",
+   *			            "flow_id":"65",
+   *			            "flow_name":"weka.RandomForest",
+   *			            "full_name":"weka.RandomForest(1)_K",
+   *			            "parameter_name":"K",
+   *			            "data_type":"option",
+   *			            "default_value":"0",
+   *			            "value":"0"
+   *			          }
+   *			        }
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n670 - Please specify at least one filter.\n671 - Illegal filter.\n672 - Illegal filter input.\n673 - Result set too big. Please use one of the filters or the limit option.\n674 - No results, please check the filter.\n675 - Cannot specify offset without limit.\n676 - Requested result limit too high.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   function setup_list($segs) { 
     $result_limit = 1000;
     $legal_filters = array('flow', 'setup', 'limit', 'offset', 'tag');
@@ -280,6 +538,55 @@ private function setup_exists($partial) {
     }
   }
 
+  /**
+   *@OA\Delete(
+   *	path="/setup/{id}",
+   *	tags={"setup"},
+   *	summary="Delete setup",
+   *	description="Deletes a setup. Upon success, it returns the ID of the deleted setup.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the setup.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted setup",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="study_delete",
+   *				ref="#/components/schemas/inline_response_200_14_study_delete",
+   *			),
+   *			example={
+   *			  "setup_delete": {
+   *			    "id": "1"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n401 - Authentication failed. Please provide API key. In order to remove your content, please authenticate.\n402 - Setup does not exists. The setup ID could not be linked to an existing setup.\n404 - Setup deletion failed. Setup is in use by other content (runs, schedules, etc). Can not be deleted.\n405 - Setup deletion failed. Please try again later.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function setup_delete($setup_id) {
 
     $setup = $this->Algorithm_setup->getById( $setup_id );
diff --git a/openml_OS/models/api/v1/Api_study.php b/openml_OS/models/api/v1/Api_study.php
index 8bb45dfb6..d4edd56c3 100644
--- a/openml_OS/models/api/v1/Api_study.php
+++ b/openml_OS/models/api/v1/Api_study.php
@@ -21,7 +21,7 @@ function bootstrap($format, $segments, $request_type, $user_id) {
       $this->study_list($segments);
       return;
     }
-    
+
     if (count($segments) == 0) {
       $this->study_create();
       return;
@@ -31,19 +31,19 @@ function bootstrap($format, $segments, $request_type, $user_id) {
       $this->status_update($this->input->post('study_id'), $this->input->post('status'));
       return;
     }
-    
+
     if (count($segments) == 1 && is_numeric($segments[0]) && $request_type == 'delete') {
       $this->study_delete($segments[0]);
       return;
     }
-    
+
     if (count($segments) == 2 && is_numeric($segments[0]) && $segments[1] == 'attach' && $request_type == 'post') {
-      $this->study_attach_detach($segments[0], true);
+      $this->study_attach($segments[0]);
       return;
     }
-    
+
     if (count($segments) == 2 && is_numeric($segments[0]) && $segments[1] == 'detach' && $request_type == 'post') {
-      $this->study_attach_detach($segments[0], false);
+      $this->study_detach($segments[0]);
       return;
     }
     
@@ -52,19 +52,62 @@ function bootstrap($format, $segments, $request_type, $user_id) {
       if (count($segments) == 2) {
         $type = $segments[1];
       }
-
-      if (is_numeric($segments[0])) {
-        $this->study_by_id($segments[0], $type);
-        return;
-      } else {
-        $this->study_by_alias($segments[0], $type);
-        return;
-      }
+      $this->study_get($segments[0], $type);
+      return;
     }
 
     $this->returnError(100, $this->version);
   }
-  
+
+  /**
+   *@OA\Post(
+   *	path="/study",
+   *	tags={"study"},
+   *	summary="Create new study",
+   *	description="Creates a new study. Upon success, it returns the study id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the study. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.study.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/study).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the uploaded study",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_study",
+   *				ref="#/components/schemas/inline_response_200_25_upload_study",
+   *			),
+   *			example={
+   *			  "upload_study": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n1031 - Description file not present. Please upload the study description.\n1032 - Problem validating uploaded description file. The XML description format does not meet the standards. See the XSD schema.\n1033 - Illegal main entity type. Currently only collections of tasks and can be created.\n1034 - Linked entities are not of the correct type fot this study.\n1035 - Benchmark suites can only be linked to run studies.\n1036 - Referred benchmark suite cannot be found.\n1037 - Referred benchmark suite should be a task collection.\n1038 - Study alias is not unique.\n1039 - Dataset insertion problem. Please contact the administrators.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function study_create() {
     $xsdFile = xsd('openml.study.upload', $this->controller, $this->version);
 
@@ -206,7 +249,135 @@ private function status_update($study_id, $status) {
     
     $this->xmlContents('study-status-update', $this->version, $template_vars);
   }
-  
+
+  /**
+   *@OA\Post(
+   *	path="/study/{id}/attach",
+   *	tags={"study"},
+   *	summary="Attach a new entity to a study",
+   *	description="Attach a new entity to an exising study. Upon success, it returns the study id, type, and linked entities.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the study. Supplied in the URL path.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="ids",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Comma-separated list of entity IDs to be attached to the study. For instance, if this is a run study, the list of run IDs that need to be added (attached) to the study. Must be supplied as a POST variable.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Properties of the updated study",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="study_attach",
+   *				ref="#/components/schemas/inline_response_200_26_study_attach",
+   *			),
+   *			example={
+   *			  "study_attach": {
+   *			    "id": "1",
+   *			    "main_entity_type": "task",
+   *			    "linked_entities": "5"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n1041 - Could not find study. Check the study ID in your request.\n1042 - Cannnot attach entities to legacy studies.\n1043 - Please provide POST field 'ids'.\n1044 - Please ensure that the 'ids' in the POST field is a list of natural numbers.\n1045 - Could not attach entities to the study. It appears as if the entity does not exist.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function study_attach($study_id) {
+    $this->study_attach_detach($study_id, true);
+  }
+
+  /**
+   *@OA\Post(
+   *	path="/study/{id}/detach",
+   *	tags={"study"},
+   *	summary="Detach an entity from a study",
+   *	description="Detach an entity from an exising study. Upon success, it returns the study id, type, and linked entities.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the study.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="ids",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Comma-separated list of entity IDs to be detached from the study. For instance, if this is a run study, the list of run IDs that need to be removed (detached) from the study. Must be supplied as a POST variable.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Properties of the updated study",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_study",
+   *				ref="#/components/schemas/inline_response_200_26_study_attach",
+   *			),
+   *			example={
+   *			  "study_detach": {
+   *			    "id": "1",
+   *			    "main_entity_type": "task",
+   *			    "linked_entities": "5"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n1041 - Could not find study. Check the study ID in your request.\n1042 - Cannot attach entities to legacy studies.\n1043 - Please provide POST field 'ids'.\n1044 - Please ensure that the 'ids' in the POST field is a list of natural numbers.\n1046 - Could not detach entities from the study. It appears as if the entity does not exist.       \n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function study_detach($study_id) {
+    $this->study_attach_detach($study_id, false);
+  }
+
   private function study_attach_detach($study_id, $attach) {
     $study = $this->Study->getById($study_id);
     if ($study === false) {
@@ -275,7 +446,56 @@ private function study_attach_detach($study_id, $attach) {
     
     $this->xmlContents('study-attach-detach', $this->version, $template_vars);
   }
-  
+
+  /**
+   *@OA\Delete(
+   *	path="/study/{id}",
+   *	tags={"study"},
+   *	summary="Delete study",
+   *	description="Deletes a study. Upon success, it returns the ID of the deleted study.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the study.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted study",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="study_delete",
+   *				ref="#/components/schemas/inline_response_200_24_study_delete",
+   *			),
+   *			example={
+   *			  "study_delete": {
+   *			    "id": "1"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n591 - Please provide API key. In order to remove your content, please authenticate.\n592 - Study does not exists. The study ID could not be linked to an existing study.\n593 - Study deletion failed. Please try again later.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function study_delete($study_id) {
 
     $study = $this->Study->getById($study_id);
@@ -313,6 +533,61 @@ private function study_delete($study_id) {
     $this->xmlContents('study-delete', $this->version, array('study' => $study));
   }
 
+  /**
+   *@OA\Get(
+   *	path="/study/list/{filters}",
+   *	tags={"study"},
+   *	summary="List all studies (collections of items)",
+   *	description="List studies, optionally filtered by a range of properties. Any number of properties can be combined by listing them one after the other in the form '/study/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all studies that match the constraints.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /main_entity_type/{type} - only return studies collecting entities of a given type (e.g. 'task' or 'run').
+  /uploader/{ids} - return only evaluations uploaded by specific users, specified as a comma-separated list of user IDs, e.g. ''1,2,3''
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, results 11..15 will be returned. Both limit and offset need to be specified.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of studies",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/StudyList",
+   *			example={
+   *			  "study_list":{
+   *			    "study":{
+   *			      {
+   *			        "id":"1",
+   *			        "alias":"Study_1",
+   *			        "name":"A large-scale comparison of classification algorithms",
+   *			        "creation_date":"2017-07-20 15:51:20",
+   *			        "creator":"2"
+   *			      },
+   *			      {
+   *			        "id":"2",
+   *			        "alias":"Study_2",
+   *			        "name":"Fast Algorithm Selection using Learning Curves",
+   *			        "creation_date":"2017-07-20 15:51:20",
+   *			        "creator":"2"
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function study_list($segs) {
     $legal_filters = array('limit', 'offset', 'main_entity_type', 'uploader', 'status', 'benchmark_suite');
     
@@ -367,6 +642,59 @@ private function study_list($segs) {
     $this->xmlContents('study-list', $this->version, array('studies' => $studies));
   }
 
+  /**
+   *@OA\Get(
+   *	path="/study/{id}",
+   *	tags={"study"},
+   *	summary="Get study description by study id or alias",
+   *	description="Returns information about the study with the given id or alias.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="ID or alias of the study.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A study description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Study",
+   *			example={
+   *			  "study": {
+   *			    "id": "99",
+   *			    "main_entity_type": "task",
+   *			    "name": "CC18 benchmark suite",
+   *			    "description": "CC18 benchmark suite",
+   *			    "creation_date": "2019-02-16T17:35:58",
+   *			    "creator": "1159",
+   *			    "data": {"data_id": {"1","2","3"}},
+   *			    "tasks": {"task_id": {"1","2","3"}}
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n601 - Unknown study. The study with the given id or alias was not found in the database\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function study_get($study_id, $type) {
+    if (is_numeric($study_id)) {
+      $this->study_by_id($study_id, $type);
+      return;
+    } else {
+      $this->study_by_alias($study_id, $type);
+      return;
+    }
+  }
+
   private function study_by_id($study_id, $entity_type) {
     $study = $this->Study->getById($study_id);
 
diff --git a/openml_OS/models/api/v1/Api_task.php b/openml_OS/models/api/v1/Api_task.php
index 6e23ff0cc..35ca249a0 100644
--- a/openml_OS/models/api/v1/Api_task.php
+++ b/openml_OS/models/api/v1/Api_task.php
@@ -3,6 +3,9 @@ class Api_task extends MY_Api_Model {
 
   protected $version = 'v1';
 
+  const ENTITY_TYPE = "task";
+  const ENTITY_SPECIAL_NAME = "task";
+
   function __construct() {
     parent::__construct();
     
@@ -54,12 +57,12 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     }
 
     if (count($segments) == 1 && $segments[0] == 'tag' && $request_type == 'post') {
-      $this->entity_tag_untag('task', $this->input->post('task_id'), $this->input->post('tag'), false, 'task');
+      $this->task_tag($this->input->post('task_id'), $this->input->post('tag'));
       return;
     }
 
     if (count($segments) == 1 && $segments[0] == 'untag' && $request_type == 'post') {
-      $this->entity_tag_untag('task', $this->input->post('task_id'), $this->input->post('tag'), true, 'task');
+      $this->task_untag($this->input->post('task_id'), $this->input->post('tag'));
       return;
     }
     
@@ -71,7 +74,253 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->returnError(100, $this->version);
   }
 
+  /**
+   *@OA\Post(
+   *	path="/task/tag",
+   *	tags={"task"},
+   *	summary="Tag a task",
+   *	description="Tags a task.",
+   *	@OA\Parameter(
+   *		name="task_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the task.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="The id of the tagged task",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="task_tag",
+   *				ref="#/components/schemas/inline_response_200_6_task_tag",
+   *			),
+   *			example={
+   *			  "task_tag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n470 - In order to add a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\n471 - Entity not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\n472 - Entity already tagged by this tag. The entity {dataset, task, flow, run} already had this tag.\n473 - Something went wrong inserting the tag. Please contact OpenML Team.\n474 - Internal error tagging the entity. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function task_tag($task_id, $tag) {
+    $this->entity_tag_untag(self::ENTITY_TYPE, $task_id, $tag, false, self::ENTITY_SPECIAL_NAME);
+  }
 
+  /**
+   *@OA\Post(
+   *	path="/task/untag",
+   *	tags={"task"},
+   *	summary="Untag a task",
+   *	description="Untags a task.",
+   *	@OA\Parameter(
+   *		name="task_id",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the task.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="tag",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Tag name",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A the features of the task",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="task_untag",
+   *				ref="#/components/schemas/inline_response_200_7_task_untag",
+   *			),
+   *			example={
+   *			  "task_untag": {
+   *			    "id": "2"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n475 - Please give entity_id {data_id, flow_id, run_id} and tag. In order to remove a tag, please upload the entity id (either data_id, task_id, flow_id, run_id) and tag (the name of the tag).\n476 - Entity {dataset, task, flow, run} not found. The provided entity_id {data_id, task_id, flow_id, run_id} does not correspond to an existing entity.\n477 - Tag not found. The provided tag is not associated with the entity {dataset, task, flow, run}.\n478 - Tag is not owned by you. The entity {dataset, flow, run} was tagged by another user. Hence you cannot delete it.\n479 - Internal error removing the tag. Please contact OpenML Team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
+  private function task_untag($task_id, $tag) {
+    $this->entity_tag_untag(self::ENTITY_TYPE, $task_id, $tag, true, self::ENTITY_SPECIAL_NAME);
+  }
+
+  /**
+   *@OA\Get(
+   *	path="/task/list/{filters}",
+   *	tags={"task"},
+   *	summary="List and filter tasks",
+   *	description="List tasks, possibly filtered by a range of properties from the task itself or from the underlying dataset. Any number of properties can be combined by listing them one after the other in the form '/task/list/{filter}/{value}/{filter}/{value}/...' Returns an array with all tasks that match the constraints.",
+   *	@OA\Parameter(
+   *		name="filters",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Any combination of these filters
+  /limit/{limit}/offset/{offset} - returns only {limit} results starting from result number {offset}. Useful for paginating results. With /limit/5/offset/10, tasks 11..15 will be returned. Both limit and offset need to be specified.
+  /status/{status} - returns only tasks with a given status, either 'active', 'deactivated', or 'in_preparation'.
+  /type/{type_id} - returns only tasks with a given task type id. See the list of task types of the ID's (e.g. 1 = Supervised Classification).
+  /tag/{tag} - returns only tasks tagged with the given tag.
+  /data_tag/{tag} - returns only tasks for which the underlying dataset is tagged with the given tag.
+  /{data_quality}/{range} - returns only tasks for which the underlying datasets have certain qualities. {data_quality} can be data_id, data_name, number_instances, number_features, number_classes, number_missing_values. {range} can be a specific value or a range in the form 'low..high'. Multiple qualities can be combined, as in 'number_instances/0..50/number_features/0..10'.
+  ",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of tasks with the given tag",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/TaskList",
+   *			example={
+   *			  "task": {
+   *			    "task": {
+   *			      {
+   *			        "task_id":"1",
+   *			        "task_type":"Supervised Classification",
+   *			        "did":"1",
+   *			        "name":"anneal",
+   *			        "status":"active",
+   *			        "format":"ARFF",
+   *			        "input":{
+   *			          {
+   *			            "name":"estimation_procedure",
+   *			            "value":"1"
+   *			          },
+   *			          {
+   *			            "name":"evaluation_measures",
+   *			            "value":"predictive_accuracy"
+   *			          },
+   *			          {
+   *			            "name":"source_data",
+   *			            "value":"1"
+   *			          },
+   *			          {
+   *			            "name":"target_feature",
+   *			            "value":"class"
+   *			          }
+   *			          },
+   *			        "quality":{
+   *			          {
+   *			            "name":"MajorityClassSize",
+   *			            "value":"684"
+   *			          },
+   *			          {
+   *			            "name":"MaxNominalAttDistinctValues",
+   *			            "value":"10.0"
+   *			          },
+   *			          {
+   *			            "name":"MinorityClassSize",
+   *			            "value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumBinaryAtts",
+   *			            "value":"14.0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfClasses",
+   *			            "value":"6"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfFeatures",
+   *			            "value":"39"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfInstances",
+   *			            "value":"898"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfInstancesWithMissingValues",
+   *			            "value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfMissingValues",
+   *			            "value":"0"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfNumericFeatures",
+   *			            "value":"6"
+   *			          },
+   *			          {
+   *			            "name":"NumberOfSymbolicFeatures",
+   *			            "value":"32"
+   *			          }
+   *			          },
+   *			        "tag":{
+   *			          "basic",
+   *			          "study_1",
+   *			          "study_7",
+   *			          "under100k",
+   *			          "under1m"
+   *			        }
+   *			      }
+   *			    }
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n480 - Illegal filter specified.\n481 - Filter values/ranges not properly specified.\n482 - No results. There where no matches for the given constraints.\n483 - Can not specify an offset without a limit.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function task_list($segs, $user_id) {
     $legal_filters = array('type', 'tag', 'data_tag', 'status', 'limit', 'offset', 'task_id', 'data_id', 'data_name', 'number_instances', 'number_features', 'number_classes', 'number_missing_values');
     
@@ -149,7 +398,116 @@ private function task_list($segs, $user_id) {
     $this->xmlContents( 'tasks', $this->version, array( 'tasks' => $tasks_res ) );
   }
 
-
+  /**
+   *@OA\Get(
+   *	path="/task/{id}",
+   *	tags={"task"},
+   *	summary="Get task description",
+   *	description="Returns information about a task. The information includes the task type, input data, train/test sets, and more.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="ID of the task.",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="A task description",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Task",
+   *			example={
+   *			  "task": {
+   *			    "task_id":"1",
+   *			    "task_type":"Supervised Classification",
+   *			    "input":{
+   *			      {
+   *			        "name":"source_data",
+   *			        "data_set":{
+   *			          "data_set_id":"1",
+   *			          "target_feature":"class"
+   *			        }
+   *			      },
+   *			      {
+   *			        "name":"estimation_procedure",
+   *			        "estimation_procedure":{
+   *			          "type":"crossvalidation",
+   *			          "data_splits_url":"https://www.openml.org/api_splits/get/1/Task_1_splits.arff",
+   *			          "parameter":{
+   *			            {
+   *			              "name":"number_repeats",
+   *			              "value":"1"
+   *			            },
+   *			            {
+   *			              "name":"number_folds",
+   *			              "value":"10"
+   *			            },
+   *			            {
+   *			              "name":"percentage"
+   *			            },
+   *			            {
+   *			              "name":"stratified_sampling",
+   *			              "value":"true"
+   *			            }
+   *			          }
+   *			        }
+   *			      },
+   *			      {
+   *			        "name":"cost_matrix",
+   *			        "cost_matrix":{}
+   *			      },
+   *			      {
+   *			        "name":"evaluation_measures",
+   *			        "evaluation_measures":
+   *			          {
+   *			            "evaluation_measure":"predictive_accuracy"
+   *			          }
+   *			      }
+   *			    },
+   *			    "output":{
+   *			      "name":"predictions",
+   *			      "predictions":{
+   *			        "format":"ARFF",
+   *			        "feature":{
+   *			          {
+   *			            "name":"repeat",
+   *			            "type":"integer"
+   *			          },
+   *			          {
+   *			            "name":"fold",
+   *			            "type":"integer"
+   *			          },
+   *			          {
+   *			            "name":"row_id",
+   *			            "type":"integer"
+   *			          },
+   *			          {
+   *			            "name":"confidence.classname",
+   *			            "type":"numeric"
+   *			          },
+   *			          {
+   *			            "name":"prediction",
+   *			            "type":"string"
+   *			          }
+   *			        }
+   *			      }
+   *			    },
+   *			    "tag":{"basic","study_1","under100k","under1m"}
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n150 - Please provide task_id.\n151 - Unknown task. The task with the given id was not found in the database\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function task($task_id) {
     $task = $this->Task->getById($task_id);
     if($task === false) {
@@ -197,6 +555,55 @@ private function task_inputs($task_id) {
     $this->xmlContents('task-inputs', $this->version, array('task' => $task, 'inputs' => $inputs));
   }
 
+  /**
+   *@OA\Delete(
+   *	path="/task/{id}",
+   *	tags={"task"},
+   *	summary="Delete task",
+   *	description="Deletes a task. Upon success, it returns the ID of the deleted task.",
+   *	@OA\Parameter(
+   *		name="id",
+   *		in="path",
+   *		@OA\Schema(
+   *          type="integer"
+   *        ),
+   *		description="Id of the task.",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="ID of the deleted task",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="task_delete",
+   *				ref="#/components/schemas/inline_response_200_4_task_delete",
+   *			),
+   *			example={
+   *			  "task_delete": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n450 - Please provide API key. In order to remove your content, please authenticate.\n451 - Authentication failed. The API key was not valid. Please try to login again, or contact api administrators.\n452 - Task does not exists. The task ID could not be linked to an existing task.\n454 - Task is executed in some runs. Delete these first.\n455 - Deleting the task failed. Please contact support team.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   private function task_delete($task_id) {
 
     $task = $this->Task->getById($task_id);
@@ -243,6 +650,55 @@ private function task_delete($task_id) {
     $this->xmlContents( 'task-delete', $this->version, array( 'task' => $task ) );
   }
 
+  /**
+   *@OA\Post(
+   *	path="/task",
+   *	tags={"task"},
+   *	summary="Upload task",
+   *	description="Uploads a task. Upon success, it returns the task id.",
+   *	@OA\Parameter(
+   *		name="description",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="file"
+   *        ),
+   *		description="An XML file describing the task. Only name, description, and task format are required. Also see the [XSD schema](https://www.openml.org/api/v1/xsd/openml.task.upload) and an [XML example](https://www.openml.org/api/v1/xml_example/task).",
+   *		required=true,
+   *	),
+   *	@OA\Parameter(
+   *		name="api_key",
+   *		in="query",
+   *		@OA\Schema(
+   *          type="string"
+   *        ),
+   *		description="Api key to authenticate the user",
+   *		required=true,
+   *	),
+   *	@OA\Response(
+   *		response=200,
+   *		description="Id of the uploaded task",
+   *		@OA\JsonContent(
+   *			type="object",
+   *			@OA\Property(
+   *				property="upload_task",
+   *				ref="#/components/schemas/inline_response_200_5_upload_task",
+   *			),
+   *			example={
+   *			  "upload_task": {
+   *			    "id": "4328"
+   *			  }
+   *			}
+   *		),
+   *	),
+   *	@OA\Response(
+   *		response=412,
+   *		description="Precondition failed. An error code and message are returned.\n530 - Description file not present. Please upload the task description.\n531 - Internal error. Please contact api support team\n532 - Problem validating uploaded description file. The XML description format does not meet the standards\n533 - Task already exists.\n534 - Error creating the task.\n",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/Error",
+   *		),
+   *	),
+   *)
+   */
   public function task_upload() {
 
     $description = isset( $_FILES['description'] ) ? $_FILES['description'] : false;
diff --git a/openml_OS/models/api/v1/Api_tasktype.php b/openml_OS/models/api/v1/Api_tasktype.php
index f95a9130d..2f6552a46 100644
--- a/openml_OS/models/api/v1/Api_tasktype.php
+++ b/openml_OS/models/api/v1/Api_tasktype.php
@@ -16,11 +16,139 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     
     $getpost = array('get','post');
 
+    /**
+     *@OA\Get(
+     *	path="/tasktype/list",
+     *	tags={"tasktype"},
+     *	summary="List all task types",
+     *	description="Returns an array with all task types in the system.",
+     *	@OA\Response(
+     *		response="default",
+     *		description="Unexpected error",
+     *		@OA\JsonContent(
+     *			ref="#/components/schemas/Error",
+     *		),
+     *	),
+     *	@OA\Response(
+     *		response=200,
+     *		description="A task description",
+     *		@OA\JsonContent(
+     *			ref="#/components/schemas/TaskTypeList",
+     *			example={
+     *			  "task_types":{
+     *			    "task_type":{
+     *			      {
+     *			        "id":"1",
+     *			        "name":"Supervised Classification",
+     *			        "description":"In supervised classification, you are given ...",
+     *			        "creator":"Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl"
+     *			      },
+     *			      {
+     *			        "id":"2",
+     *			        "name":"Supervised Regression",
+     *			        "description":"Given a dataset with a numeric target ...",
+     *			        "creator":"Joaquin Vanschoren, Jan van Rijn, Luis Torgo, Bernd Bischl"
+     *			      }
+     *			    }
+     *			  }
+     *			}
+     *		),
+     *	),
+     *)
+     */
     if (count($segments) == 1 && $segments[0] == 'list') {
       $this->tasktype_list();
       return;
     }
 
+    /**
+     *@OA\Get(
+     *	path="/tasktype/{id}",
+     *	tags={"tasktype"},
+     *	summary="Get task type description",
+     *	description="Returns information about a task type. The information includes a description, the given inputs and the expected outputs.",
+     *	@OA\Parameter(
+     *		name="id",
+     *		in="path",
+     *		@OA\Schema(
+     *          type="integer"
+     *        ),
+     *		description="ID of the task.",
+     *		required=true,
+     *	),
+     *	@OA\Response(
+     *		response=200,
+     *		description="A task type description",
+     *		@OA\JsonContent(
+     *			ref="#/components/schemas/TaskType",
+     *			example={
+     *			  "task_type": {
+     *			    "id": "1",
+     *			    "name": "Supervised Classification",
+     *			    "description": "In supervised classification, you are given an input dataset in which instances are labeled with a certain class. The goal is to build a model that predicts the class for future unlabeled instances. The model is evaluated using a train-test procedure, e.g. cross-validation.<br><br>\
+     *			\
+     *			To make results by different users comparable, you are given the exact train-test folds to be used, and you need to return at least the predictions generated by your model for each of the test instances. OpenML will use these predictions to calculate a range of evaluation measures on the server.<br><br>\
+     *			\
+     *			You can also upload your own evaluation measures, provided that the code for doing so is available from the implementation used. For extremely large datasets, it may be infeasible to upload all predictions. In those cases, you need to compute and provide the evaluations yourself.<br><br>\
+     *			\
+     *			Optionally, you can upload the model trained on all the input data. There is no restriction on the file format, but please use a well-known format or PMML.",
+     *			    "creator": {
+     *			      "Joaquin Vanschoren",
+     *			      "Jan van Rijn",
+     *			      "Luis Torgo",
+     *			      "Bernd Bischl"
+     *			    },
+     *			    "contributor": {
+     *			      "Bo Gao",
+     *			      "Simon Fischer",
+     *			      "Venkatesh Umaashankar",
+     *			      "Michael Berthold",
+     *			      "Bernd Wiswedel",
+     *			      "Patrick Winter"
+     *			    },
+     *			    "creation_date": "2013-01-24 00:00:00",
+     *			    "input": {
+     *			      {
+     *			        "name": "source_data",
+     *			        "requirement": "required",
+     *			        "data_type": "numeric"
+     *			      },
+     *			      {
+     *			        "name": "target_feature",
+     *			        "requirement": "required",
+     *			        "data_type": "string"
+     *			      },
+     *			      {
+     *			        "name": "estimation_procedure",
+     *			        "requirement": "required",
+     *			        "data_type": "numeric"
+     *			      },
+     *			      {
+     *			        "name": "cost_matrix",
+     *			        "data_type": "json"
+     *			      },
+     *			      {
+     *			        "name": "custom_testset",
+     *			        "data_type": "json"
+     *			      },
+     *			      {
+     *			        "name": "evaluation_measures",
+     *			        "data_type": "string"
+     *			      }
+     *			    }
+     *			  }
+     *			}
+     *		),
+     *	),
+     *	@OA\Response(
+     *		response=412,
+     *		description="Precondition failed. An error code and message are returned.\n240 - Please provide task type ID.\n241 - Unknown task type. The task type with the given id was not found in the database\n",
+     *		@OA\JsonContent(
+     *			ref="#/components/schemas/Error",
+     *		),
+     *	),
+     *)
+     */
     if (count($segments) == 1 && is_numeric($segments[0]) && in_array($request_type, $getpost)) {
       $this->tasktype($segments[0]);
       return;
diff --git a/openml_OS/models/api/v1/Api_user.php b/openml_OS/models/api/v1/Api_user.php
index 14a988483..71b946bab 100644
--- a/openml_OS/models/api/v1/Api_user.php
+++ b/openml_OS/models/api/v1/Api_user.php
@@ -17,12 +17,13 @@ function bootstrap($format, $segments, $request_type, $user_id) {
     $this->outputFormat = $format;
 
   # http://test.openml.org/api/v1/user/list/uploader/1,2
+
     if (count($segments) >= 1 && $segments[0] == 'list') {
       array_shift($segments);
       $this->username_list($segments);
       return;
     }
-    
+
     /*$getpost = array('get','post');
 
     if (count($segments) == 1 && is_numeric($segments[0]) && $request_type == 'delete') {
@@ -82,6 +83,35 @@ function bootstrap($format, $segments, $request_type, $user_id) {
   } */
 
 
+  /**
+   *@OA\Get(
+   *	path="/user/list",
+   *	tags={"user"},
+   *	summary="List all users by user id",
+   *	description="Returns an array with all user ids and names.",
+   *	@OA\Response(
+   *		response=200,
+   *		description="A list of users",
+   *		@OA\JsonContent(
+   *			ref="#/components/schemas/UserList",
+   *			example={
+   *			  "users":{
+   *			    "user":{
+   *			      {
+   *			        "id":"1",
+   *			        "username":"janvanrijn@gmail.com"
+   *                  },
+   *			      {
+   *			        "id":"2",
+   *			        "username":"joaquin.vanschoren@gmail.com"
+   *                  }
+   *			    }
+   *			  }
+   *			}
+   *		)
+   *	)
+   *)
+   */
   private function username_list($segs) {
     # pass uploader list to get username list
     $legal_filters = array('user_id');
diff --git a/openml_OS/views/pages/api_new/v1/xml/data-edit.tpl.php b/openml_OS/views/pages/api_new/v1/xml/data-edit.tpl.php
new file mode 100644
index 000000000..0095640d0
--- /dev/null
+++ b/openml_OS/views/pages/api_new/v1/xml/data-edit.tpl.php
@@ -0,0 +1,3 @@
+<oml:data_edit xmlns:oml="http://openml.org/openml">
+	<oml:id><?php echo $id; ?></oml:id>
+</oml:data_edit>
diff --git a/openml_OS/views/pages/api_new/v1/xml/data-fork.tpl.php b/openml_OS/views/pages/api_new/v1/xml/data-fork.tpl.php
new file mode 100644
index 000000000..ae2af9f8e
--- /dev/null
+++ b/openml_OS/views/pages/api_new/v1/xml/data-fork.tpl.php
@@ -0,0 +1,3 @@
+<oml:data_fork xmlns:oml="http://openml.org/openml">
+	<oml:id><?php echo $id; ?></oml:id>
+</oml:data_fork>
diff --git a/openml_OS/views/pages/api_new/v1/xml/pre.php b/openml_OS/views/pages/api_new/v1/xml/pre.php
index e4bbf62a7..1de54754c 100644
--- a/openml_OS/views/pages/api_new/v1/xml/pre.php
+++ b/openml_OS/views/pages/api_new/v1/xml/pre.php
@@ -509,4 +509,21 @@
 $this->apiErrors[1053] = 'Study not owned by you';
 $this->apiErrors[1054] = 'Problem inserting in database';
 
+
+//openml.data.edit
+$this->apiErrors[1060] = 'Problem validating edit_parameters xml';
+$this->apiErrors[1061] = 'Please provide edit_parameters xml';
+$this->apiErrors[1062] = 'Data ID is required';
+$this->apiErrors[1063] = 'Unknown dataset';
+$this->apiErrors[1064] = 'Please provide atleast one field among description, creator, contributor, collection_date, language, citation, original_data_url, default_target_attribute, row_id_attribute, ignore_attribute or paper_url to edit. ';
+$this->apiErrors[1065] = 'Critical features default_target_attribute, row_id_attribute and ignore_attribute can be edited only by the owner. Fork the dataset if changes are required.';
+$this->apiErrors[1066] = 'Critical features default_target_attribute, row_id_attribute and ignore_attribute can only be edited for datasets without any tasks.';
+$this->apiErrors[1067] = 'Dataset update failed';
+
+
+//openml.data.edit
+$this->apiErrors[1070] = 'Data ID is required';
+$this->apiErrors[1071] = 'Unknown dataset';
+$this->apiErrors[1072] = 'Failed to insert record in database';
+
 ?>
diff --git a/openml_OS/views/pages/api_new/v1/xsd/openml.data.edit.xsd b/openml_OS/views/pages/api_new/v1/xsd/openml.data.edit.xsd
new file mode 100644
index 000000000..18f08b122
--- /dev/null
+++ b/openml_OS/views/pages/api_new/v1/xsd/openml.data.edit.xsd
@@ -0,0 +1,89 @@
+<!-- Schema for dataset edit parameters. -->
+
+<xs:schema targetNamespace="http://openml.org/openml"
+    xmlns:oml="http://openml.org/openml"
+    xmlns:xs="http://www.w3.org/2001/XMLSchema"
+    elementFormDefault="qualified">
+<xs:element name="data_edit_parameters" type="oml:data_edit_parameters"/>
+<xs:complexType name="data_edit_parameters">
+  <xs:sequence>
+    <!-- dublin core -->
+    <xs:element name="description" minOccurs="0" type="oml:basic_latin16384"/>  <!-- Description of the dataset, given by the user who uploaded it. -->
+    <xs:element name="creator" minOccurs="0" maxOccurs="unbounded" type="oml:basic_latin128"/> <!-- The person who created the dataset -->
+    <xs:element name="contributor" minOccurs="0" maxOccurs="unbounded" type="oml:basic_latin128"/> <!-- People who contributed to the current version of the datadat (e.g. reformatting) -->
+    <xs:element name="collection_date" minOccurs="0" type="oml:basic_latin128"/>  <!-- The date the data was originally collected, given by the uploader -->
+    <xs:element name="language" minOccurs="0" type="oml:casual_string128"/> <!-- Language in which the data is represented. Starts with 1 upper case letter, rest lower case, e.g. 'English' -->
+    <!-- other -->
+    <xs:element name="default_target_attribute" minOccurs="0" type="oml:casual_string1024"/>  <!-- The default target attribute, if it exists. Can also have multiple values (comma-separated). Of course, tasks can be defined that use another attribute as target -->  
+    <xs:element name="row_id_attribute" minOccurs="0" type="oml:casual_string128"/>  <!-- The attribute that represents the row-id column, if present in the dataset. If not set, the tool should create a string rowid from 0 to (n.obs - 1). This is what is then used in the data splits object. If set, it should of course be excluded in modeling -->
+    <xs:element name="ignore_attribute" minOccurs="0" maxOccurs="unbounded" type="oml:casual_string128"/>  <!-- Attributes that should be excluded in modelling, such as identifiers and indexes. -->
+    <xs:element name="citation" minOccurs="0" type="oml:basic_latin1024"/> <!-- Reference(s) that should be cited when building on this data -->
+    <xs:element name="original_data_url" minOccurs="0" type="xs:anyURI"/> <!-- For derived data, the url to the original dataset. This can be an OpenML dataset, e.g. 'http://openml.org/d/1'. -->
+    <xs:element name="paper_url" minOccurs="0" type="xs:anyURI"/> <!-- Link to a paper describing the dataset -->
+  </xs:sequence>
+</xs:complexType>
+
+<xs:simpleType name="system_string64"> <!-- Subset on xs:string. Highly restricted form of string. URL-friendly -->
+  <xs:restriction base="xs:string">
+    <xs:pattern value="([a-zA-Z0-9_\-\.])+"/>
+    <xs:maxLength value="64" />
+     <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="casual_string1024"> <!-- Subset on xs:string. Highly restricted form of string. URL-friendly -->
+  <xs:restriction base="xs:string">
+    <xs:pattern value="([a-zA-Z0-9_\-\.\(\),])+"/>
+    <xs:maxLength value="1024" />
+     <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="casual_string128"> <!-- Subset on xs:string. Highly restricted form of string. URL-friendly -->
+  <xs:restriction base="xs:string">
+    <xs:pattern value="([a-zA-Z0-9_\-\.\(\),])+"/>
+    <xs:maxLength value="128" />
+     <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="casual_string64"> <!-- Subset on xs:string. Highly restricted form of string. URL-friendly -->
+  <xs:restriction base="xs:string">
+     <xs:pattern value="([a-zA-Z0-9_\-\.\(\),])+"/>
+     <xs:maxLength value="64" />
+     <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="basic_latin64">
+  <xs:restriction base="xs:string">
+    <xs:pattern value="\p{IsBasicLatin}*"/>
+    <xs:maxLength value="64" />
+    <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="basic_latin128">
+  <xs:restriction base="xs:string">
+    <xs:pattern value="\p{IsBasicLatin}*"/>
+    <xs:maxLength value="128" />
+    <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="basic_latin1024">
+  <xs:restriction base="xs:string">
+    <xs:pattern value="\p{IsBasicLatin}*"/>
+    <xs:maxLength value="1024" />
+    <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="basic_latin16384">
+  <xs:restriction base="xs:string">
+    <xs:pattern value="\p{IsBasicLatin}*"/>
+    <xs:maxLength value="16384" />
+    <xs:minLength value="1" />
+  </xs:restriction>
+</xs:simpleType>
+<xs:simpleType name="access_policy">
+  <xs:restriction base="xs:string">
+    <xs:enumeration value="circles" />
+    <xs:enumeration value="public" />
+    <xs:enumeration value="private" />
+  </xs:restriction>
+</xs:simpleType>
+</xs:schema>
diff --git a/openml_OS/views/pages/frontend/contact/body.php b/openml_OS/views/pages/frontend/contact/body.php
index 205e28363..49b6251a8 100644
--- a/openml_OS/views/pages/frontend/contact/body.php
+++ b/openml_OS/views/pages/frontend/contact/body.php
@@ -19,7 +19,7 @@
     <p>OpenML is an open source project on GitHub. To get help quickly, please <a href='https://docs.openml.org/developers/'>post issues in our GitHub issue trackers</a></p>.
 
     <h2 id="team-core">Core Team</h2>
-    <p>OpenML wouldn't exist without these people who put their heart and sole into making machine learning easier for all</p>
+    <p>OpenML wouldn't exist without these people who put their heart and soul into making machine learning easier for all</p>
 
     <?php
      if( $this->coreteam != false ) {
diff --git a/openml_OS/views/pages/frontend/home/body.php b/openml_OS/views/pages/frontend/home/body.php
index 70cf7adda..b87e73883 100644
--- a/openml_OS/views/pages/frontend/home/body.php
+++ b/openml_OS/views/pages/frontend/home/body.php
@@ -30,6 +30,15 @@
       <!-- Wrapper for slides -->
       <div class="carousel-inner list-group" role="listbox">
         <div class="item active">
+          <div class="list-group-item">
+              <div class="row-picture">
+                  <img class="circle" src="https://i0.wp.com/syncedreview.com/wp-content/uploads/2020/06/Imagenet.jpg?fit=1400%2C600&ssl=1" alt="icon">
+              </div>
+              <div class="row-content">
+                  <h4 class="list-group-item-heading">Challenge on Metalearning released</h4>
+                  <p class="list-group-item-text"><a href="https://competitions.codalab.org/competitions/26638">Read the details and participate</a>, and <a href="https://sites.google.com/chalearn.org/metalearning/">an associated workshop at AAAI </a> with high-level keynote speakers</p>
+              </div>
+          </div>
           <div class="list-group-item">
               <div class="row-picture">
                   <img class="circle" src="http://icons.iconarchive.com/icons/cornmanthe3rd/plex/512/Other-python-icon.png" alt="icon">
@@ -48,7 +57,7 @@
               <div class="row-content">
                   <h4 class="list-group-item-heading"><div id="colorchange"><span>H</span><span>A</span><span>C</span><span>K</span><span>A</span><span>T</span><span>H</span><span>O</span><span>N</span></div></h4>
                   <p class="list-group-item-text">Bring your own data, bring your own algorithms, or build cool new features.</p>
-                  <p class="list-group-item-text"><a href="https://meet.openml.org" target="_blank">Next workshop: March 2020, Munich, Germany</a></p>
+                  <p class="list-group-item-text"><a href="https://new.openml.org/meet#upcoming" target="_blank">Next workshop: 26-30 October 2020, Virtual</a></p>
               </div>
           </div>
         </div>

From 146ae1b457ba63585c6d50d716f120e30c8d6bfa Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Mon, 2 Nov 2020 23:00:45 +0100
Subject: [PATCH 05/15] install db fix

---
 data/sql/dataset_status.sql | 130 ------------------------------------
 data/sql/task_study.sql     | 101 ----------------------------
 2 files changed, 231 deletions(-)
 delete mode 100644 data/sql/dataset_status.sql
 delete mode 100644 data/sql/task_study.sql

diff --git a/data/sql/dataset_status.sql b/data/sql/dataset_status.sql
deleted file mode 100644
index a461b1c51..000000000
--- a/data/sql/dataset_status.sql
+++ /dev/null
@@ -1,130 +0,0 @@
-INSERT INTO `dataset_status` (`did`, `status`, `status_date`, `user_id`) VALUES
-(1, 'active', '2018-09-17 09:00:00', 1),
-(2, 'active', '2018-09-17 09:00:00', 1),
-(3, 'active', '2018-09-17 09:00:00', 1),
-(4, 'active', '2018-09-17 09:00:00', 1),
-(5, 'active', '2018-09-17 09:00:00', 1),
-(6, 'active', '2018-09-17 09:00:00', 1),
-(7, 'active', '2018-09-17 09:00:00', 1),
-(8, 'active', '2018-09-17 09:00:00', 1),
-(9, 'active', '2018-09-17 09:00:00', 1),
-(10, 'active', '2018-09-17 09:00:00', 1),
-(11, 'active', '2018-09-17 09:00:00', 1),
-(12, 'active', '2018-09-17 09:00:00', 1),
-(13, 'active', '2018-09-17 09:00:00', 1),
-(14, 'active', '2018-09-17 09:00:00', 1),
-(15, 'active', '2018-09-17 09:00:00', 1),
-(16, 'active', '2018-09-17 09:00:00', 1),
-(17, 'active', '2018-09-17 09:00:00', 1),
-(18, 'active', '2018-09-17 09:00:00', 1),
-(19, 'active', '2018-09-17 09:00:00', 1),
-(20, 'active', '2018-09-17 09:00:00', 1),
-(21, 'active', '2018-09-17 09:00:00', 1),
-(22, 'active', '2018-09-17 09:00:00', 1),
-(23, 'active', '2018-09-17 09:00:00', 1),
-(24, 'active', '2018-09-17 09:00:00', 1),
-(25, 'active', '2018-09-17 09:00:00', 1),
-(26, 'active', '2018-09-17 09:00:00', 1),
-(27, 'active', '2018-09-17 09:00:00', 1),
-(28, 'active', '2018-09-17 09:00:00', 1),
-(29, 'active', '2018-09-17 09:00:00', 1),
-(30, 'active', '2018-09-17 09:00:00', 1),
-(31, 'active', '2018-09-17 09:00:00', 1),
-(32, 'active', '2018-09-17 09:00:00', 1),
-(33, 'active', '2018-09-17 09:00:00', 1),
-(34, 'active', '2018-09-17 09:00:00', 1),
-(35, 'active', '2018-09-17 09:00:00', 1),
-(36, 'active', '2018-09-17 09:00:00', 1),
-(37, 'active', '2018-09-17 09:00:00', 1),
-(38, 'active', '2018-09-17 09:00:00', 1),
-(39, 'active', '2018-09-17 09:00:00', 1),
-(40, 'active', '2018-09-17 09:00:00', 1),
-(41, 'active', '2018-09-17 09:00:00', 1),
-(42, 'active', '2018-09-17 09:00:00', 1),
-(43, 'active', '2018-09-17 09:00:00', 1),
-(44, 'active', '2018-09-17 09:00:00', 1),
-(45, 'active', '2018-09-17 09:00:00', 1),
-(46, 'active', '2018-09-17 09:00:00', 1),
-(47, 'active', '2018-09-17 09:00:00', 1),
-(48, 'active', '2018-09-17 09:00:00', 1),
-(49, 'active', '2018-09-17 09:00:00', 1),
-(50, 'active', '2018-09-17 09:00:00', 1),
-(51, 'active', '2018-09-17 09:00:00', 1),
-(52, 'active', '2018-09-17 09:00:00', 1),
-(53, 'active', '2018-09-17 09:00:00', 1),
-(54, 'active', '2018-09-17 09:00:00', 1),
-(55, 'active', '2018-09-17 09:00:00', 1),
-(56, 'active', '2018-09-17 09:00:00', 1),
-(57, 'active', '2018-09-17 09:00:00', 1),
-(58, 'active', '2018-09-17 09:00:00', 1),
-(59, 'active', '2018-09-17 09:00:00', 1),
-(60, 'active', '2018-09-17 09:00:00', 1),
-(61, 'active', '2018-09-17 09:00:00', 1),
-(62, 'active', '2018-09-17 09:00:00', 1),
-(63, 'active', '2018-09-17 09:00:00', 1),
-(64, 'active', '2018-09-17 09:00:00', 1),
-(65, 'active', '2018-09-17 09:00:00', 1),
-(66, 'active', '2018-09-17 09:00:00', 1),
-(67, 'active', '2018-09-17 09:00:00', 1),
-(68, 'active', '2018-09-17 09:00:00', 1),
-(69, 'active', '2018-09-17 09:00:00', 1),
-(70, 'active', '2018-09-17 09:00:00', 1),
-(71, 'active', '2018-09-17 09:00:00', 1),
-(72, 'active', '2018-09-17 09:00:00', 1),
-(73, 'active', '2018-09-17 09:00:00', 1),
-(74, 'active', '2018-09-17 09:00:00', 1),
-(75, 'active', '2018-09-17 09:00:00', 1),
-(76, 'active', '2018-09-17 09:00:00', 1),
-(77, 'active', '2018-09-17 09:00:00', 1),
-(78, 'active', '2018-09-17 09:00:00', 1),
-(79, 'active', '2018-09-17 09:00:00', 1),
-(80, 'active', '2018-09-17 09:00:00', 1),
-(81, 'active', '2018-09-17 09:00:00', 1),
-(82, 'active', '2018-09-17 09:00:00', 1),
-(83, 'active', '2018-09-17 09:00:00', 1),
-(84, 'active', '2018-09-17 09:00:00', 1),
-(85, 'active', '2018-09-17 09:00:00', 1),
-(86, 'active', '2018-09-17 09:00:00', 1),
-(87, 'active', '2018-09-17 09:00:00', 1),
-(88, 'active', '2018-09-17 09:00:00', 1),
-(89, 'active', '2018-09-17 09:00:00', 1),
-(90, 'active', '2018-09-17 09:00:00', 1),
-(91, 'active', '2018-09-17 09:00:00', 1),
-(92, 'active', '2018-09-17 09:00:00', 1),
-(93, 'active', '2018-09-17 09:00:00', 1),
-(94, 'active', '2018-09-17 09:00:00', 1),
-(95, 'active', '2018-09-17 09:00:00', 1),
-(96, 'active', '2018-09-17 09:00:00', 1),
-(97, 'active', '2018-09-17 09:00:00', 1),
-(98, 'active', '2018-09-17 09:00:00', 1),
-(99, 'active', '2018-09-17 09:00:00', 1),
-(100, 'active', '2018-09-17 09:00:00', 1),
-(101, 'active', '2018-09-17 09:00:00', 1),
-(102, 'active', '2018-09-17 09:00:00', 1),
-(103, 'active', '2018-09-17 09:00:00', 1),
-(104, 'active', '2018-09-17 09:00:00', 1),
-(105, 'active', '2018-09-17 09:00:00', 1),
-(106, 'active', '2018-09-17 09:00:00', 1),
-(107, 'active', '2018-09-17 09:00:00', 1),
-(108, 'active', '2018-09-17 09:00:00', 1),
-(109, 'active', '2018-09-17 09:00:00', 1),
-(110, 'active', '2018-09-17 09:00:00', 1),
-(111, 'active', '2018-09-17 09:00:00', 1),
-(112, 'active', '2018-09-17 09:00:00', 1),
-(113, 'active', '2018-09-17 09:00:00', 1),
-(114, 'active', '2018-09-17 09:00:00', 1),
-(115, 'active', '2018-09-17 09:00:00', 1),
-(116, 'active', '2018-09-17 09:00:00', 1),
-(117, 'active', '2018-09-17 09:00:00', 1),
-(118, 'active', '2018-09-17 09:00:00', 1),
-(119, 'active', '2018-09-17 09:00:00', 1),
-(120, 'active', '2018-09-17 09:00:00', 1),
-(121, 'active', '2018-09-17 09:00:00', 1),
-(122, 'active', '2018-09-17 09:00:00', 1),
-(123, 'active', '2018-09-17 09:00:00', 1),
-(124, 'active', '2018-09-17 09:00:00', 1),
-(125, 'active', '2018-09-17 09:00:00', 1),
-(126, 'active', '2018-09-17 09:00:00', 1),
-(127, 'active', '2018-09-17 09:00:00', 1),
-(128, 'active', '2018-09-17 09:00:00', 1),
-(130, 'active', '2018-09-17 09:00:00', 1);
diff --git a/data/sql/task_study.sql b/data/sql/task_study.sql
deleted file mode 100644
index 80febbfd5..000000000
--- a/data/sql/task_study.sql
+++ /dev/null
@@ -1,101 +0,0 @@
-INSERT INTO `task_study` (`study_id`, `task_id`, `uploader`, `date`) VALUES
-(1, 1, 1, '2019-02-25 16:15:01'),
-(1, 7, 1, '2019-02-25 16:15:01'),
-(1, 13, 1, '2019-02-25 16:15:01'),
-(1, 19, 1, '2019-02-25 16:15:01'),
-(1, 25, 1, '2019-02-25 16:15:01'),
-(1, 31, 1, '2019-02-25 16:15:01'),
-(1, 37, 1, '2019-02-25 16:15:01'),
-(1, 43, 1, '2019-02-25 16:15:01'),
-(1, 49, 1, '2019-02-25 16:15:01'),
-(1, 55, 1, '2019-02-25 16:15:01'),
-(1, 61, 1, '2019-02-25 16:15:01'),
-(1, 67, 1, '2019-02-25 16:15:01'),
-(1, 73, 1, '2019-02-25 16:15:01'),
-(1, 79, 1, '2019-02-25 16:15:01'),
-(1, 85, 1, '2019-02-25 16:15:01'),
-(1, 91, 1, '2019-02-25 16:15:01'),
-(1, 97, 1, '2019-02-25 16:15:01'),
-(1, 103, 1, '2019-02-25 16:15:01'),
-(1, 109, 1, '2019-02-25 16:15:01'),
-(1, 115, 1, '2019-02-25 16:15:01'),
-(1, 121, 1, '2019-02-25 16:15:01'),
-(1, 127, 1, '2019-02-25 16:15:01'),
-(1, 133, 1, '2019-02-25 16:15:01'),
-(1, 139, 1, '2019-02-25 16:15:01'),
-(1, 145, 1, '2019-02-25 16:15:01'),
-(1, 151, 1, '2019-02-25 16:15:01'),
-(1, 157, 1, '2019-02-25 16:15:01'),
-(1, 163, 1, '2019-02-25 16:15:01'),
-(1, 169, 1, '2019-02-25 16:15:01'),
-(1, 175, 1, '2019-02-25 16:15:01'),
-(1, 181, 1, '2019-02-25 16:15:01'),
-(1, 187, 1, '2019-02-25 16:15:01'),
-(1, 193, 1, '2019-02-25 16:15:01'),
-(1, 199, 1, '2019-02-25 16:15:01'),
-(1, 205, 1, '2019-02-25 16:15:01'),
-(1, 211, 1, '2019-02-25 16:15:01'),
-(1, 217, 1, '2019-02-25 16:15:01'),
-(1, 223, 1, '2019-02-25 16:15:01'),
-(1, 229, 1, '2019-02-25 16:15:01'),
-(1, 235, 1, '2019-02-25 16:15:01'),
-(1, 241, 1, '2019-02-25 16:15:01'),
-(1, 247, 1, '2019-02-25 16:15:01'),
-(1, 253, 1, '2019-02-25 16:15:01'),
-(1, 259, 1, '2019-02-25 16:15:01'),
-(1, 265, 1, '2019-02-25 16:15:01'),
-(1, 271, 1, '2019-02-25 16:15:01'),
-(1, 277, 1, '2019-02-25 16:15:01'),
-(1, 283, 1, '2019-02-25 16:15:01'),
-(1, 289, 1, '2019-02-25 16:15:01'),
-(1, 295, 1, '2019-02-25 16:15:01'),
-(1, 301, 1, '2019-02-25 16:15:01'),
-(1, 307, 1, '2019-02-25 16:15:01'),
-(1, 313, 1, '2019-02-25 16:15:01'),
-(1, 319, 1, '2019-02-25 16:15:01'),
-(1, 325, 1, '2019-02-25 16:15:01'),
-(1, 331, 1, '2019-02-25 16:15:01'),
-(1, 337, 1, '2019-02-25 16:15:01'),
-(1, 343, 1, '2019-02-25 16:15:01'),
-(1, 349, 1, '2019-02-25 16:15:01'),
-(1, 355, 1, '2019-02-25 16:15:01'),
-(1, 361, 1, '2019-02-25 16:15:01'),
-(1, 367, 1, '2019-02-25 16:15:01'),
-(1, 373, 1, '2019-02-25 16:15:01'),
-(1, 379, 1, '2019-02-25 16:15:01'),
-(1, 385, 1, '2019-02-25 16:15:01'),
-(1, 391, 1, '2019-02-25 16:15:01'),
-(1, 397, 1, '2019-02-25 16:15:01'),
-(1, 403, 1, '2019-02-25 16:15:01'),
-(1, 409, 1, '2019-02-25 16:15:01'),
-(1, 415, 1, '2019-02-25 16:15:01'),
-(1, 421, 1, '2019-02-25 16:15:01'),
-(1, 427, 1, '2019-02-25 16:15:01'),
-(1, 433, 1, '2019-02-25 16:15:01'),
-(1, 439, 1, '2019-02-25 16:15:01'),
-(1, 445, 1, '2019-02-25 16:15:01'),
-(1, 451, 1, '2019-02-25 16:15:01'),
-(1, 457, 1, '2019-02-25 16:15:01'),
-(1, 463, 1, '2019-02-25 16:15:01'),
-(1, 469, 1, '2019-02-25 16:15:01'),
-(1, 475, 1, '2019-02-25 16:15:01'),
-(1, 481, 1, '2019-02-25 16:15:01'),
-(1, 487, 1, '2019-02-25 16:15:01'),
-(1, 493, 1, '2019-02-25 16:15:01'),
-(1, 499, 1, '2019-02-25 16:15:01'),
-(1, 505, 1, '2019-02-25 16:15:01'),
-(1, 511, 1, '2019-02-25 16:15:01'),
-(1, 517, 1, '2019-02-25 16:15:01'),
-(1, 523, 1, '2019-02-25 16:15:01'),
-(1, 529, 1, '2019-02-25 16:15:01'),
-(1, 535, 1, '2019-02-25 16:15:01'),
-(1, 541, 1, '2019-02-25 16:15:01'),
-(1, 547, 1, '2019-02-25 16:15:01'),
-(1, 553, 1, '2019-02-25 16:15:01'),
-(1, 559, 1, '2019-02-25 16:15:01'),
-(1, 565, 1, '2019-02-25 16:15:01'),
-(1, 571, 1, '2019-02-25 16:15:01'),
-(1, 577, 1, '2019-02-25 16:15:01'),
-(1, 583, 1, '2019-02-25 16:15:01'),
-(1, 589, 1, '2019-02-25 16:15:01'),
-(1, 595, 1, '2019-02-25 16:15:01');

From 9d3202a2f6dbce743dbaf42fa578d457db4ca800 Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Thu, 19 Nov 2020 13:39:03 +0100
Subject: [PATCH 06/15] clear up init_local_env

---
 openml_OS/controllers/Cron.php | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index f3fe2ae2e..4f4563b74 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -259,8 +259,8 @@ function install_database() {
   // Runs all local env init steps
   public function init_local_env() {
     $this->install_database();
-    $this->create_es_openml_index();
-    $this->initialize_es_indices();
+    // $this->create_es_openml_index();
+    // $this->initialize_es_indices();
     $this->create_local_admin();
     $this->build_es_indices();
   }

From 20490cb695332b2c19abfa7c07d8955d8c328039 Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Wed, 25 Nov 2020 11:15:52 +0100
Subject: [PATCH 07/15] working install db and es

---
 data/sql/dataset.sql            | 47 +--------------------------------
 data/sql/implementation.sql     | 30 ++++++++++-----------
 data/sql/implementation_tag.sql | 38 +++++++++++++-------------
 openml_OS/controllers/Cron.php  |  7 ++---
 4 files changed, 39 insertions(+), 83 deletions(-)

diff --git a/data/sql/dataset.sql b/data/sql/dataset.sql
index c8468e962..c2dc34c44 100644
--- a/data/sql/dataset.sql
+++ b/data/sql/dataset.sql
@@ -30,9 +30,6 @@ INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_
 (29, 1, 0, 'satimage', '1', '1', '**Author**: Ashwin Srinivasan, Department of Statistics and Data Modeling, University of Strathclyde \r\n**Source**: https://archive.ics.uci.edu/ml/datasets/Statlog+(Landsat+Satellite)\r\n**Please cite**:   \r\n\r\nThe database consists of the multi-spectral values of pixels in 3x3 neighbourhoods in a satellite image, and the classification associated with the central pixel in each neighbourhood. The aim is to predict this classification, given the multi-spectral values. In the sample database, the class of a pixel is coded as a number.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:14:59', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3619/dataset_186_satimage.arff', 'true', 29, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:14:59'),
 (30, 1, 0, 'abalone', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Abalone data\n \n 2. Sources:\n \n    (a) Original owners of database:\n 	Marine Resources Division\n 	Marine Research Laboratories - Taroona\n 	Department of Primary Industry and Fisheries, Tasmania\n 	GPO Box 619F, Hobart, Tasmania 7001, Australia\n 	(contact: Warwick Nash +61 02 277277, wnash@dpi.tas.gov.au)\n \n    (b) Donor of database:\n 	Sam Waugh (Sam.Waugh@cs.utas.edu.au)\n 	Department of Computer Science, University of Tasmania\n 	GPO Box 252C, Hobart, Tasmania 7001, Australia\n \n    (c) Date received: December 1995\n \n \n 3. Past Usage:\n \n    Sam Waugh (1995) \"Extending and benchmarking Cascade-Correlation\", PhD\n    thesis, Computer Science Department, University of Tasmania.\n \n    -- Test set performance (final 1044 examples, first 3133 used for training):\n 	24.86% Cascade-Correlation (no hidden nodes)\n 	26.25% Cascade-Correlation (5 hidden nodes)\n 	21.5%  C4.5\n 	 0.0%  Linear Discriminate Analysis\n 	 3.57% k=5 Nearest Neighbour\n       (Problem encoded as a classification task)\n \n    -- Data set samples are highly overlapped.  Further information is required\n 	to separate completely using affine combinations.  Other restrictions\n 	to data set examined.\n \n    David Clark, Zoltan Schreter, Anthony Adams \"A Quantitative Comparison of\n    Dystal and Backpropagation\", submitted to the Australian Conference on\n    Neural Networks (ACNN\'96). Data set treated as a 3-category classification\n    problem (grouping ring classes 1-8, 9 and 10, and 11 on).\n \n    -- Test set performance (3133 training, 1044 testing as above):\n 	64%    Backprop\n 	55%    Dystal\n    -- Previous work (Waugh, 1995) on same data set:\n 	61.40% Cascade-Correlation (no hidden nodes)\n 	65.61% Cascade-Correlation (5 hidden nodes)\n 	59.2%  C4.5\n 	32.57% Linear Discriminate Analysis\n 	62.46% k=5 Nearest Neighbour\n \n \n 4. Relevant Information Paragraph:\n \n    Predicting the age of abalone from physical measurements.  The age of\n    abalone is determined by cutting the shell through the cone, staining it,\n    and counting the number of rings through a microscope -- a boring and\n    time-consuming task.  Other measurements, which are easier to obtain, are\n    used to predict the age.  Further information, such as weather patterns\n    and location (hence food availability) may be required to solve the problem.\n \n    From the original data examples with missing values were removed (the\n    majority having the predicted value missing), and the ranges of the\n    continuous values have been scaled for use with an ANN (by dividing by 200).\n \n    Data comes from an original (non-machine-learning) study:\n \n 	Warwick J Nash, Tracy L Sellers, Simon R Talbot, Andrew J Cawthorn and\n 	Wes B Ford (1994) \"The Population Biology of Abalone (_Haliotis_\n 	species) in Tasmania. I. Blacklip Abalone (_H. rubra_) from the North\n 	Coast and Islands of Bass Strait\", Sea Fisheries Division, Technical\n 	Report No. 48 (ISSN 1034-3288)\n \n \n 5. Number of Instances: 4177\n \n \n 6. Number of Attributes: 8\n \n \n 7. Attribute information:\n \n    Given is the attribute name, attribute type, the measurement unit and a\n    brief description.  The number of rings is the value to predict: either\n    as a continuous value or as a classification problem.\n \n 	Name		Data Type	Meas.	Description\n 	----		---------	-----	-----------\n 	Sex		nominal			M, F, and I (infant)\n 	Length		continuous	mm	Longest shell measurement\n 	Diameter	continuous	mm	perpendicular to length\n 	Height		continuous	mm	with meat in shell\n 	Whole weight	continuous	grams	whole abalone\n 	Shucked weight	continuous	grams	weight of meat\n 	Viscera weight	continuous	grams	gut weight (after bleeding)\n 	Shell weight	continuous	grams	after being dried\n 	Rings		integer			+1.5 gives the age in years\n \n    Statistics for numeric domains:\n \n 		Length	Diam	Height	Whole	Shucked	Viscera	Shell	Rings\n 	Min	0.075	0.055	0.000	0.002	0.001	0.001	0.002	    1\n 	Max	0.815	0.650	1.130	2.826	1.488	0.760	1.005	   29\n 	Mean	0.524	0.408	0.140	0.829	0.359	0.181	0.239	9.934\n 	SD	0.120	0.099	0.042	0.490	0.222	0.110	0.139	3.224\n 	Correl	0.557	0.575	0.557	0.540	0.421	0.504	0.628	  1.0\n \n \n 8. Missing Attribute Values: None\n \n \n 9. Class Distribution:\n \n 	Class	Examples\n 	-----	--------\n 	1	1\n 	2	1\n 	3	15\n 	4	57\n 	5	115\n 	6	259\n 	7	391\n 	8	568\n 	9	689\n 	10	634\n 	11	487\n 	12	267\n 	13	203\n 	14	126\n 	15	103\n 	16	67\n 	17	58\n 	18	42\n 	19	32\n 	20	26\n 	21	14\n 	22	6\n 	23	9\n 	24	2\n 	25	1\n 	26	1\n 	27	2\n 	29	1\n 	-----	----\n 	Total	4177\n \n Num Instances:     4177\n Num Attributes:    9\n Num Continuous:    8 (Int 1 / Real 7)\n Num Discrete:      1\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'Sex\'                     Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n   2 \'Length\'                  Real   0%   0% 100%     0 /  0%   134 /  3%   0% \n   3 \'Diameter\'                Real   0%   0% 100%     0 /  0%   111 /  3%   0% \n   4 \'Height\'                  Real   0%   0% 100%     0 /  0%    51 /  1%   0% \n   5 \'Whole weight\'            Real   0%   0% 100%     0 /  0%  2429 / 58%  31% \n   6 \'Shucked weight\'          Real   0%   0% 100%     0 /  0%  1515 / 36%  10% \n   7 \'Viscera weight\'          Real   0%   0% 100%     0 /  0%   880 / 21%   3% \n   8 \'Shell weight\'            Real   0%   0% 100%     0 /  0%   926 / 22%   8% \n   9 \'Class_Rings\'             Int    0% 100%   0%     0 /  0%    28 /  1%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:15:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3620/dataset_187_abalone.arff', 'true', 30, 'Class_number_of_rings', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:15:04'),
 (31, 1, 0, 'eucalyptus', '1', '1', '**Author**: Bruce Bulloch  \n**Source**: [WEKA Dataset Collection](http://www.cs.waikato.ac.nz/ml/weka/datasets.html)   \n**Please cite**:   \n\n**Eucalyptus Soil Conservation**  \nThe objective was to determine which seedlots in a species are best for soil conservation in seasonally dry hill country. Determination is found by measurement of height, diameter by height, survival, and other contributing factors. \n \nIt is important to note that eucalypt trial methods changed over time; earlier trials included mostly 15 - 30cm tall seedling grown in peat plots and the later trials have included mostly three replications of eight trees grown. This change may contribute to less significant results.\n\nExperimental data recording procedures which require noting include:\n - instances with no data recorded due to experimental recording procedures\n   require that the absence of a species from one replicate at a site was\n   treated as a missing value, but if absent from two or more replicates at a\n   site the species was excluded from the site\'s analyses.\n - missing data for survival, vigour, insect resistance, stem form, crown form\n   and utility especially for the data recorded at the Morea Station; this \n   could indicate the death of species in these areas or a lack in collection\n   of data.  \n\n Attribute Information:  \n  1.  Abbrev - site abbreviation - enumerated\n  2.  Rep - site rep - integer\n  3.  Locality - site locality in the North Island - enumerated\n  4.  Map_Ref - map location in the North Island - enumerated\n  5.  Latitude - latitude approximation - enumerated\n  6.  Altitude - altitude approximation - integer\n  7.  Rainfall - rainfall (mm pa) - integer\n  8.  Frosts - frosts (deg. c) - integer\n  9.  Year - year of planting - integer\n  10. Sp - species code - enumerated\n  11. PMCno - seedlot number - integer\n  12. DBH - best diameter base height (cm) - real\n  13. Ht - height (m) - real\n  14. Surv - survival - integer\n  15. Vig - vigour - real\n  16. Ins_res - insect resistance - real\n  17. Stem_Fm - stem form - real\n  18. Crown_Fm - crown form - real\n  19. Brnch_Fm - branch form - real\n  Class:\n  20. Utility - utility rating - enumerated\n\n Class Distribution:\n  none    - 180\n  low     - 107\n  average - 130\n  good    - 214\n  best    - 105\n\n Contact: Bruce Bulloch, 128 Cook Street, Palmerston North, New Zealand', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:15:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3625/dataset_194_eucalyptus.arff', 'true', 31, 'Utility', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:15:28'),
-(32, 94, 0, 'isolet', '1', NULL, '**Author**: Ron Cole and Mark Fanty (cole@cse.ogi.edu, fanty@cse.ogi.edu)  \n**Donor**: Tom Dietterich (tgd@cs.orst.edu)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/ISOLET) - 1994  \n**Please cite**:   \n\n**ISOLET (Isolated Letter Speech Recognition)**  \nThis data set was generated as follows. 150 subjects spoke the name of each letter of the alphabet twice. Hence, we have 52 training examples from each speaker. The speakers are grouped into sets of 30 speakers each, 4 groups can serve as trainings set, the last group as the test set.\n\nYou will note that 3 examples are missing.  I believe they were dropped due to difficulties in recording.\n\nI believe this is a good domain for a noisy, perceptual task.  It is also a very good domain for testing the scaling abilities of algorithms. For example, C4.5 on this domain is slower than backpropagation!\n\nPast Usage:\n* Fanty, M., Cole, R. (1991).  Spoken letter recognition.  In Lippman, R. P., Moody, J., and Touretzky, D. S. (Eds). Advances in Neural Information Processing Systems 3.  San Mateo, CA: Morgan Kaufmann.\nGoal: Predict which letter-name was spoken, a simple classification task.  95.9% correct classification using the OPT backpropagation implementation.  Training on isolet1+2+3+4, testing on isolet5.  Network architecture: 56 hidden units, 26 output units (one-per-class).\n\n* Dietterich, T. G., Bakiri, G. (1991)  Error-correcting output codes: A general method for improving multiclass inductive learning programs.  Proceedings of the Ninth National Conference on Artificial Intelligence (AAAI-91), Anaheim, CA: AAAI Press.\nGoal: same as above. 95.83% correct using OPT backpropagation.  (Architecture: 78 hidden units, 26 output units, one-per-class). 96.73% correct using a 30-bit error-correcting output code with OPT (Architecture: 156 hidden units, 30 output units).\n\n**Attributes**  \nAll attributes are continuous, real-valued attributes scaled into the range -1.0 to 1.0. The features are described in the paper by Cole and Fanty cited above. The features include spectral coefficients; contour features, sonorant features, pre-sonorant features, and post-sonorant features.  Exact order of appearance of the features is not known.', 'ARFF', NULL, NULL, '1994', '2014-08-20 20:59:05', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52405/phpB0xrNj', 'true', 32, 'class', NULL, NULL, NULL, 'public', NULL, NULL, 'all numeric variables declared as factors before', '2014-09-25 23:26:09'),
-(33, 2, 0, 'vowel', '2', '2', '**Author**: Peter Turney (peter@ai.iit.nrc.ca)   \n**Source**: [UCI](https://archive.ics.uci.edu/ml/machine-learning-databases/undocumented/connectionist-bench/vowel/) - date unknown  \n**Please cite**:   \n\n**Vowel Recognition Data**  \nIn my work on context-sensitive learning, I used the \"Deterding Vowel Recognition Data\", but I found it necessary to reformulate the data. Implicit in the original data is contextual information on the speaker\'s gender and identity. For my work, it was necessary to make this information explicit. This dataset adds the speaker\'s sex and identity as new features. \n\nNotes:  \n* This is version 2. Version 1 is hidden because it includes a feature dividing the data in train and test set. In OpenML this information is explicitly available in the corresponding task.', 'ARFF', '\"Peter Turney\"', '\"\"', '', '2014-08-22 16:57:30', NULL, 'Public', '', NULL, 'https://www.openml.org/data/download/52210/phpd8EoD9', 'true', 33, 'Class', '', NULL, '', 'public', NULL, 'http://www.openml.org/d/58', NULL, '2014-08-22 16:57:30'),
-(34, 94, 0, 'scene', '1', '', '**Author**:   \n  \n**Source**: Unknown -   \n**Please cite**:   \n\nScene recognition dataset\n\nSource:\nMatthew R. Boutell, Jiebo Luo, Xipeng Shen, and Christopher M. Brown.\nLearning multi-label scene classification.\nPattern Recognition, 37(9):1757-1771, 2004. \n\n1: Description.\nScenes data set contains characteristics about images and thier classes. One image can belong to one or more classes.\n\n2: Type: Multi label \n3: Origin: Real world\n4: Instances: 2407\n5: Features: 294\n6: Labels: 6 \n7: Missing values: No', 'ARFF', '\"\"', '\"\"', '', '2014-08-25 11:43:22', NULL, 'Public', '', NULL, 'https://www.openml.org/data/download/1390080/phpuZu33P', 'true', 34, 'Urban', '', NULL, '', 'public', NULL, '', 'removed ranges after the definition of a numerical attribute', '2015-04-15 22:22:58'),
 (35, 2, 0, 'monks-problems-1', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 1\n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks. \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA  E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992\n\n4. Relevant Information:  The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991.  One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users.  There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.\n\nAttribute information: \n1. class: 0, 1  \n2. a1: 1, 2, 3 \n3. a2: 1, 2, 3 \n4. a3: 1, 2 \n5. a4: 1, 2, 3 \n6. a5: 1, 2, 3, 4 \n7. a6: 1, 2 8. \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, 'October 1992', '2014-08-26 17:11:18', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52236/phpAyyBys', 'true', 35, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:11:18'),
 (36, 2, 0, 'monks-problems-2', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 2  \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992  \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.  \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2 \n\n8.   Target Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, 'October 1992', '2014-08-26 17:29:02', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52237/php4fATLZ', 'true', 36, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:29:02'),
 (37, 2, 0, 'monks-problems-3', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) -   \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 3   \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992   \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.   \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2  \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, NULL, '2014-08-26 17:41:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52238/phphZierv', 'true', 37, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:41:07'),
@@ -59,48 +56,6 @@ INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_
 (58, 2, NULL, 'MagicTelescope', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDataset from the MLRR repository: http://axon.cs.byu.edu:5000/', 'ARFF', NULL, NULL, NULL, '2014-10-07 00:42:01', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/54003/MagicTelescope.arff', 'true', 58, 'class:', 'ID', NULL, NULL, 'public', NULL, NULL, 'ID is a row id', '2015-04-15 17:41:20'),
 (59, 2, NULL, 'Internet-Advertisements', '1', NULL, '**Author**: Nicholas Kushmerick  \n**Source**: [UCI](http://archive.ics.uci.edu/ml/datasets/Internet+Advertisements) - 1998  \n**Please cite**:   \n\nThis dataset represents a set of possible advertisements on Internet pages. The features encode the geometry of the image (if available) as well as phrases occurring in the URL, the image\'s URL and alt text, the anchor text, and words occurring near the anchor text. The task is to predict whether an image is an advertisement (\"ad\") or not (\"nonad\").\n\nRelevant Papers: N. Kushmerick (1999). \"Learning to remove Internet advertisements\", 3rd Int Conf Autonomous Agents.  \nAvailable at: http://rexa.info/paper/2fdc1cee89b7f4f2c9227d6f5d9b05d22c5ab3e9', 'ARFF', '\"Nicholas Kushmerick\"', NULL, NULL, '2014-10-30 11:15:44', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/116567/phpCzcrGG', 'true', 59, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-30 11:15:44'),
 (60, 2, NULL, 'Click_prediction_small', '5', NULL, '**Author**: Tencent Inc.  \n**Source**: [KDD Cup](https://www.kddcup2012.org/) - 2012  \n**Please cite**:   \n\n**0.1% balanced subsample of the original KDD dataset**  \n\nThis data is derived from the 2012 KDD Cup. The data is subsampled to 0.1% of the original number of instances, downsampling the majority class (click=0) so that the target feature is reasonably balanced (5 to 1).\n\nThe data is about advertisements shown alongside search results in a search engine, and whether or not people clicked on these ads. \nThe task is to build the best possible model to predict whether a user will click on a given ad.\n\nA search session contains information on user id, the query issued by the user, ads displayed to the user, and target feature indicating whether a user clicked at least one of the ads in this session. The number of ads displayed to a user in a session is called ‘depth’. The order of an ad in the displayed list is called ‘position’.  An ad is displayed as a short text called ‘title’, followed by a slightly longer text called ’description’, and a URL  called ‘display URL’.   \nTo construct this dataset each session was split into multiple instances. Each instance describes an ad displayed under a certain setting  (‘depth’, ‘position’).  Instances with the same user id, ad id, query, and setting are merged. Each ad and each user have some additional properties located in separate data files that can be looked up using ids in the instances.\n\nThe dataset has the following features:  \n* Click – binary variable indicating whether a user clicked on at least one ad. \n* Impression - the number of search sessions in which AdID was impressed by UserID who issued Query.\n* Url_hash - URL is hashed for anonymity\n* AdID \n* AdvertiserID - some advertisers consistently optimize their ads, so the title and description of their ads are more attractive than those of others’ ads.\n* Depth - number of ads displayed to a user in a session\n* Position - order of an ad in the displayed list\n* QueryID - is the key of the data file \'queryid_tokensid.txt\'. (follow the link to the original KDD Cup page, track 2)\n* KeywordID - is the key of  \'purchasedkeyword_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* TitleID - is the key of \'titleid_tokensid.txt\'\n* DescriptionID - is the key of \'descriptionid_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* UserID – is also the key of \'userid_profile.txt\' (follow the link to the original KDD Cup page, track 2). 0 is a special value denoting that the user could be identified.', 'ARFF', NULL, NULL, NULL, '2014-11-27 01:18:40', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/184157/phpfGCaQC', 'true', 60, 'click', NULL, '\"url_hash\",\"query_id\"', NULL, 'public', NULL, NULL, NULL, '2014-11-27 01:26:36'),
-(61, 64, NULL, 'artificial-characters', '1', NULL, '**Author**: H. Altay Guvenir, Burak Acar, Haldun Muderrisoglu    \n\n**Source**: UCI\n\n**Please cite**:   \n\nDataset artificially generated by using first order theory which describes structure of ten capital letters of English alphabet', 'ARFF', NULL, NULL, NULL, '2015-05-21 20:58:53', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586212/phpPQrHPH', 'true', 61, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-21 20:58:53'),
-(62, 64, NULL, 'bank-marketing', '1', NULL, '**Author**: Paulo Cortez, Sérgio Moro   \n**Source**: UCI    \n**Please cite**: S. Moro, R. Laureano and P. Cortez. Using Data Mining for Bank Direct Marketing: An Application of the CRISP-DM Methodology. In P. Novais et al. (Eds.), Proceedings of the European Simulation and Modelling Conference - ESM\'2011, pp. 117-121, Guimarães, Portugal, October, 2011. EUROSIS.       \n\nAvailable at: [pdf] http://hdl.handle.net/1822/14838\n              [bib] http://www3.dsi.uminho.pt/pcortez/bib/2011-esm-1.txt\n \n1. Title: Bank Marketing\n \n2. Sources\n    Created by: Paulo Cortez (Univ. Minho) and Sérgio Moro (ISCTE-IUL) @ 2012\n    \n3. Past Usage:\n \n   The full dataset was described and analyzed in:\n \n S. Moro, R. Laureano and P. Cortez. Using Data Mining for Bank Direct Marketing: An Application of the CRISP-DM Methodology. In P. Novais et al. (Eds.), Proceedings of the European Simulation and Modelling Conference - ESM\'2011, pp. 117-121, Guimarães, Portugal, October, 2011. EUROSIS.\n\n4. Relevant Information:\n\nThe data is related with direct marketing campaigns of a Portuguese banking institution. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required, in order to access if the product (bank term deposit) would be (or not) subscribed. \n\nThe classification goal is to predict if the client will subscribe a term deposit (variable y).\n\n5. Attribute information:\n \nFor more information, read [Moro et al., 2011].\n\nInput variables:\n# bank client data:\n1 - age (numeric)\n2 - job : type of job (categorical: \"admin.\",\"unknown\",\"unemployed\",\"management\",\"housemaid\",\"entrepreneur\", \"student\",\"blue-collar\",\"self-employed\",\"retired\",\"technician\",\"services\") \n3 - marital : marital status (categorical: \"married\",\"divorced\",\"single\"; note: \"divorced\"  means divorced or widowed)\n4 - education (categorical: \"unknown\",\"secondary\",\"primary\",\"tertiary\")\n5 - default: has credit in default? (binary: \"yes\",\"no\")\n6 - balance: average yearly balance, in euros (numeric) \n7 - housing: has housing loan? (binary: \"yes\",\"no\")\n8 - loan: has personal loan? (binary: \"yes\",\"no\")\n# related with the last contact of the current campaign:\n9 - contact: contact communication type (categorical: \"unknown\",\"telephone\",\"cellular\") \n10 - day: last contact day of the month (numeric)\n11 - month: last contact month of year (categorical: \"jan\", \"feb\", \"mar\", ..., \"nov\", \"dec\")\n12 - duration: last contact duration, in seconds (numeric)\n# other attributes:\n13 - campaign: number of contacts performed during this campaign and for this client (numeric, includes last contact)\n14 - pdays: number of days that passed by after the client was last contacted from a previous campaign (numeric, -1 means client was not previously contacted)\n15 - previous: number of contacts performed before this campaign and for this client (numeric)\n16 - poutcome: outcome of the previous marketing campaign (categorical: \"unknown\",\"other\",\"failure\",\"success\")\n \nOutput variable (desired target):\n17 - y - has the client subscribed a term deposit? (binary: \"yes\",\"no\")\n', 'ARFF', NULL, NULL, NULL, '2015-05-21 22:16:49', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586218/phpkIxskf', 'true', 62, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-21 22:16:49'),
-(63, 64, NULL, 'banknote-authentication', '1', NULL, '**Author**: Volker Lohweg  \n**Source**: UCI\n**Please cite**:   \n\nSource:\n\nOwner of database: Volker Lohweg (University of Applied Sciences, Ostwestfalen-Lippe, volker.lohweg \'@\' hs-owl.de) \n\nDonor of database: Helene Doerksen (University of Applied Sciences, Ostwestfalen-Lippe, helene.doerksen \'@\' hs-owl.de) . Date received: August, 2012 \n\nData Set Information:\n \nData were extracted from images that were taken from genuine and forged banknote-like specimens. For digitization, an industrial camera usually used for print inspection was used. The final images have 400x 400 pixels. Due to the object lens and distance to the investigated object gray-scale pictures with a resolution of about 660 dpi were gained. Wavelet Transform tool were used to extract features from images.\n\nAttribute Information:\n \n1. variance of Wavelet Transformed image (continuous) \n2. skewness of Wavelet Transformed image (continuous) \n3. curtosis of Wavelet Transformed image (continuous) \n4. entropy of image (continuous) \n5. class (integer)', 'ARFF', '\"Volker Lohweg\"', NULL, NULL, '2015-05-21 22:40:57', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586223/php50jXam', 'true', 63, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-01-18 18:45:44'),
-(64, 64, NULL, 'blood-transfusion-service-center', '1', NULL, '**Author**: Prof. I-Cheng Yeh  \n**Source**: UCI\n**Please cite**: Yeh, I-Cheng, Yang, King-Jang, and Ting, Tao-Ming, \"Knowledge discovery on RFM model using Bernoulli sequence, \"Expert Systems with Applications, 2008 (doi:10.1016/j.eswa.2008.07.018).   \n\nTitle: Blood Transfusion Service Center Data Set\n\nAbstract: Data taken from the Blood Transfusion Service Center in Hsin-Chu City in Taiwan -- this is a classification problem.\n\n-----------------------------------------------------\nDate Donated: 2008-10-03\n-----------------------------------------------------\n  \nSource:\n  \nOriginal Owner and Donor\nProf. I-Cheng Yeh, Department of Information Management, Chung-Hua University,  Hsin Chu, Taiwan 30067, R.O.C.\ne-mail:icyeh \'at\' chu.edu.tw, TEL:886-3-5186511\nDate Donated: October 3, 2008 \n\n-----------------------------------------------------\n  \nData Set Information:\n  \nTo demonstrate the RFMTC marketing model (a modified version of RFM), this study adopted the donor database of Blood Transfusion Service Center in Hsin-Chu City in Taiwan. The center passes their blood transfusion service bus to one university in Hsin-Chu City to gather blood donated about every three months. To build a FRMTC model, we selected 748 donors at random from the donor database. These 748 donor data, each one included R (Recency - months since last donation), F (Frequency - total number of donation), M (Monetary - total blood donated in c.c.), T (Time - months since first donation), and a binary variable representing whether he/she donated blood in March 2007 (1 stand for donating blood; 0 stands for not donating blood).\n\n-----------------------------------------------------\n  \nAttribute Information:\n  \nGiven is the variable name, variable type, the measurement unit and a brief description. The \"Blood Transfusion Service Center\" is a classification problem. The order of this listing corresponds to the order of numerals along the rows of the database.\n\nR (Recency - months since last donation),\nF (Frequency - total number of donation),\nM (Monetary - total blood donated in c.c.),\nT (Time - months since first donation), and a binary variable representing whether he/she donated blood in March 2007 (1 stand for donating blood; 0 stands for not donating blood).\n \nCitation Request:\n  \nNOTE: Reuse of this database is unlimited with retention of copyright notice for Prof. I-Cheng Yeh and the following published paper:\n\nYeh, I-Cheng, Yang, King-Jang, and Ting, Tao-Ming, \"Knowledge discovery on RFM model using Bernoulli sequence, \"Expert Systems with Applications, 2008 (doi:10.1016/j.eswa.2008.07.018).', 'ARFF', NULL, NULL, NULL, '2015-05-21 22:49:48', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586225/php0iVrYT', 'true', 64, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 21:02:59'),
-(65, 64, NULL, 'cardiotocography', '1', NULL, '**Author**: J. P. Marques de Sá, J. Bernardes, D. Ayers de Campos.  \n**Source**: UCI  \n**Please cite**:     \n\n* Source:\n\nMarques de Sá, J.P., jpmdesa \'@\' gmail.com, Biomedical Engineering Institute, Porto, Portugal. \nBernardes, J., joaobern \'@\' med.up.pt, Faculty of Medicine, University of Porto, Portugal. \nAyres de Campos, D., sisporto \'@\' med.up.pt, Faculty of Medicine, University of Porto, Portugal.\n\n\n* Data Set Information:\n\n2126 fetal cardiotocograms (CTGs) were automatically processed and the respective diagnostic features measured. The CTGs were also classified by three expert obstetricians and a consensus classification label assigned to each of them. Classification was both with respect to a morphologic pattern (A, B, C. ...) and to a fetal state (N, S, P). Therefore the dataset can be used either for 10-class or 3-class experiments.\n\n\n* Attribute Information:\n\nLB - FHR baseline (beats per minute) \nAC - # of accelerations per second \nFM - # of fetal movements per second \nUC - # of uterine contractions per second \nDL - # of light decelerations per second \nDS - # of severe decelerations per second \nDP - # of prolongued decelerations per second \nASTV - percentage of time with abnormal short term variability \nMSTV - mean value of short term variability \nALTV - percentage of time with abnormal long term variability \nMLTV - mean value of long term variability \nWidth - width of FHR histogram \nMin - minimum of FHR histogram \nMax - Maximum of FHR histogram \nNmax - # of histogram peaks \nNzeros - # of histogram zeros \nMode - histogram mode \nMean - histogram mean \nMedian - histogram median \nVariance - histogram variance \nTendency - histogram tendency \nCLASS - FHR pattern class code (1 to 10) \nNSP - fetal state class code (N=normal; S=suspect; P=pathologic)\n\n\n* Relevant Papers:\n\nAyres de Campos et al. (2000) SisPorto 2.0 A Program for Automated Analysis of Cardiotocograms. J Matern Fetal Med 5:311-318 ', 'ARFF', NULL, NULL, NULL, '2015-05-21 23:05:18', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586231/php9HX2u8', 'true', 65, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:49:14'),
-(66, 64, NULL, 'climate-model-simulation-crashes', '1', NULL, '**Author**:   \n**Source**: UCI\n\n**Please cite**:   \n\nSource:\n\nD. Lucas (ddlucas .at. alum.mit.edu), Lawrence Livermore National Laboratory; R. Klein (rklein .at. astron.berkeley.edu), Lawrence Livermore National Laboratory & U.C. Berkeley; J. Tannahill (tannahill1 .at. llnl.gov), Lawrence Livermore National Laboratory; D. Ivanova (ivanova2 .at. llnl.gov), Lawrence Livermore National Laboratory; S. Brandon (brandon1 .at. llnl.gov), Lawrence Livermore National Laboratory; D. Domyancic (domyancic1 .at. llnl.gov), Lawrence Livermore National Laboratory; Y. Zhang (zhang24 .at. llnl.gov), Lawrence Livermore National Laboratory .\n\nThis data was constructed using LLNL\'s UQ Pipeline, was created under the auspices of the US Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344, was funded by LLNL\'s Uncertainty Quantification Strategic Initiative Laboratory Directed Research and Development Project under tracking code 10-SI-013, and is released under UCRL number LLNL-MISC-633994.\n\n\nData Set Information:\n\nThis dataset contains records of simulation crashes encountered during climate model uncertainty quantification (UQ) ensembles. Ensemble members were constructed using a Latin hypercube method in LLNL\'s UQ Pipeline software system to sample the uncertainties of 18 model parameters within the Parallel Ocean Program (POP2) component of the Community Climate System Model (CCSM4). Three separate Latin hypercube ensembles were conducted, each containing 180 ensemble members. 46 out of the 540 simulations failed for numerical reasons at combinations of parameter values. The goal is to use classification to predict simulation outcomes (fail or succeed) from input parameter values, and to use sensitivity analysis and feature selection to determine the causes of simulation crashes. Further details about the data and methods are given in the publication \'Failure Analysis of Parameter-Induced Simulation Crashes in Climate Models,\' Geoscientific Model Development ([Web Link]).\n\n\nAttribute Information:\n\nThe goal is to predict climate model simulation outcomes (column 21, fail or succeed) given scaled values of climate model input parameters (columns 3-20). \n\nColumn 1: Latin hypercube study ID (study 1 to study 3) \nColumn 2: simulation ID (run 1 to run 180) \nColumns 3-20: values of 18 climate model parameters scaled in the interval [0, 1] \nColumn 21: simulation outcome (0 = failure, 1 = success)\n\nRelevant Papers:\n\nLucas, D. D., Klein, R., Tannahill, J., Ivanova, D., Brandon, S., Domyancic, D., and Zhang, Y.: Failure analysis of parameter-induced simulation crashes in climate models, Geosci. Model Dev. Discuss., 6, 585-623, [Web Link], 2013.', 'ARFF', NULL, NULL, NULL, '2015-05-21 23:11:45', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586232/phpXeun7q', 'true', 66, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:49:00'),
-(67, 64, NULL, 'cnae-9', '1', NULL, '**Author**: Patrick Marques Ciarelli, Elias Oliviera   \n**Source**: UCI\n\n**Please cite**:   \n\nSource:\n\nPatrick Marques Ciarelli, pciarelli \'@\' lcad.inf.ufes.br, Department of Electrical Engineering, Federal University of Espirito Santo \nElias Oliveira, elias \'@\' lcad.inf.ufes.br, Department of Information Science, Federal University of Espirito Santo\n\n\nData Set Information:\n\nThis is a data set containing 1080 documents of free text business descriptions of Brazilian companies categorized into a \nsubset of 9 categories cataloged in a table called National Classification of Economic Activities (Classificação Nacional de \nAtividade Econômicas - CNAE). The original texts were pre-processed to obtain the current data set: initially, it was kept only \nletters and then it was removed prepositions of the texts. Next, the words were transformed to their canonical form. Finally, \neach document was represented as a vector, where the weight of each word is its frequency in the document. This data set is \nhighly sparse (99.22% of the matrix is filled with zeros).\n\n\nAttribute Information:\n\nIn the data set there are 857 attributes, 1 attributes with the class of instance and 856 with word frequency: \n1. category: range 1 - 9 (integer) \n2 - 857. word frequency: (integer)\n\n\nRelevant Papers:\n\nPatrick Marques Ciarelli, Elias Oliveira, \'Agglomeration and Elimination of Terms for Dimensionality Reduction\', \nNinth International Conference on Intelligent Systems Design and Applications, pp.547-552, 2009 \n\nPatrick Marques Ciarelli, Elias Oliveira, Evandro O. T. Salles, \'An Evolving System Based on Probabilistic Neural Network\', \nBrazilian Symposium on Artificial Neural Network, 2010', 'ARFF', NULL, NULL, NULL, '2015-05-21 23:19:32', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1586233/phpmcGu2X', 'true', 67, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-21 23:19:32'),
-(68, 64, NULL, 'eeg-eye-state', '1', NULL, '**Author**: Oliver Roesler, it12148\'@\'lehre.dhbw-stuttgart.de  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/EEG+Eye+State), Baden-Wuerttemberg, Cooperative State University (DHBW), Stuttgart, Germany  \n**Please cite**:   \n\nAll data is from one continuous EEG measurement with the Emotiv EEG Neuroheadset. The duration of the measurement was 117 seconds. The eye state was detected via a camera during the EEG measurement and added later manually to the file after analysing the video frames. \'1\' indicates the eye-closed and \'0\' the eye-open state. All values are in chronological order with the first measured value at the top of the data.\n', 'ARFF', NULL, NULL, NULL, '2015-05-22 16:40:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1587924/phplE7q6h', 'true', 68, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:42:10'),
-(69, 64, NULL, 'first-order-theorem-proving', '1', NULL, '**Author**: James P Bridge, Sean B Holden and Lawrence C Paulson \n  \n**Source**: UCI  \n\n**Please cite**: James P Bridge, Sean B Holden and Lawrence C Paulson . Machine learning for first-order theorem proving: learning to select a good heuristic. Journal of Automated Reasoning, Springer 2012/13. \n\nSource:\n\nJames P Bridge, Sean B Holden and Lawrence C Paulson \n\nUniversity of Cambridge \nComputer Laboratory \nWilliam Gates Building \n15 JJ Thomson Avenue \nCambridge CB3 0FD \nUK \n\n+44 (0)1223 763500 \nforename.surname \'@\' cl.cam.ac.uk\n\n\nData Set Information:\n\nSee the file dataset file.\n\n\nAttribute Information:\n\nThe attributes are a mixture of static and dynamic features derived from theorems to be proved. See the paper for full details.\n', 'ARFF', NULL, NULL, NULL, '2015-05-22 17:30:40', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1587932/phpPbCMyg', 'true', 69, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:36:55'),
-(70, 64, NULL, 'gas-drift', '1', NULL, '**Author**: Alexander Vergara\n\n**Source**: UCI  \n\n**Please cite**: Alexander Vergara and Shankar Vembu and Tuba Ayhan and Margaret A. Ryan and Margie L. Homer and Ramón Huerta, Chemical gas sensor drift compensation using classifier ensembles, Sensors and Actuators B: Chemical (2012) doi: 10.1016/j.snb.2012.01.074. \n\nTitle: Gas Sensor Array Drift Dataset Data Set\n\nSource:\n\nCreators: Alexander Vergara (vergara \'@\' ucsd.edu) \nBioCircutis Institute \nUniversity of California San Diego \nSan Diego, California, USA \nDonors of the Dataset: Alexander Vergara (vergara \'@\' ucsd.edu) \nRamon Huerta (rhuerta \'@\' ucsd.edu) \n\n\nData Set Information:\n\nThis archive contains 13910 measurements from 16 chemical sensors utilized in simulations for drift compensation in a discrimination task of 6 gases at various levels of concentrations. The goal is to achieve good performance (or as low degradation as possible) over time, as reported in the paper mentioned below in Section 2: Data collection. The primary purpose of providing this dataset is to make it freely accessible on-line to the chemo-sensor research community and artificial intelligence to develop strategies to cope with sensor/concept drift. The dataset can be used exclusively for research purposes. Commercial purposes are fully excluded.\nThe dataset was gathered within January 2007 to February 2011 (36 months) in a gas delivery platform facility situated at the ChemoSignals Laboratory in the BioCircuits Institute, University of California San Diego. Being completely operated by a fully computerized environment —controlled by a LabVIEW–National Instruments software on a PC fitted with the appropriate serial data acquisition boards. The measurement system platform provides versatility for obtaining the desired concentrations of the chemical substances of interest with high accuracy and in a highly reproducible manner, minimizing thereby the common mistakes caused by human intervention and making it possible to exclusively concentrate on the chemical sensors for compensating real drift.\nThe resulting dataset comprises recordings from six distinct pure gaseous substances, namely Ammonia, Acetaldehyde, Acetone, Ethylene, Ethanol, and Toluene, each dosed at a wide variety of concentration values ranging from 5 to 1000 ppmv. See Tables 1 and 2 of the below cited manuscript for details on the gas identity name, concentration values, and time distribution sequence of the measurement recordings considered in this dataset.\n\nBatch10.dat was updated on 10/14/2013 to correct some corrupted values in the last 120 lines of the file.\n\nAn extension of this dataset with the concentration values is available at Gas Sensor Array Drift Dataset at Different Concentrations Data Set [Web Link]\n\n\nAttribute Information:\n\nThe response of the said sensors is read-out in the form of the resistance across the active layer of each sensor; hence each measurement produced a 16-channel time series, each of which represented by an aggregate of features reflecting all the dynamic processes occurring at the sensor surface in reaction to the chemical substance being evaluated. In particular, two distinct types of features were considered in the creation of this dataset: (i) The so-called steady-state feature (ΔR), defined as the difference of the maximal resistance change and the baseline and its normalized version expressed by the ratio of the maximal resistance and the baseline values when the chemical vapor is present in the test chamber. And (ii), an aggregate of features reflecting the sensor dynamics of the increasing/decaying transient portion of the sensor response during the entire measurement procedure under controlled conditions, namely the exponential moving average (emaα). These aggregate of features is a transform, borrowed from the field of econometrics originally introduced to the chemo-sensing community by Muezzinoglu et al. (2009), that converts the said transient portion into a real scalar, by estimating the maximum value —minimum for the decaying portion of the sensor response— of its exponential moving average (emaα), with an initial condition set to zero and a scalar smoothing parameter of the operator, α, that defines both the quality of the feature and the time of its occurrence along the time series the scalar, set to range between 0 and 1. In particular, three different values for α were set to obtain three different feature values from the pre-recorded rising portion of the sensor response and three additional features with the same α values but for the decaying portion of the sensor response, covering thus the entire sensor response dynamics. For a more detailed analysis and discussion on these features as well as a graphical illustration of them please refer to Section 2.3 and Figure 2, respectively of the annotated manuscript.\nOnce the abovementioned features are calculated, one is to form a feature vector containing the 8 features extracted from each particular sensor multiplied by the 16 sensors considered here. In the end, the resulting 128-dimensional feature vector containing all the features indicated above (8 features × 16 sensors) is organized as follows:\nΔR_1, |ΔR|_1, EMAi0.001_1, EMAi0.01_1, EMAi0.1_1, EMAd0.001_1, EMAd0.01_1, EMAd0.1_1, ΔR_2, |ΔR|_2, EMAi0.001_2, EMAi0.01_2, EMAi0.1_2, EMAd0.001_2, EMAd0.01_2, EMAd0.1_2,..., ΔR_16, |ΔR|_16, EMAi0.001_16, EMAi0.01_16, EMAi0.1_16, EMAd0.001_16, EMAd0.01_16, EMAd0.1_16,\n\nwhere: “ΔR_1” and “|ΔR|_1” is the ΔR and the normalized ΔR feature, respectively, “EMAi0.001_1”, “EMAi0.01_1”, and “EMAi0.1_1”, the emaα of the rising transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, and “EMAd0.001_1”, “EMAd0.01_1”, and “EMAd0.1_1”, the emaα of the decaying transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, all corresponding to sensor # 1; “ΔR_2” and “|ΔR|_2” is the ΔR and the normalized ΔR feature, respectively, “EMAi0.001_2”, “EMAi0.01_2”, and “EMAi0.1_2”, the emaα of the rising transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, and “EMAd0.001_2”, “EMAd0.01_2”, and “EMAd0.1_2”, the emaα of the decaying transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, all corresponding to sensor # 2; and so forth up until sensor # 16, forming thus the 128-dimensional feature vector that is to be fetched to the classifiers for training.\nFor processing purposes, the data is organized into ten batches, each containing the number of measurements per class and month indicated in the table below. This reorganization of data was done to ensure having a sufficient and as uniformly distributed as possible number of experiments in each class and month when training the classifier.\n\nDataset organization details. Each row corresponds to months that were combined to form a batch:\nBatch ID Month IDs\nBatch 1 Months 1 and 2\nBatch 2 Months 3, 4, 8, 9 and 10\nBatch 3 Months 11, 12, and 13\nBatch 4 Months 14 and 15\nBatch 5 Month 16\nBatch 6 Months 17, 18, 19, and 20\nBatch 7 Month 21\nBatch 8 Months 22 and 23\nBatch 9 Months 24 and 30\nBatch 10 Month 36\n\nThe data format follows the same coding style as in libsvm, in which one indicates the class each data point belongs to (1: Ethanol; 2: Ethylene; 3:Ammonia; 4: Acetaldehyde; 5: Acetone; 6: Toluene), and, then, the collection of features in a format x:v, where x stands for the feature number and v for the actual value of the feature. For example, in \n\n1 1:15596.162100 2:1.868245 3:2.371604 4:2.803678 5:7.512213 … 128:-2.654529 \n\nThe number “1” stands for the class number (in this case Ethanol), whereas the remaining 128 columns list the actual feature values for each measurement recording organized as described above. \nFinally, to make the results presented in the associated article reproducible for the reader, please use the following parameter values in the training task:\n• folds: 10\n• log2c = -5, 10, 1\n• log2g = -10, 5, 1\n• Scale the features in the training set appropriately to lie between -1 and +1.\n• And use the following cross validation parameters:\n\nBatch C Gamma (ɤ) Rate\n1 256.0 0.03125 98.8764\n2 64.0 0.00390625 99.7588\n3 128.0 0.03125 100.0\n4 1.0 0.25 100.0\n5 2.0 0.015625 99.4924\n6 256.0 0.0009765625 99.5217\n7 64.0 0.0625 99.9723\n8 1024.0 0.0078125 99.6599\n9 2.0 0.00390625 100.0', 'ARFF', NULL, NULL, NULL, '2015-05-22 20:11:33', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1588715/phpbL6t4U', 'true', 70, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-22 20:11:33'),
-(71, 64, NULL, 'har', '1', NULL, '**Author**: Jorge L. Reyes-Ortiz, Davide Anguita, Alessandro Ghio, Luca Oneto and Xavier Parra\n\n**Source**: UCI\n\n**Please cite**: Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. A Public Domain Dataset for Human Activity Recognition Using Smartphones. 21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013. Bruges, Belgium 24-26 April 2013.   \n\nTitle: Human Activity Recognition Using Smartphones\n\nAbstract: Human Activity Recognition database built from the recordings of 30 subjects performing activities of daily living (ADL) while carrying a waist-mounted smartphone with embedded inertial sensors.\n\nSource:\n\nJorge L. Reyes-Ortiz(1,2), Davide Anguita(1), Alessandro Ghio(1), Luca Oneto(1) and Xavier Parra(2)\n1 - Smartlab - Non-Linear Complex Systems Laboratory DITEN - Università degli Studi di Genova, Genoa (I-16145), Italy. \n2 - CETpD - Technical Research Centre for Dependency Care and Autonomous Living Universitat Politècnica de Catalunya (BarcelonaTech). Vilanova i la Geltrú (08800), Spain, activityrecognition \'@\' smartlab.ws\n\n\nData Set Information:\n\nThe experiments have been carried out with a group of 30 volunteers within an age bracket of 19-48 years. Each person performed six activities (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING) wearing a smartphone (Samsung Galaxy S II) on the waist. Using its embedded accelerometer and gyroscope, we captured 3-axial linear acceleration and 3-axial angular velocity at a constant rate of 50Hz. The experiments have been video-recorded to label the data manually. The obtained dataset has been randomly partitioned into two sets, where 70% of the volunteers was selected for generating the training data and 30% the test data. \n\nThe sensor signals (accelerometer and gyroscope) were pre-processed by applying noise filters and then sampled in fixed-width sliding windows of 2.56 sec and 50% overlap (128 readings/window). The sensor acceleration signal, which has gravitational and body motion components, was separated using a Butterworth low-pass filter into body acceleration and gravity. The gravitational force is assumed to have only low frequency components, therefore a filter with 0.3 Hz cutoff frequency was used. From each window, a vector of features was obtained by calculating variables from the time and frequency domain.\n\nCheck the README.txt file for further details about this dataset. \n\nA video of the experiment including an example of the 6 recorded activities with one of the participants can be seen in the following link: [Web Link]\n\n\nAttribute Information:\n\nFor each record in the dataset it is provided: \n- Triaxial acceleration from the accelerometer (total acceleration) and the estimated body acceleration. \n- Triaxial Angular velocity from the gyroscope. \n- A 561-feature vector with time and frequency domain variables. \n- Its activity label. \n- An identifier of the subject who carried out the experiment.\n\n\nRelevant Papers:\n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. International Workshop of Ambient Assisted Living (IWAAL 2012). Vitoria-Gasteiz, Spain. Dec 2012 \n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra, Jorge L. Reyes-Ortiz. Energy Efficient Smartphone-Based Activity Recognition using Fixed-Point Arithmetic. Journal of Universal Computer Science. Special Issue in Ambient Assisted Living: Home Care. Volume 19, Issue 9. May 2013\n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. 4th International Workshop of Ambient Assited Living, IWAAL 2012, Vitoria-Gasteiz, Spain, December 3-5, 2012. Proceedings. Lecture Notes in Computer Science 2012, pp 216-223. \n\nJorge Luis Reyes-Ortiz, Alessandro Ghio, Xavier Parra-Llanas, Davide Anguita, Joan Cabestany, Andreu Català. Human Activity and Motion Disorder Recognition: Towards Smarter Interactive Cognitive Environments. 21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013. Bruges, Belgium 24-26 April 2013.\n\n', 'ARFF', NULL, NULL, NULL, '2015-05-22 20:38:11', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1589271/php88ZB4Q', 'true', 71, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-22 20:38:11'),
-(72, 64, NULL, 'hill-valley', '1', NULL, '**Author**: Lee Graham, Franz Oppacher\n**Source**: UCI\n\n**Please cite**:   \n\n1. Source:\n\nLee Graham (lee \'@\' stellaralchemy.com) \n\nFranz Oppacher (oppacher \'@\' scs.carleton.ca) \nCarleton University, Department of Computer Science \nIntelligent Systems Research Unit \n1125 Colonel By Drive, Ottawa, Ontario, Canada, K1S5B6 \n\n\n2. Attribute Information:\n\n1-100: Labeled “X##”. Floating point values (numeric) \n101: Labeled “class”. Binary {0, 1} representing {valley, hill} ', 'ARFF', NULL, NULL, NULL, '2015-05-22 21:11:58', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1590101/php3isjYz', 'true', 72, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-22 21:11:58'),
-(73, 64, NULL, 'ilpd', '1', NULL, '**Author**: Bendi Venkata Ramana, M. Surendra Prasad Babu, N. B. Venkateswarlu \n  \n**Source**: UCI  \n\n**Please cite**:   \n\nSource:\n\n1. Bendi Venkata Ramana,  \nramana.bendi \'@\' gmail.com \nAssociate Professor, \nDepartment of Information Technology, \nAditya Instutute of Technology and Management, \nTekkali - 532201, Andhra Pradesh, India. \n\n2. Prof. M. Surendra Prasad Babu \ndrmsprasadbabu \'@\' yahoo.co.in \nDeptartment of Computer Science & Systems Engineering, \nAndhra University College of Engineering, \nVisakhapatnam-530 003 Andhra Pradesh, India. \n\n3.Prof. N. B. Venkateswarlu \nvenkat_ritch \'@\' yahoo.com \nDepartment of Computer Science and Engineering, \nAditya Instutute of Technology and Management, \nTekkali - 532201, Andhra Pradesh, India.\n\n\nData Set Information:\n\nThis data set contains 416 liver patient records and 167 non liver patient records.The data set was collected from north east of Andhra Pradesh, India. Selector is a class label used to divide into groups(liver patient or not). This data set contains 441 male patient records and 142 female patient records. \n\nAny patient whose age exceeded 89 is listed as being of age \"90\".\n\nAttribute Information:\n\n1. Age Age of the patient \n2. Gender Gender of the patient \n3. TB Total Bilirubin \n4. DB Direct Bilirubin \n5. Alkphos Alkaline Phosphotase \n6. Sgpt Alamine Aminotransferase \n7. Sgot Aspartate Aminotransferase \n8. TP Total Protiens \n9. ALB Albumin \n10. A/G Ratio Albumin and Globulin Ratio \n11. Selector field used to split the data into two sets (labeled by the experts) \n\n\nRelevant Papers:\n\n1. Bendi Venkata Ramana, Prof. M. S. Prasad Babu and Prof. N. B. Venkateswarlu, “A Critical Comparative Study of Liver Patients from USA and INDIA: An Exploratory Analysis”, International Journal of Computer Science Issues, ISSN :1694-0784, May 2012. \n2. Bendi Venkata Ramana, Prof. M. S. Prasad Babu and Prof. N. B. Venkateswarlu, “A Critical Study of Selected Classification Algorithms for Liver Disease Diagnosis”, International Journal of Database Management Systems (IJDMS), Vol.3, No.2, ISSN : 0975-5705, PP 101-114, May 2011.', 'ARFF', NULL, NULL, NULL, '2015-05-22 22:40:56', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1590565/phpOJxGL9', 'true', 73, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:32:35'),
-(74, 64, NULL, 'madelon', '1', NULL, '**Author**: Isabelle Guyon  \n**Source**: UCI  \n**Please cite**: Isabelle Guyon, Steve R. Gunn, Asa Ben-Hur, Gideon Dror, 2004. Result analysis of the NIPS 2003 feature selection challenge.\n\nAbstract: \n\nMADELON is an artificial dataset, which was part of the NIPS 2003 feature selection challenge. This is a two-class classification problem with continuous input variables. The difficulty is that the problem is multivariate and highly non-linear.\n\nSource:\n\nIsabelle Guyon \nClopinet \n955 Creston Road \nBerkeley, CA 90708 \nisabelle \'@\' clopinet.com \n\n\nData Set Information:\n\nMADELON is an artificial dataset containing data points grouped in 32 clusters placed on the vertices of a five dimensional hypercube and randomly labeled +1 or -1. The five dimensions constitute 5 informative features. 15 linear combinations of those features were added to form a set of 20 (redundant) informative features. Based on those 20 features one must separate the examples into the 2 classes (corresponding to the +-1 labels). We added a number of distractor feature called \'probes\' having no predictive power. The order of the features and patterns were randomized. \n\nThis dataset is one of five datasets used in the NIPS 2003 feature selection challenge. Our website is still open for post-challenge submissions.  All details about the preparation of the data are found in our technical report: Design of experiments for the NIPS 2003 variable selection benchmark, Isabelle Guyon, July 2003. Such information was made available only after the end of the challenge. \n\nThe data are split into training, validation, and test set. Target values are provided only for the 2 first sets. Test set performance results are obtained by submitting prediction results to: [Web Link]. \n\nThe data are in the following format: \ndataname.param: Parameters and statistics about the data \ndataname.feat: Identities of the features (in the order the features are found in the data). \ndataname_train.data: Training set (a space-delimited regular matrix, patterns in lines, features in columns). \ndataname_valid.data: Validation set. \ndataname_test.data: Test set. \ndataname_train.labels: Labels (truth values of the classes) for training examples. \ndataname_valid.labels: Validation set labels (withheld during the benchmark, but provided now). \ndataname_test.labels: Test set labels (withheld, so the data can still be use as a benchmark). \n\n\nAttribute Information:\nWe do not provide attribute information, to avoid biasing the feature selection process.\n\n\nRelevant Papers:\n\nThe best challenge entrants wrote papers collected in the book: \nIsabelle Guyon, Steve Gunn, Masoud Nikravesh, Lofti Zadeh (Eds.), Feature Extraction, Foundations and Applications. Studies in Fuzziness and Soft Computing. Physica-Verlag, Springer.\n\nSee also: \nIsabelle Guyon, et al, 2007. Competitive baseline methods set new standards for the NIPS 2003 feature selection benchmark. Pattern Recognition Letters 28 (2007) 1438–1444. \n\nand the associated technical report: \nIsabelle Guyon, et al. 2006. Feature selection with the CLOP package. Technical Report.\n', 'ARFF', NULL, NULL, NULL, '2015-05-22 23:46:18', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1590986/phpfLuQE4', 'true', 74, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-22 23:46:18'),
-(75, 64, NULL, 'nomao', '1', NULL, '**Author**:   \n**Source**: UCI\n**Please cite**:   \n\nRelevant Papers: Laurent Candillier and Vincent Lemaire. Design and Analysis of the Nomao Challenge - Active Learning in the Real-World. In: Proceedings of the ALRA : Active Learning in Real-world Applications, Workshop ECML-PKDD 2012, Friday, September 28, 2012, Bristol, UK.\n\n1. Data set title:\nNomao Data Set \n\n\n2. Abstract: \nNomao collects data about places (name, phone, localization...) from many sources. Deduplication consists in detecting what data refer to the same place. Instances in the dataset compare 2 spots.\n\n3. Data Set Characteristics:  \n\n- Univariate\n- Area: Computer\n- Attribute Characteristics: Real\n- Associated Tasks: Classification\n- Missing Values?: Yes\n\n\n4. Source:\n\n(a) Original owner of database (name / phone / snail address / email address) \nNomao / 00 33 5 62 48 33 90 / 1 avenue Jean Rieux, 31500 Toulouse / challenge \'@\' nomao.com \n(b) Donor of database (name / phone / snail address / email address) \nLaurent Candillier / - / 1 avenue Jean Rieux, 31500 Toulouse / laurent \'@\' nomao.com\n\n\n5. Data Set Information:\n\nThe dataset has been enriched during the Nomao Challenge: organized along with the ALRA workshop (Active Learning in Real-world Applications): held at the ECML-PKDD 2012 conference.', 'ARFF', NULL, NULL, NULL, '2015-05-25 19:09:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592278/phpDYCOet', 'true', 75, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-05-25 19:09:04'),
-(76, 64, NULL, 'ozone-level-8hr', '1', NULL, '**Author**: Kun Zhang, Wei Fan, XiaoJing Yuan\n\n**Source**: UCI\n\n**Please cite**:   \n\n\n1 . Abstract: \nTwo ground ozone level data sets are included in this collection. One is the eight hour peak set (eighthr.data), the other is the one hour peak set (onehr.data). Those data were collected from 1998 to 2004 at the Houston, Galveston and Brazoria area.\n\n2. Source:\n\nKun Zhang, zhang.kun05 \'@\' gmail.com, Department of Computer Science, Xavier University of Lousiana \nWei Fan, wei.fan \'@\' gmail.com, IBM T.J.Watson Research \nXiaoJing Yuan, xyuan \'@\' uh.edu, Engineering Technology Department, College of Technology, University of Houston \n\n\n3. Data Set Information:\n\nAll the attribute start with T means the temperature measured at different time throughout the day; and those starts with WS indicate the wind speed at various time. \n\nWSR_PK: continuous. peek wind speed -- resultant (meaning average of wind vector) \nWSR_AV: continuous. average wind speed \nT_PK: continuous. Peak T \nT_AV: continuous. Average T \nT85: continuous. T at 850 hpa level (or about 1500 m height) \nRH85: continuous. Relative Humidity at 850 hpa \nU85: continuous. (U wind - east-west direction wind at 850 hpa) \nV85: continuous. V wind - N-S direction wind at 850 \nHT85: continuous. Geopotential height at 850 hpa, it is about the same as height at low altitude \nT70: continuous. T at 700 hpa level (roughly 3100 m height) \n\nRH70: continuous. \nU70: continuous. \nV70: continuous. \nHT70: continuous. \n\nT50: continuous. T at 500 hpa level (roughly at 5500 m height) \n\nRH50: continuous. \nU50: continuous. \nV50: continuous. \nHT50: continuous. \n\nKI: continuous. K-Index [Web Link] \nTT: continuous. T-Totals [Web Link] \nSLP: continuous. Sea level pressure \nSLP_: continuous. SLP change from previous day \n\nPrecp: continuous. -- precipitation\n\n\n4. Attribute Information:\n\nThe following are specifications for several most important attributes that are highly valued by Texas Commission on Environmental Quality (TCEQ). More details can be found in the two relevant papers. \n\nO 3 - Local ozone peak prediction \nUpwind - Upwind ozone background level \nEmFactor - Precursor emissions related factor \nTmax - Maximum temperature in degrees F \nTb - Base temperature where net ozone production begins (50 F) \nSRd - Solar radiation total for the day \nWSa - Wind speed near sunrise (using 09-12 UTC forecast mode) \nWSp - Wind speed mid-day (using 15-21 UTC forecast mode) \n\n\n5. Relevant Papers:\n\nForecasting skewed biased stochastic ozone days: analyses, solutions and beyond, Knowledge and Information Systems, Vol. 14, No. 3, 2008. \n\nIt Discusses details about the dataset, its use as well as various experiments (both cross-validation and streaming) using many state-of-the-art methods. \nA shorter version of the paper (does not contain some detailed experiments as the journal paper above) is in: \nForecasting Skewed Biased Stochastic Ozone Days: Analyses and Solutions. ICDM 2006: 753-764 \n\n', 'ARFF', NULL, NULL, NULL, '2015-05-25 19:22:29', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592279/phpdReP6S', 'true', 76, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:25:43'),
-(77, 64, NULL, 'phoneme', '1', NULL, '**Author**:   \n**Source**: KEEL\n**Please cite**:   \n\n* Title:\n\nPhoneme dataset\n\n* Abstract:\n\nThe aim of this dataset is to distinguish between nasal (class 0) and oral sounds (class 1). The class distribution is 3,818 samples in class 0 and 1,586 samples in class 1. The phonemes are transcribed as follows: sh as in she, dcl as in dark, iy as the vowel in she, aa as the vowel in dark, and ao as the first vowel in water.\n\n* Attributes information:\n\n@relation phoneme\n@attribute Aa real [-1.7, 4.107]\n@attribute Ao real [-1.327, 4.378]\n@attribute Dcl real [-1.823, 3.199]\n@attribute Iy real [-1.581, 2.826]\n@attribute Sh real [-1.284, 2.719]\n@attribute Class {0, 1}\n@inputs Aa, Ao, Dcl, Iy, Sh\n@outputs Class', 'ARFF', NULL, NULL, NULL, '2015-05-25 19:34:17', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592281/php8Mz7BG', 'true', 77, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:25:20'),
-(78, 64, NULL, 'one-hundred-plants-margin', '1', NULL, '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \n\n**Source**: UCI \n\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \n\n1. One-hundred plant species leaves data set (class = margin).\n \n2. Sources:\n   (a) Original owners of colour Leaves Samples:\n\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\n The colour images are not included in this submission.\n The Leaves were collected in the Royal Botanic Gardens, Kew, UK.\n email: james.cope@kingston.ac.uk\n   \n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\n\n   (c) Date received 03/12/2012\n\n3. Past Usage:\n\n   (a) This is a new data set, provisional paper: \n\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\n\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \n\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\n Plant texture classification using gabor cooccurrences.\n Advances in Visual Computing,\n pages 669ñ677, 2010.\n\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\n Shape and texture based plant leaf classification. In\n Advanced Concepts for Intelligent Vision Systems,\n pages 345ñ353. Springer, 2010.\n\n4. Relevant Information Paragraph:\n The data directory contains the binary images (masks) of the leaf samples. The colour images are not included.\n The data set features are organised as the following:\n * \'data_Sha_64.txt\'\n  * \'data_Tex_64.txt\'\n * [THIS DATASET]\'data_Mar_64.txt\'\n\n One file for each 64-element feature vectors. Each row begins with the class label.\n The remaining 64 elements is the feature vector.\n\n5. Number of Instances\n\n 1600 samples each of three features (16 samples per leaf class).\n\n6. Number of Attributes\n\n    Three 64 element feature vectors per sample.\n\n7. Vectors\n There are three features: Shape, Margin and Texture. As discussed in the paper(s) above.\n For Each feature, a 64 element vector is given per sample of leaf.\n These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin).\n\n8. Missing Attribute Values: none\n\n9. Class Distribution: 16 instances per class', 'ARFF', NULL, NULL, NULL, '2015-05-25 20:51:03', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592283/phpCsX3fx', 'true', 78, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:21:22'),
-(79, 64, NULL, 'one-hundred-plants-shape', '1', NULL, '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \n\n**Source**: UCI \n\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \n\n1. One-hundred plant species leaves data set (class = shape).\n \n2. Sources:\n   (a) Original owners of colour Leaves Samples:\n\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\n The colour images are not included in this submission.\n The Leaves were collected in the Royal Botanic Gardens, Kew, UK.\n email: james.cope@kingston.ac.uk\n   \n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\n\n   (c) Date received 03/12/2012\n\n3. Past Usage:\n\n   (a) This is a new data set, provisional paper: \n\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\n\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \n\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\n Plant texture classification using gabor cooccurrences.\n Advances in Visual Computing,\n pages 669ñ677, 2010.\n\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\n Shape and texture based plant leaf classification. In\n Advanced Concepts for Intelligent Vision Systems,\n pages 345ñ353. Springer, 2010.\n\n4. Relevant Information Paragraph:\n The data directory contains the binary images (masks) of the leaf samples. The colour images are not included.\n The data set features are organised as the following:\n * [THIS DATASET]\'data_Sha_64.txt\'\n  * \'data_Tex_64.txt\'\n * \'data_Mar_64.txt\'\n\n One file for each 64-element feature vectors. Each row begins with the class label.\n The remaining 64 elements is the feature vector.\n\n5. Number of Instances\n\n 1600 samples each of three features (16 samples per leaf class).\n\n6. Number of Attributes\n\n    Three 64 element feature vectors per sample.\n\n7. Vectors\n There are three features: Shape, Margin and Texture. As discussed in the paper(s) above.\n For Each feature, a 64 element vector is given per sample of leaf.\n These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin).\n\n8. Missing Attribute Values: none\n\n9. Class Distribution: 16 instances per class', 'ARFF', NULL, NULL, NULL, '2015-05-25 20:58:37', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592284/php0FyS2T', 'true', 79, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:21:11'),
-(80, 64, NULL, 'one-hundred-plants-texture', '1', NULL, '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/One-hundred+plant+species+leaves+data+set) - 2012  \r\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \r\n\r\n**One-hundred plant species leaves data set (class = texture)**  \r\nThe data directory contains the binary images (masks) of the leaf samples. The colour images are not included. There are three features: Shape, Margin and Texture. As discussed in the paper(s) above. For Each feature, a 64 element vector is given per sample of leaf. These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin). The colour images are not included in this submission. The Leaves were collected in the Royal Botanic Gardens, Kew, UK. email: james.cope@kingston.ac.uk\r\n\r\n**Sources**  \r\n   (a) Original owners of colour Leaves Samples:\r\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\r\n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\r\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\r\n   (c) Date received 03/12/2012\r\n\r\n**Usage**\r\n   (a) This is a new data set, provisional paper: \r\n\r\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\r\n\r\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \r\n\r\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\r\n Plant texture classification using gabor cooccurrences.\r\n Advances in Visual Computing,\r\n pages 669ñ677, 2010.\r\n\r\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\r\n Shape and texture based plant leaf classification. In\r\n Advanced Concepts for Intelligent Vision Systems,\r\n pages 345ñ353. Springer, 2010.', 'ARFF', NULL, NULL, NULL, '2015-05-25 20:59:55', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592285/phpoOxxNn', 'true', 80, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:20:59'),
-(81, 64, NULL, 'qsar-biodeg', '1', NULL, '**Author**:   \n**Source**: UCI  \n**Please cite**: Mansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878 \n\n\nQSAR biodegradation Data Set \n\n* Abstract: \n\nData set containing values for 41 attributes (molecular descriptors) used to classify 1055 chemicals into 2 classes (ready and not ready biodegradable).\n\n\n* Source:\n\nKamel Mansouri, Tine Ringsted, Davide Ballabio (davide.ballabio \'@\' unimib.it), Roberto Todeschini, Viviana Consonni, Milano Chemometrics and QSAR Research Group (http://michem.disat.unimib.it/chm/), Università degli Studi Milano – Bicocca, Milano (Italy)\n\n\n* Data Set Information:\n\nThe QSAR biodegradation dataset was built in the Milano Chemometrics and QSAR Research Group (Università degli Studi Milano – Bicocca, Milano, Italy). The research leading to these results has received funding from the European Community’s Seventh Framework Programme [FP7/2007-2013] under Grant Agreement n. 238701 of Marie Curie ITN Environmental Chemoinformatics (ECO) project. \nThe data have been used to develop QSAR (Quantitative Structure Activity Relationships) models for the study of the relationships between chemical structure and biodegradation of molecules. Biodegradation experimental values of 1055 chemicals were collected from the webpage of the National Institute of Technology and Evaluation of Japan (NITE). Classification models were developed in order to discriminate ready (356) and not ready (699) biodegradable molecules by means of three different modelling methods: k Nearest Neighbours, Partial Least Squares Discriminant Analysis and Support Vector Machines. Details on attributes (molecular descriptors) selected in each model can be found in the quoted reference: Mansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878.\n\n\n* Attribute Information:\n\n41 molecular descriptors and 1 experimental class: \n1) SpMax_L: Leading eigenvalue from Laplace matrix \n2) J_Dz(e): Balaban-like index from Barysz matrix weighted by Sanderson electronegativity \n3) nHM: Number of heavy atoms \n4) F01[N-N]: Frequency of N-N at topological distance 1 \n5) F04[C-N]: Frequency of C-N at topological distance 4 \n6) NssssC: Number of atoms of type ssssC \n7) nCb-: Number of substituted benzene C(sp2) \n8) C%: Percentage of C atoms \n9) nCp: Number of terminal primary C(sp3) \n10) nO: Number of oxygen atoms \n11) F03[C-N]: Frequency of C-N at topological distance 3 \n12) SdssC: Sum of dssC E-states \n13) HyWi_B(m): Hyper-Wiener-like index (log function) from Burden matrix weighted by mass \n14) LOC: Lopping centric index \n15) SM6_L: Spectral moment of order 6 from Laplace matrix \n16) F03[C-O]: Frequency of C - O at topological distance 3 \n17) Me: Mean atomic Sanderson electronegativity (scaled on Carbon atom) \n18) Mi: Mean first ionization potential (scaled on Carbon atom) \n19) nN-N: Number of N hydrazines \n20) nArNO2: Number of nitro groups (aromatic) \n21) nCRX3: Number of CRX3 \n22) SpPosA_B(p): Normalized spectral positive sum from Burden matrix weighted by polarizability \n23) nCIR: Number of circuits \n24) B01[C-Br]: Presence/absence of C - Br at topological distance 1 \n25) B03[C-Cl]: Presence/absence of C - Cl at topological distance 3 \n26) N-073: Ar2NH / Ar3N / Ar2N-Al / R..N..R \n27) SpMax_A: Leading eigenvalue from adjacency matrix (Lovasz-Pelikan index) \n28) Psi_i_1d: Intrinsic state pseudoconnectivity index - type 1d \n29) B04[C-Br]: Presence/absence of C - Br at topological distance 4 \n30) SdO: Sum of dO E-states \n31) TI2_L: Second Mohar index from Laplace matrix \n32) nCrt: Number of ring tertiary C(sp3) \n33) C-026: R--CX--R \n34) F02[C-N]: Frequency of C - N at topological distance 2 \n35) nHDon: Number of donor atoms for H-bonds (N and O) \n36) SpMax_B(m): Leading eigenvalue from Burden matrix weighted by mass \n37) Psi_i_A: Intrinsic state pseudoconnectivity index - type S average \n38) nN: Number of Nitrogen atoms \n39) SM6_B(m): Spectral moment of order 6 from Burden matrix weighted by mass \n40) nArCOOR: Number of esters (aromatic) \n41) nX: Number of halogen atoms \n42) experimental class: ready biodegradable (RB) and not ready biodegradable (NRB)\n\n\n* Relevant Papers:\n\nMansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878\n', 'ARFF', NULL, NULL, NULL, '2015-05-25 21:14:53', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592286/phpGUrE90', 'true', 81, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:20:47'),
-(82, 64, NULL, 'wall-robot-navigation', '1', NULL, '**Author**: Ananda Freire, Marcus Veloso and Guilherme Barreto     \n**Source**: UCI   \n**Please cite**:   \n\n* Dataset Title: \n\nWall-Following Robot Navigation Data Data Set \n\n* Abstract:  \n\nThe data were collected as the SCITOS G5 robot navigates through the room following the wall in a clockwise direction, for 4 rounds, using 24 ultrasound sensors arranged circularly around its \'waist\'.\n\n* Source:\n\n(a) Creators: Ananda Freire, Marcus Veloso and Guilherme Barreto \nDepartment of Teleinformatics Engineering \nFederal University of Ceará \nFortaleza, Ceará, Brazil \n\n(b) Donors of database: Ananda Freire (anandalf \'@\' gmail.com) \nGuilherme Barreto (guilherme \'@\' deti.ufc.br)\n\n* Data Set Information:\n\nThe provided file contain the raw values of the measurements of all 24 ultrasound sensors and the corresponding class label. Sensor readings are sampled at a rate of 9 samples per second. \n\nIt is worth mentioning that the 24 ultrasound readings and the simplified distances were collected at the same time step, so each file has the same number of rows (one for each sampling time step). \n\nThe wall-following task and data gathering were designed to test the hypothesis that this apparently simple navigation task is indeed a non-linearly separable classification task. Thus, linear classifiers, such as the Perceptron network, are not able to learn the task and command the robot around the room without collisions. Nonlinear neural classifiers, such as the MLP network, are able to learn the task and command the robot successfully without collisions. \n\nIf some kind of short-term memory mechanism is provided to the neural classifiers, their performances are improved in general. For example, if past inputs are provided together with current sensor readings, even the Perceptron becomes able to learn the task and command the robot successfully. If a recurrent neural network, such as the Elman network, is used to learn the task, the resulting dynamical classifier is able to learn the task using less hidden neurons than the MLP network. \n\n* Attribute Information:\n\nNumber of Attributes: sensor_readings_24.data: 24 numeric attributes and the class. \n\nFor Each Attribute: \n-- File sensor_readings_24.data: \n1. US1: ultrasound sensor at the front of the robot (reference angle: 180°) - (numeric: real)  \n2. US2: ultrasound reading (reference angle: -165°) - (numeric: real)   \n3. US3: ultrasound reading (reference angle: -150°) - (numeric: real)   \n4. US4: ultrasound reading (reference angle: -135°) - (numeric: real)     \n5. US5: ultrasound reading (reference angle: -120°) - (numeric: real)   \n6. US6: ultrasound reading (reference angle: -105°) - (numeric: real)   \n7. US7: ultrasound reading (reference angle: -90°) - (numeric: real)   \n8. US8: ultrasound reading (reference angle: -75°) - (numeric: real)   \n9. US9: ultrasound reading (reference angle: -60°) - (numeric: real)   \n10. US10: ultrasound reading (reference angle: -45°) - (numeric: real)   \n11. US11: ultrasound reading (reference angle: -30°) - (numeric: real)   \n12. US12: ultrasound reading (reference angle: -15°) - (numeric: real)   \n13. US13: reading of ultrasound sensor situated at the back of the robot (reference angle: 0°) - (numeric: real)   \n14. US14: ultrasound reading (reference angle: 15°) - (numeric: real)   \n15. US15: ultrasound reading (reference angle: 30°) - (numeric: real)   \n16. US16: ultrasound reading (reference angle: 45°) - (numeric: real)   \n17. US17: ultrasound reading (reference angle: 60°) - (numeric: real)   \n18. US18: ultrasound reading (reference angle: 75°) - (numeric: real)   \n19. US19: ultrasound reading (reference angle: 90°) - (numeric: real)   \n20. US20: ultrasound reading (reference angle: 105°) - (numeric: real)   \n21. US21: ultrasound reading (reference angle: 120°) - (numeric: real)   \n22. US22: ultrasound reading (reference angle: 135°) - (numeric: real)   \n23. US23: ultrasound reading (reference angle: 150°) - (numeric: real)   \n24. US24: ultrasound reading (reference angle: 165°) - (numeric: real)   \n25. Class: {Move-Forward, Slight-Right-Turn, Sharp-Right-Turn, Slight-Left-Turn}   \n\n\n* Relevant Papers:\n\nAnanda L. Freire, Guilherme A. Barreto, Marcus Veloso and Antonio T. Varela (2009), \n\'Short-Term Memory Mechanisms in Neural Network Learning of Robot Navigation \nTasks: A Case Study\'. Proceedings of the 6th Latin American Robotics Symposium (LARS\'2009), \nValparaíso-Chile, pages 1-6, DOI: 10.1109/LARS.2009.5418323\n', 'ARFF', NULL, NULL, NULL, '2015-05-25 21:50:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592289/phpVeNa5j', 'true', 82, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:19:18'),
-(83, 64, NULL, 'semeion', '1', NULL, '**Author**: Semeion Research Center of Sciences of Communication     \n**Source**: UCI     \n**Please cite**: Semeion Research Center of Sciences of Communication, via Sersale 117, 00128 Rome, Italy \nTattile Via Gaetano Donizetti, 1-3-5,25030 Mairano (Brescia), Italy.    \n\n* Title:   \nSemeion Handwritten Digit Data Set \n\n* Abstract:   \n1593 handwritten digits from around 80 persons were scanned, stretched in a rectangular box 16x16 in a gray scale of 256 values.\n\n\n* Source:\n\nThe dataset was created by Tactile Srl, Brescia, Italy (http://www.tattile.it) and donated in 1994 to Semeion Research Center of Sciences of Communication, Rome, Italy (http://www.semeion.it), for machine learning research. \n\nFor any questions, e-mail Massimo Buscema (m.buscema \'@\' semeion.it) or Stefano Terzi (s.terzi \'@\' semeion.it)\n\n\n* Data Set Information:\n\n1593 handwritten digits from around 80 persons were scanned, stretched in a rectangular box 16x16 in a gray scale of 256 values.Then each pixel of each image was scaled into a bolean (1/0) value using a fixed threshold. \n\nEach person wrote on a paper all the digits from 0 to 9, twice. The commitment was to write the digit the first time in the normal way (trying to write each digit accurately) and the second time in a fast way (with no accuracy). \n\nThe best validation protocol for this dataset seems to be a 5x2CV, 50% Tune (Train +Test) and completly blind 50% Validation\n\n\n* Attribute Information:\n\nThis dataset consists of 1593 records (rows) and 256 attributes (columns). Each record represents a handwritten digit, orginally scanned with a resolution of 256 grays scale (28). Each pixel of the each original scanned image was first stretched, and after scaled between 0 and 1 (setting to 0 every pixel whose value was under tha value 127 of the grey scale (127 included) and setting to 1 each pixel whose orinal value in the grey scale was over 127). Finally, each binary image was scaled again into a 16x16 square box (the final 256 binary attributes).\n\n\n* Relevant Papers:\n\nM Buscema, MetaNet: The Theory of Independent Judges, in Substance Use & Misuse 33(2)1998, pp 439-461.', 'ARFF', NULL, NULL, NULL, '2015-05-25 22:22:34', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592293/phptd5jYj', 'true', 83, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:18:33'),
-(84, 64, NULL, 'steel-plates-fault', '1', NULL, '**Author**: Semeion, Research Center of Sciences of Communication, Rome, Italy.     \n**Source**: UCI     \n**Please cite**: Dataset provided by Semeion, Research Center of Sciences of Communication, Via Sersale 117, 00128, Rome, Italy. \n(www.semeion.it)   \n\n\n* Title: Steel Plates Faults Data Set \n\n* Abstract: \n\nA dataset of steel plates\' faults, classified into 7 different types. The goal was to train machine learning for automatic pattern recognition.\n\n* Source:\n\nSemeion, Research Center of Sciences of Communication, Via Sersale 117, 00128, Rome, Italy. \nwww.semeion.it\n\n\n* Data Set Information:\n\nType of dependent variables (7 Types of Steel Plates Faults): \n1.Pastry \n2.Z_Scratch \n3.K_Scatch \n4.Stains \n5.Dirtiness \n6.Bumps \n7.Other_Faults \n\n\nAttribute Information:\n\n27 independent variables: \nX_Minimum \nX_Maximum \nY_Minimum \nY_Maximum \nPixels_Areas \nX_Perimeter \nY_Perimeter \nSum_of_Luminosity \nMinimum_of_Luminosity \nMaximum_of_Luminosity \nLength_of_Conveyer \nTypeOfSteel_A300 \nTypeOfSteel_A400 \nSteel_Plate_Thickness \nEdges_Index \nEmpty_Index \nSquare_Index \nOutside_X_Index \nEdges_X_Index \nEdges_Y_Index \nOutside_Global_Index \nLogOfAreas \nLog_X_Index \nLog_Y_Index \nOrientation_Index \nLuminosity_Index \nSigmoidOfAreas \n\n\n* Relevant Papers:\n\n1.M Buscema, S Terzi, W Tastle, A New Meta-Classifier,in NAFIPS 2010, Toronto (CANADA),26-28 July 2010, 978-1-4244-7858-6/10 ©2010 IEEE \n2.M Buscema, MetaNet: The Theory of Independent Judges, in Substance Use & Misuse, 33(2), 439-461,1998\n\n', 'ARFF', NULL, NULL, NULL, '2015-05-25 22:42:40', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592296/php9xWOpn', 'true', 84, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:17:37'),
-(85, 64, NULL, 'tamilnadu-electricity', '1', NULL, '**Author**: K.Kalyani.    \n**Source**: UCI    \n**Please cite**:   \n\n* Title: Tamilnadu Electricity Board Hourly Readings Data Set \n\n* Abstract: \n\nThis data can be effectively produced the result to fewer parameter of the Load profile can be reduced in the Database\n\n* Source:\n\nK.Kalyani ,kkalyanims \'@\' gmail.com,T.U.K Arts College,Karanthai,Thanjavur.\n\n\n* Data Set Information:\n\nCollect the real time readings for residential,commercial,industrial,agriculure,to find the accuracy consumption in Tamil Nadu Around Thanajvur.\n\n\n* Attribute Information:\n\nforkva,forkw,type,sector,service\n\n\n* Relevant Papers:\n\nEfficient Electricity Utilization By IHBMO\n', 'ARFF', NULL, NULL, NULL, '2015-05-25 22:52:37', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592297/phpvFYBg2', 'true', 85, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:17:26'),
-(86, 64, NULL, 'wdbc', '1', NULL, '**Author**: William H. Wolberg, W. Nick Street, Olvi L. Mangasarian    \n**Source**: UCI   \n**Please cite**:   \n\n* Title: \n\nBreast Cancer Wisconsin (Diagnostic) Data Set (WDBC)\n\n* Abstract: \n\nDiagnostic Wisconsin Breast Cancer Database\n\n* Source:\n\nCreators: \n\n1. Dr. William H. Wolberg, General Surgery Dept. \nUniversity of Wisconsin, Clinical Sciences Center \nMadison, WI 53792 \nwolberg \'@\' eagle.surgery.wisc.edu \n\n2. W. Nick Street, Computer Sciences Dept. \nUniversity of Wisconsin, 1210 West Dayton St., Madison, WI 53706 \nstreet \'@\' cs.wisc.edu 608-262-6619 \n\n3. Olvi L. Mangasarian, Computer Sciences Dept. \nUniversity of Wisconsin, 1210 West Dayton St., Madison, WI 53706 \nolvi \'@\' cs.wisc.edu \n\nDonor: Nick Street\n\n* Data Set Information:\n\nFeatures are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. They describe characteristics of the cell nuclei present in the image. A few of the images can be found at [Web Link] \n\nSeparating plane described above was obtained using Multisurface Method-Tree (MSM-T) [K. P. Bennett, \"Decision Tree Construction Via Linear Programming.\" Proceedings of the 4th Midwest Artificial Intelligence and Cognitive Science Society, pp. 97-101, 1992], a classification method which uses linear programming to construct a decision tree. Relevant features were selected using an exhaustive search in the space of 1-4 features and 1-3 separating planes. \n\nThe actual linear program used to obtain the separating plane in the 3-dimensional space is that described in: [K. P. Bennett and O. L. Mangasarian: \"Robust Linear Programming Discrimination of Two Linearly Inseparable Sets\", Optimization Methods and Software 1, 1992, 23-34]. \n\nThis database is also available through the UW CS ftp server: \nftp ftp.cs.wisc.edu \ncd math-prog/cpo-dataset/machine-learn/WDBC/\n\n\n* Attribute Information:\n\n1) ID number \n2) Diagnosis (M = malignant, B = benign) \n3-32) \n\nTen real-valued features are computed for each cell nucleus: \n\na) radius (mean of distances from center to points on the perimeter) \nb) texture (standard deviation of gray-scale values) \nc) perimeter \nd) area \ne) smoothness (local variation in radius lengths) \nf) compactness (perimeter^2 / area - 1.0) \ng) concavity (severity of concave portions of the contour) \nh) concave points (number of concave portions of the contour) \ni) symmetry \nj) fractal dimension (\"coastline approximation\" - 1)\n\n\n* Relevant Papers:\n\nW.N. Street, W.H. Wolberg and O.L. Mangasarian. Nuclear feature extraction for breast tumor diagnosis. IS&T/SPIE 1993 International Symposium on Electronic Imaging: Science and Technology, volume 1905, pages 861-870, San Jose, CA, 1993. \n[Web Link] \n\nO.L. Mangasarian, W.N. Street and W.H. Wolberg. Breast cancer diagnosis and prognosis via linear programming. Operations Research, 43(4), pages 570-577, July-August 1995. \n[Web Link] \n\nW.H. Wolberg, W.N. Street, and O.L. Mangasarian. Machine learning techniques to diagnose breast cancer from fine-needle aspirates. Cancer Letters 77 (1994) 163-171. \n\nW.H. Wolberg, W.N. Street, and O.L. Mangasarian. Image analysis and machine learning applied to breast cancer diagnosis and prognosis. Analytical and Quantitative Cytology and Histology, Vol. 17 No. 2, pages 77-87, April 1995. \n\nW.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. Computerized breast cancer diagnosis and prognosis from fine needle aspirates. Archives of Surgery 1995;130:511-516. \n\nW.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. Computer-derived nuclear features distinguish malignant from benign breast cytology. Human Pathology, 26:792--796, 1995. \n\n', 'ARFF', NULL, NULL, NULL, '2015-05-26 16:24:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1592318/phpAmSP4g', 'true', 86, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-09 20:15:56'),
-(87, 64, NULL, 'micro-mass', '2', '2', '**Author**: Pierre Mahé, Jean-Baptiste Veyrieras  \n**Source**: [original](http://www.openml.org/d/1514) - UCI   \n**Please cite**:   \n\n* Dataset Title: MicroMass - Pure (pure spectra version)   \n\n* Abstract:   \nA dataset to explore machine learning approaches for the identification of microorganisms from mass-spectrometry data.  \n\n* Source:\n\nPierre Mahé, pierre.mahe \'@\' biomerieux.com, bioMérieux\nJean-Baptiste Veyrieras, jean-baptiste.veyrieras \'@\' biomerieux.com, bioMérieux\n\n* Data Set Information:\n\nThis MALDI-TOF dataset consists in:\nA) A reference panel of 20 Gram positive and negative bacterial species covering 9 genera among which several species are known to be hard to discriminate by mass spectrometry (MALDI-TOF). Each species was represented by 11 to 60 mass spectra obtained from 7 to 20 bacterial strains, constituting altogether a dataset of 571 spectra obtained from 213 strains. The spectra were obtained according to the standard culture-based workflow used in clinical routine in which the microorganism was first grown on an agar plate for 24 to 48 hours, before a portion of colony was picked, spotted on a MALDI slide and a mass spectrum was acquired. \nB) Based on this reference panel, a dedicated in vitro mock-up mixture dataset was constituted. For that purpose we considered 10 pairs of species of various taxonomic proximity:\n* 4 mixtures, labelled A, B, C and D, involved species that belong to the same genus,  \n* 2 mixtures, labelled E and F, involved species that belong to distinct genera, but to the same Gram type,  \n* 4 mixtures, labelled G, H, I and J, involved species that belong to distinct Gram types.  \nEach mixture was represented by 2 pairs of strains, which were mixed according to the following 9 concentration ratios : 1:0, 10:1, 5:1, 2:1, 1:1, 1:2, 1:5, 1:10, 0:1. Two replicate spectra were acquired for each concentration ratio and each couple of strains, leading altogether to a dataset of 360 spectra, among which 80 are actually pure sample spectra.\n\n* Relevant Papers:\n\nMahé et al. (2014). Automatic identification of mixed bacterial species fingerprints in a MALDI-TOF mass-spectrum. Bioinformatics.\n\nVervier et al., A benchmark of support vector machines strategies for microbial identification by mass-spectrometry data, submitted\n\n\n\n\n\n', 'ARFF', '\"Pierre Mahé\",\"Jean-Baptiste Veyrieras\"', NULL, NULL, '2015-06-01 16:57:25', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1593707/phpHyLSNF', 'true', 87, 'Class', NULL, NULL, NULL, 'public', NULL, 'http://www.openml.org/d/1514', NULL, '2015-06-01 16:57:25'),
-(88, 64, NULL, 'wilt', '1', NULL, '**Author**: Brian Johnson     \n**Source**: [UCI] (https://archive.ics.uci.edu/ml/datasets/Wilt)  \n**Please cite**: Johnson, B., Tateishi, R., Hoan, N., 2013. A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees. International Journal of Remote Sensing, 34 (20), 6969-6982.   \n\n\n* Dataset:  \nWilt Data Set \n\n* Abstract:    \nHigh-resolution Remote Sensing data set (Quickbird). Small number of training samples of diseased trees, large number for other land cover. Testing data set from stratified random sample of image.\n\n* Source:\n  \nBrian Johnson; \nInstitute for Global Environmental Strategies; \n2108-11 Kamiyamaguchi, Hayama, Kanagawa,240-0115 Japan; \nEmail: Johnson \'@\' iges.or.jp \n\n\n* Data Set Information:  \n\nThis data set contains some training and testing data from a remote sensing study by Johnson et al. (2013) that involved detecting diseased trees in Quickbird imagery. There are few training samples for the \'diseased trees\' class (74) and many for \'other land cover\' class (4265). \n\nThe data set consists of image segments, generated by segmenting the pansharpened image. The segments contain spectral information from the Quickbird multispectral image bands and texture information from the panchromatic (Pan) image band. The testing data set is for the row with “Segmentation scale 15” segments and “original multi-spectral image” Spectral information in Table 2 of the reference (i.e. row 5). Please see the reference below for more information on the data set, and please cite the reference if you use this data set. Enjoy! \n\n* Attribute Information:\n\nclass: \'w\' (diseased trees), \'n\' (all other land cover)   \nGLCM_Pan: GLCM mean texture (Pan band)   \nMean_G: Mean green value   \nMean_R: Mean red value   \nMean_NIR: Mean NIR value   \nSD_Pan: Standard deviation (Pan band)   \n\n\n* Relevant Papers:\n\nJohnson, B., Tateishi, R., Hoan, N., 2013. A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees. International Journal of Remote Sensing, 34 (20), 6969-6982.\n', 'ARFF', NULL, NULL, NULL, '2015-06-01 23:34:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1675983/phpnThNfi', 'true', 88, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-10-08 14:43:09'),
-(89, 2, NULL, 'adult', '2', '2', '**Author**: Ronny Kohavi and Barry Becker  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Adult) - 1996-05-01  \n**Please cite**: Ron Kohavi, \"Scaling Up the Accuracy of Naive-Bayes Classifiers: a Decision-Tree Hybrid\", Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, 1996  \n\n**Note: this is the original version from the UCI repository, with training and test sets merged.**\n\nPrediction task is to determine whether a person makes over 50K a year. Extraction was done by Barry Becker from the 1994 Census database. A set of reasonably clean records was extracted using the following conditions: ((AAGE>16) && (AGI>100) && (AFNLWGT>1)&& (HRSWK>0))\n\nRonny Kohavi and Barry Becker. Data Mining and Visualization, Silicon Graphics.  \ne-mail: ronnyk \'@\' live.com for questions. \n', 'ARFF', NULL, NULL, NULL, '2015-06-09 16:39:06', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1595261/phpMawTba', 'true', 89, 'class', NULL, NULL, NULL, 'public', NULL, NULL, 'added target attribute', '2015-06-09 16:56:26'),
-(90, 2, NULL, 'Bioresponse', '1', NULL, 'Data from the Kaggle Bioresponse challenge:\nhttps://www.kaggle.com/c/bioresponse\n\nThe objective of the competition is to help us build as good a model as possible so that we can, as optimally as this data allows, relate molecular information, to an actual biological response.\n\nWe have shared the data in the comma separated values (CSV) format. Each row in this data set represents a molecule. The first column contains experimental data describing an actual biological response; the molecule was seen to elicit this response (1), or not (0). The remaining columns represent molecular descriptors (d1 through d1776), these are calculated properties that can capture some of the characteristics of the molecule - for example size, shape, or elemental constitution. The descriptor matrix has been normalized.\n\nThe original training and test set were merged.', 'ARFF', NULL, NULL, NULL, '2015-11-04 23:59:56', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1681097/phpSSK7iA', 'true', 90, 'target', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-04 23:59:56'),
-(91, 2, NULL, 'Amazon_employee_access', '1', NULL, 'Data from the Kaggle Amazon Employee Access Challenge:\nhttps://www.kaggle.com/c/amazon-employee-access-challenge\n\nWhen an employee at any company starts work, they first need to obtain the computer access necessary to fulfill their role. This access may allow an employee to read/manipulate resources through various applications or web portals. It is assumed that employees fulfilling the functions of a given role will access the same or similar resources. It is often the case that employees figure out the access they need as they encounter roadblocks during their daily work (e.g. not able to log into a reporting portal). A knowledgeable supervisor then takes time to manually grant the needed access in order to overcome access obstacles. As employees move throughout a company, this access discovery/recovery cycle wastes a nontrivial amount of time and money.\n\nThere is a considerable amount of data regarding an employee&rsquo;s role within an organization and the resources to which they have access. Given the data related to current employees and their provisioned access, models can be built that automatically determine access privileges as employees enter and leave roles within a company. These auto-access models seek to minimize the human involvement required to grant or revoke employee access.\n\nThe original training and test set were merged.', 'ARFF', NULL, NULL, NULL, '2015-11-05 00:06:06', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1681098/phpmPOD5A', 'true', 91, 'target', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2015-11-05 00:06:06'),
-(92, 874, NULL, 'PhishingWebsites', '1', NULL, '**Author**: Rami Mustafa A Mohammad ( University of Huddersfield\",\"rami.mohammad \'@\' hud.ac.uk\",\"rami.mustafa.a \'@\' gmail.com) Lee McCluskey (University of Huddersfield\",\"t.l.mccluskey \'@\' hud.ac.uk )  Fadi Thabtah (Canadian University of Dubai\",\"fadi \'@\' cud.ac.ae)  \n**Source**: UCI \n**Please cite**: Please refer to the Machine Learning Repository\'s citation policy  \n\nSource:\n\nRami Mustafa A Mohammad ( University of Huddersfield, rami.mohammad \'@\' hud.ac.uk, rami.mustafa.a \'@\' gmail.com)\nLee McCluskey (University of Huddersfield,t.l.mccluskey \'@\' hud.ac.uk )\n\nFadi Thabtah (Canadian University of Dubai,fadi \'@\' cud.ac.ae)\n\n\nData Set Information:\n\nOne of the challenges faced by our research was the unavailability of reliable training datasets. In fact this challenge faces any researcher in the field. However, although plenty of articles about predicting phishing websites have been disseminated these days, no reliable training dataset has been published publically, may be because there is no agreement in literature on the definitive features that characterize phishing webpages, hence it is difficult to shape a dataset that covers all possible features. \nIn this dataset, we shed light on the important features that have proved to be sound and effective in predicting phishing websites. In addition, we propose some new features.\n\n\nAttribute Information:\n\nFor Further information about the features see the features file in the data folder.\n\n\nRelevant Papers:\n\nMohammad, Rami, McCluskey, T.L. and Thabtah, Fadi (2012) An Assessment of Features Related to Phishing Websites using an Automated Technique. In: International Conferece For Internet Technology And Secured Transactions. ICITST 2012 . IEEE, London, UK, pp. 492-497. ISBN 978-1-4673-5325-0\n\nMohammad, Rami, Thabtah, Fadi Abdeljaber and McCluskey, T.L. (2014) Predicting phishing websites based on self-structuring neural network. Neural Computing and Applications, 25 (2). pp. 443-458. ISSN 0941-0643\n\nMohammad, Rami, McCluskey, T.L. and Thabtah, Fadi Abdeljaber (2014) Intelligent Rule based Phishing Websites Classification. IET Information Security, 8 (3). pp. 153-160. ISSN 1751-8709\n\n \n\nCitation Request:\n\nPlease refer to the Machine Learning Repository\'s citation policy', 'ARFF', '\"Rami Mustafa A Mohammad ( University of Huddersfield\",\"rami.mohammad \'@\' hud.ac.uk\",\"rami.mustafa.a \'@\' gmail.com) Lee McCluskey (University of Huddersfield\",\"t.l.mccluskey \'@\' hud.ac.uk )  Fadi Thabtah (Canadian University of Dubai\",\"fadi \'@\' cud.ac.ae)\"', NULL, NULL, '2016-02-16 15:30:33', NULL, 'Public', 'Please refer to the Machine Learning Repository\'s citation policy', NULL, 'https://www.openml.org/data/download/1798106/phpV5QYya', 'true', 92, 'Result', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-02-16 15:30:33'),
-(93, 874, NULL, 'GesturePhaseSegmentationProcessed', '1', NULL, '**Author**: enata Cristina Barros Madeo (Madeo\",\"R. C. B.)  Priscilla Koch Wagner (Wagner\",\"P. K.)  Sarajane Marques Peres (Peres\",\"S. M.)  {renata.si\",\"priscilla.wagner\",\"sarajane} at usp.br  http://each.uspnet.usp.br/sarajane/  \n**Source**: UCI \n**Please cite**: Please refer to the Machine Learning Repository\'s citation policy. \nAdditionally, the authors require a citation to one or more publications from those cited as relevant papers.  \n\nSource:\n\nCreators: \nRenata Cristina Barros Madeo (Madeo, R. C. B.) \nPriscilla Koch Wagner (Wagner, P. K.) \nSarajane Marques Peres (Peres, S. M.) \n{renata.si, priscilla.wagner, sarajane} at usp.br \nhttp://each.uspnet.usp.br/sarajane/ \n\nDonor: \nUniversity of Sao Paulo - Brazil\n\n\nData Set Information:\n\nThe dataset is composed by features extracted from 7 videos with people gesticulating, aiming at studying Gesture Phase Segmentation. \nEach video is represented by two files: a raw file, which contains the position of hands, wrists, head and spine of the user in each frame; and a processed file, which contains velocity and acceleration of hands and wrists. See the data set description for more information on the dataset.\n\n\nAttribute Information:\n\nRaw files: 18 numeric attributes (double), a timestamp and a class attribute (nominal). \nProcessed files: 32 numeric attributes (double) and a class attribute (nominal). \nA feature vector with up to 50 numeric attributes can be generated with the two files mentioned above.\n\n\nRelevant Papers:\n\n1. Madeo, R. C. B. ; Lima, C. A. M. ; PERES, S. M. . Gesture Unit Segmentation using Support Vector Machines: Segmenting \nGestures from Rest Positions. In: Symposium on Applied Computing (SAC), 2013, Coimbra. Proceedings of the 28th Annual \nACM Symposium on Applied Computing (SAC), 2013. p. 46-52. \n* In this paper, the videos A1 and A2 were studied. \n\n2. Wagner, P. K. ; PERES, S. M. ; Madeo, R. C. B. ; Lima, C. A. M. ; Freitas, F. A. . Gesture Unit Segmentation Using \nSpatial-Temporal Information and Machine Learning. In: 27th Florida Artificial Intelligence Research Society Conference \n(FLAIRS), 2014, Pensacola Beach. Proceedings of the 27th Florida Artificial Intelligence Research Society Conference \n(FLAIRS). Palo Alto : The AAAI Press, 2014. p. 101-106. \n* In this paper, the videos A1, A2, A3, B1, B3, C1 and C3 were studied. \n\n3. Madeo, R. C. B.. Support Vector Machines and Gesture Analysis: incorporating temporal aspects (in Portuguese). Master \nThesis - Universidade de Sao Paulo, Sao Paulo Researcher Foundation. 2013. \n* In this document, the videos named B1 and B3 in the document correspond to videos C1 and C3 in this dataset. Only \nfive videos were explored in this document: A1, A2, A3, C1 and C3. \n\n4. Wagner, P. K. ; Madeo, R. C. B. ; PERES, S. M. ; Lima, C. A. M. . Segmenta&Atilde;&sect;ao de Unidades Gestuais com Multilayer \nPerceptrons (in Portuguese). In: Encontro Nacional de Inteligencia Artificial e Computacional (ENIAC), 2013, Fortaleza. \nAnais do X Encontro Nacional de Inteligencia Artificial e Computacional (ENIAC), 2013. \n* In this paper, the videos A1, A2 and A3 were studied.\n\n\n\nCitation Request:\n\nPlease refer to the Machine Learning Repository\'s citation policy. \nAdditionally, the authors require a citation to one or more publications from those cited as relevant papers.', 'ARFF', '\"enata Cristina Barros Madeo (Madeo\",\"R. C. B.)  Priscilla Koch Wagner (Wagner\",\"P. K.)  Sarajane Marques Peres (Peres\",\"S. M.)  {renata.si\",\"priscilla.wagner\",\"sarajane} at usp.br  http://each.uspnet.usp.br/sarajane/\"', NULL, NULL, '2016-02-17 11:42:33', NULL, 'Public', 'Please refer to the Machine Learning Repository\'s citation policy. \nAdditionally, the authors require a citation to one or more publications from those cited as relevant papers.', NULL, 'https://www.openml.org/data/download/1798765/phpYLeydd', 'true', 93, 'Phase', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-02-20 13:08:52'),
-(94, 874, NULL, 'MiceProtein', '1', NULL, '**Author**: lara Higuera Department of Software Engineering and Artificial Intelligence\",\"Faculty of Informatics and the Department of Biochemistry and Molecular Biology\",\"Faculty of Chemistry\",\"University Complutense\",\"Madrid\",\"Spain.  Email: clarahiguera \'@\' ucm.es   Katheleen J. Gardiner\",\"creator and owner of the protein expression data\",\"is currently with the Linda Crnic Institute for Down Syndrome\",\"Department of Pediatrics\",\"Department of Biochemistry and Molecular Genetics\",\"Human Medical Genetics and Genomics\",\"and Neuroscience Programs\",\"University of Colorado\",\"School of Medicine\",\"Aurora\",\"Colorado\",\"USA.  Email: katheleen.gardiner \'@\' ucdenver.edu   Krzysztof J. Cios is currently with the Department of Computer Science\",\"Virginia Commonwealth University\",\"Richmond\",\"Virginia\",\"USA\",\"and IITiS Polish Academy of Sciences\",\"Poland.  Email: kcios \'@\' vcu.edu  \n**Source**: UCI  \n**Please cite**: Higuera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126  \n\nAbstract: Expression levels of 77 proteins measured in the cerebral cortex of 8 classes of control and Down syndrome mice exposed to context fear conditioning, a task used to assess associative learning.\nSource:\n\nClara Higuera Department of Software Engineering and Artificial Intelligence, Faculty of Informatics and the Department of Biochemistry and Molecular Biology, Faculty of Chemistry, University Complutense, Madrid, Spain. \nEmail: clarahiguera \'@\' ucm.es \n\nKatheleen J. Gardiner, creator and owner of the protein expression data, is currently with the Linda Crnic Institute for Down Syndrome, Department of Pediatrics, Department of Biochemistry and Molecular Genetics, Human Medical Genetics and Genomics, and Neuroscience Programs, University of Colorado, School of Medicine, Aurora, Colorado, USA. \nEmail: katheleen.gardiner \'@\' ucdenver.edu \n\nKrzysztof J. Cios is currently with the Department of Computer Science, Virginia Commonwealth University, Richmond, Virginia, USA, and IITiS Polish Academy of Sciences, Poland. \nEmail: kcios \'@\' vcu.edu \n\n\nData Set Information:\n\nThe data set consists of the expression levels of 77 proteins/protein modifications that produced detectable signals in the nuclear fraction of cortex. There are 38 control mice and 34 trisomic mice (Down syndrome), for a total of 72 mice. In the experiments, 15 measurements were registered of each protein per sample/mouse. Therefore, for control mice, there are 38x15, or 570 measurements, and for trisomic mice, there are 34x15, or 510 measurements. The dataset contains a total of 1080 measurements per protein. Each measurement can be considered as an independent sample/mouse. \n\nThe eight classes of mice are described based on features such as genotype, behavior and treatment. According to genotype, mice can be control or trisomic. According to behavior, some mice have been stimulated to learn (context-shock) and others have not (shock-context) and in order to assess the effect of the drug memantine in recovering the ability to learn in trisomic mice, some mice have been injected with the drug and others have not. \n\nClasses: \nc-CS-s: control mice, stimulated to learn, injected with saline (9 mice) \nc-CS-m: control mice, stimulated to learn, injected with memantine (10 mice) \nc-SC-s: control mice, not stimulated to learn, injected with saline (9 mice) \nc-SC-m: control mice, not stimulated to learn, injected with memantine (10 mice) \n\nt-CS-s: trisomy mice, stimulated to learn, injected with saline (7 mice) \nt-CS-m: trisomy mice, stimulated to learn, injected with memantine (9 mice) \nt-SC-s: trisomy mice, not stimulated to learn, injected with saline (9 mice) \nt-SC-m: trisomy mice, not stimulated to learn, injected with memantine (9 mice) \n\nThe aim is to identify subsets of proteins that are discriminant between the classes. \n\n\nAttribute Information:\n\n1 Mouse ID \n2..78 Values of expression levels of 77 proteins; the names of proteins are followed by &acirc;&euro;&oelig;_n&acirc;&euro; indicating that they were measured in the nuclear fraction. For example: DYRK1A_n \n79 Genotype: control (c) or trisomy (t) \n80 Treatment type: memantine (m) or saline (s) \n81 Behavior: context-shock (CS) or shock-context (SC) \n82 Class: c-CS-s, c-CS-m, c-SC-s, c-SC-m, t-CS-s, t-CS-m, t-SC-s, t-SC-m \n\n\nRelevant Papers:\n\nThe posted data were analyzed by: \nHiguera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126 \n\nThe data are a subset of the data analyzed by: \nAhmed MM, Dhanasekaran AR, Block A, Tong S, Costa ACS, Stasko M, et al. (2015) Protein Dynamics Associated with Failed and Rescued Learning in the Ts65Dn Mouse Model of Down Syndrome. PLoS ONE 10(3): e0119491. [Web Link] \n\n\n\n\nCitation Request:\n\nHiguera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126', 'ARFF', '\"lara Higuera Department of Software Engineering and Artificial Intelligence\",\"Faculty of Informatics and the Department of Biochemistry and Molecular Biology\",\"Faculty of Chemistry\",\"University Complutense\",\"Madrid\",\"Spain.  Email: clarahiguera \'@\' ucm.es   Katheleen J. Gardiner\",\"creator and owner of the protein expression data\",\"is currently with the Linda Crnic Institute for Down Syndrome\",\"Department of Pediatrics\",\"Department of Biochemistry and Molecular Genetics\",\"Human Medical Genetics and Genomics\",\"and Neuroscience Programs\",\"University of Colorado\",\"School of Medicine\",\"Aurora\",\"Colorado\",\"USA.  Email: katheleen.gardiner \'@\' ucdenver.edu   Krzysztof J. Cios is currently with the Department of Computer Science\",\"Virginia Commonwealth University\",\"Richmond\",\"Virginia\",\"USA\",\"and IITiS Polish Academy of Sciences\",\"Poland.  Email: kcios \'@\' vcu.edu\"', NULL, NULL, '2016-02-17 14:32:49', NULL, 'Public', 'Higuera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126', NULL, 'https://www.openml.org/data/download/1804243/phpO6CpB2', 'true', 94, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-02-21 21:09:16'),
-(95, 2, NULL, 'cylinder-bands', '2', '2', '**Author**: Bob Evans, RR Donnelley & Sons Co.  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Cylinder+Bands) - August, 1995  \n**Please cite**:   \n\n**Cylinder bands**  \nProcess delays known as cylinder banding in rotogravure printing were substantially mitigated using control rules discovered by decision tree induction.\n \nAttribute Information:\n>  \n   1. timestamp: numeric;19500101 - 21001231  \n   2. cylinder number: nominal  \n   3. customer: nominal;  \n   4. job number: nominal;   \n   5. grain screened: nominal; yes, no  \n   6. ink color: nominal;  key, type  \n   7. proof on ctd ink:  nominal;  yes, no   \n   8. blade mfg: nominal;  benton, daetwyler, uddeholm  \n   9. cylinder division: nominal;  gallatin, warsaw, mattoon  \n  10. paper type: nominal;  uncoated, coated, super  \n  11. ink type: nominal;  uncoated, coated, cover  \n  12. direct steam: nominal; use; yes, no *  \n  13. solvent type: nominal;  xylol, lactol, naptha, line, other  \n  14. type on cylinder:  nominal;  yes, no   \n  15. press type: nominal; use; 70 wood hoe, 70 motter, 70 albert, 94 motter  \n  16. press: nominal;  821, 802, 813, 824, 815, 816, 827, 828  \n  17. unit number: nominal;  1, 2, 3, 4, 5, 6, 7, 8, 9, 10  \n  18. cylinder size: nominal;  catalog, spiegel, tabloid  \n  19. paper mill location: nominal; north us, south us, canadian, \n      scandanavian, mid european  \n  20. plating tank: nominal; 1910, 1911, other  \n  21. proof cut: numeric;  0-100  \n  22. viscosity: numeric;  0-100  \n  23. caliper: numeric;  0-1.0  \n  24. ink temperature: numeric;  5-30  \n  25. humifity: numeric;  5-120  \n  26. roughness: numeric;  0-2  \n  27. blade pressure: numeric;  10-75  \n  28. varnish pct: numeric;  0-100  \n  29. press speed: numeric;  0-4000  \n  30. ink pct: numeric;  0-100  \n  31. solvent pct: numeric;  0-100  \n  32. ESA Voltage: numeric;  0-16  \n  33. ESA Amperage: numeric;  0-10  \n  34. wax: numeric ;  0-4.0  \n  35. hardener:  numeric; 0-3.0  \n  36. roller durometer:  numeric;  15-120  \n  37. current density:  numeric;  20-50  \n  38. anode space ratio:  numeric;  70-130  \n  39. chrome content: numeric; 80-120  \n  40. band type: nominal; class; band, no band  \n\nNotes:  \n* cylinder number is an identifier and should be ignored when modelling the data', 'ARFF', NULL, NULL, NULL, '2016-03-30 22:09:17', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1854224/phpAz9Len', 'true', 95, 'band_type', NULL, '\"timestamp\",\"cylinder_number\"', NULL, 'public', NULL, NULL, NULL, '2016-03-30 22:09:17'),
-(96, 2, NULL, 'cjs', '3', NULL, '**Author**: Dr. Fernando Camacho  \n**Source**: Unknown - 1995  \n**Please cite**: Camacho, F. and Arron, G. (1995)  Effects of the regulators paclobutrazol and flurprimidol on the growth of terminal sprouts formed on trimmed silver maple trees. Canadian Journal of Statistics 3(23).\n\nData on tree growth used in the Case Study published in the September, 1995 issue of the Canadian Journal of Statistics. This data set was been provided by Dr. Fernando Camacho, Ontario Hydro Technologies, 800 Kipling Ave, Toronto Canada M3Z 5S4. It forms the basis of the Case Study in Data Analysis published in the Canadian Journal of Statistics, September 1995. It can be freely used for noncommercial purposes, as long as proper acknowledgement to the source and to the Canadian Journal of Statistics is made.\n\n\nDescription\n\n\nThe effects of the Growth Regulators Paclobutrazol (PP 333)\nand Flurprimidol (EL-500) on the Number and Length of Internodes\nin Terminal Sprouts Formed on Trimmed Silver Maple Trees.\n \nIntroduction:\n \nThe trimming of trees under distribution lines on city streets and\nin rural areas is a major problem and expense for electrical\nutilities.  Such operations are routinely performed at intervals of\none to eight years depending upon the individual species growth rate\nand the amount of clearance required.  Ontario Hydro trims about\n500,000 trees per year at a cost of about $25 per tree.\n \nMuch effort has been spent in developing chemicals for the horticultural\nindustry to retard the growth of woody and herbaceous plants.  Recently,\na group of new growth regulators was introduced which was shown to be\neffective in controlling the growth of trees without producing\nnoticeable injury symptoms.  In this group are PP 333 ( common name\npaclobutrazol) (2RS, 3RS - 1 -(4-chlorophenyl) - 4,4 - dimethyl - 2 -\n(1,2,4-triazol-l-yl) pentan - 3- ol and EL-500 (common name flurprimidol\nand composition alpha - (1-methylethyl) - alpha - [4-(trifluromethoxyl)\nphenyl] - 5- pyrimidine - methanol).  Both EL-500 and PP-333 have been\nreported to control excessive sprout growth in a number of species\nwhen applied as a foliar spray, as a soil drench, or by trunk injection.\nSprout length is a function of both the number of internodes and\nthe length of the individual internodes in the sprout.  While there\nhave been many reports that both PP 333 and EL-500 cause a reduction\nin the length of internodes formed in sprouts on woody plants treated\nwith the growth regulators, there has been but one report that EL-500\napplication to apple trees resulted in a reduction of the number\nof internodes formed per sprout.\n \nThe purpose of the present study was to investigate the length of the\nterminal sprouts, the length of the individual internodes in those\nsprouts, and the number of internodes in trimmed silver maple trees\nfollowing trunk injection with the growth regulators PP 333 and EL-500.\n \nExperimental Details.\n \nMultistemmed 12-year-old silver maple trees growing at Wesleyville,\nOntario were trunk injected with methanolic solutions of EL-500\nand PP-333 in May of 1985 using a third generation Asplundh\ntree injector.\n \nTwo different application rates (20 g/L and 4 g/L) were used for each\nchemical.  The volume of solution (and hence the amount of active\ningredient) injected into each tree was determined from the diameter\nof the tree, using the formula: vol(mL) = (dbh)*(dbh)*.492 where dbh\nis the diameter at breast height.  Two sets of control trees were\nincluded in the experiment.  In one set, tree received no injection\n(control) and in a second set, the trees were injected with\nmethanol, the carrier in the growth regulator solutions.  Ten trees,\nchosen at random, were used in each of the control and experimental\nsets.  Prior to injection, all the trees were trimmed by a forestry\ncrew, with their heights being reduced by about one third.\n \nIn January 1987, twenty months after the trees were injected, between\nsix and eight limbs were removed at random from the bottom two-thirds\nof the canopy of each of the ten trees in each experimental and control\nset.  The limbs were returned to the laboratory and the length of all\nthe terminal sprouts, the lengths of the individual internodes, and\nthe number of internodes recorded.  Between one and 25 terminal\nsprouts were found on each limb collected.  Sprouts which had a\nlength of 1 cm or less were recorded as being 1 cm in length.\nIn such spouts, the internode lengths were not measured, but were\ncalculated from the total length of the sprout and the number\nof internodes counted.  Internode lengths were then expressed to one\ndecimal place.  In two instances, one of the ten trees in a set\ncould not be sampled because limb removal would have jeopardized the\nhealth of the tree over the long-term.\n \nData set:\n \nEach of the records represents a terminal sprout and contains the\nfollowing information:\n   N   the sprout number\n   TR  treatment 1  control\n                  2  methanol control\n                  3  PP 333 20g/L\n                  4  PP 333  4g/L\n                  5  EL 500 20g/L\n                  6  EL 500  4g/L\n   TREE  tree id\n   BR    branch id\n   TL    total sprout length (cm)\n   IN    number of internodes on the sprout\n   INTER a list of the lengths of the internodes in the sprout,\n          starting from the base of the sprout (129 entries)\n \nSprouts 1868 to 1879 do not have branch identification data.', 'ARFF', NULL, NULL, NULL, '2016-04-11 18:32:46', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1910442/phpDAC5gS', 'true', 96, 'TR', 'N', NULL, NULL, 'public', NULL, NULL, NULL, '2016-04-11 18:32:46'),
-(97, 64, NULL, 'dresses-sales', '2', NULL, '**Author**: Muhammad Usman & Adeel Ahmed   \r\n**Source**: origin source at [UCI](https://archive.ics.uci.edu/ml/datasets/Dresses_Attribute_Sales)  \r\n**Please cite**:   \r\n\r\n####1. Summary\r\n\r\nThis dataset contain attributes of dresses and their recommendations according to their sales.Sales are monitor on the basis of alternate days. \r\n\r\nThe attributes present analyzed are: Style, Price, Rating, Size, Season, NeckLine, SleeveLength, waiseline, Material, FabricType, Decoration, Pattern, Type, Recommendation.\r\n\r\nContact:\r\n```\r\nMuhammad Usman & Adeel Ahmed, usman.madspot \'@\' gmail.com adeel.ahmed92 \'@\' gmail.com, Air University, Students at Air University.\r\n```\r\n\r\n####2: Attribute Information:\r\n\r\n```\r\nStyle: Bohemia,brief,casual,cute,fashion,flare,novelty,OL,party,sexy,vintage,work. \r\nPrice:Low,Average,Medium,High,Very-High \r\nRating:1-5 \r\nSize:S,M,L,XL,Free \r\nSeason:Autumn,winter,Spring,Summer \r\nNeckLine:O-neck,backless,board-neck,Bowneck,halter,mandarin-collor,open,peterpan-collor,ruffled,scoop,slash-neck,square-collar,sweetheart,turndowncollar,V-neck. \r\nSleeveLength:full,half,halfsleeves,butterfly,sleveless,short,threequarter,turndown,null \r\nwaiseline:dropped,empire,natural,princess,null. \r\nMaterial:wool,cotton,mix etc \r\nFabricType:shafoon,dobby,popline,satin,knitted,jersey,flannel,corduroy etc \r\nDecoration:applique,beading,bow,button,cascading,crystal,draped,embroridary,feathers,flowers etc \r\nPattern type: solid,animal,dot,leapard etc \r\nRecommendation:0,1 \r\n```\r\n', 'ARFF', NULL, NULL, NULL, '2016-04-11 19:57:27', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1910507/phpcFPMhq', 'true', 97, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-04-11 19:57:27'),
-(98, 64, NULL, 'LED-display-domain-7digit', '1', '1', '1. Title of Database: LED display domain\n\n2. Sources:\n   (a) Breiman,L., Friedman,J.H., Olshen,R.A., &amp; Stone,C.J. (1984). \n       Classification and Regression Trees.  Wadsworth International\n       Group: Belmont, California.  (see pages 43-49).\n   (b) Donor: David Aha \n   (c) Date: 11/10/1988\n\n3. Past Usage: (many)\n     1. CART book (above):\n        -- Optimal Bayes classification rate: 74%\n        -- CART decision tree algorithm: 71% (resubstitution estimate)\n        -- Nearest Neighbor Algorithm: 71%\n           -- 200 training and 5000 test instances\n     \n     2. Quinlan,J.R. (1987). Simplifying Decision Trees.  In International\n        Journal of Man-Machine Studies (to appear).\n        -- C4 decision tree algorithm: 72.6% (using pessimistic pruning)\n           -- 2000 training and 500 test instances\n     3. Tan,M. &amp; Eshelman,L. (1988). Using Weighted Networks to Represent\n        Classification Knowledge in Noisy Domains.  In Proceedings of the\n        5th International Conference on Machine Learning, 121-134, Ann\n        Arbor, Michigan: Morgan Kaufmann.  \n        -- IWN system: 73.3% (using the And-OR classification algorithm)\n           -- 400 training and 500 test cases\n\n4. Relevant Information Paragraph:\n     This simple domain contains 7 Boolean attributes and 10 concepts,\n     the set of decimal digits.  Recall that LED displays contain 7\n     light-emitting diodes -- hence the reason for 7 attributes.  The\n     problem would be easy if not for the introduction of noise.  In\n     this case, each attribute value has the 10% probability of having\n     its value inverted.  \n\n     It\'s valuable to know the optimal Bayes rate for these databases.\n     In this case, the misclassification rate is 26% (74% classification\n     accuracy).\n        \n5. Number of Instances: 500. But in the original URL you can find a C script and run it choosing the number of instances to be generated.\n\n6. Number of Attributes: 7 (all Boolean-valued)\n\n7. Attribute Information:\n   -- All attribute values are either 0 or 1, according to whether\n      the corresponding light is on or not for the decimal digit.\n   -- Each attribute (excluding the class attribute, which is an\n      integer ranging between 0 and 9 inclusive) has a 10% percent\n      chance of being inverted.\n\n8. Missing Attribute Values: None\n\n9. Class Distribution: 10% (Theoretical)\n   -- Each concept (digit) has the same theoretical probability\n      distribution.  The program randomly selects the attribute.', 'ARFF', NULL, NULL, NULL, '2016-07-29 20:36:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/4535757/phpSj3fWL', 'true', 98, 'Class', NULL, NULL, NULL, 'public', NULL, 'http://sci2s.ugr.es/keel/dataset.php?cod=63, https://archive.ics.uci.edu/ml/datasets/LED+Display+Domain', NULL, '2016-07-29 20:36:10'),
-(99, 64, NULL, 'texture', '1', '1', '**Author**: Laboratory of Image Processing and Pattern Recognition (INPG-LTIRF), Grenoble - France.  \r\n**Source**: [original](https://www.elen.ucl.ac.be/neural-nets/Research/Projects/ELENA/databases/REAL/texture/) - ELENA project   \r\n**Please cite**:   \r\n\r\n####1. Summary\r\n\r\nThis database was generated by the Laboratory of Image Processing and Pattern Recognition (INPG-LTIRF) in the development of the Esprit project ELENA No. 6891 and the Esprit working group ATHOS No. 6620.\r\n```\r\n (a) Original source:\r\n\r\n   P. Brodatz \"Textures: A Photographic Album for Artists and Designers\",\r\n   Dover Publications,Inc.,New York, 1966.\r\n\r\n (b) Creation: Laboratory of Image Processing and Pattern Recognition\r\n\r\n   Institut National Polytechnique de Grenoble INPG\r\n   Laboratoire de Traitement d\'Image et de Reconnaissance de Formes LTIRF\r\n   Av. Felix Viallet, 46\r\n   F-38031 Grenoble Cedex\r\n   France\r\n\r\n (c) Contact: Dr. A. Guerin-Dugue, INPG-LTIRF, guerin@tirf.inpg.fr\r\n```\r\n\r\n####2. Past Usage:\r\n\r\nThis database has a private usage at the TIRF laboratory. It has been created in order to study the textures discrimination with high order statistics.\r\n\r\n```\r\nA.Guerin-Dugue, C. Aviles-Cruz, \"High Order Statistics from Natural Textured Images\",\r\nIn ATHOS workshop on System Identification and High Order Statistics, Sophia-Antipolis, France, September 1993.\r\n\r\nGuerin-Dugue, A. and others, Deliverable R3-B4-P - Task B4: Benchmarks, Technical report,\r\nElena-NervesII \"Enhanced Learning for Evolutive Neural Architecture\", ESPRIT-Basic Research Project  Number 6891,\r\nJune 1995.\r\n```\r\n\r\n####3. Relevant Information:\r\n\r\nThe aim is to distinguish between 11 different textures (Grass lawn, Pressed calf leather, Handmade paper, Raffia looped to a high pile, Cotton canvas, ...), each pattern (pixel) being characterised by 40 attributes built by the estimation of fourth order modified moments in four orientations: 0, 45, 90 and 135 degrees.\r\n\r\nA statistical method based on the extraction of fourth order moments for the characterization of natural micro-textures was developed called \"fourth order modified moments\" (mm4) [Guerin93], this method measures the deviation from first-order Gauss-Markov process, for each texture. The features were estimated in four directions to take into account the possible orientations of the textures (0, 45, 90 and 135 degrees). Only correlation between the current pixel, the first neighbourhood and the second neighbourhood are taken into account. This small neighbourhood is adapted to the fine grain property of the textures.\r\n\r\nThe data set contains 11 classes of 500 instances and each class refers to a type of texture in the Brodatz album.\r\n\r\nThe database dimension is 40 plus one for the class label. The 40 attributes were build respectively by the estimation of the following fourth order modified moments in four orientations: 0, 45, 90 and 135 degrees: mm4(000), mm4(001), mm4(002), mm4(011), mm4(012), mm4(022), mm4(111), mm4(112), mm4(122) and mm4(222).\r\n\r\n!! Patterns are always sorted by class and are presented in the increasing order of their class label in each dataset relative to the texture database (texture.dat, texture_CR.dat, texture_PCA.dat, texture_DFA.dat)\r\n\r\n####4. Class:\r\n\r\nThe class label is a code for the following classes:\r\n```\r\n                Class         Class label\r\n  2   Grass lawn                      (D09)  \r\n  3   Pressed calf leather            (D24) \r\n  4   Handmade paper                  (D57) \r\n  6   Raffia looped to a high pile:   (D84) \r\n  7   Cotton canvas                   (D77) \r\n  8   Pigskin                         (D92) \r\n  9   Beach sand:                     (D28) \r\n  10  Beach sand                      (D29) \r\n  12  Oriental straw cloth            (D53) \r\n  13  Oriental straw cloth            (D78) \r\n  14  Oriental grass fiber cloth      (D79) \r\n```\r\n\r\n####5. Summary Statistics:\r\n\r\nTable here below provides for each attribute of the database the dynamic (Min and Max values), the mean value and the standard deviation.\r\n\r\n```\r\nAttribute  Min      Max       Mean     Standard    \r\n                                       deviation   \r\n\r\n    1   -1.4495    0.7741   -1.0983    0.2034\r\n    2   -1.2004    0.3297   -0.5867    0.2055\r\n    3   -1.3099    0.3441   -0.5838    0.3135\r\n    4   -1.1104    0.5878   -0.4046    0.2302\r\n    5   -1.0534    0.4387   -0.3307    0.2360\r\n    6   -1.0029    0.4515   -0.2422    0.2225\r\n    7   -1.2076    0.5246   -0.6026    0.2003\r\n    8   -1.0799    0.3980   -0.4322    0.2210\r\n    9   -1.0570    0.4369   -0.3317    0.2361\r\n   10   -1.2580    0.3546   -0.5978    0.3268\r\n   11   -1.4495    0.7741   -1.0983    0.2034\r\n   12   -1.0831    0.3715   -0.5929    0.2056\r\n   13   -1.1194    0.6347   -0.4019    0.3368\r\n   14   -1.0182    0.1573   -0.6270    0.1390\r\n   15   -0.9435    0.1642   -0.4482    0.1952\r\n   16   -0.9944    0.0357   -0.5763    0.1587\r\n   17   -1.1722    0.0201   -0.7331    0.1955\r\n   18   -1.0174    0.1155   -0.4919    0.2335\r\n   19   -1.0044    0.0833   -0.4727    0.2257\r\n   20   -1.1800    0.4392   -0.4831    0.3484\r\n   21   -1.4495    0.7741   -1.0983    0.2034\r\n   22   -1.2275    0.5963   -0.7363    0.2220\r\n   23   -1.3412    0.4464   -0.7771    0.3290\r\n   24   -1.1774    0.6882   -0.5770    0.2646\r\n   25   -1.1369    0.4098   -0.5085    0.2538\r\n   26   -1.1099    0.3725   -0.4038    0.2515\r\n   27   -1.2393    0.6120   -0.7279    0.2278\r\n   28   -1.1540    0.4221   -0.5863    0.2446\r\n   29   -1.1323    0.3916   -0.5090    0.2526\r\n   30   -1.4224    0.4718   -0.7708    0.3264\r\n   31   -1.4495    0.7741   -1.0983    0.2034\r\n   32   -1.1789    0.5647   -0.6463    0.1890\r\n   33   -1.1473    0.6755   -0.4919    0.3304\r\n   34   -1.1228    0.3132   -0.6435    0.1441\r\n   35   -1.0145    0.3396   -0.4918    0.1922\r\n   36   -1.0298    0.1560   -0.5934    0.1704\r\n   37   -1.2534    0.0899   -0.7795    0.1641\r\n   38   -1.0966    0.1944   -0.5541    0.2111\r\n   39   -1.0765    0.2019   -0.5230    0.2015\r\n   40   -1.2155    0.4647   -0.5677    0.3091\r\n```\r\n\r\nThe dynamic of the attributes is in [-1.45 - 0.775]. The database resulting from the centering and reduction by attribute of the Texture database is on the ftp server in the `REAL/texture/texture_CR.dat.Z\' file.\r\n\r\n####6. Confusion matrix.\r\n\r\nThe following confusion matrix of the k_NN classifier was obtained with a Leave_One_Out error counting method on the texture_CR.dat database. k was set to 1 in order to reach the minimum mean error rate : 1.0 +/- 0.8%.\r\n\r\n```\r\n Class    2      3      4      6      7      8      9      10     12     13     14 \r\n 2      97.0    1.0    0.4    0.0    0.0    0.0    1.6    0.0    0.0    0.0    0.0   \r\n 3       0.2   99.0    0.0    0.0    0.0    0.0    0.4    0.0    0.0    0.0    0.4   \r\n 4       1.0    0.0   98.8    0.0    0.0    0.0    0.2    0.0    0.0    0.0    0.0   \r\n 6       0.0    0.0    0.0   99.4    0.0    0.0    0.0    0.6    0.0    0.0    0.0   \r\n 7       0.0    0.0    0.0    0.0  100.0    0.0    0.0    0.0    0.0    0.0    0.0   \r\n 8       0.0    0.0    0.0    0.0    0.0   98.6    0.0    1.4    0.0    0.0    0.0   \r\n 9       0.4    0.0    0.2    0.0    0.0    0.2   98.8    0.4    0.0    0.0    0.0   \r\n 10      0.0    0.0    0.0    0.0    0.0    1.4    0.0   98.6    0.0    0.0    0.0   \r\n 12      0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  100.0    0.0    0.0   \r\n 13      0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   99.8    0.2   \r\n 14      0.0    0.4    0.0    0.0    0.0    0.4    0.0    0.0    0.2    0.0   99.0   \r\n```\r\n\r\n7. Result of the Principal Component Analysis:\r\n\r\nThe Principal Components Analysis is a very classical method in pattern recognition [Duda73]. PCA reduces the sample dimension in a linear way for the best representation in lower dimensions keeping the maximum of inertia. The best axe for the representation is however not necessary the best axe for the discrimination. After PCA, features are selected according to the percentage of initial inertia which is covered by the different axes and the number of features is determined according to the percentage of initial inertia to keep for the classification process.\r\n\r\nThis selection method has been applied on the texture_CR database. When quasi-linear correlations exists between some initial features, these redundant dimensions are removed by PCA and this preprocessing is then recommended. In this case, before a PCA, the determinant of the data covariance matrix is near zero; this database is thus badly conditioned for all process which use this information (the quadratic classifier for example).\r\n\r\nThe following file is available for the texture database: \'\'texture_PCA.dat.Z\'\', it is the projection of the \'\'texture_CR\'\' database on its principal components (sorted in a decreasing order of the related inertia percentage; so, if you desire to work on the database projected on its x first principal components you only have to keep the x first attributes of the texture_PCA.dat database and the class labels (last attribute)).\r\n\r\nTable here below provides the inertia percentages associated to the eigenvalues corresponding to the principal component axis sorted in the decreasing order of the associated inertia percentage. 99.85 percent of the total database inertia will remain if the 20 first principal components are kept.\r\n\r\n```\r\n       Eigen Value   Inertia      Cumulated\r\n         value      percentage      inertia\r\n\r\n  1   30.267500000 75.6687000000  75.6687000000 \r\n  2   3.6512500000  9.1281300000  84.7969000000 \r\n  3   2.2937000000  5.7342400000  90.5311000000 \r\n  4   1.7039700000  4.2599300000  94.7910000000 \r\n  5   0.6716540000  1.6791300000  96.4702000000 \r\n  6   0.5015290000  1.2538200000  97.7240000000 \r\n  7   0.1922830000  0.4807070000  98.2047000000 \r\n  8   0.1561070000  0.3902670000  98.5950000000 \r\n  9   0.1099570000  0.2748920000  98.8699000000 \r\n  10  0.0890891000  0.2227230000  99.0926000000 \r\n  11  0.0656016000  0.1640040000  99.2566000000 \r\n  12  0.0489988000  0.1224970000  99.3791000000 \r\n  13  0.0433819000  0.1084550000  99.4875000000 \r\n  14  0.0345022000  0.0862554000  99.5738000000 \r\n  15  0.0299203000  0.0748007000  99.6486000000 \r\n  16  0.0248857000  0.0622141000  99.7108000000 \r\n  17  0.0167901000  0.0419752000  99.7528000000 \r\n  18  0.0161633000  0.0404083000  99.7932000000 \r\n  19  0.0128898000  0.0322246000  99.8254000000 \r\n  20  0.0113884000  0.0284710000  99.8539000000 \r\n  21  0.0078481400  0.0196204000  99.8735000000 \r\n  22  0.0071527800  0.0178820000  99.8914000000 \r\n  23  0.0067661400  0.0169153000  99.9083000000 \r\n  24  0.0053149500  0.0132874000  99.9216000000 \r\n  25  0.0051102600  0.0127757000  99.9344000000 \r\n  26  0.0047116600  0.0117792000  99.9461000000 \r\n  27  0.0036193700  0.0090484300  99.9552000000 \r\n  28  0.0033222000  0.0083054900  99.9635000000 \r\n  29  0.0030722400  0.0076806100  99.9712000000 \r\n  30  0.0026373300  0.0065933300  99.9778000000 \r\n  31  0.0020996800  0.0052492000  99.9830000000 \r\n  32  0.0019376500  0.0048441200  99.9879000000 \r\n  33  0.0015642300  0.0039105700  99.9918000000 \r\n  34  0.0009679080  0.0024197700  99.9942000000 \r\n  35  0.0009578000  0.0023945000  99.9966000000 \r\n  36  0.0007379780  0.0018449400  99.9984000000 \r\n  37  0.0006280250  0.0015700600  100.000000000\r\n  38  0.0000000040  0.0000000099  100.000000000 \r\n  39  0.0000000001  0.0000000003  100.000000000 \r\n  40  0.0000000008  0.0000000019  100.000000000 \r\n\r\n```\r\n\r\nThis matrix can be found in the texture_EV.dat file.\r\n\r\nThe Discriminant Factorial Analysis (DFA) can be applied to a learning database where each learning sample belongs to a particular class [Duda73]. The number of discriminant features selected by DFA is fixed in function of the number of classes (c) and of the number of input dimensions (d); this number is equal to the minimum between d and c-1. In the usual case where d is greater than c, the output dimension is fixed equal to the number of classes minus one and the discriminant axes are selected in order to maximize the between-variance and to minimize the within-variance of the classes.\r\n\r\nThe discrimination power (ratio of the projected between-variance over the projected within-variance) is not the same for each discriminant axis: this ratio decreases for each axis. So for a problem with many classes, this preprocessing will not be always efficient as the last output features will not be so discriminant. This analysis uses the information of the inverse of the global covariance matrix, so the covariance matrix must be well conditioned (for example, a preliminary PCA must be applied to remove the linearly correlated dimensions).\r\n\r\nThe Discriminant Factorial Analysis (DFA) has been applied on the 18 first principal components of the texture_PCA database (thus by keeping only the 18 first attributes of these databases before to apply the DFA preprocessing) in order to build the texture_DFA.dat.Z database file, having 10 dimensions (the texture database having 11 classes). In the case of the texture database, experiments shown that a DFA preprocessing is very useful and most of the time improved the classifiers performances.\r\n\r\n[Duda73] Duda, R.O. and Hart, P.E.,Pattern Classification and Scene Analysis, John Wiley & Sons, 1973.\r\n', 'ARFF', NULL, NULL, NULL, '2016-07-29 21:03:14', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/4535764/phpBDgUyY', 'true', 99, 'Class', NULL, NULL, NULL, 'public', NULL, 'http://sci2s.ugr.es/keel/dataset.php?cod=72', NULL, '2016-07-29 21:03:14'),
-(100, 2, NULL, 'Australian', '3', '2', '**Author**: Confidential. Donated by Ross Quinlan  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Statlog+(Australian+Credit+Approval))/LibSVM - 2014-11-14  \r\n**Please cite**:   \r\n\r\nThis is the famous Australian dataset, retrieved 2014-11-14 from the libSVM site.\r\nIt was normalized.\r\n\r\nThe original version is from [UCI](https://archive.ics.uci.edu/ml/datasets/Statlog+(Australian+Credit+Approval)).\r\n\r\nThis file concerns credit card applications. All attribute names and values have been changed to meaningless symbols to protect confidentiality of the data. \r\n\r\nThis dataset is interesting because there is a good mix of attributes -- continuous, nominal with small numbers of values, and nominal with larger numbers of values. There are also a few missing values. \r\n\r\n\r\nSource: Statlog / Australian\r\n# of classes: 2\r\n# of data: 690\r\n# of features: 14', 'ARFF', NULL, NULL, NULL, '2016-09-21 15:43:02', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/4599631/phpjP8PU3', 'true', 100, 'Y', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2016-09-21 15:43:02'),
-(101, 2, NULL, 'UNIX_user_data', '1', NULL, '**Author**: Terran Lane (terran@ecn.purdue.edu)   \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/UNIX+User+Data) - Date unknown  \n**Please cite**:   \n\nThis file contains 9 sets of sanitized user data drawn from the command histories of 8 UNIX computer users at Purdue over the course of up to 2 years (USER0 and USER1 were generated by the same person, working on different platforms and different projects).  The data is drawn from tcsh(1) history files and has been parsed and sanitized to remove filenames, user names, directory structures, web addresses, host names, and other possibly identifying items.  Command names, flags, and shell metacharacters have been preserved.  Additionally, **SOF** and **EOF** tokens have been inserted at the start and end of\nshell sessions, respectively.  Sessions are concatenated by date order and tokens appear in the order issued within the shell session, but no timestamps are included in this data.  For example, the two sessions:\n\ncd ~/private/docs  \nls -laF | more  \ncat foo.txt bar.txt zorch.txt > somewhere  \nexit  \n\ncd ~/games/  \nxquake &  \nfg  \nvi scores.txt  \nmailx john_doe@somewhere.com  \nexit  \n\nwould be represented by the token stream  \n\n**SOF**  \ncd  \n\\<1\\>                  (one \"file name\" argument)  \nls  \n-laF  \n|  \nmore  \ncat  \n\\<3\\>                     (three \"file\" arguments)\n\\>  \n\\<1\\>  \nexit  \n**EOF**  \n**SOF**  \ncd  \n\\<1\\>  \nxquake  \n&  \nfg  \nvi  \n\\<1\\>  \nmailx  \n\\<1\\>  \nexit  \n**EOF**\n\nThis data is made available under conditions of anonymity for the contributing users and may be used for research purposes only. Summaries and research results employing this data may be published, but literal tokens or token sequences from the data may not be published except with express consent of the originators of the data. No portion of this data may be released with or included in a commercial product, nor may any portion of this data be sold or redistributed for profit or as part of of a profit-making endeavor.', 'ARFF', NULL, NULL, NULL, '2014-09-27 10:55:59', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52413/UNIX_user_data.arff', 'true', 101, 'user', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-09-27 10:55:59'),
-(102, 402, NULL, 'fourclass_scale', '1', NULL, '**Author**: Tin Kam Ho and Eugene M. Kleinberg.  \nlibSVM\",\"AAD group  \n**Source**: [original](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary.html) - Date unknown  \n**Please cite**:   \n\n#Dataset from the LIBSVM data repository.\n\nPreprocessing: transform to two-class', 'Sparse_ARFF', '\"Tin Kam Ho and Eugene M. Kleinberg.\"', '\"libSVM\",\"AAD group\"', NULL, '2015-06-02 11:17:44', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1594035/phpeLklTK', 'true', 102, 'class', NULL, NULL, NULL, 'public', NULL, 'http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary.html', NULL, '2015-06-02 11:17:44'),
 (103, 1, 0, 'kin8nm', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis is data set is concerned with the forward kinematics of an 8 link\n robot arm. Among the existing variants of this data set we have used\n the variant 8nm, which is known to be highly non-linear and medium\n noisy.\n\n Original source: DELVE repository of data. \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 8192 cases, 9 attributes (0 nominal, 9 continuous).', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3626/dataset_2175_kin8nm.arff', 'true', 103, 'y', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:04'),
 (104, 1, 0, 'mbagrade', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3627/dataset_2176_mbagrade.arff', 'true', 104, 'grade_point_average', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:07'),
 (105, 1, 0, 'wisconsin', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Wisconsin Prognostic Breast Cancer (WPBC)\n \n 2. Source Information\n \n a) Creators: \n \n 	Dr. William H. Wolberg, General Surgery Dept., University of\n 	Wisconsin,  Clinical Sciences Center, Madison, WI 53792\n 	wolberg@eagle.surgery.wisc.edu\n \n 	W. Nick Street, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	street@cs.wisc.edu  608-262-6619\n \n 	Olvi L. Mangasarian, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	olvi@cs.wisc.edu \n \n b) Donor: Nick Street\n \n c) Date: December 1995\n \n 3. Past Usage:\n \n 	Various versions of this data have been used in the following\n 	publications: \n \n 	(i) W. N. Street, O. L. Mangasarian, and W.H. Wolberg. \n 	An inductive learning approach to prognostic prediction. \n 	In A. Prieditis and S. Russell, editors, Proceedings of the\n 	Twelfth International Conference on Machine Learning, pages\n 	522--530, San Francisco, 1995. Morgan Kaufmann.\n \n 	(ii) O.L. Mangasarian, W.N. Street and W.H. Wolberg. \n 	Breast cancer diagnosis and prognosis via linear programming. \n 	Operations Research, 43(4), pages 570-577, July-August 1995. \n \n 	(iii) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computerized breast cancer diagnosis and prognosis from fine\n 	needle aspirates.  Archives of Surgery 1995;130:511-516. \n \n 	(iv) W.H. Wolberg, W.N. Street, and O.L. Mangasarian. \n 	Image analysis and machine learning applied to breast cancer\n 	diagnosis and prognosis. Analytical and Quantitative Cytology\n 	and Histology, Vol. 17 No. 2, pages 77-87, April 1995.\n \n 	(v) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computer-derived nuclear ``grade\'\' and breast cancer prognosis. \n 	Analytical and Quantitative Cytology and Histology, Vol. 17,\n 	pages 257-264, 1995. \n \n See also:\n 	http://www.cs.wisc.edu/~olvi/uwmp/mpml.html\n 	http://www.cs.wisc.edu/~olvi/uwmp/cancer.html\n \n Results:\n \n 	Two possible learning problems:\n \n 	1) Predicting field 2, outcome: R = recurrent, N = nonrecurrent\n 	- Dataset should first be filtered to reflect a particular\n 	endpoint; e.g., recurrences before 24 months = positive,\n 	nonrecurrence beyond 24 months = negative.\n 	- 86.3% accuracy estimated accuracy on 2-year recurrence using\n 	previous version of this data.  Learning method: MSM-T (see\n 	below) in the 4-dimensional space of Mean Texture, Worst Area,\n 	Worst Concavity, Worst Fractal Dimension.\n \n 	2) Predicting Time To Recur (field 3 in recurrent records)\n 	- Estimated mean error 13.9 months using Recurrence Surface\n 	Approximation. (See references (i) and (ii) above)\n \n 4. Relevant information\n \n 	Each record represents follow-up data for one breast cancer\n 	case.  These are consecutive patients seen by Dr. Wolberg\n 	since 1984, and include only those cases exhibiting invasive\n 	breast cancer and no evidence of distant metastases at the\n 	time of diagnosis. \n \n 	The first 30 features are computed from a digitized image of a\n 	fine needle aspirate (FNA) of a breast mass.  They describe\n 	characteristics of the cell nuclei present in the image.\n 	A few of the images can be found at\n 	http://www.cs.wisc.edu/~street/images/\n \n 	The separation described above was obtained using\n 	Multisurface Method-Tree (MSM-T) [K. P. Bennett, \"Decision Tree\n 	Construction Via Linear Programming.\" Proceedings of the 4th\n 	Midwest Artificial Intelligence and Cognitive Science Society,\n 	pp. 97-101, 1992], a classification method which uses linear\n 	programming to construct a decision tree.  Relevant features\n 	were selected using an exhaustive search in the space of 1-4\n 	features and 1-3 separating planes.\n \n 	The actual linear program used to obtain the separating plane\n 	in the 3-dimensional space is that described in:\n 	[K. P. Bennett and O. L. Mangasarian: \"Robust Linear\n 	Programming Discrimination of Two Linearly Inseparable Sets\",\n 	Optimization Methods and Software 1, 1992, 23-34].\n \n 	The Recurrence Surface Approximation (RSA) method is a linear\n 	programming model which predicts Time To Recur using both\n 	recurrent and nonrecurrent cases.  See references (i) and (ii)\n 	above for details of the RSA method. \n \n 	This database is also available through the UW CS ftp server:\n \n 	ftp ftp.cs.wisc.edu\n 	cd math-prog/cpo-dataset/machine-learn/WPBC/\n \n 5. Number of instances: 198\n \n 6. Number of attributes: 34 (ID, outcome, 32 real-valued input features)\n \n 7. Attribute information\n \n 1) ID number\n 2) Outcome (R = recur, N = nonrecur)\n 3) Time (recurrence time if field 2 = R, disease-free time if \n 	field 2	= N)\n 4-33) Ten real-valued features are computed for each cell nucleus:\n \n 	a) radius (mean of distances from center to points on the perimeter)\n 	b) texture (standard deviation of gray-scale values)\n 	c) perimeter\n 	d) area\n 	e) smoothness (local variation in radius lengths)\n 	f) compactness (perimeter^2 / area - 1.0)\n 	g) concavity (severity of concave portions of the contour)\n 	h) concave points (number of concave portions of the contour)\n 	i) symmetry \n 	j) fractal dimension (\"coastline approximation\" - 1)\n \n Several of the papers listed above contain detailed descriptions of\n how these features are computed. \n \n The mean, standard error, and \"worst\" or largest (mean of the three\n largest values) of these features were computed for each image,\n resulting in 30 features.  For instance, field 4 is Mean Radius, field\n 14 is Radius SE, field 24 is Worst Radius.\n \n Values for features 4-33 are recoded with four significant digits.\n \n 34) Tumor size - diameter of the excised tumor in centimeters\n 35) Lymph node status - number of positive axillary lymph nodes\n observed at time of surgery\n \n 8. Missing attribute values: \n 	Lymph node status is missing in 4 cases.\n \n 9. Class distribution: 151 nonrecur, 47 recur\n\n-----------------------------------------------------------------------------------------------------------\n Luis Torgo\'s version: (reconstructed)\n - removed the four instances with unknown values of the last attribute\n - exchanged the attribute position of attributes n.3 (Time) and n.35\n   (Lymph node).\n - removed the attribute outcome as it is the class attribute if the\n   problem is treated as a classification one\n-----------------------------------------------------------------------------------------------------------', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3628/dataset_2177_wisconsin.arff', 'true', 105, 'time', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:10'),
@@ -128,5 +83,5 @@ INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_
 (127, 1, 0, 'pharynx', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Case number deleted. \n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Name:  Pharynx (A clinical Trial in the Trt. of Carcinoma of the Oropharynx).\n SIZE:  195 observations, 13 variables.\n \n \n \n DESCRIPTIVE ABSTRACT:\n \n The .dat file gives the data for a part of a large clinical trial\n carried out by the Radiation Therapy Oncology Group in the United States. \n The full study included patients with squamous carcinoma of 15 sites in \n the mouth and throat, with 16 participating institutions, though only data \n on three sites in the oropharynx reported by the six largest institutions \n are considered here. Patients entering the study were randomly assigned to \n one of two treatment groups, radiation therapy alone or radiation therapy \n together with a chemotherapeutic agent.  One objective of the study was to \n compare the two treatment policies with respect to patient survival.\n \n \n \n SOURCE:  The Statistical Analysis of Failure Time Data, by JD Kalbfleisch\n          & RL Prentice, (1980),  Published by John Wiley & Sons \n \n \n \n VARIABLE DESCRIPTIONS:\n \n The data are in free format.  That is, at least one blank space separates \n each variable in the .dat file.  The variables are as follows:\n \n \n Case:         Case Number\n Inst:         Participating Institution\n sex:          1=male, 2=female\n Treatment:    1=standard, 2=test\n Grade:        1=well differentiated, 2=moderately differentiated, \n               3=poorly differentiated,  9=missing\n Age:          In years at time of diagnosis\n Condition:    1=no disability, 2=restricted work, 3=requires assistance with\n               self care, 4=bed confined,  9=missing\n Site:         1=faucial arch, 2=tonsillar fossa, 3=posterior pillar,\n               4=pharyngeal tongue, 5=posterior wall\n T staging:    1=primary tumor measuring 2 cm or less in largest diameter,\n               2=primary tumor measuring 2 cm to 4 cm in largest diameter with\n               minimal infiltration in depth, 3=primary tumor measuring more \n               than 4 cm, 4=massive invasive tumor\n N staging:    0=no clinical evidence of node metastases, 1=single positive\n               node 3 cm or less in diameter, not fixed, 2=single positive\n               node more than 3 cm in diameter, not fixed, 3=multiple\n               positive nodes or fixed positive nodes \n Entry Date:   Date of study entry: Day of year and year\n Status:       0=censored,  1=dead\n Time:         Survival time in days from day of diagnosis \n \n \n \n \n \n \n STORY BEHIND THE DATA:\n \n Approximately 30% of the survival times are censored owing primarily to \n patients surviving to the time of analysis. Some patients were lost\n to follow-up because the patient moved or transferred to an institution not\n participating in the study, though these cases were relatively rare. From \n a statistical point of view, an important feature of these data is the \n considerable lack of homogeneity between individuals being studied. \n Of course, as part of the study design, certain criteria for patient \n eligibility had to be met which eliminated extremes in the extent of disease, \n but still many factors are not controlled.\n \n This study included measurements of many covariates which would be expected \n to relate to survival experience. Six such variables are given in the data \n (sex, T staging, N staging, age, general condition, and grade).   The site \n of the primary tumor and possible differences between participating \n institutions require consideration as well.\n      \n The T,N staging classification gives a measure of the extent of the tumor at \n the primary site and at regional lymph nodes. T=1, refers to a small primary \n tumor, 2 centimeters or less in largest diameter, whereas T=4 is a massive \n tumor with extension to adjoining tissue. T=2 and T=3 refer to intermediate\n cases. N=0 refers to there being no clinical evidence of a lymph node \n metastasis and N=1, N=2, N=3 indicate, in increasing magnitude, the extent of \n existing lymph node involvement. Patients with classifications T=1,N=0; \n T=1,N=1;  T=2,N=0; or T=2,N=1, or with distant metastases were excluded \n from study.\n \n The variable general condition gives a measure of the functional capacity of \n the patient at the time of diagnosis (1 refers to no disability whereas\n 4 denotes bed confinement; 2 and 3 measure intermediate levels). The variable\n grade is a measure of the degree of differentiation of the tumor (the degree\n to which the tumor cell resembles the host cell) from 1 (well differentiated) \n to 3 (poorly differentiated)\n \n In addition to the primary question whether the combined treatment mode is\n preferable to the conventional radiation therapy, it is of considerable \n interest to determine the extent to which the several covariates relate to\n subsequent survival.  It is also imperative in answering the primary question \n to adjust the survivals for possible imbalance that may be present in the \n study with regard to the other covariates. Such problems are similar to those \n encountered in the classical theory of linear regression and the analysis of \n covariance.  Again, the need to accommodate censoring is an important \n distinguishing point. In many situations it is also important to develop \n nonparametric and robust procedures since there is frequently little empirical\n or theoretical work to support a particular family of failure time \n distributions.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:32', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3650/dataset_2199_pharynx.arff', 'true', 127, 'class', NULL, '\"Entry\"', NULL, 'public', NULL, NULL, 'set ignore feature', '2014-10-05 00:03:24'),
 (128, 1, 0, 'iris', '1', '1', '', 'ARFF', 'R.A. Fisher', NULL, '1936', '2014-04-06 23:23:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/61/dataset_61_iris.arff', 'true', 128, 'class', NULL, NULL, 'http://digital.library.adelaide.edu.au/dspace/handle/2440/15227', 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/Iris', NULL, '2014-04-06 23:23:39'),
 (129, 1, NULL, 'iris-challenge', '1', '1', '', 'ARFF', NULL, NULL, NULL, '2018-05-29 19:06:15', NULL, 'CC0', NULL, NULL, 'https://www.openml.org/data/download/19330175/iris-challenge.arff', 'true', 129, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2018-05-29 19:06:15'),
-(130, 16, 0, 'iris', '2', '1', '', 'ARFF', 'R.A. Fisher', NULL, '1936', '2014-04-06 23:23:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/61/dataset_61_iris.arff', 'true', 130, 'class', NULL, NULL, 'http://digital.library.adelaide.edu.au/dspace/handle/2440/15227', 'private', NULL, 'https://archive.ics.uci.edu/ml/datasets/Iris', NULL, '2014-04-06 23:23:39');
+(130, 1, 0, 'iris', '2', '1', '', 'ARFF', 'R.A. Fisher', NULL, '1936', '2014-04-06 23:23:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/61/dataset_61_iris.arff', 'true', 130, 'class', NULL, NULL, 'http://digital.library.adelaide.edu.au/dspace/handle/2440/15227', 'private', NULL, 'https://archive.ics.uci.edu/ml/datasets/Iris', NULL, '2014-04-06 23:23:39');
 
diff --git a/data/sql/implementation.sql b/data/sql/implementation.sql
index 5262d1bf8..9239f777f 100644
--- a/data/sql/implementation.sql
+++ b/data/sql/implementation.sql
@@ -1,17 +1,17 @@
 INSERT INTO `implementation` (`id`, `fullName`, `uploader`, `name`, `custom_name`, `class_name`, `version`, `external_version`, `creator`, `contributor`, `uploadDate`, `licence`, `language`, `description`, `fullDescription`, `installationNotes`, `dependencies`, `implements`, `binary_file_id`, `source_file_id`, `visibility`, `citation`) VALUES
-(1, 'weka.ZeroR(1)', 16, 'weka.ZeroR', NULL, 'weka.classifiers.rules.ZeroR', 1, 'Weka_3.9.0_12024', NULL, NULL, '2017-03-24 14:26:38', NULL, 'English', 'Weka implementation of ZeroR', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(2, 'weka.OneR(1)', 16, 'weka.OneR', NULL, 'weka.classifiers.rules.OneR', 1, 'Weka_3.9.0_10153', NULL, NULL, '2017-03-24 14:26:39', NULL, 'English', 'R.C. Holte (1993). Very simple classification rules perform well on most commonly used datasets. Machine Learning. 11:63-91.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(3, 'weka.JRip(1)', 16, 'weka.JRip', NULL, 'weka.classifiers.rules.JRip', 1, 'Weka_3.9.0_10153', NULL, NULL, '2017-03-24 14:26:40', NULL, 'English', 'William W. Cohen: Fast Effective Rule Induction. In: Twelfth International Conference on Machine Learning, 115-123, 1995.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(4, 'weka.J48(1)', 16, 'weka.J48', NULL, 'weka.classifiers.trees.J48', 1, 'Weka_3.9.0_11194', NULL, NULL, '2017-03-24 14:26:40', NULL, 'English', 'Ross Quinlan (1993). C4.5: Programs for Machine Learning. Morgan Kaufmann Publishers, San Mateo, CA.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(5, 'weka.REPTree(1)', 16, 'weka.REPTree', NULL, 'weka.classifiers.trees.REPTree', 1, 'Weka_3.9.0_11615', NULL, NULL, '2017-03-24 14:26:41', NULL, 'English', 'Weka implementation of REPTree', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(6, 'weka.HoeffdingTree(1)', 16, 'weka.HoeffdingTree', NULL, 'weka.classifiers.trees.HoeffdingTree', 1, 'Weka_3.9.0_11006', NULL, NULL, '2017-03-24 14:26:41', NULL, 'English', 'Geoff Hulten, Laurie Spencer, Pedro Domingos: Mining time-changing data streams. In: ACM SIGKDD Intl. Conf. on Knowledge Discovery and Data Mining, 97-106, 2001.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(7, 'weka.LMT(1)', 16, 'weka.LMT', NULL, 'weka.classifiers.trees.LMT', 1, 'Weka_3.9.0_11568', NULL, NULL, '2017-03-24 14:26:42', NULL, 'English', 'Niels Landwehr, Mark Hall, Eibe Frank (2005). Logistic Model Trees. Machine Learning. 95(1-2):161-205.\n\nMarc Sumner, Eibe Frank, Mark Hall: Speeding up Logistic Model Tree Induction. In: 9th European Conference on Principles and Practice of Knowledge Discovery in Databases, 675-683, 2005.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(8, 'weka.NaiveBayes(1)', 16, 'weka.NaiveBayes', NULL, 'weka.classifiers.bayes.NaiveBayes', 1, 'Weka_3.9.0_11741', NULL, NULL, '2017-03-24 14:26:43', NULL, 'English', 'George H. John, Pat Langley: Estimating Continuous Distributions in Bayesian Classifiers. In: Eleventh Conference on Uncertainty in Artificial Intelligence, San Mateo, 338-345, 1995.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(9, 'weka.IBk(1)', 16, 'weka.IBk', NULL, 'weka.classifiers.lazy.IBk', 1, 'Weka_3.9.0_10141', NULL, NULL, '2017-03-24 14:26:43', NULL, 'English', 'D. Aha, D. Kibler (1991). Instance-based learning algorithms. Machine Learning. 6:37-66.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(10, 'weka.SMO(1)', 16, 'weka.SMO', NULL, 'weka.classifiers.functions.SMO', 1, 'Weka_3.9.0_12558', NULL, NULL, '2017-03-24 14:26:44', NULL, 'English', 'J. Platt: Fast Training of Support Vector Machines using Sequential Minimal Optimization. In B. Schoelkopf and C. Burges and A. Smola, editors, Advances in Kernel Methods - Support Vector Learning, 1998.\n\nS.S. Keerthi, S.K. Shevade, C. Bhattacharyya, K.R.K. Murthy (2001). Improvements to Platt''s SMO Algorithm for SVM Classifier Design. Neural Computation. 13(3):637-649.\n\nTrevor Hastie, Robert Tibshirani: Classification by Pairwise Coupling. In: Advances in Neural Information Processing Systems, 1998.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(11, 'weka.Logistic(1)', 16, 'weka.Logistic', NULL, 'weka.classifiers.functions.Logistic', 1, 'Weka_3.9.0_12617', NULL, NULL, '2017-03-24 14:26:45', NULL, 'English', 'le Cessie, S., van Houwelingen, J.C. (1992). Ridge Estimators in Logistic Regression. Applied Statistics. 41(1):191-201.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(12, 'weka.MultilayerPerceptron(1)', 16, 'weka.MultilayerPerceptron', NULL, 'weka.classifiers.functions.MultilayerPerceptron', 1, 'Weka_3.9.0_12449', NULL, NULL, '2017-03-24 14:26:45', NULL, 'English', 'Weka implementation of MultilayerPerceptron', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(13, 'weka.RandomForest(1)', 16, 'weka.RandomForest', NULL, 'weka.classifiers.trees.RandomForest', 1, 'Weka_3.9.0_12645', NULL, NULL, '2017-03-24 14:26:46', NULL, 'English', 'Leo Breiman (2001). Random Forests. Machine Learning. 45(1):5-32.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(14, 'weka.Bagging(1)', 16, 'weka.Bagging', NULL, 'weka.classifiers.meta.Bagging', 1, 'Weka_3.9.0_12579', NULL, NULL, '2017-03-24 14:26:46', NULL, 'English', 'Leo Breiman (1996). Bagging predictors. Machine Learning. 24(2):123-140.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
-(15, 'weka.AdaBoostM1(1)', 16, 'weka.AdaBoostM1', NULL, 'weka.classifiers.meta.AdaBoostM1', 1, 'Weka_3.9.0_10969', NULL, NULL, '2017-03-24 14:26:47', NULL, 'English', 'Yoav Freund, Robert E. Schapire: Experiments with a new boosting algorithm. In: Thirteenth International Conference on Machine Learning, San Francisco, 148-156, 1996.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL);
+(1, 'weka.ZeroR(1)', 1, 'weka.ZeroR', NULL, 'weka.classifiers.rules.ZeroR', 1, 'Weka_3.9.0_12024', NULL, NULL, '2017-03-24 14:26:38', NULL, 'English', 'Weka implementation of ZeroR', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(2, 'weka.OneR(1)', 1, 'weka.OneR', NULL, 'weka.classifiers.rules.OneR', 1, 'Weka_3.9.0_10153', NULL, NULL, '2017-03-24 14:26:39', NULL, 'English', 'R.C. Holte (1993). Very simple classification rules perform well on most commonly used datasets. Machine Learning. 11:63-91.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(3, 'weka.JRip(1)', 1, 'weka.JRip', NULL, 'weka.classifiers.rules.JRip', 1, 'Weka_3.9.0_10153', NULL, NULL, '2017-03-24 14:26:40', NULL, 'English', 'William W. Cohen: Fast Effective Rule Induction. In: Twelfth International Conference on Machine Learning, 115-123, 1995.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(4, 'weka.J48(1)', 1, 'weka.J48', NULL, 'weka.classifiers.trees.J48', 1, 'Weka_3.9.0_11194', NULL, NULL, '2017-03-24 14:26:40', NULL, 'English', 'Ross Quinlan (1993). C4.5: Programs for Machine Learning. Morgan Kaufmann Publishers, San Mateo, CA.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(5, 'weka.REPTree(1)', 1, 'weka.REPTree', NULL, 'weka.classifiers.trees.REPTree', 1, 'Weka_3.9.0_1115', NULL, NULL, '2017-03-24 14:26:41', NULL, 'English', 'Weka implementation of REPTree', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(6, 'weka.HoeffdingTree(1)', 1, 'weka.HoeffdingTree', NULL, 'weka.classifiers.trees.HoeffdingTree', 1, 'Weka_3.9.0_11006', NULL, NULL, '2017-03-24 14:26:41', NULL, 'English', 'Geoff Hulten, Laurie Spencer, Pedro Domingos: Mining time-changing data streams. In: ACM SIGKDD Intl. Conf. on Knowledge Discovery and Data Mining, 97-106, 2001.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(7, 'weka.LMT(1)', 1, 'weka.LMT', NULL, 'weka.classifiers.trees.LMT', 1, 'Weka_3.9.0_11568', NULL, NULL, '2017-03-24 14:26:42', NULL, 'English', 'Niels Landwehr, Mark Hall, Eibe Frank (2005). Logistic Model Trees. Machine Learning. 95(1-2):11-205.\n\nMarc Sumner, Eibe Frank, Mark Hall: Speeding up Logistic Model Tree Induction. In: 9th European Conference on Principles and Practice of Knowledge Discovery in Databases, 675-683, 2005.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(8, 'weka.NaiveBayes(1)', 1, 'weka.NaiveBayes', NULL, 'weka.classifiers.bayes.NaiveBayes', 1, 'Weka_3.9.0_11741', NULL, NULL, '2017-03-24 14:26:43', NULL, 'English', 'George H. John, Pat Langley: Estimating Continuous Distributions in Bayesian Classifiers. In: Eleventh Conference on Uncertainty in Artificial Intelligence, San Mateo, 338-345, 1995.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(9, 'weka.IBk(1)', 1, 'weka.IBk', NULL, 'weka.classifiers.lazy.IBk', 1, 'Weka_3.9.0_10141', NULL, NULL, '2017-03-24 14:26:43', NULL, 'English', 'D. Aha, D. Kibler (1991). Instance-based learning algorithms. Machine Learning. 6:37-66.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(10, 'weka.SMO(1)', 1, 'weka.SMO', NULL, 'weka.classifiers.functions.SMO', 1, 'Weka_3.9.0_12558', NULL, NULL, '2017-03-24 14:26:44', NULL, 'English', 'J. Platt: Fast Training of Support Vector Machines using Sequential Minimal Optimization. In B. Schoelkopf and C. Burges and A. Smola, editors, Advances in Kernel Methods - Support Vector Learning, 1998.\n\nS.S. Keerthi, S.K. Shevade, C. Bhattacharyya, K.R.K. Murthy (2001). Improvements to Platt''s SMO Algorithm for SVM Classifier Design. Neural Computation. 13(3):637-649.\n\nTrevor Hastie, Robert Tibshirani: Classification by Pairwise Coupling. In: Advances in Neural Information Processing Systems, 1998.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(11, 'weka.Logistic(1)', 1, 'weka.Logistic', NULL, 'weka.classifiers.functions.Logistic', 1, 'Weka_3.9.0_12617', NULL, NULL, '2017-03-24 14:26:45', NULL, 'English', 'le Cessie, S., van Houwelingen, J.C. (1992). Ridge Estimators in Logistic Regression. Applied Statistics. 41(1):191-201.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(12, 'weka.MultilayerPerceptron(1)', 1, 'weka.MultilayerPerceptron', NULL, 'weka.classifiers.functions.MultilayerPerceptron', 1, 'Weka_3.9.0_12449', NULL, NULL, '2017-03-24 14:26:45', NULL, 'English', 'Weka implementation of MultilayerPerceptron', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(13, 'weka.RandomForest(1)', 1, 'weka.RandomForest', NULL, 'weka.classifiers.trees.RandomForest', 1, 'Weka_3.9.0_12645', NULL, NULL, '2017-03-24 14:26:46', NULL, 'English', 'Leo Breiman (2001). Random Forests. Machine Learning. 45(1):5-32.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(14, 'weka.Bagging(1)', 1, 'weka.Bagging', NULL, 'weka.classifiers.meta.Bagging', 1, 'Weka_3.9.0_12579', NULL, NULL, '2017-03-24 14:26:46', NULL, 'English', 'Leo Breiman (1996). Bagging predictors. Machine Learning. 24(2):123-140.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL),
+(15, 'weka.AdaBoostM1(1)', 1, 'weka.AdaBoostM1', NULL, 'weka.classifiers.meta.AdaBoostM1', 1, 'Weka_3.9.0_10969', NULL, NULL, '2017-03-24 14:26:47', NULL, 'English', 'Yoav Freund, Robert E. Schapire: Experiments with a new boosting algorithm. In: Thirteenth International Conference on Machine Learning, San Francisco, 148-156, 1996.', NULL, NULL, 'Weka_3.9.0', NULL, NULL, NULL, 'public', NULL);
 
diff --git a/data/sql/implementation_tag.sql b/data/sql/implementation_tag.sql
index e8c767539..c4c00ed78 100644
--- a/data/sql/implementation_tag.sql
+++ b/data/sql/implementation_tag.sql
@@ -10,23 +10,23 @@ INSERT INTO `implementation_tag` (`id`, `tag`, `uploader`, `date`) VALUES
 (5, 'OpenmlWeka', 16, '2017-03-24 14:26:41'),
 (5, 'weka', 16, '2017-03-24 14:26:41'),
 (6, 'OpenmlWeka', 16, '2017-03-24 14:26:42'),
-(6, 'weka', 16, '2017-03-24 14:26:42'),
-(7, 'OpenmlWeka', 16, '2017-03-24 14:26:42'),
-(7, 'weka', 16, '2017-03-24 14:26:42'),
-(8, 'OpenmlWeka', 16, '2017-03-24 14:26:43'),
-(8, 'weka', 16, '2017-03-24 14:26:43'),
-(9, 'OpenmlWeka', 16, '2017-03-24 14:26:43'),
-(9, 'weka', 16, '2017-03-24 14:26:43'),
-(10, 'OpenmlWeka', 16, '2017-03-24 14:26:44'),
-(10, 'weka', 16, '2017-03-24 14:26:44'),
-(11, 'OpenmlWeka', 16, '2017-03-24 14:26:45'),
-(11, 'weka', 16, '2017-03-24 14:26:45'),
-(12, 'OpenmlWeka', 16, '2017-03-24 14:26:45'),
-(12, 'weka', 16, '2017-03-24 14:26:45'),
-(13, 'OpenmlWeka', 16, '2017-03-24 14:26:46'),
-(13, 'weka', 16, '2017-03-24 14:26:46'),
-(14, 'OpenmlWeka', 16, '2017-03-24 14:26:46'),
-(14, 'weka', 16, '2017-03-24 14:26:47'),
-(15, 'OpenmlWeka', 16, '2017-03-24 14:26:47'),
-(15, 'weka', 16, '2017-03-24 14:26:47');
+(6, 'weka', 1, '2017-03-24 14:26:42'),
+(7, 'OpenmlWeka', 1, '2017-03-24 14:26:42'),
+(7, 'weka', 1, '2017-03-24 14:26:42'),
+(8, 'OpenmlWeka', 1, '2017-03-24 14:26:43'),
+(8, 'weka', 1, '2017-03-24 14:26:43'),
+(9, 'OpenmlWeka', 1, '2017-03-24 14:26:43'),
+(9, 'weka', 1, '2017-03-24 14:26:43'),
+(10, 'OpenmlWeka', 1, '2017-03-24 14:26:44'),
+(10, 'weka', 1, '2017-03-24 14:26:44'),
+(11, 'OpenmlWeka', 1, '2017-03-24 14:26:45'),
+(11, 'weka', 1, '2017-03-24 14:26:45'),
+(12, 'OpenmlWeka', 1, '2017-03-24 14:26:45'),
+(12, 'weka', 1, '2017-03-24 14:26:45'),
+(13, 'OpenmlWeka', 1, '2017-03-24 14:26:46'),
+(13, 'weka', 1, '2017-03-24 14:26:46'),
+(14, 'OpenmlWeka', 1, '2017-03-24 14:26:46'),
+(14, 'weka', 1, '2017-03-24 14:26:47'),
+(15, 'OpenmlWeka', 1, '2017-03-24 14:26:47'),
+(15, 'weka', 1, '2017-03-24 14:26:47');
 
diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index 4f4563b74..25c5fe9e7 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -101,7 +101,7 @@ public function create_local_admin() {
   
     $username="admin";
     $email="admin";
-    $session_hash=md5(rand());
+    $session_hash='0123456789abcdef0123456789abcdef';
     $additional_data = array(
       'first_name' => "Admin",
       'last_name'  => "",
@@ -124,7 +124,7 @@ public function create_local_admin() {
 
     # Add to admin group
     $group = array('1');
-    $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
+    //$adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
     
     echo "\r\nDone! ";
     
@@ -226,7 +226,8 @@ function install_database() {
     echo "\r\nInstall database...";
     // note that this one does not come from DATA folder, as they are stored in github
     $models = directory_map('data/sql/', 1);
-    $manipulated_order = array('file.sql', 'implementation.sql', 'algorithm_setup.sql', 'dataset.sql', 'task_type.sql', 'task.sql', 'study.sql', 'groups.sql', 'users.sql');
+    $manipulated_order = array('file.sql', 'groups.sql', 'users.sql', 'implementation.sql', 'algorithm_setup.sql', 
+      'dataset.sql', 'task_type.sql', 'task.sql', 'study.sql');
 
     // moves elements of $manipulated_order to the start of the models array
     foreach (array_reverse($manipulated_order) as $name) {

From 3a6dc6c0a8766cf7dadcf11fad9c2c1305540929 Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Thu, 26 Nov 2020 13:58:02 +0100
Subject: [PATCH 08/15] fewer datasets and local admin changes

---
 data/sql/dataset.sql            |  78 +-----------------
 data/sql/dataset_tag.sql        |  92 +--------------------
 data/sql/file.sql               | 140 +++-----------------------------
 data/sql/implementation_tag.sql |  22 ++---
 openml_OS/controllers/Cron.php  |  47 ++++++-----
 5 files changed, 48 insertions(+), 331 deletions(-)

diff --git a/data/sql/dataset.sql b/data/sql/dataset.sql
index c2dc34c44..654edd1cd 100644
--- a/data/sql/dataset.sql
+++ b/data/sql/dataset.sql
@@ -8,80 +8,4 @@ INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_
 (7, 1, 0, 'breast-w', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:20', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52350/openml_phpJNxH0q', 'true', 7, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, 'added special attributes', '2014-09-21 23:04:47'),
 (8, 1, 0, 'mfeat-karhunen', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:30', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/16/dataset_16_mfeat-karhunen.arff', 'true', 8, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:30'),
 (9, 1, 0, 'mfeat-morphological', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (0 - 9)\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class \'0\', followed by sets of 200 patterns\n for each of the classes \'1\' - \'9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:37', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/18/dataset_18_mfeat-morphological.arff', 'true', 9, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:37'),
-(10, 1, 0, 'mfeat-pixel', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:48', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 'true', 10, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:48'),
-(11, 1, 0, 'car', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Car Evaluation Database\n \n 2. Sources:\n    (a) Creator: Marko Bohanec\n    (b) Donors: Marko Bohanec   (marko.bohanec@ijs.si)\n                Blaz Zupan      (blaz.zupan@ijs.si)\n    (c) Date: June, 1997\n \n 3. Past Usage:\n \n    The hierarchical decision model, from which this dataset is\n    derived, was first presented in \n \n    M. Bohanec and V. Rajkovic: Knowledge acquisition and explanation for\n    multi-attribute decision making. In 8th Intl Workshop on Expert\n    Systems and their Applications, Avignon, France. pages 59-78, 1988.\n \n    Within machine-learning, this dataset was used for the evaluation\n    of HINT (Hierarchy INduction Tool), which was proved to be able to\n    completely reconstruct the original hierarchical model. This,\n    together with a comparison with C4.5, is presented in\n \n    B. Zupan, M. Bohanec, I. Bratko, J. Demsar: Machine learning by\n    function decomposition. ICML-97, Nashville, TN. 1997 (to appear)\n \n 4. Relevant Information Paragraph:\n \n    Car Evaluation Database was derived from a simple hierarchical\n    decision model originally developed for the demonstration of DEX\n    (M. Bohanec, V. Rajkovic: Expert system for decision\n    making. Sistemica 1(1), pp. 145-157, 1990.). The model evaluates\n    cars according to the following concept structure:\n \n    CAR                      car acceptability\n    . PRICE                  overall price\n    . . buying               buying price\n    . . maint                price of the maintenance\n    . TECH                   technical characteristics\n    . . COMFORT              comfort\n    . . . doors              number of doors\n    . . . persons            capacity in terms of persons to carry\n    . . . lug_boot           the size of luggage boot\n    . . safety               estimated safety of the car\n \n    Input attributes are printed in lowercase. Besides the target\n    concept (CAR), the model includes three intermediate concepts:\n    PRICE, TECH, COMFORT. Every concept is in the original model\n    related to its lower level descendants by a set of examples (for\n    these examples sets see http://www-ai.ijs.si/BlazZupan/car.html).\n \n    The Car Evaluation Database contains examples with the structural\n    information removed, i.e., directly relates CAR to the six input\n    attributes: buying, maint, doors, persons, lug_boot, safety.\n \n    Because of known underlying concept structure, this database may be\n    particularly useful for testing constructive induction and\n    structure discovery methods.\n \n 5. Number of Instances: 1728\n    (instances completely cover the attribute space)\n \n 6. Number of Attributes: 6\n \n 7. Attribute Values:\n \n    buying       v-high, high, med, low\n    maint        v-high, high, med, low\n    doors        2, 3, 4, 5-more\n    persons      2, 4, more\n    lug_boot     small, med, big\n    safety       low, med, high\n \n 8. Missing Attribute Values: none\n \n 9. Class Distribution (number of instances per class)\n \n    class      N          N[%]\n    -----------------------------\n    unacc     1210     (70.023 %) \n    acc        384     (22.222 %) \n    good        69     ( 3.993 %) \n    v-good      65     ( 3.762 %) \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:52', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/21/dataset_21_car.arff', 'true', 11, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:52'),
-(12, 1, 0, 'mfeat-zernike', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:00', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/22/dataset_22_mfeat-zernike.arff', 'true', 12, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:00'),
-(13, 1, 0, 'cmc', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Contraceptive Method Choice\n \n 2. Sources:\n    (a) Origin:  This dataset is a subset of the 1987 National Indonesia\n                 Contraceptive Prevalence Survey\n    (b) Creator: Tjen-Sien Lim (limt@stat.wisc.edu)\n    (c) Donor:   Tjen-Sien Lim (limt@stat.wisc.edu)\n    (c) Date:    June 7, 1997\n \n 3. Past Usage:\n    Lim, T.-S., Loh, W.-Y. & Shih, Y.-S. (1999). A Comparison of\n    Prediction Accuracy, Complexity, and Training Time of Thirty-three\n    Old and New Classification Algorithms. Machine Learning. Forthcoming.\n    (ftp://ftp.stat.wisc.edu/pub/loh/treeprogs/quest1.7/mach1317.pdf or\n    (http://www.stat.wisc.edu/~limt/mach1317.pdf)\n \n 4. Relevant Information:\n    This dataset is a subset of the 1987 National Indonesia Contraceptive\n    Prevalence Survey. The samples are married women who were either not \n    pregnant or do not know if they were at the time of interview. The \n    problem is to predict the current contraceptive method choice \n    (no use, long-term methods, or short-term methods) of a woman based \n    on her demographic and socio-economic characteristics.\n \n 5. Number of Instances: 1473\n \n 6. Number of Attributes: 10 (including the class attribute)\n \n 7. Attribute Information:\n \n    1. Wife\'s age                     (numerical)\n    2. Wife\'s education               (categorical)      1=low, 2, 3, 4=high\n    3. Husband\'s education            (categorical)      1=low, 2, 3, 4=high\n    4. Number of children ever born   (numerical)\n    5. Wife\'s religion                (binary)           0=Non-Islam, 1=Islam\n    6. Wife\'s now working?            (binary)           0=Yes, 1=No\n    7. Husband\'s occupation           (categorical)      1, 2, 3, 4\n    8. Standard-of-living index       (categorical)      1=low, 2, 3, 4=high\n    9. Media exposure                 (binary)           0=Good, 1=Not good\n    10. Contraceptive method used     (class attribute)  1=No-use \n                                                         2=Long-term\n                                                         3=Short-term\n \n 8. Missing Attribute Values: None\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:03', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/23/dataset_23_cmc.arff', 'true', 13, 'Contraceptive_method_used', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:03'),
-(14, 1, 0, 'mushroom', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Mushroom Database\n \n 2. Sources: \n     (a) Mushroom records drawn from The Audubon Society Field Guide to North\n         American Mushrooms (1981). G. H. Lincoff (Pres.), New York: Alfred\n         A. Knopf\n     (b) Donor: Jeff Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n     (c) Date: 27 April 1987\n \n 3. Past Usage:\n     1. Schlimmer,J.S. (1987). Concept Acquisition Through Representational\n        Adjustment (Technical Report 87-19).  Doctoral disseration, Department\n        of Information and Computer Science, University of California, Irvine.\n        --- STAGGER: asymptoted to 95% classification accuracy after reviewing\n            1000 instances.\n     2. Iba,W., Wogulis,J., & Langley,P. (1988).  Trading off Simplicity\n        and Coverage in Incremental Concept Learning. In Proceedings of \n        the 5th International Conference on Machine Learning, 73-79.\n        Ann Arbor, Michigan: Morgan Kaufmann.  \n        -- approximately the same results with their HILLARY algorithm    \n     3. In the following references a set of rules (given below) were\n         learned for this data set which may serve as a point of\n         comparison for other researchers.\n \n         Duch W, Adamczak R, Grabczewski K (1996) Extraction of logical rules\n         from training data using backpropagation networks, in: Proc. of the\n         The 1st Online Workshop on Soft Computing, 19-30.Aug.1996, pp. 25-30,\n         available on-line at: http://www.bioele.nuee.nagoya-u.ac.jp/wsc1/\n \n         Duch W, Adamczak R, Grabczewski K, Ishikawa M, Ueda H, Extraction of\n         crisp logical rules using constrained backpropagation networks -\n         comparison of two new approaches, in: Proc. of the European Symposium\n         on Artificial Neural Networks (ESANN\'97), Bruge, Belgium 16-18.4.1997,\n         pp. xx-xx\n \n         Wlodzislaw Duch, Department of Computer Methods, Nicholas Copernicus\n         University, 87-100 Torun, Grudziadzka 5, Poland\n         e-mail: duch@phys.uni.torun.pl\n         WWW     http://www.phys.uni.torun.pl/kmk/\n         \n         Date: Mon, 17 Feb 1997 13:47:40 +0100\n         From: Wlodzislaw Duch <duch@phys.uni.torun.pl>\n         Organization: Dept. of Computer Methods, UMK\n \n         I have attached a file containing logical rules for mushrooms.\n         It should be helpful for other people since only in the last year I\n         have seen about 10 papers analyzing this dataset and obtaining quite\n         complex rules. We will try to contribute other results later.\n \n         With best regards, Wlodek Duch\n         ________________________________________________________________\n \n         Logical rules for the mushroom data sets.\n \n         Logical rules given below seem to be the simplest possible for the\n         mushroom dataset and therefore should be treated as benchmark results.\n \n         Disjunctive rules for poisonous mushrooms, from most general\n         to most specific:\n \n         P_1) odor=NOT(almond.OR.anise.OR.none)\n              120 poisonous cases missed, 98.52% accuracy\n \n         P_2) spore-print-color=green\n              48 cases missed, 99.41% accuracy\n          \n         P_3) odor=none.AND.stalk-surface-below-ring=scaly.AND.\n                   (stalk-color-above-ring=NOT.brown) \n              8 cases missed, 99.90% accuracy\n          \n         P_4) habitat=leaves.AND.cap-color=white\n                  100% accuracy     \n \n         Rule P_4) may also be\n \n         P_4\') population=clustered.AND.cap_color=white\n \n         These rule involve 6 attributes (out of 22). Rules for edible\n         mushrooms are obtained as negation of the rules given above, for\n         example the rule:\n \n         odor=(almond.OR.anise.OR.none).AND.spore-print-color=NOT.green\n \n         gives 48 errors, or 99.41% accuracy on the whole dataset.\n \n         Several slightly more complex variations on these rules exist,\n         involving other attributes, such as gill_size, gill_spacing,\n         stalk_surface_above_ring, but the rules given above are the simplest\n         we have found.\n \n \n 4. Relevant Information:\n     This data set includes descriptions of hypothetical samples\n     corresponding to 23 species of gilled mushrooms in the Agaricus and\n     Lepiota Family (pp. 500-525).  Each species is identified as\n     definitely edible, definitely poisonous, or of unknown edibility and\n     not recommended.  This latter class was combined with the poisonous\n     one.  The Guide clearly states that there is no simple rule for\n     determining the edibility of a mushroom; no rule like ``leaflets\n     three, let it be\'\' for Poisonous Oak and Ivy.\n \n 5. Number of Instances: 8124\n \n 6. Number of Attributes: 22 (all nominally valued)\n \n 7. Attribute Information: (classes: edible=e, poisonous=p)\n      1. cap-shape:                bell=b,conical=c,convex=x,flat=f,\n                                   knobbed=k,sunken=s\n      2. cap-surface:              fibrous=f,grooves=g,scaly=y,smooth=s\n      3. cap-color:                brown=n,buff=b,cinnamon=c,gray=g,green=r,\n                                   pink=p,purple=u,red=e,white=w,yellow=y\n      4. bruises?:                 bruises=t,no=f\n      5. odor:                     almond=a,anise=l,creosote=c,fishy=y,foul=f,\n                                   musty=m,none=n,pungent=p,spicy=s\n      6. gill-attachment:          attached=a,descending=d,free=f,notched=n\n      7. gill-spacing:             close=c,crowded=w,distant=d\n      8. gill-size:                broad=b,narrow=n\n      9. gill-color:               black=k,brown=n,buff=b,chocolate=h,gray=g,\n                                   green=r,orange=o,pink=p,purple=u,red=e,\n                                   white=w,yellow=y\n     10. stalk-shape:              enlarging=e,tapering=t\n     11. stalk-root:               bulbous=b,club=c,cup=u,equal=e,\n                                   rhizomorphs=z,rooted=r,missing=?\n     12. stalk-surface-above-ring: ibrous=f,scaly=y,silky=k,smooth=s\n     13. stalk-surface-below-ring: ibrous=f,scaly=y,silky=k,smooth=s\n     14. stalk-color-above-ring:   brown=n,buff=b,cinnamon=c,gray=g,orange=o,\n                                   pink=p,red=e,white=w,yellow=y\n     15. stalk-color-below-ring:   brown=n,buff=b,cinnamon=c,gray=g,orange=o,\n                                   pink=p,red=e,white=w,yellow=y\n     16. veil-type:                partial=p,universal=u\n     17. veil-color:               brown=n,orange=o,white=w,yellow=y\n     18. ring-number:              none=n,one=o,two=t\n     19. ring-type:                cobwebby=c,evanescent=e,flaring=f,large=l,\n                                   none=n,pendant=p,sheathing=s,zone=z\n     20. spore-print-color:        black=k,brown=n,buff=b,chocolate=h,green=r,\n                                   orange=o,purple=u,white=w,yellow=y\n     21. population:               abundant=a,clustered=c,numerous=n,\n                                   scattered=s,several=v,solitary=y\n     22. habitat:                  grasses=g,leaves=l,meadows=m,paths=p,\n                                   urban=u,waste=w,woods=d\n \n 8. Missing Attribute Values: 2480 of them (denoted by \"?\"), all for\n    attribute #11.\n \n 9. Class Distribution: \n     --    edible: 4208 (51.8%)\n     -- poisonous: 3916 (48.2%)\n     --     total: 8124 instances', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:11', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/24/dataset_24_mushroom.arff', 'true', 14, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:11'),
-(15, 1, 0, 'optdigits', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Optical Recognition of Handwritten Digits\n \n 2. Source:\n 	E. Alpaydin, C. Kaynak\n 	Department of Computer Engineering\n 	Bogazici University, 80815 Istanbul Turkey\n 	alpaydin@boun.edu.tr\n 	July 1998\n \n 3. Past Usage:\n 	C. Kaynak (1995) Methods of Combining Multiple Classifiers and Their\n 	Applications to Handwritten Digit Recognition, \n 	MSc Thesis, Institute of Graduate Studies in Science and \n 	Engineering, Bogazici University.\n \n 	E. Alpaydin, C. Kaynak (1998) Cascading Classifiers, Kybernetika,\n 	to appear. ftp://ftp.icsi.berkeley.edu/pub/ai/ethem/kyb.ps.Z\n \n 4. Relevant Information:\n 	We used preprocessing programs made available by NIST to extract\n 	normalized bitmaps of handwritten digits from a preprinted form. From\n 	a total of 43 people, 30 contributed to the training set and different\n 	13 to the test set. 32x32 bitmaps are divided into nonoverlapping \n 	blocks of 4x4 and the number of on pixels are counted in each block.\n 	This generates an input matrix of 8x8 where each element is an \n 	integer in the range 0..16. This reduces dimensionality and gives \n 	invariance to small distortions.\n \n 	For info on NIST preprocessing routines, see \n 	M. D. Garris, J. L. Blue, G. T. Candela, D. L. Dimmick, J. Geist, \n 	P. J. Grother, S. A. Janet, and C. L. Wilson, NIST Form-Based \n 	Handprint Recognition System, NISTIR 5469, 1994.\n \n 5. Number of Instances\n 	optdigits.tra	Training	3823\n 	optdigits.tes	Testing		1797\n 	\n 	The way we used the dataset was to use half of training for \n 	actual training, one-fourth for validation and one-fourth\n 	for writer-dependent testing. The test set was used for \n 	writer-independent testing and is the actual quality measure.\n \n 6. Number of Attributes\n 	64 input+1 class attribute\n \n 7. For Each Attribute:\n 	All input attributes are integers in the range 0..16.\n 	The last attribute is the class code 0..9\n \n 8. Missing Attribute Values\n 	None\n \n 9. Class Distribution\n 	Class:	No of examples in training set\n 	0:  376\n 	1:  389\n 	2:  380\n 	3:  389\n 	4:  387\n 	5:  376\n 	6:  377\n 	7:  387\n 	8:  380\n 	9:  382\n \n 	Class: No of examples in testing set\n 	0:  178\n 	1:  182\n 	2:  177\n 	3:  183\n 	4:  181\n 	5:  182\n 	6:  181\n 	7:  179\n 	8:  174\n 	9:  180\n \n Accuracy on the testing set with k-nn \n using Euclidean distance as the metric\n \n  k =  1   : 98.00\n  k =  2   : 97.38\n  k =  3   : 97.83\n  k =  4   : 97.61\n  k =  5   : 97.89\n  k =  6   : 97.77\n  k =  7   : 97.66\n  k =  8   : 97.66\n  k =  9   : 97.72\n  k = 10   : 97.55\n  k = 11   : 97.89', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:34', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/28/dataset_28_optdigits.arff', 'true', 15, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:34'),
-(16, 1, 0, 'credit-a', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Credit Approval\n \n 2. Sources: \n     (confidential)\n     Submitted by quinlan@cs.su.oz.au\n \n 3.  Past Usage:\n \n     See Quinlan,\n     * \"Simplifying decision trees\", Int J Man-Machine Studies 27,\n       Dec 1987, pp. 221-234.\n     * \"C4.5: Programs for Machine Learning\", Morgan Kaufmann, Oct 1992\n   \n 4.  Relevant Information:\n \n     This file concerns credit card applications.  All attribute names\n     and values have been changed to meaningless symbols to protect\n     confidentiality of the data.\n   \n     This dataset is interesting because there is a good mix of\n     attributes -- continuous, nominal with small numbers of\n     values, and nominal with larger numbers of values.  There\n     are also a few missing values.\n   \n 5.  Number of Instances: 690\n \n 6.  Number of Attributes: 15 + class attribute\n \n 7.  Attribute Information:\n \n     A1:	b, a.\n     A2:	continuous.\n     A3:	continuous.\n     A4:	u, y, l, t.\n     A5:	g, p, gg.\n     A6:	c, d, cc, i, j, k, m, r, q, w, x, e, aa, ff.\n     A7:	v, h, bb, j, n, z, dd, ff, o.\n     A8:	continuous.\n     A9:	t, f.\n     A10:	t, f.\n     A11:	continuous.\n     A12:	t, f.\n     A13:	g, p, s.\n     A14:	continuous.\n     A15:	continuous.\n     A16: +,-         (class attribute)\n \n 8.  Missing Attribute Values:\n     37 cases (5%) have one or more missing values.  The missing\n     values from particular attributes are:\n \n     A1:  12\n     A2:  12\n     A4:   6\n     A5:   6\n     A6:   9\n     A7:   9\n     A14: 13\n \n 9.  Class Distribution\n   \n     +: 307 (44.5%)\n     -: 383 (55.5%)', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:38', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/29/dataset_29_credit-a.arff', 'true', 16, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:38'),
-(17, 1, 0, 'credit-g', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDescription of the German credit dataset.\n \n 1. Title: German Credit data\n \n 2. Source Information\n \n Professor Dr. Hans Hofmann  \n Institut f\"ur Statistik und \"Okonometrie  \n Universit\"at Hamburg  \n FB Wirtschaftswissenschaften  \n Von-Melle-Park 5    \n 2000 Hamburg 13 \n \n 3. Number of Instances:  1000\n \n Two datasets are provided.  the original dataset, in the form provided\n by Prof. Hofmann, contains categorical/symbolic attributes and\n is in the file \"german.data\".   \n  \n For algorithms that need numerical attributes, Strathclyde University \n produced the file \"german.data-numeric\".  This file has been edited \n and several indicator variables added to make it suitable for \n algorithms which cannot cope with categorical variables.   Several\n attributes that are ordered categorical (such as attribute 17) have\n been coded as integer.    This was the form used by StatLog.\n \n \n 6. Number of Attributes german: 20 (7 numerical, 13 categorical)\n    Number of Attributes german.numer: 24 (24 numerical)\n \n \n 7.  Attribute description for german\n \n Attribute 1:  (qualitative)\n 	       Status of existing checking account\n                A11 :      ... <    0 DM\n 	       A12 : 0 <= ... <  200 DM\n 	       A13 :      ... >= 200 DM /\n 		     salary assignments for at least 1 year\n                A14 : no checking account\n \n Attribute 2:  (numerical)\n 	      Duration in month\n \n Attribute 3:  (qualitative)\n 	      Credit history\n 	      A30 : no credits taken/\n 		    all credits paid back duly\n               A31 : all credits at this bank paid back duly\n 	      A32 : existing credits paid back duly till now\n               A33 : delay in paying off in the past\n 	      A34 : critical account/\n 		    other credits existing (not at this bank)\n \n Attribute 4:  (qualitative)\n 	      Purpose\n 	      A40 : car (new)\n 	      A41 : car (used)\n 	      A42 : furniture/equipment\n 	      A43 : radio/television\n 	      A44 : domestic appliances\n 	      A45 : repairs\n 	      A46 : education\n 	      A47 : (vacation - does not exist?)\n 	      A48 : retraining\n 	      A49 : business\n 	      A410 : others\n \n Attribute 5:  (numerical)\n 	      Credit amount\n \n Attibute 6:  (qualitative)\n 	      Savings account/bonds\n 	      A61 :          ... <  100 DM\n 	      A62 :   100 <= ... <  500 DM\n 	      A63 :   500 <= ... < 1000 DM\n 	      A64 :          .. >= 1000 DM\n               A65 :   unknown/ no savings account\n \n Attribute 7:  (qualitative)\n 	      Present employment since\n 	      A71 : unemployed\n 	      A72 :       ... < 1 year\n 	      A73 : 1  <= ... < 4 years  \n 	      A74 : 4  <= ... < 7 years\n 	      A75 :       .. >= 7 years\n \n Attribute 8:  (numerical)\n 	      Installment rate in percentage of disposable income\n \n Attribute 9:  (qualitative)\n 	      Personal status and sex\n 	      A91 : male   : divorced/separated\n 	      A92 : female : divorced/separated/married\n               A93 : male   : single\n 	      A94 : male   : married/widowed\n 	      A95 : female : single\n \n Attribute 10: (qualitative)\n 	      Other debtors / guarantors\n 	      A101 : none\n 	      A102 : co-applicant\n 	      A103 : guarantor\n \n Attribute 11: (numerical)\n 	      Present residence since\n \n Attribute 12: (qualitative)\n 	      Property\n 	      A121 : real estate\n 	      A122 : if not A121 : building society savings agreement/\n 				   life insurance\n               A123 : if not A121/A122 : car or other, not in attribute 6\n 	      A124 : unknown / no property\n \n Attribute 13: (numerical)\n 	      Age in years\n \n Attribute 14: (qualitative)\n 	      Other installment plans \n 	      A141 : bank\n 	      A142 : stores\n 	      A143 : none\n \n Attribute 15: (qualitative)\n 	      Housing\n 	      A151 : rent\n 	      A152 : own\n 	      A153 : for free\n \n Attribute 16: (numerical)\n               Number of existing credits at this bank\n \n Attribute 17: (qualitative)\n 	      Job\n 	      A171 : unemployed/ unskilled  - non-resident\n 	      A172 : unskilled - resident\n 	      A173 : skilled employee / official\n 	      A174 : management/ self-employed/\n 		     highly qualified employee/ officer\n \n Attribute 18: (numerical)\n 	      Number of people being liable to provide maintenance for\n \n Attribute 19: (qualitative)\n 	      Telephone\n 	      A191 : none\n 	      A192 : yes, registered under the customers name\n \n Attribute 20: (qualitative)\n 	      foreign worker\n 	      A201 : yes\n 	      A202 : no\n \n \n \n 8.  Cost Matrix\n \n This dataset requires use of a cost matrix (see below)\n \n \n       1        2\n ----------------------------\n   1   0        1\n -----------------------\n   2   5        0\n \n (1 = Good,  2 = Bad)\n \n the rows represent the actual classification and the columns\n the predicted classification.\n \n It is worse to class a customer as good when they are bad (5), \n than it is to class a customer as bad when they are good (1).\n \n\n\n\n\n Relabeled values in attribute checking_status\n    From: A11                     To: \'<0\'                \n    From: A12                     To: \'0<=X<200\'          \n    From: A13                     To: \'>=200\'             \n    From: A14                     To: \'no checking\'       \n\n\n Relabeled values in attribute credit_history\n    From: A30                     To: \'no credits/all paid\'\n    From: A31                     To: \'all paid\'          \n    From: A32                     To: \'existing paid\'     \n    From: A33                     To: \'delayed previously\'\n    From: A34                     To: \'critical/other existing credit\'\n\n\n Relabeled values in attribute purpose\n    From: A40                     To: \'new car\'           \n    From: A41                     To: \'used car\'          \n    From: A42                     To: furniture/equipment \n    From: A43                     To: radio/tv            \n    From: A44                     To: \'domestic appliance\'\n    From: A45                     To: repairs             \n    From: A46                     To: education           \n    From: A47                     To: vacation            \n    From: A48                     To: retraining          \n    From: A49                     To: business            \n    From: A410                    To: other               \n\n\n Relabeled values in attribute savings_status\n    From: A61                     To: \'<100\'              \n    From: A62                     To: \'100<=X<500\'        \n    From: A63                     To: \'500<=X<1000\'       \n    From: A64                     To: \'>=1000\'            \n    From: A65                     To: \'no known savings\'  \n\n\n Relabeled values in attribute employment\n    From: A71                     To: unemployed          \n    From: A72                     To: \'<1\'                \n    From: A73                     To: \'1<=X<4\'            \n    From: A74                     To: \'4<=X<7\'            \n    From: A75                     To: \'>=7\'               \n\n\n Relabeled values in attribute personal_status\n    From: A91                     To: \'male div/sep\'      \n    From: A92                     To: \'female div/dep/mar\'\n    From: A93                     To: \'male single\'       \n    From: A94                     To: \'male mar/wid\'      \n    From: A95                     To: \'female single\'     \n\n\n Relabeled values in attribute other_parties\n    From: A101                    To: none                \n    From: A102                    To: \'co applicant\'      \n    From: A103                    To: guarantor           \n\n\n Relabeled values in attribute property_magnitude\n    From: A121                    To: \'real estate\'       \n    From: A122                    To: \'life insurance\'    \n    From: A123                    To: car                 \n    From: A124                    To: \'no known property\' \n\n\n Relabeled values in attribute other_payment_plans\n    From: A141                    To: bank                \n    From: A142                    To: stores              \n    From: A143                    To: none                \n\n\n Relabeled values in attribute housing\n    From: A151                    To: rent                \n    From: A152                    To: own                 \n    From: A153                    To: \'for free\'          \n\n\n Relabeled values in attribute job\n    From: A171                    To: \'unemp/unskilled non res\'\n    From: A172                    To: \'unskilled resident\'\n    From: A173                    To: skilled             \n    From: A174                    To: \'high qualif/self emp/mgmt\'\n\n\n Relabeled values in attribute own_telephone\n    From: A191                    To: none                \n    From: A192                    To: yes                 \n\n\n Relabeled values in attribute foreign_worker\n    From: A201                    To: yes                 \n    From: A202                    To: no                  \n\n\n Relabeled values in attribute class\n    From: 1                       To: good                \n    From: 2                       To: bad', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:47', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/31/dataset_31_credit-g.arff', 'true', 17, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:47'),
-(18, 1, 0, 'pendigits', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Pen-Based Recognition of Handwritten Digits\n \n 2. Source:\n 	E. Alpaydin, F. Alimoglu\n 	Department of Computer Engineering\n 	Bogazici University, 80815 Istanbul Turkey\n 	alpaydin@boun.edu.tr\n 	July 1998\n \n 3. Past Usage:\n 	F. Alimoglu (1996) Combining Multiple Classifiers for Pen-Based\n 	Handwritten Digit Recognition, \n 	MSc Thesis, Institute of Graduate Studies in Science and \n 	Engineering, Bogazici University.\n 	http://www.cmpe.boun.edu.tr/~alimoglu/alimoglu.ps.gz\n \n 	F. Alimoglu, E. Alpaydin, \"Methods of Combining Multiple Classifiers \n 	Based on Different Representations for Pen-based Handwriting\n 	Recognition,\" Proceedings of the Fifth Turkish Artificial \n 	Intelligence and Artificial Neural Networks Symposium (TAINN 96), \n 	June 1996, Istanbul, Turkey.\n 	http://www.cmpe.boun.edu.tr/~alimoglu/tainn96.ps.gz\n \n 	\n 4. Relevant Information:\n \n 	We create a digit database by collecting 250 samples from 44 writers.\n 	The samples written by 30 writers are used for training,\n 	cross-validation and writer dependent testing, and the digits \n 	written by the other 14 are used for writer independent testing. This\n 	database is also available in the UNIPEN format.\n \n 	We use a WACOM PL-100V pressure sensitive tablet with an integrated \n 	LCD display and a cordless stylus. The input and display areas are\n 	located in the same place. Attached to the serial port of an Intel \n 	486 based PC, it allows us to collect handwriting samples. The tablet\n 	sends $x$ and $y$ tablet coordinates and pressure level values of the\n 	pen at fixed time intervals (sampling rate) of 100 miliseconds. \n \n 	These writers are asked to write 250 digits in random order inside \n 	boxes of 500 by 500 tablet pixel resolution.  Subject are monitored \n 	only during the first entry screens. Each screen contains five boxes\n 	with the digits to be written displayed above. Subjects are told to\n 	write only inside these boxes.  If they make a mistake or are unhappy\n 	with their writing, they are instructed to clear the content of a box \n 	by using an on-screen button. The first ten digits are ignored \n 	because most writers are not familiar with this type of input devices,\n 	but subjects are not aware of this. \n \n 	In our study, we use only ($x, y$) coordinate information. The stylus\n 	pressure level values are ignored. First we apply normalization to \n 	make our representation invariant to translations and scale \n 	distortions. The raw data that we capture from the tablet consist of\n 	integer values between 0 and 500 (tablet input box resolution). The \n 	new coordinates are such that the coordinate which has the maximum \n 	range varies between 0 and 100. Usually $x$ stays in this range, since\n 	most characters are taller than they are wide.  \n \n 	In order to train and test our classifiers, we need to represent \n 	digits as constant length feature vectors. A commonly used technique\n 	leading to good results is resampling the ( x_t, y_t) points. \n 	Temporal resampling (points regularly spaced in time) or spatial\n 	resampling (points regularly spaced in arc length) can be used here. \n 	Raw point data are already regularly spaced in time but the distance\n 	between them is variable. Previous research showed that spatial\n 	resampling to obtain a constant number of regularly spaced points \n 	on the trajectory yields much better performance, because it provides \n 	a better alignment between points. Our resampling algorithm uses \n 	simple linear interpolation between pairs of points. The resampled\n 	digits are represented as a sequence of T points ( x_t, y_t )_{t=1}^T,\n 	regularly spaced in arc length, as opposed to the input sequence, \n 	which is regularly spaced in time.\n \n 	So, the input vector size is 2*T, two times the number of points\n 	resampled. We considered spatial resampling to T=8,12,16 points in our\n 	experiments and found that T=8 gave the best trade-off between \n 	accuracy and complexity.\n \n \n 5. Number of Instances\n 	pendigits.tra	Training	7494\n 	pendigits.tes	Testing		3498\n 	\n 	The way we used the dataset was to use first half of training for \n 	actual training, one-fourth for validation and one-fourth\n 	for writer-dependent testing. The test set was used for \n 	writer-independent testing and is the actual quality measure.\n \n 6. Number of Attributes\n 	16 input+1 class attribute\n \n 7. For Each Attribute:\n 	All input attributes are integers in the range 0..100.\n 	The last attribute is the class code 0..9\n \n 8. Missing Attribute Values\n 	None\n \n 9. Class Distribution\n 	Class: No of examples in training set\n 	0:  780\n 	1:  779\n 	2:  780\n 	3:  719\n 	4:  780\n 	5:  720\n 	6:  720\n 	7:  778\n 	8:  719\n 	9:  719\n 	Class: No of examples in testing set\n 	0:  363\n 	1:  364\n 	2:  364\n 	3:  336\n 	4:  364\n 	5:  335\n 	6:  336\n 	7:  364\n 	8:  336\n 	9:  336\n \n Accuracy on the testing set with k-nn \n using Euclidean distance as the metric\n \n  k =  1 : 97.74\n  k =  2 : 97.37\n  k =  3 : 97.80\n  k =  4 : 97.66\n  k =  5 : 97.60\n  k =  6 : 97.57\n  k =  7 : 97.54\n  k =  8 : 97.54\n  k =  9 : 97.46\n  k = 10 : 97.48\n  k = 11 : 97.34', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:21:54', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/32/dataset_32_pendigits.arff', 'true', 18, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:21:54'),
-(19, 1, 0, 'segment', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Image Segmentation data\n \n 2. Source Information\n    -- Creators: Vision Group, University of Massachusetts\n    -- Donor: Vision Group (Carla Brodley, brodley@cs.umass.edu)\n    -- Date: November, 1990\n  \n 3. Past Usage: None yet published\n \n 4. Relevant Information:\n \n    The instances were drawn randomly from a database of 7 outdoor \n    images.  The images were handsegmented to create a classification\n    for every pixel.  \n \n    Each instance is a 3x3 region.\n \n 5. Number of Instances: Training data: 210  Test data: 2100\n \n 6. Number of Attributes: 19 continuous attributes\n \n 7. Attribute Information:\n \n     1.  region-centroid-col:  the column of the center pixel of the region.\n     2.  region-centroid-row:  the row of the center pixel of the region.\n     3.  region-pixel-count:  the number of pixels in a region = 9.\n     4.  short-line-density-5:  the results of a line extractoin algorithm that \n          counts how many lines of length 5 (any orientation) with\n          low contrast, less than or equal to 5, go through the region.\n     5.  short-line-density-2:  same as short-line-density-5 but counts lines\n          of high contrast, greater than 5.\n     6.  vedge-mean:  measure the contrast of horizontally\n          adjacent pixels in the region.  There are 6, the mean and \n          standard deviation are given.  This attribute is used as\n         a vertical edge detector.\n     7.  vegde-sd:  (see 6)\n     8.  hedge-mean:  measures the contrast of vertically adjacent\n           pixels. Used for horizontal line detection. \n     9.  hedge-sd: (see 8).\n     10. intensity-mean:  the average over the region of (R + G + B)/3\n     11. rawred-mean: the average over the region of the R value.\n     12. rawblue-mean: the average over the region of the B value.\n     13. rawgreen-mean: the average over the region of the G value.\n     14. exred-mean: measure the excess red:  (2R - (G + B))\n     15. exblue-mean: measure the excess blue:  (2B - (G + R))\n     16. exgreen-mean: measure the excess green:  (2G - (R + B))\n     17. value-mean:  3-d nonlinear transformation\n          of RGB. (Algorithm can be found in Foley and VanDam, Fundamentals\n          of Interactive Computer Graphics)\n     18. saturatoin-mean:  (see 17)\n     19. hue-mean:  (see 17)\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: \n \n    Classes:  brickface, sky, foliage, cement, window, path, grass.\n \n    30 instances per class for training data.\n    300 instances per class for test data.\n \n\n\n\n\n Relabeled values in attribute class\n    From: 1                       To: brickface           \n    From: 2                       To: sky                 \n    From: 3                       To: foliage             \n    From: 4                       To: cement              \n    From: 5                       To: window              \n    From: 6                       To: path                \n    From: 7                       To: grass', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/36/dataset_36_segment.arff', 'true', 19, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:10'),
-(20, 1, 0, 'diabetes', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Pima Indians Diabetes Database\n \n 2. Sources:\n    (a) Original owners: National Institute of Diabetes and Digestive and\n                         Kidney Diseases\n    (b) Donor of database: Vincent Sigillito (vgs@aplcen.apl.jhu.edu)\n                           Research Center, RMI Group Leader\n                           Applied Physics Laboratory\n                           The Johns Hopkins University\n                           Johns Hopkins Road\n                           Laurel, MD 20707\n                           (301) 953-6231\n    (c) Date received: 9 May 1990\n \n 3. Past Usage:\n     1. Smith,~J.~W., Everhart,~J.~E., Dickson,~W.~C., Knowler,~W.~C., &\n        Johannes,~R.~S. (1988). Using the ADAP learning algorithm to forecast\n        the onset of diabetes mellitus.  In {it Proceedings of the Symposium\n        on Computer Applications and Medical Care} (pp. 261--265).  IEEE\n        Computer Society Press.\n \n        The diagnostic, binary-valued variable investigated is whether the\n        patient shows signs of diabetes according to World Health Organization\n        criteria (i.e., if the 2 hour post-load plasma glucose was at least \n        200 mg/dl at any survey  examination or if found during routine medical\n        care).   The population lives near Phoenix, Arizona, USA.\n \n        Results: Their ADAP algorithm makes a real-valued prediction between\n        0 and 1.  This was transformed into a binary decision using a cutoff of \n        0.448.  Using 576 training instances, the sensitivity and specificity\n        of their algorithm was 76% on the remaining 192 instances.\n \n 4. Relevant Information:\n       Several constraints were placed on the selection of these instances from\n       a larger database.  In particular, all patients here are females at\n       least 21 years old of Pima Indian heritage.  ADAP is an adaptive learning\n       routine that generates and executes digital analogs of perceptron-like\n       devices.  It is a unique algorithm; see the paper for details.\n \n 5. Number of Instances: 768\n \n 6. Number of Attributes: 8 plus class \n \n 7. For Each Attribute: (all numeric-valued)\n    1. Number of times pregnant\n    2. Plasma glucose concentration a 2 hours in an oral glucose tolerance test\n    3. Diastolic blood pressure (mm Hg)\n    4. Triceps skin fold thickness (mm)\n    5. 2-Hour serum insulin (mu U/ml)\n    6. Body mass index (weight in kg/(height in m)^2)\n    7. Diabetes pedigree function\n    8. Age (years)\n    9. Class variable (0 or 1)\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: (class value 1 is interpreted as \"tested positive for\n    diabetes\")\n \n    Class Value  Number of instances\n    0            500\n    1            268\n \n 10. Brief statistical analysis:\n \n     Attribute number:    Mean:   Standard Deviation:\n     1.                     3.8     3.4\n     2.                   120.9    32.0\n     3.                    69.1    19.4\n     4.                    20.5    16.0\n     5.                    79.8   115.2\n     6.                    32.0     7.9\n     7.                     0.5     0.3\n     8.                    33.2    11.8\n \n \n\n\n\n\n Relabeled values in attribute \'class\'\n    From: 0                       To: tested_negative     \n    From: 1                       To: tested_positive', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:13', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/37/dataset_37_diabetes.arff', 'true', 20, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:13'),
-(21, 1, 0, 'sick', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n;\n ; Thyroid disease records supplied by the Garavan Institute and J. Ross\n ; Quinlan, New South Wales Institute, Syndney, Australia.\n ;\n ; 1987.\n ;\n \n sick, negative.			|  classes\n \n age:				continuous.\n sex:				M, F.\n on thyroxine:			f, t.\n query on thyroxine:		f, t.\n on antithyroid medication:	f, t.\n sick:				f, t.\n pregnant:			f, t.\n thyroid surgery:		f, t.\n I131 treatment:			f, t.\n query hypothyroid:		f, t.\n query hyperthyroid:		f, t.\n lithium:			f, t.\n goitre:				f, t.\n tumor:				f, t.\n hypopituitary:			f, t.\n psych:				f, t.\n TSH measured:			f, t.\n TSH:				continuous.\n T3 measured:			f, t.\n T3:				continuous.\n TT4 measured:			f, t.\n TT4:				continuous.\n T4U measured:			f, t.\n T4U:				continuous.\n FTI measured:			f, t.\n FTI:				continuous.\n TBG measured:			f, t.\n TBG:				continuous.\n referral source:		WEST, STMW, SVHC, SVI, SVHD, other.\n \n\n Num Instances:     3772\n Num Attributes:    30\n Num Continuous:    7 (Int 1 / Real 6)\n Num Discrete:      23\n Missing values:    6064 /  5.4%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'age\'                     Int    0% 100%   0%     1 /  0%    93 /  2%   0% \n   2 \'sex\'                     Enum  96%   0%   0%   150 /  4%     2 /  0%   0% \n   3 \'on thyroxine\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'query on thyroxine\'      Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'on antithyroid medicati  Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'sick\'                    Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'pregnant\'                Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'thyroid surgery\'         Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'I131 treatment\'          Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'query hypothyroid\'       Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'query hyperthyroid\'      Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'lithium\'                 Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'goitre\'                  Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'tumor\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'hypopituitary\'           Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  16 \'psych\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'TSH measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'TSH\'                     Real   0%  11%  79%   369 / 10%   287 /  8%   2% \n  19 \'T3 measured\'             Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'T3\'                      Real   0%   9%  71%   769 / 20%    69 /  2%   0% \n  21 \'TT4 measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'TT4\'                     Real   0%  94%   0%   231 /  6%   241 /  6%   1% \n  23 \'T4U measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'T4U\'                     Real   0%   2%  87%   387 / 10%   146 /  4%   1% \n  25 \'FTI measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'FTI\'                     Real   0%  90%   0%   385 / 10%   234 /  6%   2% \n  27 \'TBG measured\'            Enum 100%   0%   0%     0 /  0%     1 /  0%   0% \n  28 \'TBG\'                     Real   0%   0%   0%  3772 /100%     0 /  0%   0% \n  29 \'referral source\'         Enum 100%   0%   0%     0 /  0%     5 /  0%   0% \n  30 \'Class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:19', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/38/dataset_38_sick.arff', 'true', 21, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:19'),
-(22, 1, 0, 'soybean', '1', '1', '**Author**: R.S. Michalski and R.L. Chilausky (Donors: Ming Tan & Jeff Schlimmer)  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Soybean+(Large)) - 1988  \r\n**Please cite**: R.S. Michalski and R.L. Chilausky \"Learning by Being Told and Learning from Examples: An Experimental Comparison of the Two Methods of Knowledge Acquisition in the Context of Developing an Expert System for Soybean Disease Diagnosis\", International Journal of Policy Analysis and Information Systems, Vol. 4, No. 2, 1980.  \r\n\r\n**Large Soybean Database**  \r\nThis is the large soybean database from the UCI repository, with its training and test database combined into a single file. There are 19 classes, only the first 15 of which have been used in prior work.  The folklore seems to be that the last four classes are unjustified by the data since they have so few examples. There are 35 categorical attributes, some nominal and some ordered.  The value \'dna\' means does not apply.  The values for attributes are encoded numerically, with the first value encoded as \'0\', the second as \'1\', and so forth.  An unknown values is encoded as \'?\'.', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:32', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/42/dataset_42_soybean.arff', 'true', 22, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:32'),
-(23, 1, 0, 'spambase', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title:  SPAM E-mail Database\n \n 2. Sources:\n    (a) Creators: Mark Hopkins, Erik Reeber, George Forman, Jaap Suermondt\n         Hewlett-Packard Labs, 1501 Page Mill Rd., Palo Alto, CA 94304\n    (b) Donor: George Forman (gforman at nospam hpl.hp.com)  650-857-7835\n    (c) Generated: June-July 1999\n \n 3. Past Usage:\n    (a) Hewlett-Packard Internal-only Technical Report. External forthcoming.\n    (b) Determine whether a given email is spam or not.\n    (c) ~7% misclassification error.\n        False positives (marking good mail as spam) are very undesirable.\n        If we insist on zero false positives in the training/testing set,\n        20-25% of the spam passed through the filter.\n \n 4. Relevant Information:\n         The \"spam\" concept is diverse: advertisements for products/web\n         sites, make money fast schemes, chain letters, pornography...\n 	Our collection of spam e-mails came from our postmaster and \n 	individuals who had filed spam.  Our collection of non-spam \n 	e-mails came from filed work and personal e-mails, and hence\n 	the word \'george\' and the area code \'650\' are indicators of \n 	non-spam.  These are useful when constructing a personalized \n 	spam filter.  One would either have to blind such non-spam \n 	indicators or get a very wide collection of non-spam to \n 	generate a general purpose spam filter.\n \n         For background on spam:\n         Cranor, Lorrie F., LaMacchia, Brian A.  Spam! \n         Communications of the ACM, 41(8):74-83, 1998.\n \n 5. Number of Instances: 4601 (1813 Spam = 39.4%)\n \n 6. Number of Attributes: 58 (57 continuous, 1 nominal class label)\n \n 7. Attribute Information:\n The last column of \'spambase.data\' denotes whether the e-mail was \n considered spam (1) or not (0), i.e. unsolicited commercial e-mail.  \n Most of the attributes indicate whether a particular word or\n character was frequently occuring in the e-mail.  The run-length\n attributes (55-57) measure the length of sequences of consecutive \n capital letters.  For the statistical measures of each attribute, \n see the end of this file.  Here are the definitions of the attributes:\n \n 48 continuous real [0,100] attributes of type word_freq_WORD \n = percentage of words in the e-mail that match WORD,\n i.e. 100 * (number of times the WORD appears in the e-mail) / \n total number of words in e-mail.  A \"word\" in this case is any \n string of alphanumeric characters bounded by non-alphanumeric \n characters or end-of-string.\n \n 6 continuous real [0,100] attributes of type char_freq_CHAR\n = percentage of characters in the e-mail that match CHAR,\n i.e. 100 * (number of CHAR occurences) / total characters in e-mail\n \n 1 continuous real [1,...] attribute of type capital_run_length_average\n = average length of uninterrupted sequences of capital letters\n \n 1 continuous integer [1,...] attribute of type capital_run_length_longest\n = length of longest uninterrupted sequence of capital letters\n \n 1 continuous integer [1,...] attribute of type capital_run_length_total\n = sum of length of uninterrupted sequences of capital letters\n = total number of capital letters in the e-mail\n \n 1 nominal {0,1} class attribute of type spam\n = denotes whether the e-mail was considered spam (1) or not (0), \n i.e. unsolicited commercial e-mail.  \n \n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution:\n 	Spam	  1813  (39.4%)\n 	Non-Spam  2788  (60.6%)\n \n \n Attribute Statistics:\n    Min: Max:   Average:  Std.Dev: Coeff.Var_%: \n 1  0    4.54   0.10455   0.30536  292          \n 2  0    14.28  0.21301   1.2906   606          \n 3  0    5.1    0.28066   0.50414  180          \n 4  0    42.81  0.065425  1.3952   2130         \n 5  0    10     0.31222   0.67251  215          \n 6  0    5.88   0.095901  0.27382  286          \n 7  0    7.27   0.11421   0.39144  343          \n 8  0    11.11  0.10529   0.40107  381          \n 9  0    5.26   0.090067  0.27862  309          \n 10 0    18.18  0.23941   0.64476  269          \n 11 0    2.61   0.059824  0.20154  337          \n 12 0    9.67   0.5417    0.8617   159          \n 13 0    5.55   0.09393   0.30104  320          \n 14 0    10     0.058626  0.33518  572          \n 15 0    4.41   0.049205  0.25884  526          \n 16 0    20     0.24885   0.82579  332          \n 17 0    7.14   0.14259   0.44406  311          \n 18 0    9.09   0.18474   0.53112  287          \n 19 0    18.75  1.6621    1.7755   107          \n 20 0    18.18  0.085577  0.50977  596          \n 21 0    11.11  0.80976   1.2008   148          \n 22 0    17.1   0.1212    1.0258   846          \n 23 0    5.45   0.10165   0.35029  345          \n 24 0    12.5   0.094269  0.44264  470          \n 25 0    20.83  0.5495    1.6713   304          \n 26 0    16.66  0.26538   0.88696  334          \n 27 0    33.33  0.7673    3.3673   439          \n 28 0    9.09   0.12484   0.53858  431          \n 29 0    14.28  0.098915  0.59333  600          \n 30 0    5.88   0.10285   0.45668  444          \n 31 0    12.5   0.064753  0.40339  623          \n 32 0    4.76   0.047048  0.32856  698          \n 33 0    18.18  0.097229  0.55591  572          \n 34 0    4.76   0.047835  0.32945  689          \n 35 0    20     0.10541   0.53226  505          \n 36 0    7.69   0.097477  0.40262  413          \n 37 0    6.89   0.13695   0.42345  309          \n 38 0    8.33   0.013201  0.22065  1670         \n 39 0    11.11  0.078629  0.43467  553          \n 40 0    4.76   0.064834  0.34992  540          \n 41 0    7.14   0.043667  0.3612   827          \n 42 0    14.28  0.13234   0.76682  579          \n 43 0    3.57   0.046099  0.22381  486          \n 44 0    20     0.079196  0.62198  785          \n 45 0    21.42  0.30122   1.0117   336          \n 46 0    22.05  0.17982   0.91112  507          \n 47 0    2.17   0.0054445 0.076274 1400         \n 48 0    10     0.031869  0.28573  897          \n 49 0    4.385  0.038575  0.24347  631          \n 50 0    9.752  0.13903   0.27036  194          \n 51 0    4.081  0.016976  0.10939  644          \n 52 0    32.478 0.26907   0.81567  303          \n 53 0    6.003  0.075811  0.24588  324          \n 54 0    19.829 0.044238  0.42934  971          \n 55 1    1102.5 5.1915    31.729   611          \n 56 1    9989   52.173    194.89   374          \n 57 1    15841  283.29    606.35   214          \n 58 0    1      0.39404   0.4887   124          \n \n \n This file: \'spambase.DOCUMENTATION\' at the UCI Machine Learning Repository\n http://www.ics.uci.edu/~mlearn/MLRepository.html\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:41', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/44/dataset_44_spambase.arff', 'true', 23, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:41'),
-(24, 1, 0, 'splice', '1', '1', '**Author**: Genbank  \nDonor: G. Towell, M. Noordewier, and J. Shavlik  \n**Source**: [Genbank 64.1](genbank.bio.net) - 1/1/92  \n**Please cite**:   \n\nPrimate splice-junction gene sequences (DNA) with associated imperfect domain theory. All examples taken from Genbank 64.1. Categories \"ei\" and \"ie\" include every \"split-gene\" for primates in Genbank 64.1. Non-splice examples taken from sequences known not to include a splicing site.\n \nProblem Description:  \nSplice junctions are points on a DNA sequence at which \'superfluous\' DNA is removed during the process of protein creation in higher organisms. The problem posed in this dataset is to recognize, given a sequence of DNA, the boundaries between exons (the parts of the DNA sequence retained after splicing) and introns (the parts of the DNA sequence that are spliced out). This problem consists of two subtasks: recognizing exon/intron boundaries (referred to as EI sites), and recognizing intron/exon boundaries (IE sites). (In the biological community, IE borders are referred to a \'\'acceptors\'\' while EI borders are referred to as \'\'donors\'\'.)\n \nThis dataset has been developed to help evaluate a \"hybrid\" learning algorithm (KBANN) that uses examples to inductively refine preexisting knowledge.\n        \nAttributes:  \n>\n              1   One of {n ei ie}, indicating the class.\n              2   The instance name.\n           3-62   The remaining 60 fields are the sequence, starting at \n                  position -30 and ending at position +30. Each of\n                  these fields is almost always filled by one of \n                  {a, g, t, c}. Other characters indicate ambiguity among\n                  the standard characters according to the following table:\n    character: meaning\n        D: A or G or T\n        N: A or G or C or T\n        S: C or G\n        R: A or G\n\nNotes:  \n* Instance_name is an identifier and should be ignored for modelling', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:49', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/46/dataset_46_splice.arff', 'true', 24, 'Class', NULL, 'Instance_name', NULL, 'public', NULL, NULL, 'Instance_name is an identifier and should be ignored for modelling', '2014-09-19 17:06:29'),
-(25, 1, 0, 'tic-tac-toe', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Tic-Tac-Toe Endgame database\n \n 2. Source Information\n    -- Creator: David W. Aha (aha@cs.jhu.edu)\n    -- Donor: David W. Aha (aha@cs.jhu.edu)\n    -- Date: 19 August 1991\n  \n 3. Known Past Usage: \n    1. Matheus,~C.~J., & Rendell,~L.~A. (1989).  Constructive\n       induction on decision trees.  In {it Proceedings of the\n       Eleventh International Joint Conference on Artificial Intelligence} \n       (pp. 645--650).  Detroit, MI: Morgan Kaufmann.\n       -- CITRE was applied to 100-instance training and 200-instance test\n          sets.  In a study using various amounts of domain-specific\n          knowledge, its highest average accuracy was 76.7% (using the\n          final decision tree created for testing).\n \n    2. Matheus,~C.~J. (1990). Adding domain knowledge to SBL through\n       feature construction.  In {it Proceedings of the Eighth National\n       Conference on Artificial Intelligence} (pp. 803--808). \n       Boston, MA: AAAI Press.\n       -- Similar experiments with CITRE, includes learning curves up\n          to 500-instance training sets but used _all_ instances in the\n          database for testing.  Accuracies reached above 90%, but specific\n          values are not given (see Chris\'s dissertation for more details).\n \n    3. Aha,~D.~W. (1991). Incremental constructive induction: An instance-based\n       approach.  In {it Proceedings of the Eighth International Workshop\n       on Machine Learning} (pp. 117--121).  Evanston, ILL: Morgan Kaufmann.\n       -- Used 70% for training, 30% of the instances for testing, evaluated\n          over 10 trials.  Results reported for six algorithms:\n          -- NewID:   84.0%\n          -- CN2:     98.1%  \n          -- MBRtalk: 88.4%\n          -- IB1:     98.1% \n          -- IB3:     82.0%\n          -- IB3-CI:  99.1%\n       -- Results also reported when adding an additional 10 irrelevant \n          ternary-valued attributes; similar _relative_ results except that\n          IB1\'s performance degraded more quickly than the others.\n \n 4. Relevant Information:\n \n    This database encodes the complete set of possible board configurations\n    at the end of tic-tac-toe games, where \"x\" is assumed to have played\n    first.  The target concept is \"win for x\" (i.e., true when \"x\" has one\n    of 8 possible ways to create a \"three-in-a-row\").  \n \n    Interestingly, this raw database gives a stripped-down decision tree\n    algorithm (e.g., ID3) fits.  However, the rule-based CN2 algorithm, the\n    simple IB1 instance-based learning algorithm, and the CITRE \n    feature-constructing decision tree algorithm perform well on it.\n \n 5. Number of Instances: 958 (legal tic-tac-toe endgame boards)\n \n 6. Number of Attributes: 9, each corresponding to one tic-tac-toe square\n \n 7. Attribute Information: (x=player x has taken, o=player o has taken, b=blank)\n \n     1. top-left-square: {x,o,b}\n     2. top-middle-square: {x,o,b}\n     3. top-right-square: {x,o,b}\n     4. middle-left-square: {x,o,b}\n     5. middle-middle-square: {x,o,b}\n     6. middle-right-square: {x,o,b}\n     7. bottom-left-square: {x,o,b}\n     8. bottom-middle-square: {x,o,b}\n     9. bottom-right-square: {x,o,b}\n    10. Class: {positive,negative}\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: About 65.3% are positive (i.e., wins for \"x\")\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:22:59', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/50/dataset_50_tic-tac-toe.arff', 'true', 25, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:22:59'),
-(26, 1, 0, 'vehicle', '1', '1', '**Author**:  Peter Mowforth  \n**Source**: UCI -   \n**Please cite**: Siebert,JP. Turing Institute Research Memorandum TIRM-87-018 \"Vehicle Recognition Using Rule Based Methods\" (March 1987)  \n\n NAME\n         vehicle silhouettes\n \n PURPOSE\n         to classify a given silhouette as one of four types of vehicle,\n         using  a set of features extracted from the silhouette. The\n         vehicle may be viewed from one of many different angles.  \n \n PROBLEM TYPE\n         classification\n         \n SOURCE\n         Drs.Pete Mowforth and Barry Shepherd\n         Turing Institute\n         George House\n         36 North Hanover St.\n         Glasgow\n         G1 2AD\n \n CONTACT\n         Alistair Sutherland\n         Statistics Dept.\n         Strathclyde University\n         Livingstone Tower\n         26 Richmond St.\n         GLASGOW G1 1XH\n         Great Britain\n         \n         Tel: 041 552 4400 x3033\n         \n         Fax: 041 552 4711 \n         \n         e-mail: alistair@uk.ac.strathclyde.stams\n \n HISTORY\n         This data was originally gathered at the TI in 1986-87 by\n         JP Siebert. It was partially financed by Barr and Stroud Ltd.\n         The original purpose was to find a method of distinguishing\n         3D objects within a 2D image by application of an ensemble of\n         shape feature extractors to the 2D silhouettes of the objects.\n         Measures of shape features extracted from example silhouettes\n         of objects to be discriminated were used to generate a class-\n         ification rule tree by means of computer induction.\n          This object recognition strategy was successfully used to \n         discriminate between silhouettes of model cars, vans and buses\n         viewed from constrained elevation but all angles of rotation.\n          The rule tree classification performance compared favourably\n         to MDC (Minimum Distance Classifier) and k-NN (k-Nearest Neigh-\n         bour) statistical classifiers in terms of both error rate and\n         computational efficiency. An investigation of these rule trees\n         generated by example indicated that the tree structure was \n         heavily influenced by the orientation of the objects, and grouped\n         similar object views into single decisions.\n \n DESCRIPTION\n          The features were extracted from the silhouettes by the HIPS\n         (Hierarchical Image Processing System) extension BINATTS, which \n         extracts a combination of scale independent features utilising\n         both classical moments based measures such as scaled variance,\n         skewness and kurtosis about the major/minor axes and heuristic\n         measures such as hollows, circularity, rectangularity and\n         compactness.\n          Four \"Corgie\" model vehicles were used for the experiment:\n         a double decker bus, Cheverolet van, Saab 9000 and an Opel Manta 400.\n         This particular combination of vehicles was chosen with the \n         expectation that the bus, van and either one of the cars would\n         be readily distinguishable, but it would be more difficult to\n         distinguish between the cars.\n          The images were acquired by a camera looking downwards at the\n         model vehicle from a fixed angle of elevation (34.2 degrees\n         to the horizontal). The vehicles were placed on a diffuse\n         backlit surface (lightbox). The vehicles were painted matte black\n         to minimise highlights. The images were captured using a CRS4000\n         framestore connected to a vax 750. All images were captured with\n         a spatial resolution of 128x128 pixels quantised to 64 greylevels.\n         These images were thresholded to produce binary vehicle silhouettes,\n         negated (to comply with the processing requirements of BINATTS) and\n         thereafter subjected to shrink-expand-expand-shrink HIPS modules to\n         remove \"salt and pepper\" image noise.\n          The vehicles were rotated and their angle of orientation was measured\n         using a radial graticule beneath the vehicle. 0 and 180 degrees\n         corresponded to \"head on\" and \"rear\" views respectively while 90 and\n         270 corresponded to profiles in opposite directions. Two sets of\n         60 images, each set covering a full 360 degree rotation, were captured\n         for each vehicle. The vehicle was rotated by a fixed angle between \n         images. These datasets are known as e2 and e3 respectively.\n          A further two sets of images, e4 and e5, were captured with the camera \n         at elevations of 37.5 degs and 30.8 degs respectively. These sets\n         also contain 60 images per vehicle apart from e4.van which contains\n         only 46 owing to the difficulty of containing the van in the image\n         at some orientations.\n \n ATTRIBUTES\n         \n         COMPACTNESS     (average perim)**2/area\n         \n         CIRCULARITY     (average radius)**2/area\n         \n         DISTANCE CIRCULARITY    area/(av.distance from border)**2\n         \n         RADIUS RATIO    (max.rad-min.rad)/av.radius\n         \n         PR.AXIS ASPECT RATIO    (minor axis)/(major axis)\n         \n         MAX.LENGTH ASPECT RATIO (length perp. max length)/(max length)\n         \n         SCATTER RATIO   (inertia about minor axis)/(inertia about major axis)\n         \n         ELONGATEDNESS           area/(shrink width)**2\n         \n         PR.AXIS RECTANGULARITY  area/(pr.axis length*pr.axis width)\n         \n         MAX.LENGTH RECTANGULARITY area/(max.length*length perp. to this)\n         \n         SCALED VARIANCE         (2nd order moment about minor axis)/area\n         ALONG MAJOR AXIS\n         \n         SCALED VARIANCE         (2nd order moment about major axis)/area\n         ALONG MINOR AXIS \n         \n         SCALED RADIUS OF GYRATION       (mavar+mivar)/area\n         \n         SKEWNESS ABOUT  (3rd order moment about major axis)/sigma_min**3\n         MAJOR AXIS\n         \n         SKEWNESS ABOUT  (3rd order moment about minor axis)/sigma_maj**3\n         MINOR AXIS\n                 \n         KURTOSIS ABOUT  (4th order moment about major axis)/sigma_min**4\n         MINOR AXIS  \n                 \n         KURTOSIS ABOUT  (4th order moment about minor axis)/sigma_maj**4\n         MAJOR AXIS\n         \n         HOLLOWS RATIO   (area of hollows)/(area of bounding polygon)\n         \n          Where sigma_maj**2 is the variance along the major axis and\n         sigma_min**2 is the variance along the minor axis, and\n         \n         area of hollows= area of bounding poly-area of object \n         \n          The area of the bounding polygon is found as a side result of\n         the computation to find the maximum length. Each individual\n         length computation yields a pair of calipers to the object\n         orientated at every 5 degrees. The object is propagated into\n         an image containing the union of these calipers to obtain an\n         image of the bounding polygon. \n         \n NUMBER OF CLASSES\n \n         4       OPEL, SAAB, BUS, VAN\n \n NUMBER OF EXAMPLES\n \n                 Total no. = 946\n                 \n                 No. in each class\n                 \n                   opel 240\n                   saab 240\n                   bus  240\n                   van  226\n                 \n                 \n                 100 examples are being kept by Strathclyde for validation.\n                 So StatLog partners will receive 846 examples.\n \n NUMBER OF ATTRIBUTES\n \n                 No. of atts. = 18', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:23:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/54/dataset_54_vehicle.arff', 'true', 26, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:23:10'),
-(27, 1, 0, 'waveform-5000', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Waveform Database Generator (written in C)\n  \n 2. Source:\n    (a) Breiman,L., Friedman,J.H., Olshen,R.A., & Stone,C.J. (1984). \n        Classification and Regression Trees.  Wadsworth International\n        Group: Belmont, California.  (see pages 43-49).\n    (b) Donor: David Aha \n    (c) Date: 11/10/1988\n \n 3. Past Usage:\n      1. CART book (above):\n         -- Optimal Bayes classification rate: 86% accuracy\n         -- CART decision tree algorithm: 72%\n         -- Nearest Neighbor Algorithm: 78%\n            --	300 training and 5000 test instances\n \n 4. Relevant Information:\n      -- 3 classes of waves\n      -- 21 attributes, all of which include noise\n      -- See the book for details (49-55, 169)\n      -- waveform.data.Z contains 5000 instances\n \n 5. Number of Instances: chosen by user\n \n 6. Number of Attributes:\n     -- 21 attributes with continuous values between 0 and 6\n \n 7. Attribute Information:\n     -- Each class is generated from a combination of 2 of 3 \"base\" waves\n     -- Each instance is generated f added noise (mean 0, variance 1) in \n        each attribute\n      -- See the book for details (49-55, 169)\n     \n 8. Missing Attribute Values: none\n \n 9. Class Distribution: 33% for each of 3 classes', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:23:37', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/60/dataset_60_waveform-5000.arff', 'true', 27, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:23:37'),
-(28, 1, 0, 'electricity', '1', '1', '**Author**: M. Harries, J. Gama, A. Bifet  \n**Source**: [Gama](http://www.inescporto.pt/~jgama/ales/ales_5.html) - 2009  \n**Please cite**:   \n\n**Electricity** is a widely used dataset described by M. Harries and analysed by Gama. This data was collected from the Australian New South Wales Electricity Market. In this market, prices are not fixed and are affected by demand and supply of the market. They are set every five minutes. The ELEC dataset contains 45, 312 instances. The class label identifies the change of the price relative to a moving average of the last 24 hours.', 'ARFF', NULL, NULL, NULL, '2014-04-10 02:42:23', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/2419/electricity-normalized.arff', 'true', 28, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-10 02:42:23'),
-(29, 1, 0, 'satimage', '1', '1', '**Author**: Ashwin Srinivasan, Department of Statistics and Data Modeling, University of Strathclyde \r\n**Source**: https://archive.ics.uci.edu/ml/datasets/Statlog+(Landsat+Satellite)\r\n**Please cite**:   \r\n\r\nThe database consists of the multi-spectral values of pixels in 3x3 neighbourhoods in a satellite image, and the classification associated with the central pixel in each neighbourhood. The aim is to predict this classification, given the multi-spectral values. In the sample database, the class of a pixel is coded as a number.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:14:59', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3619/dataset_186_satimage.arff', 'true', 29, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:14:59'),
-(30, 1, 0, 'abalone', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Abalone data\n \n 2. Sources:\n \n    (a) Original owners of database:\n 	Marine Resources Division\n 	Marine Research Laboratories - Taroona\n 	Department of Primary Industry and Fisheries, Tasmania\n 	GPO Box 619F, Hobart, Tasmania 7001, Australia\n 	(contact: Warwick Nash +61 02 277277, wnash@dpi.tas.gov.au)\n \n    (b) Donor of database:\n 	Sam Waugh (Sam.Waugh@cs.utas.edu.au)\n 	Department of Computer Science, University of Tasmania\n 	GPO Box 252C, Hobart, Tasmania 7001, Australia\n \n    (c) Date received: December 1995\n \n \n 3. Past Usage:\n \n    Sam Waugh (1995) \"Extending and benchmarking Cascade-Correlation\", PhD\n    thesis, Computer Science Department, University of Tasmania.\n \n    -- Test set performance (final 1044 examples, first 3133 used for training):\n 	24.86% Cascade-Correlation (no hidden nodes)\n 	26.25% Cascade-Correlation (5 hidden nodes)\n 	21.5%  C4.5\n 	 0.0%  Linear Discriminate Analysis\n 	 3.57% k=5 Nearest Neighbour\n       (Problem encoded as a classification task)\n \n    -- Data set samples are highly overlapped.  Further information is required\n 	to separate completely using affine combinations.  Other restrictions\n 	to data set examined.\n \n    David Clark, Zoltan Schreter, Anthony Adams \"A Quantitative Comparison of\n    Dystal and Backpropagation\", submitted to the Australian Conference on\n    Neural Networks (ACNN\'96). Data set treated as a 3-category classification\n    problem (grouping ring classes 1-8, 9 and 10, and 11 on).\n \n    -- Test set performance (3133 training, 1044 testing as above):\n 	64%    Backprop\n 	55%    Dystal\n    -- Previous work (Waugh, 1995) on same data set:\n 	61.40% Cascade-Correlation (no hidden nodes)\n 	65.61% Cascade-Correlation (5 hidden nodes)\n 	59.2%  C4.5\n 	32.57% Linear Discriminate Analysis\n 	62.46% k=5 Nearest Neighbour\n \n \n 4. Relevant Information Paragraph:\n \n    Predicting the age of abalone from physical measurements.  The age of\n    abalone is determined by cutting the shell through the cone, staining it,\n    and counting the number of rings through a microscope -- a boring and\n    time-consuming task.  Other measurements, which are easier to obtain, are\n    used to predict the age.  Further information, such as weather patterns\n    and location (hence food availability) may be required to solve the problem.\n \n    From the original data examples with missing values were removed (the\n    majority having the predicted value missing), and the ranges of the\n    continuous values have been scaled for use with an ANN (by dividing by 200).\n \n    Data comes from an original (non-machine-learning) study:\n \n 	Warwick J Nash, Tracy L Sellers, Simon R Talbot, Andrew J Cawthorn and\n 	Wes B Ford (1994) \"The Population Biology of Abalone (_Haliotis_\n 	species) in Tasmania. I. Blacklip Abalone (_H. rubra_) from the North\n 	Coast and Islands of Bass Strait\", Sea Fisheries Division, Technical\n 	Report No. 48 (ISSN 1034-3288)\n \n \n 5. Number of Instances: 4177\n \n \n 6. Number of Attributes: 8\n \n \n 7. Attribute information:\n \n    Given is the attribute name, attribute type, the measurement unit and a\n    brief description.  The number of rings is the value to predict: either\n    as a continuous value or as a classification problem.\n \n 	Name		Data Type	Meas.	Description\n 	----		---------	-----	-----------\n 	Sex		nominal			M, F, and I (infant)\n 	Length		continuous	mm	Longest shell measurement\n 	Diameter	continuous	mm	perpendicular to length\n 	Height		continuous	mm	with meat in shell\n 	Whole weight	continuous	grams	whole abalone\n 	Shucked weight	continuous	grams	weight of meat\n 	Viscera weight	continuous	grams	gut weight (after bleeding)\n 	Shell weight	continuous	grams	after being dried\n 	Rings		integer			+1.5 gives the age in years\n \n    Statistics for numeric domains:\n \n 		Length	Diam	Height	Whole	Shucked	Viscera	Shell	Rings\n 	Min	0.075	0.055	0.000	0.002	0.001	0.001	0.002	    1\n 	Max	0.815	0.650	1.130	2.826	1.488	0.760	1.005	   29\n 	Mean	0.524	0.408	0.140	0.829	0.359	0.181	0.239	9.934\n 	SD	0.120	0.099	0.042	0.490	0.222	0.110	0.139	3.224\n 	Correl	0.557	0.575	0.557	0.540	0.421	0.504	0.628	  1.0\n \n \n 8. Missing Attribute Values: None\n \n \n 9. Class Distribution:\n \n 	Class	Examples\n 	-----	--------\n 	1	1\n 	2	1\n 	3	15\n 	4	57\n 	5	115\n 	6	259\n 	7	391\n 	8	568\n 	9	689\n 	10	634\n 	11	487\n 	12	267\n 	13	203\n 	14	126\n 	15	103\n 	16	67\n 	17	58\n 	18	42\n 	19	32\n 	20	26\n 	21	14\n 	22	6\n 	23	9\n 	24	2\n 	25	1\n 	26	1\n 	27	2\n 	29	1\n 	-----	----\n 	Total	4177\n \n Num Instances:     4177\n Num Attributes:    9\n Num Continuous:    8 (Int 1 / Real 7)\n Num Discrete:      1\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'Sex\'                     Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n   2 \'Length\'                  Real   0%   0% 100%     0 /  0%   134 /  3%   0% \n   3 \'Diameter\'                Real   0%   0% 100%     0 /  0%   111 /  3%   0% \n   4 \'Height\'                  Real   0%   0% 100%     0 /  0%    51 /  1%   0% \n   5 \'Whole weight\'            Real   0%   0% 100%     0 /  0%  2429 / 58%  31% \n   6 \'Shucked weight\'          Real   0%   0% 100%     0 /  0%  1515 / 36%  10% \n   7 \'Viscera weight\'          Real   0%   0% 100%     0 /  0%   880 / 21%   3% \n   8 \'Shell weight\'            Real   0%   0% 100%     0 /  0%   926 / 22%   8% \n   9 \'Class_Rings\'             Int    0% 100%   0%     0 /  0%    28 /  1%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:15:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3620/dataset_187_abalone.arff', 'true', 30, 'Class_number_of_rings', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:15:04'),
-(31, 1, 0, 'eucalyptus', '1', '1', '**Author**: Bruce Bulloch  \n**Source**: [WEKA Dataset Collection](http://www.cs.waikato.ac.nz/ml/weka/datasets.html)   \n**Please cite**:   \n\n**Eucalyptus Soil Conservation**  \nThe objective was to determine which seedlots in a species are best for soil conservation in seasonally dry hill country. Determination is found by measurement of height, diameter by height, survival, and other contributing factors. \n \nIt is important to note that eucalypt trial methods changed over time; earlier trials included mostly 15 - 30cm tall seedling grown in peat plots and the later trials have included mostly three replications of eight trees grown. This change may contribute to less significant results.\n\nExperimental data recording procedures which require noting include:\n - instances with no data recorded due to experimental recording procedures\n   require that the absence of a species from one replicate at a site was\n   treated as a missing value, but if absent from two or more replicates at a\n   site the species was excluded from the site\'s analyses.\n - missing data for survival, vigour, insect resistance, stem form, crown form\n   and utility especially for the data recorded at the Morea Station; this \n   could indicate the death of species in these areas or a lack in collection\n   of data.  \n\n Attribute Information:  \n  1.  Abbrev - site abbreviation - enumerated\n  2.  Rep - site rep - integer\n  3.  Locality - site locality in the North Island - enumerated\n  4.  Map_Ref - map location in the North Island - enumerated\n  5.  Latitude - latitude approximation - enumerated\n  6.  Altitude - altitude approximation - integer\n  7.  Rainfall - rainfall (mm pa) - integer\n  8.  Frosts - frosts (deg. c) - integer\n  9.  Year - year of planting - integer\n  10. Sp - species code - enumerated\n  11. PMCno - seedlot number - integer\n  12. DBH - best diameter base height (cm) - real\n  13. Ht - height (m) - real\n  14. Surv - survival - integer\n  15. Vig - vigour - real\n  16. Ins_res - insect resistance - real\n  17. Stem_Fm - stem form - real\n  18. Crown_Fm - crown form - real\n  19. Brnch_Fm - branch form - real\n  Class:\n  20. Utility - utility rating - enumerated\n\n Class Distribution:\n  none    - 180\n  low     - 107\n  average - 130\n  good    - 214\n  best    - 105\n\n Contact: Bruce Bulloch, 128 Cook Street, Palmerston North, New Zealand', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:15:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3625/dataset_194_eucalyptus.arff', 'true', 31, 'Utility', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:15:28'),
-(35, 2, 0, 'monks-problems-1', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 1\n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks. \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA  E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992\n\n4. Relevant Information:  The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991.  One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users.  There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.\n\nAttribute information: \n1. class: 0, 1  \n2. a1: 1, 2, 3 \n3. a2: 1, 2, 3 \n4. a3: 1, 2 \n5. a4: 1, 2, 3 \n6. a5: 1, 2, 3, 4 \n7. a6: 1, 2 8. \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, 'October 1992', '2014-08-26 17:11:18', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52236/phpAyyBys', 'true', 35, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:11:18'),
-(36, 2, 0, 'monks-problems-2', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 2  \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992  \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.  \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2 \n\n8.   Target Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, 'October 1992', '2014-08-26 17:29:02', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52237/php4fATLZ', 'true', 36, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:29:02'),
-(37, 2, 0, 'monks-problems-3', '1', NULL, '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) -   \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 3   \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992   \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.   \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2  \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 'ARFF', '\"Sebastian Thrun\"', NULL, NULL, '2014-08-26 17:41:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52238/phphZierv', 'true', 37, 'class', NULL, NULL, NULL, 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems', NULL, '2014-08-26 17:41:07'),
-(38, 2, NULL, 'JapaneseVowels', '1', NULL, '**Author**: Mineichi Kudo, Jun Toyama, Masaru Shimbo ({mine,jun,shimbo}@main.eng.hokudai.ac.jp)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Japanese+Vowels) - 2000  \n**Please cite**:   \n\n**Japanese vowels**  \nThis dataset records 640 time series of 12 LPC cepstrum coefficients taken from nine male speakers.\n\nThe data was collected for examining our newly developed classifier for multidimensional curves (multidimensional time series). Nine male speakers uttered two Japanese vowels /ae/ successively. For each utterance, with the analysis parameters described below, we applied 12-degree linear prediction analysis to it to obtain a discrete-time series with 12 LPC cepstrum coefficients. This means that one utterance by a speaker forms a time series whose length is in the range 7-29 and each point of a time series is of 12 features (12 coefficients).\n\nSimilar data are available for different utterances /ei/, /iu/, /uo/, /oa/ in addition to /ae/. Please contact the donor if you are interested in using this data.\n\nThe number of the time series is 640 in total. We used one set of 270 time series for training and the other set of 370 time series for testing.\n\nAnalysis parameters:  \n* Sampling rate : 10kHz\n* Frame length : 25.6 ms\n* Shift length : 6.4ms\n* Degree of LPC coefficients : 12\n\nEach line represents 12 LPC coefficients in the increasing order separated by spaces. This corresponds to one analysis\nframe. Lines are organized into blocks, which are a set of 7-29 lines separated by blank lines and corresponds to a single speech utterance of /ae/ with 7-29 frames.\n\nEach speaker is a set of consecutive blocks. In ae.train there are 30 blocks for each speaker. Blocks 1-30 represent speaker 1, blocks 31-60 represent speaker 2, and so on up to speaker 9. In ae.test, speakers 1 to 9 have the corresponding number of blocks: 31 35 88 44 29 24 40 50 29. Thus, blocks 1-31 represent speaker 1 (31 utterances of /ae/), blocks 32-66 represent speaker 2 (35 utterances of /ae/), and so on.\n\n**Past Usage**\n\nM. Kudo, J. Toyama and M. Shimbo. (1999). \"Multidimensional Curve Classification Using Passing-Through Regions\". Pattern Recognition Letters, Vol. 20, No. 11--13, pages 1103--1111.\n\nIf you publish any work using the dataset, please inform the donor. Use for commercial purposes requires donor permission.\n\nReferences  \n\n1. http://ips9.main.eng.hokudai.ac.jp/index_e.html\n2. mailto:mine@main.eng.hokudai.ac.jp\n3. mailto:jun@main.eng.hokudai.ac.jp\n4. mailto:shimbo@main.eng.hokudai.ac.jp\n5. http://kdd.ics.uci.edu/\n6. http://www.ics.uci.edu/\n7. http://www.uci.edu/', 'ARFF', NULL, NULL, NULL, '2014-09-27 10:56:03', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52415/JapaneseVowels.arff', 'true', 38, 'speaker', NULL, NULL, NULL, 'public', NULL, NULL, 'set target feature', '2014-09-27 11:19:36'),
-(39, 2, NULL, 'synthetic_control', '1', NULL, '**Author**: Dr Robert Alcock (rob@skyblue.csd.auth.gr)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Synthetic+Control+Chart+Time+Series) - 1999  \n**Please cite**:   \n\n**Synthetic Control Chart Time Series**  \nThis data consists of synthetically generated control charts. This dataset contains 600 examples of control charts synthetically generated by the process in Alcock and Manolopoulos (1999). There are six different classes of control charts:\n\n1. Normal\n2. Cyclic\n3. Increasing trend\n4. Decreasing trend\n5. Upward shift\n6. Downward shift\n\n**Past Usage**  \nAlcock R.J. and Manolopoulos Y. Time-Series Similarity Queries Employing a Feature-Based Approach. 7th Hellenic Conference on\nInformatics. August 27-29. Ioannina,Greece 1999.\n\n**References and Further Information**  \nD.T. Pham and A.B. Chan \"Control Chart Pattern Recognition using a New Type of Self Organizing Neural Network\" Proc. Instn, Mech, Engrs. Vol 212, No 1, pp 115-127 1998.\n\nReferences  \n\n1. http://skyblue.csd.auth.gr/~rob/\n2. mailto:rob@skyblue.csd.auth.gr\n3. http://kdd.ics.uci.edu/\n4. http://www.ics.uci.edu/\n5. http://www.uci.edu/', 'ARFF', NULL, NULL, NULL, '2014-09-27 10:56:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52417/synthetic_control.arff', 'true', 39, 'class', 'index', NULL, NULL, 'public', NULL, NULL, 'set index feature', '2014-09-27 11:35:53'),
-(40, 2, NULL, 'irish', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nIrish Educational Transitions Data\n\nBelow are shown data on educational transitions for a sample of 500\nIrish schoolchildren aged 11 in 1967. The data were collected by\nGreaney and Kelleghan (1984), and reanalyzed by Raftery and Hout (1985, 1993).\n\nThe data were also used, in a simplified form, as an example to illustrate\nBayesian model selection methods by Raftery (1988) and Kass and Raftery (1993).\nIn that simplified form, primary terminal leavers and cases with any missing\ndata were removed, leaving 441 cases. The Leaving Certificate variable\nwas used as the dependent variable in a logistic regression analysis.\n\nThe variables shown are as follows:\n\n1. Sex: 1=male; 2=female.\n\n2. DVRT (Drumcondra Verbal Reasoning Test Score).\n\n3. Educational level attained:\n1		Primary terminal leaver\n2	 	Junior cycle incomplete: vocational school\n3		Junior cycle incomplete: secondary school\n4		Junior cycle terminal leaver: vocational school\n5		Junior cycle terminal leaver: secondary school\n6		Senior cycle incomplete: vocational school\n7		Senior cycle incomplete: secondary school\n8		Senior cycle terminal leaver: vocational school\n9		Senior cycle terminal leaver: secondary school\n10		3rd level incomplete\n11		3rd level complete\n\n4. Leaving Certificate. 1 if Leaving Certificate not taken; 2 if taken.\n\n5. Prestige score for father\'s occupation\n(calculated by Raftery and Hout, 1985). 0 if missing.\n\n6. Type of school: 1=secondary; 2=vocational; 9=primary terminal leaver.\n\nREFERENCES\n\nGreaney, V. and Kelleghan, T. (1984). Equality of Opportunity in Irish\nSchools. Dublin: Educational Company.\n\nKass, R.E. and Raftery, A.E. (1993). Bayes factors and model uncertainty.\nTechnical Report no. 254, Department of Statistics, University of Washington.\nRevised version to appear in Journal of the American Statistical\nAssociation.\n\nRaftery, A.E. (1988). Approximate Bayes factors for generalized linear models.\nTechnical Report no. 121, Department of Statistics, University of Washington.\n\nRaftery, A.E. and Hout, M. (1985). Does Irish education approach the\nmeritocratic ideal? A logistic analysis.\nEconomic and Social Review, 16, 115-140.\n\nRaftery, A.E. and Hout, M. (1993). Maximally maintained inequality:\nExpansion, reform and opportunity in Irish schools.\nSociology of Education, 66, 41-62.\n\n\nOWNERSHIP STATEMENT\n\nThis data belongs to Vincent Greaney and Thomas Kelleghan,\nEducational Research Centre, St. Patrick\'s College, Drumcondra,\nDublin 9, Ireland, who retain the copyright.\n\nIn the form given here, it may be used solely as an example for research\non the development of statistical methods. For any other use of the data,\npermission must be obtained from the owners.  Subject to this statement,\npermission is hereby given to StatLib to distribute this data freely.\n\nSubmitted by Adrian Raftery (raftery@stat.washington.edu).\n\nCopyright 1984 Vincent Greaney and Thomas Kelleghan.\n\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: 4', 'ARFF', NULL, NULL, NULL, '2014-09-28 23:50:51', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52563/irish.arff', 'true', 40, 'Leaving_Certificate', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-09-28 23:50:51'),
-(41, 2, NULL, 'analcatdata_authorship', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nanalcatdata    A collection of data sets used in the book \"Analyzing Categorical Data,\"\nby Jeffrey S. Simonoff, Springer-Verlag, New York, 2003. The submission\nconsists of a zip file containing two versions of each of 84 data sets,\nplus this README file. Each data set is given in comma-delimited ASCII\n(.csv) form, and Microsoft Excel (.xls) form.\n\nNOTICE: These data sets may be used freely for scientific, educational and/or\nnoncommercial purposes, provided suitable acknowledgment is given (by citing\nthe above-named reference).\n\nFurther details concerning the book, including information on statistical software\n(including sample S-PLUS/R and SAS code), are available at the web site\n\nhttp://www.stern.nyu.edu/~jsimonof/AnalCatData\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last\n\n\nNote: Quotes, Single-Quotes and Backslashes were removed, Blanks replaced\nwith Underscores', 'ARFF', NULL, NULL, NULL, '2014-09-28 23:51:06', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52570/analcatdata_authorship.arff', 'true', 41, 'Author', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-09-28 23:51:06'),
-(42, 2, NULL, 'analcatdata_dmft', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nanalcatdata    A collection of data sets used in the book \"Analyzing Categorical Data,\"\nby Jeffrey S. Simonoff, Springer-Verlag, New York, 2003. The submission\nconsists of a zip file containing two versions of each of 84 data sets,\nplus this README file. Each data set is given in comma-delimited ASCII\n(.csv) form, and Microsoft Excel (.xls) form.\n\nNOTICE: These data sets may be used freely for scientific, educational and/or\nnoncommercial purposes, provided suitable acknowledgment is given (by citing\nthe above-named reference).\n\nFurther details concerning the book, including information on statistical software\n(including sample S-PLUS/R and SAS code), are available at the web site\n\nhttp://www.stern.nyu.edu/~jsimonof/AnalCatData\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last\n\n\nNote: Quotes, Single-Quotes and Backslashes were removed, Blanks replaced\nwith Underscores', 'ARFF', NULL, NULL, NULL, '2014-09-28 23:51:25', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52581/analcatdata_dmft.arff', 'true', 42, 'Prevention', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-09-28 23:51:25'),
-(43, 2, NULL, 'profb', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nPRO FOOTBALL SCORES  (raw data appears after the description below)\n\nHow well do the oddsmakers of Las Vegas predict the outcome of\nprofessional football games?  Is there really a home field advantage - if\nso how large is it?  Are teams that play the Monday Night game at a\ndisadvantage when they play again the following Sunday?  Do teams benefit\nfrom having a \"bye\" week off in the current schedule?  These questions and\na host of others can be investigated using this data set.\n\nHal Stern from the Statistics Department at Harvard University has\nmade available his compilation of scores for all National Football League\ngames from the 1989, 1990, and 1991 seasons.  Dr. Stern used these data as\npart of his presentation \"Who\'s Number One?\" in the special \"Best of\nBoston\" session at the 1992 Joint Statistics Meetings.\n\nSeveral variables in the data are keyed to the oddsmakers \"point\nspread\" for each game.  The point spread is a value assigned before each\ngame to serve as a handicap for whichever is perceived to be the better\nteam.  Thus, to win against the point spread, the \"favorite\" team must beat\nthe \"underdog\" team by more points than the spread.  The underdog \"wins\"\nagainst the spread if it wins the game outright or manages to lose by fewer\npoints than the spread.  In theory, the point spread should represent the\n\"expert\" prediction as to the game\'s outcome.  In practice, it more usually\ndenotes a point at which an equal amount of money will be wagered both for\nand against the favored team.\n\nRaw data below contains 672 cases (all 224 regular season games in\neach season and informatino on the following 9 varialbes:     .\n\nHome/Away       = Favored team is at home (1) or away (0)\nFavorite Points = Points scored by the favored team\nUnderdog Points = Points scored by the underdog team\nPointspread     = Oddsmaker\'s points to handicap the favored team\nFavorite Name   = Code for favored team\'s name\nUnderdog name   = Code for underdog\'s name\nYear            = 89, 90, or 91\nWeek            = 1, 2, ... 17\nSpecial         = Mon.night (M), Sat. (S), Thur. (H), Sun. night (N)\not - denotes an overtime game\n\n\nData were submitted by: Robin Lock  (rlock@stlawu.bitnet)\nMathematics Department, St. Lawrence University\n\nData were compiled by: Hal Stern, Dept. of Statistics, Harvard University\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: 1', 'ARFF', NULL, NULL, NULL, '2014-09-28 23:51:27', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52582/profb.arff', 'true', 43, 'Home/Away', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-09-28 23:51:27'),
-(44, 2, NULL, 'collins', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nThe following are data used in an analysis of the Brown and Frown corpora for my doctoral dissertation titled ``Variations in Written English: Characterizing Authors\' Rhetorical Language Choices Across Corpora of Published Texts\" (Completed at Carnegie Mellon Univ, 2003).  The source of the corpora was the ICAME CD-ROM  (get info at <http://www.hit.uib.no/icame/cd>).\n\nThe data were generated from the texts using tagging and visualization software, Docuscope.\n\nThe first row is the variable names. The genre of each text (assigned by the Brown corpus compilers) is in \'Genre\' column and the corpus is listed in the \'corpus\' column with 1=Brown and 2=Frown corpus.\n\nThe dataset may be freely used and distributed for non-commercial purposes.\n\nJeff Collins <jeff.collins@acm.org> 11 July 2003\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-09-28 23:51:43', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52590/collins.arff', 'true', 44, 'Corp.Genre', 'Counter', '\"Text\"', NULL, 'public', NULL, NULL, 'attribute counter is a row id', '2015-04-15 17:08:50'),
-(45, 2, NULL, 'sylva_agnostic', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\nModified by TunedIT (converted to ARFF format)\n\n\nSYLVA is the ecology database\n\n\nThe task of SYLVA is to classify forest cover types. The forest cover type for 30 x 30 meter cells is obtained from US Forest Service (USFS) Region 2 Resource Information System (RIS) data. We brought it back to a two-class classification problem (classifying Ponderosa pine vs. everything else). The \"agnostic learning track\" data consists in 216 input variables. Each pattern is composed of 4 records: 2 true records matching the target and 2 records picked at random. Thus 1/2 of the features are distracters.\n\nData type: non-sparse\nNumber of features: 216\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	  805	12281	13086	238271607.00\nValid	   81	 1228	 1309	23817234.00\n\n\nThis dataset contains samples from both training and validation datasets.', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:55:56', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53919/sylva_agnostic.arff', 'true', 45, 'label', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:55:56'),
-(46, 2, NULL, 'gina_agnostic', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\n\nModified by TunedIT (converted to ARFF format)\n\n\nGINA is digit recognition database\n\nThe task of GINA is handwritten digit recognition. For the \"agnostic learning track\" we chose the problem of separating two-digit odd numbers from two-digit even numbers. Only the unit digit is informative for that task, therefore at least 1/2 of the features are distracters. Additionally, the pixels that are almost always blank were removed and the pixel order was randomized to hide the feature identity.  This is a two class classification problem with sparse continuous input variables, in which each class is composed of several clusters. It is a problem with heterogeneous classes.\n\nData type: non-sparse\nNumber of features: 970\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	 1550	 1603	 3153	164947945.00\nValid	  155	  160	  315	16688946.00\n\n\nThis dataset contains samples from both training and validation datasets.', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:56:01', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53921/gina_agnostic.arff', 'true', 46, 'label', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:56:01'),
-(47, 2, NULL, 'ada_agnostic', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\n\nModified by TunedIT (converted to ARFF format)\n\n\nADA is the marketing database\n\nThe task of ADA is to discover high revenue people from census data. This is a two-class classification problem. The raw data from the census bureau is known as the Adult database in the UCI machine-learning repository. The 14 original attributes (features) include age, workclass,  education,\nmarital status, occupation, native country, etc. It contains continuous, binary and categorical features. This dataset is from the \"agnostic learning track\", i.e. has access to a preprocessed numeric representation eliminating categorical variables, but the identity of the features is not revealed.\n\n\n\nData type: non-sparse\nNumber of features: 48\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	 1029	 3118	 4147	6798109.00\nValid	  103	  312	  415	681151.00\n\n\nThis dataset contains samples from both training and validation datasets.', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:56:15', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53926/ada_agnostic.arff', 'true', 47, 'label', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:56:15'),
-(48, 2, NULL, 'mozilla4', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promisedata.org/repository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n(c) 2007  A. Gunes Koru\nContact: gkoru AT umbc DOT edu Phone: +1 (410) 455 8843\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n\n\n1. Title: Recurrent event (defect fix) and size data for Mozilla Classes\nThis one includes a binary attribute (event) to show defect fix.\nThe data is at the \"observation\" level. Each modification made to\na C++ class was entered as an observation. A newly added class\ncreated an observation. The observation period was between\nMay 29, 2002 and Feb 22, 2006.\n\n2. Sources\n(a) Creator: A. Gunes Koru\n(b) Date: February 23, 2007\n(c) Contact: gkoru AT umbc DOT edu Phone: +1 (410) 455 8843\n\n3. Donor: A. Gunes Koru\n\n4. Past Usage: This data set was used for:\n\nA. Gunes Koru, Dongsong Zhang, and Hongfang Liu, \"Modeling the\nEffect of Size on Defect Proneness for Open-Source Software\",\nPredictive Models in Software Engineering Workshop, PROMISE 2007,\nMay 20th 2007, Minneapolis, Minnesota, US.\n\nAbstract:\nQuality is becoming increasingly important with the continuous\nadoption of open-source software.  Previous research has found that\nthere is generally a positive relationship between module size and\ndefect proneness. Therefore, in open-source software development, it\nis important to monitor module size and understand its impact on\ndefect proneness. However, traditional approaches to quality\nmodeling, which measure specific system snapshots and obtain future\ndefect counts, are not well suited because open-source modules\nusually evolve and their size changes over time. In this study, we\nused Cox proportional hazards modeling with recurrent events to\nstudy the effect of class size on defect-proneness in the Mozilla\nproduct. We found that the effect of size was significant, and we\nquantified this effect on defect proneness.\n\nThe full paper can be downloaded from A. Gunes Koru\'s Website\nhttp://umbc.edu/~gkoru\nby following the Publications link or from the Web site of PROMISE 2007.\n\n5. Features:\n\nThis data set is used to create a conditional Cox Proportional\nHazards Model\n\nid: A numeric identification assigned to each separate C++ class\n(Note that the id\'s do not increment from the first to the last\ndata row)\n\nstart: A time infinitesimally greater than the time of the modification\nthat created this observation (practically, modification time). When a\nclass is introduced to a system, a new observation is entered with start=0\n\nend: Either the time of the next modification, or the end of the\nobservation period, or the time of deletion, whichever comes first.\n\nevent: event is set to 1 if a defect fix takes place\nat the time represented by \'end\', or 0 otherwise.  A class deletion\nis handled easily by entering a final observation whose event is set\nto 1 if the class is deleted for corrective maintenance, or 0 otherwise.\n\nsize: It is a time-dependent covariate and its column carries the\nnumber of source Lines of Code of the C++ classes\nat time \'start\'. Blank and comment lines are not counted.\n\nstate: Initially set to 0, and it becomes 1 after the class\nexperiences an event, and remains at 1 thereafter.', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53929/mozilla4.arff', 'true', 48, 'state', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:07'),
-(49, 2, NULL, 'pc4', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE data set made publicly available in order to encourage\nrepeatable, verifiable, refutable, and/or improvable predictive models\nof software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC4/ defect prediction\n\n(c) 2007 : Tim Menzies  : tim@menzies.us\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\nFor more deatils on this data set, see\nhttp://promisedata.org/repository/data/kc2/kc2.arff', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:12', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53932/pc4.arff', 'true', 49, 'c', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:12'),
-(50, 2, NULL, 'pc3', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE data set made publicly available in order to encourage\nrepeatable, verifiable, refutable, and/or improvable predictive models\nof software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC3/software defect prediction\n\n(c) 2007 : Tim Menzies  : tim@menzies.us\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\nFor more deatils on this data set, see\nhttp://promisedata.org/repository/data/kc2/kc2.arff', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:13', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53933/pc3.arff', 'true', 50, 'c', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:13'),
-(51, 2, NULL, 'jm1', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nThis is a PROMISE data set made publicly available in order to encourage repeatable, verifiable, refutable, and/or improvable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please follow the acknowledgment guidelines posted on [the PROMISE repository web page](http://promise.site.uottawa.ca/SERepository).\n\n## Title/Topic\nJM1/software defect prediction\n\n\n## Sources\n* **Creators:**  NASA, then the NASA Metrics Data Program, http://mdp.ivv.nasa.gov. \n* **Contacts:** \n  * Mike Chapman, Galaxy Global Corporation (Robert.Chapman@ivv.nasa.gov) +1-304-367-8341\n  * Pat Callis, NASA, NASA project manager for MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n* **Donor:** Tim Menzies (tim@barmag.net)\n\n* **Date:**  December 2nd, 2004\n\n## Past usage\n**_How Good is Your Blind Spot Sampling Policy?_; 2003; Tim Menzies and Justin S. Di Stefano; 2004 IEEE Conference on High Assurance Software Engineering (http://menzies.us/pdf/03blind.pdf).**\n\n### Results:\n* Very simple learners (ROCKY) perform as well in this domain as more sophisticated methods (e.g. J48, model trees, model trees) for predicting detects\n* Many learners have very low false alarm rates.\n* Probability of detection (PD) rises with effort and rarely rises above it.\n* High PDs are associated with high PFs (probability of failure)\n* PD, PF, effort can change significantly while accuracy remains essentially stable\n* With two notable exceptions, detectors learned from one data set (e.g. KC2) have nearly they same properties when applied to another (e.g. PC2, KC2). Exceptions:\n* LinesOfCode measures generate wider inter-data-set variances;\n* Precision\'s inter-data-set variances vary wildly\n\n**_\"Assessing Predictors of Software Defects\"_, T. Menzies and J. DiStefano and A. Orrego and R. Chapman, 2004,**\n**Proceedings, workshop on Predictive Software Models, Chicago, Available from http://menzies.us/pdf/04psm.pdf.**\n\n### Results:\n* From JM1, Naive Bayes generated PDs of 25% with PF of 20%\n* Naive Bayes out-performs J48 for defect detection\n* When learning on more and more data, little improvement is seen after processing 300 examples.\n* PDs are much higher from data collected below the sub-sub-system level.\n* Accuracy is a surprisingly uninformative measure of success for a defect detector. Two detectors with the same accuracy can have widely varying PDs and PFs.\n\n## Relevant information\n* JM1 is written in \"C\" and is a real-time predictive ground system: Uses simulations to generate predictions\n* Data comes from McCabe and Halstead features extractors of source code.  These features were defined in the 70s in an attempt to objectively characterize code features that are associated with software quality. The nature of association is under dispute.\n\nNotes on McCabe and Halstead follow.\n\n* The McCabe and Halstead measures are \"module\"-based where a \"module\" is the smallest unit of functionality. In C or Smalltalk, \"modules\" would be called \"function\" or \"method\" respectively.\n\n* Defect detectors can be assessed according to the following measures:\n\n    module actually has defects\n    +-------------+------------+\n    |     no      |     yes    |\n    +-----+-------------+------------+\n    classifier predicts no defects |  no |      a      |      b     |\n    +-----+-------------+------------+\n    classifier predicts some defects | yes |      c      |      d     |\n    +-----+-------------+------------+\n\n\n    accuracy                   = acc          = (a+d)/(a+b+c+d\n    probability of detection   = pd  = recall = d/(b+d)\n    probability of false alarm = pf           = c/(a+c)\n    precision                  = prec         = d/(c+d)\n    effort                     = amount of code selected by detector = (c.LOC + d.LOC)/(Total LOC)\n\n\nIdeally, detectors have high PDs, low PFs, and low effort. This ideal state rarely happens:\n\n* PD and effort are linked. The more modules that trigger the detector, the higher the PD. However, effort also gets increases\n* High PD or low PF comes at the cost of high PF or low PD (respectively). This linkage can be seen in a standard receiver operator curve (ROC).  Suppose, for example, LOC> x is used as the detector (i.e. we assume large modules have more errors). LOC > x represents a family of detectors. At x=0, EVERY module is predicted to have errors. This detector has a high PD but also a high false alarm rate. At x=0, NO module is predicted to have errors. This detector has a low false alarm rate but won\'t detect anything at all. At 0<x<1, a set of detectors are generated as shown below:\n\n    pd\n    1 |           x  x  x                KEY:\n    |        x     .                   \".\"  denotes the line PD=PF\n    |     x      .                     \"x\"  denotes the roc curve\n    |   x      .                            for a set of detectors\n    |  x     .\n    | x    .\n    | x  .\n    |x .\n    |x\n    x------------------ pf\n    0                   1\n\nNote that:\n* The only way to make no mistakes (PF=0) is to do nothing (PD=0)\n* The only way to catch more detects is to make more mistakes (increasing PD means increasing PF).\n* Our detector bends towards the \"sweet spot\" of <PD=1,PF=0> but does not reach it.\n* The line pf=pd on the above graph represents the \"no information\" line. If pf=pd then the detector is pretty useless. The better the detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more error-prone.  His metrics therefore reflect the pathways within a code module.\n\n    @Article{mccabe76,\n    title  = \"A Complexity Measure\",\n    author  = \"T.J. McCabe\",\n    pages  = \"308--320\",\n    journal = \"IEEE Transactions on Software Engineering\",\n    year  = \"1976\",\n    volume  = \"2\",\n    month  = \"December\",\n    number  = \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be fault prone. Halstead estimates reading complexity by counting the number of concepts in a module; e.g. number of unique operators.\n\n    @Book{halstead77,\n    Author    = \"M.H. Halstead\",\n    Title    = \"Elements of Software Science\",\n    Publisher = \"Elsevier \",\n    Year    = 1977}\n\nWe study these static code measures since they are useful, easy to use, and widely used:\n\n* USEFUL: E.g. this data set can generate highly accurate predictors for defects\n* EASY TO USE: Static code measures (e.g. lines of code, the McCabe/Halstead measures) can be automatically and cheaply collected.\n* WIDELY USED: Many researchers use static measures to guide software quality predictions (see the reference list in the above \"blind spot\" paper. Verification and validation (V\\&V) textbooks advise using static code complexity measures to decide which modules are worthy of manual inspections.  Further, we know of several large government software contractors that won\'t review software modules _unless_ tools like McCabe predict that they are fault prone.  Hence, defect detectors have a major economic impact when they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely criticized.  Static code measures are hardly a complete characterization of the internals of a function. Fenton offers an insightful example where the same functionality is achieved using different programming language constructs resulting in different static measurements for that module. Fenton uses this example to argue the uselessness of static code measures.\n\n    @Book{fenton97,\n    author    = \"N.E. Fenton and S.L. Pfleeger\",\n    title     = {Software metrics: a Rigorous \\& Practical Approach},\n    publisher = {International Thompson Press},\n    year      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static measures can never be a definite and certain indicator of the presence of a fault.  Rather, defect detectors based on static measures are best viewed as probabilistic statements that the frequency of faults tends to increase in code modules that trigger the detector.  By definition, such probabilistic statements will are not categorical claims with some a non-zero false alarm rate. The research challenge for data miners is to ensure that these false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics: essential complexity, cyclomatic complexity, design complexity and LOC, Lines of Code.\n\n* Cyclomatic Complexity, or \"v(G)\", measures the number of \"linearly independent paths\". A set of paths is said to be linearly independent if no path in the set is a linear combination of any other paths in the set through a program\'s \"flowgraph\". A flowgraph is a directed graph where each node corresponds to a program statement, and each arc indicates the flow of control from one statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\" where \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in the flowgraph, and \"n\" is the number of nodes in the flowgraph. The standard McCabes rules (\"v(G)\">10), are used to identify fault-prone module.\n* Essential Complexity, or \"ev(G)$\" is the extent to which a flowgraph can be \"reduced\" by decomposing all the subflowgraphs of $G$ that are \"D-structured primes\". Such \"D-structured primes\" are also sometimes referred to as \"proper one-entry one-exit subflowgraphs\" (for a more thorough discussion of D-primes, see the Fenton text referenced above). \"ev(G)\" is calculated using \"ev(G)= v(G) - m\" where $m$ is the number of subflowgraphs of \"G\" that are D-structured primes.\n* Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a module\'s reduced flowgraph.  The flowgraph, \"G\", of a module is reduced to eliminate any complexity which does not influence the interrelationship between design modules.  According to McCabe, this complexity measurement reflects the modules calling patterns to its immediate subordinate modules.\n* Lines of code is measured according to McCabe\'s line counting conventions.\n\nThe Halstead falls into three groups: the base measures, the derived measures, and lines of code measures.\n\n* Base measures:\n  * mu1             = number of unique operators\n  * mu2             = number of unique operands\n  * N1              = total occurrences of operators\n  * N2              = total occurrences of operands\n  * length     = N  = N1 + N2\n  * vocabulary = mu = mu1 + mu2\n  * Constants set for each function:\n  * mu1\' =  2 = potential operator count (just the function name and the \"return\" operator)\n  * mu2\'      = potential operand count. (the number of arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\" operators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y), \"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique operands (w,x,y).\n\n* Derived measures:\n  * P = volume = V = N * log2(mu) (the number of mental comparisons needed to write\na program of length N)\n  * V* = volume on minimal implementation = (2 + mu2\')*log2(2 + mu2\')\n  * L  = program length = V*/N\n  * D  = difficulty = 1/L\n  * L\' = 1/D\n  * I  = intelligence = L\'*V\'\n  * E  = effort to write program = V/L\n  * T  = time to write program = E/18 seconds\n\n## Number of instances\n10885\n\n## Number of attributes\n22 (5 different lines of code measure, 3 McCabe metrics, 4 base Halstead measures, 8 derived Halstead measures, a branch-count, and 1 goal field)\n\n## Attribute Information\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. defects         : {false,true} % module has/has not one or more reported defects\n\n## Missing attributes\nNone\n\n## Class Distribution\nThe class value (defects) is discrete\nfalse: 2106 = 19.35%\ntrue:  8779 = 80.65%', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:19', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53936/jm1.arff', 'true', 51, 'defects', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:19'),
-(52, 2, NULL, 'kc2', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: KC2/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From KC2, Naive Bayes generated PDs of 50% with PF of 10%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n-- Data from C++ functions. Science data processing; another part\nof the same project as KC1; different personnel than KC1.  Shared\nsome third-party software libraries with KC1, but no other software\noverlap.\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 522\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. problems        : {no,yes}% module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (problems) is discrete\nyes: 105 = 20.5%\nno:  415 = 79.5%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:36', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53946/kc2.arff', 'true', 52, 'problems', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:36'),
-(53, 2, NULL, 'kc1', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: KC1/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From KC2, Naive Bayes generated PDs of 45% with PF of 10%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n\n-- KC1 is a \"C++\" system implementing storage management for\nreceiving and processing ground data\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 2109\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. problems        : {false,true}% module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (problems) is discrete\nyes:  326 = 15.45%\nno:  1783 = 84.54%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:43', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53950/kc1.arff', 'true', 53, 'defects', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:43'),
-(54, 2, NULL, 'pc1', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC1/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From JM1, Naive Bayes generated PDs of 20% with PF of 25%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n-- Data from C  functions. flight software\nfor earth orbiting satellite.\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 1109\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. defects         : {false,true} % module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (defects) is discrete\nfalse:   77 =  6.94%\ntrue:  1032 = 93.05%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 'ARFF', NULL, NULL, NULL, '2014-10-06 23:57:45', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53951/pc1.arff', 'true', 54, 'defects', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-06 23:57:45'),
-(55, 2, NULL, 'KDDCup09_churn', '1', NULL, '**Author**: Orange Telecom  \n**Source**: [ACM KDD Cup](http://www.sigkdd.org/kddcup/index.php) - 2009  \n**Please cite**: \n\nThe KDD Cup 2009 offers the opportunity to work on large marketing databases from the French Telecom company Orange to predict the propensity of customers to switch provider (churn). \n\nChurn (wikipedia definition): Churn rate is also sometimes called attrition rate. It is one of two primary factors that determine\nthe steady-state level of customers a business will support. In its broadest sense, churn rate is a measure of the number\nof individuals or items moving into or out of a collection over a specific period of time.\n\nThe term is used in many contexts, but is most widely applied in business with respect to a contractual customer base. For instance, it is an important factor for any business with a subscriber-based service model, including mobile telephone networks and pay TV operators. The term is also used to refer to participant turnover in peer-to-peer networks.\n\nThe training set contains 50,000 examples.\nThe first predictive 190 variables are numerical and the last 40 predictive variables are categorical.\nThe last target variable is binary {-1,1}.', 'ARFF', NULL, NULL, NULL, '2014-10-07 00:08:02', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53995/KDDCup09_churn.arff', 'true', 55, 'CHURN', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-07 00:08:02'),
-(56, 2, NULL, 'KDDCup09_upselling', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from ACM KDD Cup (http://www.sigkdd.org/kddcup/index.php)\n\nKDD Cup 2009\nhttp://www.kddcup-orange.com\n\nConverted to ARFF format by TunedIT\nCustomer Relationship Management (CRM) is a key element of modern marketing strategies. The KDD Cup 2009 offers the opportunity to work on large marketing databases from the French Telecom company Orange to predict the propensity of customers to switch provider (churn), buy new products or services (appetency), or buy upgrades or add-ons proposed to them to make the sale more profitable (up-selling).\nThe most practical way, in a CRM system, to build knowledge on customer is to produce scores. A score (the output of a model) is an evaluation for all instances of a target variable to explain (i.e. churn, appetency or up-selling). Tools which produce scores allow to project, on a given population, quantifiable information. The score is computed using input variables which describe instances. Scores are then used by the information system (IS), for example, to personalize the customer relationship. An industrial customer analysis platform able to build prediction models with a very large number of input variables has been developed by Orange Labs. This platform implements several processing methods for instances and variables selection, prediction and indexation based on an efficient model combined with variable selection regularization and model averaging method. The main characteristic of this platform is its ability to scale on very large datasets with hundreds of thousands of instances and thousands of variables. The rapid and robust detection of the variables that have most contributed to the output prediction can be a key factor in a marketing application.\nUp-selling (wikipedia definition): Up-selling is a sales technique whereby a salesman attempts to have the customer purchase more expensive\nitems, upgrades, or other add-ons in an attempt to make a more profitable sale.\nUp-selling usually involves marketing more profitable services or products, but up-selling can also be simply exposing the customer\nto other options he or she may not have considered previously.\nUp-selling can imply selling something additional, or selling something that is more profitable or otherwise preferable for the seller instead of the original sale.\nThe training set contains 50,000 examples.\nThe first predictive 190 variables are numerical and the last 40 predictive variables are categorical.\nThe last target variable is binary {-1,1}.', 'ARFF', NULL, NULL, NULL, '2014-10-07 00:08:27', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53997/KDDCup09_upselling.arff', 'true', 56, 'UPSELLING', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-07 00:08:27'),
-(57, 2, NULL, 'musk', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDataset from the MLRR repository: http://axon.cs.byu.edu:5000/', 'ARFF', NULL, NULL, NULL, '2014-10-07 00:41:54', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/53999/musk.arff', 'true', 57, 'class', 'ID', '\"conformation_name\"', NULL, 'public', NULL, NULL, 'ID is a row id', '2015-04-15 17:37:23'),
-(58, 2, NULL, 'MagicTelescope', '1', NULL, '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDataset from the MLRR repository: http://axon.cs.byu.edu:5000/', 'ARFF', NULL, NULL, NULL, '2014-10-07 00:42:01', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/54003/MagicTelescope.arff', 'true', 58, 'class:', 'ID', NULL, NULL, 'public', NULL, NULL, 'ID is a row id', '2015-04-15 17:41:20'),
-(59, 2, NULL, 'Internet-Advertisements', '1', NULL, '**Author**: Nicholas Kushmerick  \n**Source**: [UCI](http://archive.ics.uci.edu/ml/datasets/Internet+Advertisements) - 1998  \n**Please cite**:   \n\nThis dataset represents a set of possible advertisements on Internet pages. The features encode the geometry of the image (if available) as well as phrases occurring in the URL, the image\'s URL and alt text, the anchor text, and words occurring near the anchor text. The task is to predict whether an image is an advertisement (\"ad\") or not (\"nonad\").\n\nRelevant Papers: N. Kushmerick (1999). \"Learning to remove Internet advertisements\", 3rd Int Conf Autonomous Agents.  \nAvailable at: http://rexa.info/paper/2fdc1cee89b7f4f2c9227d6f5d9b05d22c5ab3e9', 'ARFF', '\"Nicholas Kushmerick\"', NULL, NULL, '2014-10-30 11:15:44', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/116567/phpCzcrGG', 'true', 59, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-10-30 11:15:44'),
-(60, 2, NULL, 'Click_prediction_small', '5', NULL, '**Author**: Tencent Inc.  \n**Source**: [KDD Cup](https://www.kddcup2012.org/) - 2012  \n**Please cite**:   \n\n**0.1% balanced subsample of the original KDD dataset**  \n\nThis data is derived from the 2012 KDD Cup. The data is subsampled to 0.1% of the original number of instances, downsampling the majority class (click=0) so that the target feature is reasonably balanced (5 to 1).\n\nThe data is about advertisements shown alongside search results in a search engine, and whether or not people clicked on these ads. \nThe task is to build the best possible model to predict whether a user will click on a given ad.\n\nA search session contains information on user id, the query issued by the user, ads displayed to the user, and target feature indicating whether a user clicked at least one of the ads in this session. The number of ads displayed to a user in a session is called ‘depth’. The order of an ad in the displayed list is called ‘position’.  An ad is displayed as a short text called ‘title’, followed by a slightly longer text called ’description’, and a URL  called ‘display URL’.   \nTo construct this dataset each session was split into multiple instances. Each instance describes an ad displayed under a certain setting  (‘depth’, ‘position’).  Instances with the same user id, ad id, query, and setting are merged. Each ad and each user have some additional properties located in separate data files that can be looked up using ids in the instances.\n\nThe dataset has the following features:  \n* Click – binary variable indicating whether a user clicked on at least one ad. \n* Impression - the number of search sessions in which AdID was impressed by UserID who issued Query.\n* Url_hash - URL is hashed for anonymity\n* AdID \n* AdvertiserID - some advertisers consistently optimize their ads, so the title and description of their ads are more attractive than those of others’ ads.\n* Depth - number of ads displayed to a user in a session\n* Position - order of an ad in the displayed list\n* QueryID - is the key of the data file \'queryid_tokensid.txt\'. (follow the link to the original KDD Cup page, track 2)\n* KeywordID - is the key of  \'purchasedkeyword_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* TitleID - is the key of \'titleid_tokensid.txt\'\n* DescriptionID - is the key of \'descriptionid_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* UserID – is also the key of \'userid_profile.txt\' (follow the link to the original KDD Cup page, track 2). 0 is a special value denoting that the user could be identified.', 'ARFF', NULL, NULL, NULL, '2014-11-27 01:18:40', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/184157/phpfGCaQC', 'true', 60, 'click', NULL, '\"url_hash\",\"query_id\"', NULL, 'public', NULL, NULL, NULL, '2014-11-27 01:26:36'),
-(103, 1, 0, 'kin8nm', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis is data set is concerned with the forward kinematics of an 8 link\n robot arm. Among the existing variants of this data set we have used\n the variant 8nm, which is known to be highly non-linear and medium\n noisy.\n\n Original source: DELVE repository of data. \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 8192 cases, 9 attributes (0 nominal, 9 continuous).', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3626/dataset_2175_kin8nm.arff', 'true', 103, 'y', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:04'),
-(104, 1, 0, 'mbagrade', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3627/dataset_2176_mbagrade.arff', 'true', 104, 'grade_point_average', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:07'),
-(105, 1, 0, 'wisconsin', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Wisconsin Prognostic Breast Cancer (WPBC)\n \n 2. Source Information\n \n a) Creators: \n \n 	Dr. William H. Wolberg, General Surgery Dept., University of\n 	Wisconsin,  Clinical Sciences Center, Madison, WI 53792\n 	wolberg@eagle.surgery.wisc.edu\n \n 	W. Nick Street, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	street@cs.wisc.edu  608-262-6619\n \n 	Olvi L. Mangasarian, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	olvi@cs.wisc.edu \n \n b) Donor: Nick Street\n \n c) Date: December 1995\n \n 3. Past Usage:\n \n 	Various versions of this data have been used in the following\n 	publications: \n \n 	(i) W. N. Street, O. L. Mangasarian, and W.H. Wolberg. \n 	An inductive learning approach to prognostic prediction. \n 	In A. Prieditis and S. Russell, editors, Proceedings of the\n 	Twelfth International Conference on Machine Learning, pages\n 	522--530, San Francisco, 1995. Morgan Kaufmann.\n \n 	(ii) O.L. Mangasarian, W.N. Street and W.H. Wolberg. \n 	Breast cancer diagnosis and prognosis via linear programming. \n 	Operations Research, 43(4), pages 570-577, July-August 1995. \n \n 	(iii) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computerized breast cancer diagnosis and prognosis from fine\n 	needle aspirates.  Archives of Surgery 1995;130:511-516. \n \n 	(iv) W.H. Wolberg, W.N. Street, and O.L. Mangasarian. \n 	Image analysis and machine learning applied to breast cancer\n 	diagnosis and prognosis. Analytical and Quantitative Cytology\n 	and Histology, Vol. 17 No. 2, pages 77-87, April 1995.\n \n 	(v) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computer-derived nuclear ``grade\'\' and breast cancer prognosis. \n 	Analytical and Quantitative Cytology and Histology, Vol. 17,\n 	pages 257-264, 1995. \n \n See also:\n 	http://www.cs.wisc.edu/~olvi/uwmp/mpml.html\n 	http://www.cs.wisc.edu/~olvi/uwmp/cancer.html\n \n Results:\n \n 	Two possible learning problems:\n \n 	1) Predicting field 2, outcome: R = recurrent, N = nonrecurrent\n 	- Dataset should first be filtered to reflect a particular\n 	endpoint; e.g., recurrences before 24 months = positive,\n 	nonrecurrence beyond 24 months = negative.\n 	- 86.3% accuracy estimated accuracy on 2-year recurrence using\n 	previous version of this data.  Learning method: MSM-T (see\n 	below) in the 4-dimensional space of Mean Texture, Worst Area,\n 	Worst Concavity, Worst Fractal Dimension.\n \n 	2) Predicting Time To Recur (field 3 in recurrent records)\n 	- Estimated mean error 13.9 months using Recurrence Surface\n 	Approximation. (See references (i) and (ii) above)\n \n 4. Relevant information\n \n 	Each record represents follow-up data for one breast cancer\n 	case.  These are consecutive patients seen by Dr. Wolberg\n 	since 1984, and include only those cases exhibiting invasive\n 	breast cancer and no evidence of distant metastases at the\n 	time of diagnosis. \n \n 	The first 30 features are computed from a digitized image of a\n 	fine needle aspirate (FNA) of a breast mass.  They describe\n 	characteristics of the cell nuclei present in the image.\n 	A few of the images can be found at\n 	http://www.cs.wisc.edu/~street/images/\n \n 	The separation described above was obtained using\n 	Multisurface Method-Tree (MSM-T) [K. P. Bennett, \"Decision Tree\n 	Construction Via Linear Programming.\" Proceedings of the 4th\n 	Midwest Artificial Intelligence and Cognitive Science Society,\n 	pp. 97-101, 1992], a classification method which uses linear\n 	programming to construct a decision tree.  Relevant features\n 	were selected using an exhaustive search in the space of 1-4\n 	features and 1-3 separating planes.\n \n 	The actual linear program used to obtain the separating plane\n 	in the 3-dimensional space is that described in:\n 	[K. P. Bennett and O. L. Mangasarian: \"Robust Linear\n 	Programming Discrimination of Two Linearly Inseparable Sets\",\n 	Optimization Methods and Software 1, 1992, 23-34].\n \n 	The Recurrence Surface Approximation (RSA) method is a linear\n 	programming model which predicts Time To Recur using both\n 	recurrent and nonrecurrent cases.  See references (i) and (ii)\n 	above for details of the RSA method. \n \n 	This database is also available through the UW CS ftp server:\n \n 	ftp ftp.cs.wisc.edu\n 	cd math-prog/cpo-dataset/machine-learn/WPBC/\n \n 5. Number of instances: 198\n \n 6. Number of attributes: 34 (ID, outcome, 32 real-valued input features)\n \n 7. Attribute information\n \n 1) ID number\n 2) Outcome (R = recur, N = nonrecur)\n 3) Time (recurrence time if field 2 = R, disease-free time if \n 	field 2	= N)\n 4-33) Ten real-valued features are computed for each cell nucleus:\n \n 	a) radius (mean of distances from center to points on the perimeter)\n 	b) texture (standard deviation of gray-scale values)\n 	c) perimeter\n 	d) area\n 	e) smoothness (local variation in radius lengths)\n 	f) compactness (perimeter^2 / area - 1.0)\n 	g) concavity (severity of concave portions of the contour)\n 	h) concave points (number of concave portions of the contour)\n 	i) symmetry \n 	j) fractal dimension (\"coastline approximation\" - 1)\n \n Several of the papers listed above contain detailed descriptions of\n how these features are computed. \n \n The mean, standard error, and \"worst\" or largest (mean of the three\n largest values) of these features were computed for each image,\n resulting in 30 features.  For instance, field 4 is Mean Radius, field\n 14 is Radius SE, field 24 is Worst Radius.\n \n Values for features 4-33 are recoded with four significant digits.\n \n 34) Tumor size - diameter of the excised tumor in centimeters\n 35) Lymph node status - number of positive axillary lymph nodes\n observed at time of surgery\n \n 8. Missing attribute values: \n 	Lymph node status is missing in 4 cases.\n \n 9. Class distribution: 151 nonrecur, 47 recur\n\n-----------------------------------------------------------------------------------------------------------\n Luis Torgo\'s version: (reconstructed)\n - removed the four instances with unknown values of the last attribute\n - exchanged the attribute position of attributes n.3 (Time) and n.35\n   (Lymph node).\n - removed the attribute outcome as it is the class attribute if the\n   problem is treated as a classification one\n-----------------------------------------------------------------------------------------------------------', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3628/dataset_2177_wisconsin.arff', 'true', 105, 'time', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:10'),
-(106, 1, 0, 'vineyard', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:12', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3629/dataset_2178_vineyard.arff', 'true', 106, 'lugs_1991', 'row_number', NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:12'),
-(107, 1, 0, 'bolts', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n \n SUMMARY:\n \n Data from an experiment on the affects of machine adjustments on\n the time to count bolts.  Data appear as the STATS (Issue 10) Challenge.\n \n DATA:\n \n Submitted by W. Robert Stephenson, Iowa State University\n                             email: wrstephe@iastate.edu\n \n A manufacturer of automotive accessories provides hardware, e.g. nuts,\n bolts, washers and screws, to fasten the accessory to the car or truck.\n Hardware is counted and packaged automatically.  Specifically, bolts\n are dumped into a large metal dish.  A plate that forms the bottom of\n the dish rotates counterclockwise.  This rotation forces bolts to the\n outside of the dish and up along a narrow ledge.  Due to the vibration\n of the dish caused by the spinning bottom plate, some bolts fall off \n the ledge and back into the dish.  The ledge spirals up to a point \n where the bolts are allowed to drop into a pan on a conveyor belt.  \n As a bolt drops, it passes by an electronic eye that counts it.  When \n the electronic counter reaches the preset number of bolts, the\n rotation is stopped and the conveyor belt is moved forward.  \n \n There are several adjustments on the machine that affect its operation.  \n These include; a speed setting that controls the speed of rotation\n (SPEED1) of the plate at the bottom of the dish, a total number of \n bolts (TOTAL) to be counted, a second speed setting (SPEED2) that is \n used to change the speed of rotation (usually slowing it down) for the\n last few bolts, the number of bolts to be counted at this second speed\n (NUMBER2), and the sensitivity of the electronic eye (SENS).  The \n sensitivity setting is to insure that the correct number of bolts are \n counted.  Too few bolts packaged causes customer complaints.  Too many\n bolts packaged increases costs.  For each run conducted in this \n experiment the correct number of bolts was counted.  From an\n engineering standpoint if the correct number of bolts is counted, the \n sensitivity should not affect the time to count bolts.  The measured \n response is the time (TIME), in seconds, it takes to count the desired\n number of bolts.  In order to put times on a equal footing the\n response to be analyzed is the time to count 20 bolts (T20BOLT).\n Below are the data for 40 combinations of settings.  RUN is the order \n in which the data were collected.\n \n Analyze the data.  What adjustments have the greatest effect on the \n time to count 20 bolts?  How would you adjust the machine to get\n the shortest time to count 20 bolts?  Are there any unusual features\n to the data?\n \n The data description and data may be freely used for non-commercial\n purposes and can be freely distributed.  Copyright remains with the\n author and STATS Magazine.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:14', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3630/dataset_2179_bolts.arff', 'true', 107, 'T20BOLT', 'RUN', NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:14'),
-(108, 1, 0, 'cleveland', '1', '1', '**Author**: Andras Janosi, M.D.  \nDonor: David W. Aha (aha@ics.uci.edu)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Heart+Disease) - July 1988  \n**Please cite**: The author request that any publications resulting from the use of the data include the name of the author.\n\n**Heart Disease Databases: Cleveland**  \nThis database contains 76 attributes, but all published experiments refer to using a subset of 14 of them.  In particular, the Cleveland database is the only one that has been used by ML researchers to this date.  The \"goal\" field refers to the presence of heart disease in the patient.  It is integer valued from 0 (no presence) to 4. Experiments with the Cleveland database have concentrated on simply attempting to distinguish presence (values 1,2,3,4) from absence (value 0).  \n    \nThe names and social security numbers of the patients were recently removed from the database, replaced with dummy values.\nOne file has been \"processed\", that one containing the Cleveland database. \n     \nAttribute documentation:  \n>\n       1 id: patient identification number\n       2 ccf: social security number (I replaced this with a dummy value of 0)\n       3 age: age in years\n       4 sex: sex (1 = male; 0 = female)\n       5 painloc: chest pain location (1 = substernal; 0 = otherwise)\n       6 painexer (1 = provoked by exertion; 0 = otherwise)\n       7 relrest (1 = relieved after rest; 0 = otherwise)\n       8 pncaden (sum of 5, 6, and 7)\n       9 cp: chest pain type\n         -- Value 1: typical angina\n         -- Value 2: atypical angina\n         -- Value 3: non-anginal pain\n         -- Value 4: asymptomatic\n      10 trestbps: resting blood pressure (in mm Hg on admission to the \n         hospital)\n      11 htn\n      12 chol: serum cholestoral in mg/dl\n      13 smoke: I believe this is 1 = yes; 0 = no (is or is not a smoker)\n      14 cigs (cigarettes per day)\n      15 years (number of years as a smoker)\n      16 fbs: (fasting blood sugar > 120 mg/dl)  (1 = true; 0 = false)\n      17 dm (1 = history of diabetes; 0 = no such history)\n      18 famhist: family history of coronary artery disease (1 = yes; 0 = no)\n      19 restecg: resting electrocardiographic results\n         -- Value 0: normal\n         -- Value 1: having ST-T wave abnormality (T wave inversions and/or ST \n                     elevation or depression of > 0.05 mV)\n         -- Value 2: showing probable or definite left ventricular hypertrophy\n                     by Estes\' criteria\n      20 ekgmo (month of exercise ECG reading)\n      21 ekgday(day of exercise ECG reading)\n      22 ekgyr (year of exercise ECG reading)\n      23 dig (digitalis used furing exercise ECG: 1 = yes; 0 = no)\n      24 prop (Beta blocker used during exercise ECG: 1 = yes; 0 = no)\n      25 nitr (nitrates used during exercise ECG: 1 = yes; 0 = no)\n      26 pro (calcium channel blocker used during exercise ECG: 1 = yes; 0 = no)\n      27 diuretic (diuretic used used during exercise ECG: 1 = yes; 0 = no)\n      28 proto: exercise protocol\n           1 = Bruce     \n           2 = Kottus\n           3 = McHenry\n           4 = fast Balke\n           5 = Balke\n           6 = Noughton \n           7 = bike 150 kpa min/min  (Not sure if \"kpa min/min\" is what was \n               written!)\n           8 = bike 125 kpa min/min  \n           9 = bike 100 kpa min/min\n          10 = bike 75 kpa min/min\n          11 = bike 50 kpa min/min\n          12 = arm ergometer\n      29 thaldur: duration of exercise test in minutes\n      30 thaltime: time when ST measure depression was noted\n      31 met: mets achieved\n      32 thalach: maximum heart rate achieved\n      33 thalrest: resting heart rate\n      34 tpeakbps: peak exercise blood pressure (first of 2 parts)\n      35 tpeakbpd: peak exercise blood pressure (second of 2 parts)\n      36 dummy\n      37 trestbpd: resting blood pressure\n      38 exang: exercise induced angina (1 = yes; 0 = no)\n      39 xhypo: (1 = yes; 0 = no)\n      40 oldpeak = ST depression induced by exercise relative to rest\n      41 slope: the slope of the peak exercise ST segment\n         -- Value 1: upsloping\n         -- Value 2: flat\n         -- Value 3: downsloping\n      42 rldv5: height at rest\n      43 rldv5e: height at peak exercise\n      44 ca: number of major vessels (0-3) colored by flourosopy\n      45 restckm: irrelevant\n      46 exerckm: irrelevant\n      47 restef: rest raidonuclid (sp?) ejection fraction\n      48 restwm: rest wall (sp?) motion abnormality\n         0 = none\n         1 = mild or moderate\n         2 = moderate or severe\n         3 = akinesis or dyskmem (sp?)\n      49 exeref: exercise radinalid (sp?) ejection fraction\n      50 exerwm: exercise wall (sp?) motion \n      51 thal: 3 = normal; 6 = fixed defect; 7 = reversable defect\n      52 thalsev: not used\n      53 thalpul: not used\n      54 earlobe: not used\n      55 cmo: month of cardiac cath (sp?)  (perhaps \"call\")\n      56 cday: day of cardiac cath (sp?)\n      57 cyr: year of cardiac cath (sp?)\n      58 num: diagnosis of heart disease (angiographic disease status)\n         -- Value 0: < 50% diameter narrowing\n         -- Value 1: > 50% diameter narrowing\n         (in any major vessel: attributes 59 through 68 are vessels)\n      59 lmt\n      60 ladprox\n      61 laddist\n      62 diag\n      63 cxmain\n      64 ramus\n      65 om1\n      66 om2\n      67 rcaprox\n      68 rcadist\n      69 lvx1: not used\n      70 lvx2: not used\n      71 lvx3: not used\n      72 lvx4: not used\n      73 lvf: not used\n      74 cathef: not used\n      75 junk: not used\n      76 name: last name of patient \n         (I replaced this with the dummy string \"name\")', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:17', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3631/dataset_2180_cleveland.arff', 'true', 108, 'num', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:17'),
-(109, 1, 0, 'auto_price', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis data set consists of three types of entities:\n (a) the specification of an auto in terms of various characteristics;\n (b) its assigned insurance risk rating,;\n (c) its normalized losses in use as compared to other cars. \n The second rating corresponds to the degree to which the auto is more risky than its price indicates. Cars are initially\n assigned a risk factor symbol associated with its price. Then, if it is more risky (or less), this symbol is adjusted by\n moving it up (or down) the scale. Actuarians call this process \"symboling\". A value of +3 indicates that the auto is\n risky, -3 that it is probably pretty safe.The third factor is the relative average loss payment per insured vehicle year.\n This value is normalized for all autos within a particular size classification (two-door small, station wagons,\n sports/speciality, etc...), and represents the average loss per car per year.\n - Note: Several of the attributes in the database could be used as a \"class\" attribute.\n The original data (from the UCI repository) (http://www.ics.uci.edu/~mlearn/MLSummary.html) has 205 instances\n described by 26 attributes :\n - 15 continuous\n - 1 integer\n - 10 nominal\n The following provides more information on these attributes:\n \n   1. symboling:                 -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback,convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n The original data also has some missing attribute values denoted by \"?\" : \n \n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4\n \n I\'ve changed the original data in the following way :\n - All instances with unknowns were removed giving 159 instances.\n - The goal variable is \"price\"\n - All nominal attributes (10) were removed.\n \n Original source: UCI machine learning repository. (http://www.ics.uci.edu/~mlearn/MLSummary.html). \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 159 cases; 14 continuous variables; 1 nominal vars..', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:20', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3632/dataset_2181_auto_price.arff', 'true', 109, 'price', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:20'),
-(110, 1, 0, 'autoMpg', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identifier attribute deleted.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n\n 1. Title: Auto-Mpg Data\n \n 2. Sources:\n    (a) Origin:  This dataset was taken from the StatLib library which is\n                 maintained at Carnegie Mellon University. The dataset was \n                 used in the 1983 American Statistical Association Exposition.\n    (c) Date: July 7, 1993\n \n 3. Past Usage:\n     -  See 2b (above)\n     -  Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning.\n        In Proceedings on the Tenth International Conference of Machine \n        Learning, 236-243, University of Massachusetts, Amherst. Morgan\n        Kaufmann.\n \n 4. Relevant Information:\n \n    This dataset is a slightly modified version of the dataset provided in\n    the StatLib library.  In line with the use by Ross Quinlan (1993) in\n    predicting the attribute \"mpg\", 8 of the original instances were removed \n    because they had unknown values for the \"mpg\" attribute.  The original \n    dataset is available in the file \"auto-mpg.data-original\".\n \n    \"The data concerns city-cycle fuel consumption in miles per gallon,\n     to be predicted in terms of 3 multivalued discrete and 5 continuous\n     attributes.\" (Quinlan, 1993)\n \n 5. Number of Instances: 398\n \n 6. Number of Attributes: 9 including the class attribute\n \n 7. Attribute Information:\n \n     1. mpg:           continuous\n     2. cylinders:     multi-valued discrete\n     3. displacement:  continuous\n     4. horsepower:    continuous\n     5. weight:        continuous\n     6. acceleration:  continuous\n     7. model year:    multi-valued discrete\n     8. origin:        multi-valued discrete\n     9. car name:      string (unique for each instance)\n \n 8. Missing Attribute Values:  horsepower has 6 missing values', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:22', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3633/dataset_2182_autoMpg.arff', 'true', 110, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:22'),
-(111, 1, 0, 'cpu_act', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe Computer Activity databases are a collection of computer systems\n activity measures. The data was collected from a Sun Sparcstation\n 20/712 with 128 Mbytes of memory running in a multi-user university\n department. Users would typically be doing a large variety of tasks\n ranging from accessing the internet, editing files or running very\n cpu-bound programs.  The data was collected continuously on two\n separate occasions. On both occassions, system activity was gathered\n every 5 seconds. The final dataset is taken from both occasions with\n equal numbers of observations coming from each collection epoch.\n \n System measures used:\n 1. lread - Reads (transfers per second ) between system memory and user memory.\n 2. lwrite - writes (transfers per second) between system memory and user memory.\n 3. scall - Number of system calls of all types per second.\n 4. sread - Number of system read calls per second.\n 5. swrite - Number of system write calls per second . \n 6. fork - Number of system fork calls per second. \n 7. exec - Number of system exec calls per second. \n 8. rchar - Number of characters transferred per second by system read calls.\n 9. wchar - Number of characters transfreed per second by system write calls. \n 10. pgout - Number of page out requests per second.\n 11. ppgout - Number of pages, paged out per second. \n 12. pgfree - Number of pages per second placed on the free list. \n 13. pgscan - Number of pages checked if they can be freed per second.\n 14. atch - Number of page attaches (satisfying a page fault by reclaiming a page in memory) per second.\n 15. pgin - Number of page-in requests per second.\n 16. ppgin - Number of pages paged in per second.\n 17. pflt - Number of page faults caused by protection errors (copy-on-writes). \n 18. vflt - Number of page faults caused by address translation. \n 19. runqsz - Process run queue size.\n 20. freemem - Number of memory pages available to user processes.\n 21. freeswap - Number of disk blocks available for page swapping. \n 22. usr - Portion of time (%) that cpus run in user mode.\n 23. sys - Portion of time (%) that cpus run in system mode.\n 24. wio - Portion of time (%) that cpus are idle waiting for block IO.\n 25. idle - Portion of time (%) that cpus are otherwise idle.\n \n The two different regression tasks obtained from these databases are:\n \n CompAct \n Predict usr, the portion of time that cpus run in user mode from all attributes 1-21.\n \n CompAct(s) \n Predict usr using a restricted number (excluding the paging information (10-18)\n \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original source: DELVE repository of data. \n Characteristics: 8192 cases, 22 continuous attributes', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:32', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3634/dataset_2183_cpu_act.arff', 'true', 111, 'usr', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:32'),
-(112, 1, 0, 'delta_elevators', '1', '1', '**Author**: Rui Camacho (rcamacho@garfield.fe.up.pt)  \n**Source**: [Regression datasets collection Luis Torgo](http://www.dcc.fc.up.pt/~ltorgo/Regression/DataSets.html)  \n**Please cite**:   \n\nThis data set is also obtained from the task of controlling the ailerons of a F16 aircraft, although the target variable and attributes are different from the ailerons domain. The target variable here is a variation instead of an absolute value, and there was some pre-selection of the attributes.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52348/openml_phpmlDSTj', 'true', 112, 'Se', NULL, NULL, NULL, 'public', NULL, NULL, 'set target feature', '2014-09-22 16:13:44'),
-(113, 1, 0, 'fruitfly', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identifier attribute deleted.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  Sexual activity and the lifespan of male fruitflies\n TYPE:  Designed (almost factorial) experiment\n SIZE:  125 observations, 5 variables\n \n DESCRIPTIVE ABSTRACT:\n A cost of increased reproduction in terms of reduced longevity has been\n shown for female fruitflies, but not for males.  The flies used were an\n outbred stock.  Sexual activity was manipulated by supplying individual\n males with one or eight receptive virgin females per day.  The\n longevity of these males was compared with that of two control types.\n The first control consisted of two sets of individual males kept with\n one or eight newly inseminated females.  Newly inseminated females will\n not usually remate for at least two days, and thus served as a control\n for any effect of competition with the male for food or space.  The\n second control was a set of individual males kept with no females.\n There were 25 males in each of the five groups, which were treated\n identically in number of anaesthetizations (using CO2) and provision of\n fresh food medium.\n \n SOURCE:\n Figure 2 in the article \"Sexual Activity and the Lifespan of Male\n Fruitflies\" by Linda Partridge and Marion Farquhar.  _Nature_, 294,\n 580-581, 1981.\n \n VARIABLE DESCRIPTIONS:\n Columns  Variable    Description\n -------  --------    -----------\n  1- 2    ID          Serial No. (1-25) within each group of 25\n                      (the order in which data points were abstracted)\n \n  4       PARTNERS    Number of companions (0, 1 or 8)\n \n  6       TYPE        Type of companion\n                        0: newly pregnant female\n                        1: virgin female\n                        9: not applicable (when PARTNERS=0)\n \n  8- 9    LONGEVITY   Lifespan, in days\n \n 11-14    THORAX      Length of thorax, in mm (x.xx)\n \n 16-17    SLEEP       Percentage of each day spent sleeping\n \n \n SPECIAL NOTES:\n `Compliance\' of the males in the two experimental groups was documented\n as follows:  On two days per week throughout the life of each\n experimental male, the females that had been supplied as virgins to\n that male were kept and examined for fertile eggs.  The insemination\n rate declined from approximately 7 females/day at age one week to just\n under 2/day at age eight weeks in the males supplied with eight virgin\n females per day, and from just under 1/day at age one week to\n approximately 0.6/day at age eight weeks in the males supplied with one\n virgin female per day.  These `compliance\' data were not supplied for\n individual males, but the authors say that \"There were no significant\n differences between the individual males within each experimental\n group.\"\n \n STORY BEHIND THE DATA:\n James Hanley found this dataset in _Nature_ and was attracted by the\n way the raw data were presented in classical analysis of covariance\n style in Figure 2.  He read the data points from the graphs and brought\n them to the attention of a colleague with whom he was teaching the\n applied statistics course.  Dr. Liddell thought that with only three\n explanatory variables (THORAX, plus PARTNERS and TYPE to describe the\n five groups), it would not be challenging enough as a data-analysis\n project.  He suggested adding another variable.  James Hanley added\n SLEEP, a variable not mentioned in the published article.  Teachers can\n contact us about the construction of this variable.  (We prefer to\n divulge the details at the end of the data-analysis project.)\n \n Further discussion of the background and pedagogical use of this\n dataset can be found in Hanley (1983) and in Hanley and Shapiro\n (1994).  To obtain the Hanley and Shapiro article, send the one-line\n e-mail message:\n send jse/v2n1/datasets.hanley\n to the address archive@jse.stat.ncsu.edu\n \n PEDAGOGICAL NOTES:\n This has been the most successful and the most memorable dataset we\n have used in an \"applications of statistics\" course, which we have\n taught for ten years.  The most common analysis techniques have been\n analysis of variance, classical analysis of covariance, and multiple\n regression.  Because the variable THORAX is so strong (it explains\n about 1/3 of the variance in LONGEVITY), it is important to consider it\n to increase the precision of between-group contrasts.  When students\n first check and find that the distributions of thorax length, and in\n particular, the mean thorax length, are very similar in the different\n groups, many of them are willing to say (in epidemiological\n terminology) that THORAX is not a confounding variable, and that it can\n be omitted from the analysis.\n \n There is usually lively discussion about the primary contrast.  The\n five groups and their special structure allow opportunities for\n students to understand and verbalize what we mean by the term\n \"statistical interaction.\"\n \n There is also much debate as to whether one should take the SLEEP\n variable into account.  Some students say that it is an `intermediate\'\n variable.  Some students formally test the mean level of SLEEP across\n groups, find one pair where there is a statistically significant\n difference, and want to treat it as a confounding variable.  A few\n students muse about how it was measured.\n \n There is heteroscedasticity in the LONGEVITY variable.\n \n One very observant student (now a professor) argued that THORAX cannot\n be used as a predictor or explanatory variable for the LONGEVITY\n outcome since fruitflies who die young may not be fully grown, i.e., it\n is also an intermediate variable.  One Ph.D. student who had studied\n entomology assured us that fruitflies do not grow longer after birth;\n therefore, the THORAX length is not time-dependent!\n \n Curiously, the dataset has seldom been analyzed using techniques from\n survival analysis.  The fact that there are no censored observations is\n not really an excuse, and one could easily devise a way to introduce\n censoring of LONGEVITY.\n \n REFERENCES:\n Hanley, J. A. (1983), \"Appropriate Uses of Multivariate Analysis,\"\n _Annual Review of Public Health_, 4, 155-180.\n \n Hanley, J. A., and Shapiro, S. H. (1994), \"Sexual Activity and the\n Lifespan of Male Fruitflies:  A Dataset That Gets Attention,\" _Journal\n of Statistics Education_, Volume 2, Number 1.\n \n SUBMITTED BY:\n James A. Hanley and Stanley H. Shapiro\n Department of Epidemiology and Biostatistics\n McGill University\n 1020 Pine Avenue West\n Montreal, Quebec, H3A 1A2\n Canada\n tel: +1 (514) 398-6270 (JH) \n      +1 (514) 398-6272 (SS)\n fax: +1 (514) 398-4503\n INJH@musicb.mcgill.ca, StanS@epid.lan.mcgill.ca', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:42', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3636/dataset_2185_fruitfly.arff', 'true', 113, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:42'),
-(114, 1, 0, 'pbc', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Case number deleted. X treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  PBC Data\n SIZE:  418 observations, 20 variables\n \n \n \n DESCRIPTIVE ABSTRACT:\n \n Below is a description of the variables recorded from the Mayo Clinic trial \n in primary biliary cirrhosis (PBC) of the liver conducted between 1974 and \n 1984.  A total of 424 PBC patients, referred to Mayo Clinic during\n that ten-year interval, met eligibility criteria for the randomized placebo \n controlled trial of the drug D-penicillamine. The first 312 cases in the data \n set participated in the randomized trial, and contain largely complete data. \n The additional 112 cases did not participate in the clinical trial, but \n consented to have basic measurements recorded and to be followed for survival.\n Six of those cases were lost to follow-up shortly after diagnosis, so there \n are data here on an additional 106 cases as well as the 312 randomized \n participants. Missing data items are denoted by \".\".  At least one space \n separates each variable in the .DAT file.  Censoring was due to liver \n transplantation for twenty-five subjects with the following case numbers: \n 5, 105, 111, 120, 125, 158, 183, 241, 246, 247, 254, 263, 264, 265, 274, \n 288, 291, 295, 297, 345, 361, 362, 375, 380, 383.\n \n \n \n SOURCE:  Counting Processes and Survival Analysis by T. Fleming & \n          D. Harrington, (1991),  published by John Wiley & Sons.\n \n \n \n VARIABLE DESCRIPTIONS:\n \n The data are in free format.  That is, at least one blank space separates\n each variable.  The variables contained in the .DAT are:\n \n \n N:   Case number.\n X:   The number of days between registration and the earlier of\n      death, liver transplantation, or study analysis time in July, 1986.\n D:   1 if X is time to death, 0 if time to censoring\n Z1:  Treatment Code, 1 = D-penicillamine, 2 = placebo.\n Z2:  Age in years. For the first 312 cases, age was calculated by\n      dividing the number of days between birth and study registration by 365.\n Z3:  Sex, 0 = male, 1 = female.\n Z4:  Presence of ascites, 0 = no, 1 = yes.\n Z5:  Presence of hepatomegaly, 0 = no, 1 = yes.\n Z6:  Presence of spiders 0 = no, 1 = Yes.\n Z7:  Presence of edema, 0 = no edema and no diuretic therapy for\n      edema; 0.5 = edema present for which no diuretic therapy was given, or \n      edema resolved with diuretic therapy; 1 = edema despite diuretic therapy\n Z8:  Serum bilirubin, in mg/dl.\n Z9:  Serum cholesterol, in mg/dl.\n Z10: Albumin, in gm/dl.\n Z11: Urine copper, in mg/day.\n Z12: Alkaline phosphatase, in U/liter.\n Z13: SGOT, in U/ml.\n Z14: Triglycerides, in mg/dl.\n Z15: Platelet count; coded value is number of platelets\n      per-cubic-milliliter of blood divided by 1000.\n Z16: Prothrombin time, in seconds.\n Z17: Histologic stage of disease, graded 1, 2, 3, or 4.\n \n \n \n \n STORY BEHIND THE DATA:\n \n Between January, 1974 and May, 1984, the Mayo Clinic conducted a\n double-blinded randomized trial in primary biliary cirrhosis of the liver\n (PBC), comparing the drug D-penicillamine (DPCA) with a placebo. There\n were 424 patients who met the eligibility criteria seen at the Clinic while\n the trial was open for patient registration. Both the treating physician and\n the patient agreed to participate in the randomized trial in 312 of the 424\n cases. The date of randomization and a large number of clinical, biochemical,\n serologic, and histologic parameters were recorded for each of the 312\n clinical trial patients. The data from the trial were analyzed in 1986 for\n presentation in the clinical literature. For that analysis, disease and \n survival status as of July, 1986, were recorded for as many patients as \n possible.  By that date, 125 of the 312 patients had died, with only 11 \n not attributable to PBC.  Eight patients had been lost to follow up, and 19 \n had undergone liver transplantation. \n \n PBC is a rare but fatal chronic liver disease of unknown cause,\n with a prevalence of about 50-cases-per-million population. The primary\n pathologic event appears to be the destruction of interlobular bile ducts,\n which may be mediated by immunologic mechanisms. The data discussed here are\n important in two respects. First, controlled clinical trials are difficult to\n complete in rare diseases, and this case series of patients uniformly\n diagnosed, treated, and followed is the largest existing for PBC. The\n treatment comparison in this trial is more precise than in similar trials\n having fewer participants and avoids the bias that may arise in comparing\n a case series to historical controls. Second, the data present an\n opportunity to study the natural history of the disease. We will see that, \n despite the immunosuppressive properties of DPCA, there are no detectable\n differences between the distributions of survival times for the DPCA and\n placebo treatment groups. This suggests that these groups can be combined\n in studying the association between survival time from randomization and\n clinical and other measurements. In the early to mid 1980s, the rate of \n successful liver transplant increased substantially, and transplant has \n become an effective therapy for PBC. The Mayo Clinic data set is therefore \n one of the last allowing a study of the natural history of PBC in patients \n who were treated with only supportive care or its equivalent. The PBC data \n can be used to: estimate a survival distribution; test for differences \n between two groups; and estimate covariate effects via a regression\n model.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:16:46', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3637/dataset_2186_pbc.arff', 'true', 114, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:16:46'),
-(115, 1, 0, 'pol', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis is a commercial application described in Weiss & Indurkhya (1995). \n The data describes a telecommunication problem. No further information\n is available.\n \n Characteristics: (10000+5000) cases, 49 continuous attributes \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original Source: The data in the original format can be obtained \n from http://www.cs.su.oz.au/~nitin', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:01', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3638/dataset_2187_pol.arff', 'true', 115, 'foo', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:01'),
-(116, 1, 0, 'autoHorse', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Horsepower treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n 1. Title: 1985 Auto Imports Database\n \n 2. Source Information:\n    -- Creator/Donor: Jeffrey C. Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n    -- Date: 19 May 1987\n    -- Sources:\n      1) 1985 Model Import Car and Truck Specifications, 1985 Ward\'s\n         Automotive Yearbook.\n      2) Personal Auto Manuals, Insurance Services Office, 160 Water\n         Street, New York, NY 10038 \n      3) Insurance Collision Report, Insurance Institute for Highway\n         Safety, Watergate 600, Washington, DC 20037\n\n 3. Past Usage:\n    -- Kibler,~D., Aha,~D.~W., & Albert,~M. (1989).  Instance-based prediction\n       of real-valued attributes.  {it Computational Intelligence}, {it 5},\n       51--57.\n          -- Predicted price of car using all numeric and Boolean attributes\n          -- Method: an instance-based learning (IBL) algorithm derived from a\n             localized k-nearest neighbor algorithm.  Compared with a\n             linear regression prediction...so all instances\n             with missing attribute values were discarded.  This resulted with\n             a training set of 159 instances, which was also used as a test\n             set (minus the actual instance during testing).\n          -- Results: Percent Average Deviation Error of Prediction from Actual\n             -- 11.84% for the IBL algorithm\n             -- 14.12% for the resulting linear regression equation\n \n 4. Relevant Information:\n    -- Description\n       This data set consists of three types of entities: (a) the\n       specification of an auto in terms of various characteristics, (b)\n       its assigned insurance risk rating, (c) its normalized losses in use\n       as compared to other cars.  The second rating corresponds to the\n       degree to which the auto is more risky than its price indicates.\n       Cars are initially assigned a risk factor symbol associated with its\n       price.   Then, if it is more risky (or less), this symbol is\n       adjusted by moving it up (or down) the scale.  Actuarians call this\n       process \"symboling\".  A value of +3 indicates that the auto is\n       risky, -3 that it is probably pretty safe.\n \n       The third factor is the relative average loss payment per insured\n       vehicle year.  This value is normalized for all autos within a\n       particular size classification (two-door small, station wagons,\n       sports/speciality, etc...), and represents the average loss per car\n       per year.\n \n    -- Note: Several of the attributes in the database could be used as a\n             \"class\" attribute.\n \n 5. Number of Instances: 205\n \n 6. Number of Attributes: 26 total\n    -- 15 continuous\n    -- 1 integer\n    -- 10 nominal\n \n 7. Attribute Information:     \n      Attribute:                Attribute Range:\n      ------------------        -----------------------------------------------\n   1. symboling:                -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback, convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n 8. Missing Attribute Values: (denoted by \"?\")\n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4%', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3639/dataset_2188_autoHorse.arff', 'true', 116, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:04'),
-(117, 1, 0, 'lowbwt', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identification code deleted. \n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  LOW BIRTH WEIGHT DATA\n KEYWORDS:  Logistic Regression\n SIZE:  189 observations, 11 variables\n \n NOTE:\n         These data come from Appendix 1 of Hosmer and Lemeshow (1989).\n These data are copyrighted and must be acknowledged and used accordingly.\n \n DESCRIPTIVE ABSTRACT:\n         The goal of this study was to identify risk factors associated with\n giving birth to a low birth weight baby (weighing less than 2500 grams).\n Data were collected on 189 women, 59 of which had low birth weight babies\n and 130 of which had normal birth weight babies.  Four variables which were\n thought to be of importance were age, weight of the subject at her last\n menstrual period, race, and the number of physician visits during the first\n trimester of pregnancy.\n \n \n SOURCE:\n          Data were collected at Baystate Medical Center, Springfield,\n Massachusetts, during 1986.\n \n \n NOTE:\n           This data set consists of the complete data.  A paired data set\n created from this low birth weight data may be found in plowbwt.dat and\n a 3 to 1 matched data set created from the low birth weight data may be\n found in mlowbwt.dat.\n \n \n \n Table:  Code Sheet for the Variables in the Low Birth Weight Data Set.\n \n Columns   Variable                                              Abbreviation\n -----------------------------------------------------------------------------\n 2-4     Identification Code                                     ID\n    \n 10      Low Birth Weight (0 = Birth Weight ge 2500g,            LOW\n                           l = Birth Weight < 2500g)\n   \n 17-18   Age of the Mother in Years                              AGE\n      \n 23-25   Weight in Pounds at the Last Menstrual Period           LWT\n      \n 32      Race (1 = White, 2 = Black, 3 = Other)                  RACE\n      \n 40      Smoking Status During Pregnancy (1 = Yes, 0 = No)       SMOKE\n      \n 48      History of Premature Labor (0 = None, 1 = One, etc.)    PTL\n      \n 55      History of Hypertension (1 = Yes, 0 = No)               HT\n      \n 61      Presence of Uterine Irritability (1 = Yes, 0 = No)      UI\n      \n 67      Number of Physician Visits During the First Trimester   FTV\n                 (0 = None, 1 = One, 2 = Two, etc.)\n      \n 73-76   Birth Weight in Grams                                   BWT\n -----------------------------------------------------------------------------\n \n PEDAGOGICAL NOTES:\n         These data have been used as an example of fitting a multiple\n logistic regression model.\n \n STORY BEHIND THE DATA:\n         Low birth weight is an outcome that has been of concern to physicians\n for years. This is due to the fact that infant mortality rates and birth\n defect rates are very high for low birth weight babies. A woman\'s behavior\n during pregnancy (including diet, smoking habits, and receiving prenatal care)\n can greatly alter the chances of carrying the baby to term and, consequently,\n of delivering a baby of normal birth weight.\n         The variables identified in the code sheet given in the table have been\n shown to be associated with low birth weight in the obstetrical literature. The\n goal of the current study was to ascertain if these variables were important\n in the population being served by the medical center where the data were\n collected.\n \n \n References:\n \n 1. Hosmer and Lemeshow, Applied Logistic Regression, Wiley, (1989).', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:07', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3640/dataset_2189_lowbwt.arff', 'true', 117, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:07'),
-(118, 1, 0, 'cholesterol', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Cholesterol treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Publication Request: \n    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n    This file describes the contents of the heart-disease directory.\n \n    This directory contains 4 databases concerning heart disease diagnosis.\n    All attributes are numeric-valued.  The data was collected from the\n    four following locations:\n \n      1. Cleveland Clinic Foundation (cleveland.data)\n      2. Hungarian Institute of Cardiology, Budapest (hungarian.data)\n      3. V.A. Medical Center, Long Beach, CA (long-beach-va.data)\n      4. University Hospital, Zurich, Switzerland (switzerland.data)\n \n    Each database has the same instance format.  While the databases have 76\n    raw attributes, only 14 of them are actually used.  Thus I\'ve taken the\n    liberty of making 2 copies of each database: one with all the attributes\n    and 1 with the 14 attributes actually used in past experiments.\n \n    The authors of the databases have requested:\n \n       ...that any publications resulting from the use of the data include the \n       names of the principal investigator responsible for the data collection\n       at each institution.  They would be:\n \n        1. Hungarian Institute of Cardiology. Budapest: Andras Janosi, M.D.\n        2. University Hospital, Zurich, Switzerland: William Steinbrunn, M.D.\n        3. University Hospital, Basel, Switzerland: Matthias Pfisterer, M.D.\n        4. V.A. Medical Center, Long Beach and Cleveland Clinic Foundation:\n           Robert Detrano, M.D., Ph.D.\n \n    Thanks in advance for abiding by this request.\n \n    David Aha\n    July 22, 1988\n    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n \n 1. Title: Heart Disease Databases\n \n 2. Source Information:\n    (a) Creators: \n        -- 1. Hungarian Institute of Cardiology. Budapest: Andras Janosi, M.D.\n        -- 2. University Hospital, Zurich, Switzerland: William Steinbrunn, M.D.\n        -- 3. University Hospital, Basel, Switzerland: Matthias Pfisterer, M.D.\n        -- 4. V.A. Medical Center, Long Beach and Cleveland Clinic Foundation:\n              Robert Detrano, M.D., Ph.D.\n    (b) Donor: David W. Aha (aha@ics.uci.edu) (714) 856-8779   \n    (c) Date: July, 1988\n \n 3. Past Usage:\n     1. Detrano,~R., Janosi,~A., Steinbrunn,~W., Pfisterer,~M., Schmid,~J.,\n        Sandhu,~S., Guppy,~K., Lee,~S., & Froelicher,~V. (1989).  {it \n        International application of a new probability algorithm for the \n        diagnosis of coronary artery disease.}  {it American Journal of \n        Cardiology}, {it 64},304--310.\n        -- International Probability Analysis \n        -- Address: Robert Detrano, M.D.\n                    Cardiology 111-C\n                    V.A. Medical Center\n                    5901 E. 7th Street\n                    Long Beach, CA 90028\n        -- Results in percent accuracy: (for 0.5 probability threshold)\n              Data Name:  CDF    CADENZA\n           -- Hungarian   77     74\n              Long beach  79     77\n              Swiss       81     81\n           -- Approximately a 77% correct classification accuracy with a\n              logistic-regression-derived discriminant function\n     2. David W. Aha & Dennis Kibler\n        -- \n           \n           \n           -- Instance-based prediction of heart-disease presence with the \n              Cleveland database\n              -- NTgrowth: 77.0% accuracy\n              --       C4: 74.8% accuracy\n     3. John Gennari\n        -- Gennari, J.~H., Langley, P, & Fisher, D. (1989). Models of\n           incremental concept formation. {it Artificial Intelligence, 40},\n           11--61.\n        -- Results: \n           -- The CLASSIT conceptual clustering system achieved a 78.9% accuracy\n              on the Cleveland database.\n \n 4. Relevant Information:\n      This database contains 76 attributes, but all published experiments\n      refer to using a subset of 14 of them.  In particular, the Cleveland\n      database is the only one that has been used by ML researchers to \n      this date.  The \"goal\" field refers to the presence of heart disease\n      in the patient.  It is integer valued from 0 (no presence) to 4.\n      Experiments with the Cleveland database have concentrated on simply\n      attempting to distinguish presence (values 1,2,3,4) from absence (value\n      0).  \n    \n      The names and social security numbers of the patients were recently \n      removed from the database, replaced with dummy values.\n \n      One file has been \"processed\", that one containing the Cleveland \n      database.  All four unprocessed files also exist in this directory.\n     \n 5. Number of Instances: \n         Database:    # of instances:\n           Cleveland: 303\n           Hungarian: 294\n         Switzerland: 123\n       Long Beach VA: 200\n \n 6. Number of Attributes: 76 (including the predicted attribute)\n \n 7. Attribute Information:\n    -- Only 14 used\n       -- 1. #3  (age)       \n       -- 2. #4  (sex)       \n       -- 3. #9  (cp)        \n       -- 4. #10 (trestbps)  \n       -- 5. #12 (chol)      \n       -- 6. #16 (fbs)       \n       -- 7. #19 (restecg)   \n       -- 8. #32 (thalach)   \n       -- 9. #38 (exang)     \n       -- 10. #40 (oldpeak)   \n       -- 11. #41 (slope)     \n       -- 12. #44 (ca)        \n       -- 13. #51 (thal)      \n       -- 14. #58 (num)       (the predicted attribute)\n \n    -- Complete attribute documentation:\n       1 id: patient identification number\n       2 ccf: social security number (I replaced this with a dummy value of 0)\n       3 age: age in years\n       4 sex: sex (1 = male; 0 = female)\n       5 painloc: chest pain location (1 = substernal; 0 = otherwise)\n       6 painexer (1 = provoked by exertion; 0 = otherwise)\n       7 relrest (1 = relieved after rest; 0 = otherwise)\n       8 pncaden (sum of 5, 6, and 7)\n       9 cp: chest pain type\n         -- Value 1: typical angina\n         -- Value 2: atypical angina\n         -- Value 3: non-anginal pain\n         -- Value 4: asymptomatic\n      10 trestbps: resting blood pressure (in mm Hg on admission to the \n         hospital)\n      11 htn\n      12 chol: serum cholestoral in mg/dl\n      13 smoke: I believe this is 1 = yes; 0 = no (is or is not a smoker)\n      14 cigs (cigarettes per day)\n      15 years (number of years as a smoker)\n      16 fbs: (fasting blood sugar > 120 mg/dl)  (1 = true; 0 = false)\n      17 dm (1 = history of diabetes; 0 = no such history)\n      18 famhist: family history of coronary artery disease (1 = yes; 0 = no)\n      19 restecg: resting electrocardiographic results\n         -- Value 0: normal\n         -- Value 1: having ST-T wave abnormality (T wave inversions and/or ST \n                     elevation or depression of > 0.05 mV)\n         -- Value 2: showing probable or definite left ventricular hypertrophy\n                     by Estes\' criteria\n      20 ekgmo (month of exercise ECG reading)\n      21 ekgday(day of exercise ECG reading)\n      22 ekgyr (year of exercise ECG reading)\n      23 dig (digitalis used furing exercise ECG: 1 = yes; 0 = no)\n      24 prop (Beta blocker used during exercise ECG: 1 = yes; 0 = no)\n      25 nitr (nitrates used during exercise ECG: 1 = yes; 0 = no)\n      26 pro (calcium channel blocker used during exercise ECG: 1 = yes; 0 = no)\n      27 diuretic (diuretic used used during exercise ECG: 1 = yes; 0 = no)\n      28 proto: exercise protocol\n           1 = Bruce     \n           2 = Kottus\n           3 = McHenry\n           4 = fast Balke\n           5 = Balke\n           6 = Noughton \n           7 = bike 150 kpa min/min  (Not sure if \"kpa min/min\" is what was \n               written!)\n           8 = bike 125 kpa min/min  \n           9 = bike 100 kpa min/min\n          10 = bike 75 kpa min/min\n          11 = bike 50 kpa min/min\n          12 = arm ergometer\n      29 thaldur: duration of exercise test in minutes\n      30 thaltime: time when ST measure depression was noted\n      31 met: mets achieved\n      32 thalach: maximum heart rate achieved\n      33 thalrest: resting heart rate\n      34 tpeakbps: peak exercise blood pressure (first of 2 parts)\n      35 tpeakbpd: peak exercise blood pressure (second of 2 parts)\n      36 dummy\n      37 trestbpd: resting blood pressure\n      38 exang: exercise induced angina (1 = yes; 0 = no)\n      39 xhypo: (1 = yes; 0 = no)\n      40 oldpeak = ST depression induced by exercise relative to rest\n      41 slope: the slope of the peak exercise ST segment\n         -- Value 1: upsloping\n         -- Value 2: flat\n         -- Value 3: downsloping\n      42 rldv5: height at rest\n      43 rldv5e: height at peak exercise\n      44 ca: number of major vessels (0-3) colored by flourosopy\n      45 restckm: irrelevant\n      46 exerckm: irrelevant\n      47 restef: rest raidonuclid (sp?) ejection fraction\n      48 restwm: rest wall (sp?) motion abnormality\n         0 = none\n         1 = mild or moderate\n         2 = moderate or severe\n         3 = akinesis or dyskmem (sp?)\n      49 exeref: exercise radinalid (sp?) ejection fraction\n      50 exerwm: exercise wall (sp?) motion \n      51 thal: 3 = normal; 6 = fixed defect; 7 = reversable defect\n      52 thalsev: not used\n      53 thalpul: not used\n      54 earlobe: not used\n      55 cmo: month of cardiac cath (sp?)  (perhaps \"call\")\n      56 cday: day of cardiac cath (sp?)\n      57 cyr: year of cardiac cath (sp?)\n      58 num: diagnosis of heart disease (angiographic disease status)\n         -- Value 0: < 50% diameter narrowing\n         -- Value 1: > 50% diameter narrowing\n         (in any major vessel: attributes 59 through 68 are vessels)\n      59 lmt\n      60 ladprox\n      61 laddist\n      62 diag\n      63 cxmain\n      64 ramus\n      65 om1\n      66 om2\n      67 rcaprox\n      68 rcadist\n      69 lvx1: not used\n      70 lvx2: not used\n      71 lvx3: not used\n      72 lvx4: not used\n      73 lvf: not used\n      74 cathef: not used\n      75 junk: not used\n      76 name: last name of patient \n         (I replaced this with the dummy string \"name\")\n \n 9. Missing Attribute Values: Several.  Distinguished with value -9.0.\n \n 10. Class Distribution:\n         Database:      0   1   2   3   4 Total\n           Cleveland: 164  55  36  35  13   303\n           Hungarian: 188  37  26  28  15   294\n         Switzerland:   8  48  32  30   5   123\n       Long Beach VA:  51  56  41  42  10   200', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:10', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3641/dataset_2190_cholesterol.arff', 'true', 118, 'chol', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:10'),
-(119, 1, 0, 'sleep', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n Data from which conclusions  were  drawn  in  the  article  \"Sleep  in \n Mammals: Ecological and Constitutional Correlates\" by Allison, T.  and \n Cicchetti, D. (1976), _Science_, November 12, vol. 194,  pp.  732-734. \n Includes brain and body  weight,  life  span,  gestation  time,  time \n sleeping, and predation and danger indices for 62 mammals.\n \n \n \n Variables below (from left to right) for Mammals Data Set:\n \n species of animal\n \n body weight in kg\n \n brain weight in g\n \n slow wave (\"nondreaming\") sleep (hrs/day)\n \n paradoxical (\"dreaming\") sleep (hrs/day)\n \n total sleep (hrs/day)  (sum of slow wave and paradoxical sleep)\n \n maximum life span (years)\n \n gestation time (days)\n \n predation index (1-5)\n                 1 = minimum (least likely to be preyed upon)\n                 5 = maximum (most likely to be preyed upon)\n \n sleep exposure index (1-5)\n                 1 = least exposed (e.g. animal sleeps in a \n                     well-protected den)\n                 5 = most exposed\n \n overall danger index (1-5)\n                 (based on the above two indices and other information)\n                 1 = least danger (from other animals)\n                 5 = most danger (from other animals)\n \n Note: Missing values denoted by -999.0\n \n \n For more details, see\n \n Allison, Truett and Cicchetti, Domenic V. (1976), \"Sleep  in  Mammals: \n Ecological and Constitutional  Correlates\",  _Science_,  November  12, \n vol. 194, pp. 732-734.\n \n The above data set can be freely used for non-commercial purposes  and \n can be freely distributed (permission in  writing  obtained  from  Dr. \n Truett Allison).\n \n Submitted by Roger Johnson\n rwjohnso@silver.sdsmt.edu\n\n Total sleep treated as the class attribute. Attributes for slow\n wave and paradoxical sleep have been deleted. (The animal\'s\n name has also been deleted.)', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:12', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3642/dataset_2191_sleep.arff', 'true', 119, 'danger_index', NULL, NULL, NULL, 'public', NULL, NULL, 'corrected target', '2014-10-06 22:07:54'),
-(120, 1, 0, 'triazines', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe problem is to learn a regression equation/rule/tree to predict the\n activity from the descriptive structural attributes.  The data and\n methodology is described in detail in: - King, Ross .D., Hurst,\n Jonathan. D., and Sternberg, Michael.J.E. A comparison of artificial\n intelligence methods for modelling QSARs Applied Artificial\n Intelligence, 1994 (in press).  - Hurst, Jonathan. D., King, Ross\n .D. and Sternberg, Michael.J.E. Quantitative Structure-Activity\n Relationships by neural networks and inductive logic programming:\n 2. The inhibition of dihydrofolate reductase by triazines. Journal of\n Computer Aided Molecular Design, 1994 (in press).\n \n Original source: ?. \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 186 cases; 61 continuous variables', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:16', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3643/dataset_2192_triazines.arff', 'true', 120, 'activity', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:16'),
-(121, 1, 0, 'autoPrice', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n All nominal attributes and instances with missing values are deleted.\n Price treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n 1. Title: 1985 Auto Imports Database\n \n 2. Source Information:\n    -- Creator/Donor: Jeffrey C. Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n    -- Date: 19 May 1987\n    -- Sources:\n      1) 1985 Model Import Car and Truck Specifications, 1985 Ward\'s\n         Automotive Yearbook.\n      2) Personal Auto Manuals, Insurance Services Office, 160 Water\n         Street, New York, NY 10038 \n      3) Insurance Collision Report, Insurance Institute for Highway\n         Safety, Watergate 600, Washington, DC 20037\n\n 3. Past Usage:\n    -- Kibler,~D., Aha,~D.~W., & Albert,~M. (1989).  Instance-based prediction\n       of real-valued attributes.  {it Computational Intelligence}, {it 5},\n       51--57.\n          -- Predicted price of car using all numeric and Boolean attributes\n          -- Method: an instance-based learning (IBL) algorithm derived from a\n             localized k-nearest neighbor algorithm.  Compared with a\n             linear regression prediction...so all instances\n             with missing attribute values were discarded.  This resulted with\n             a training set of 159 instances, which was also used as a test\n             set (minus the actual instance during testing).\n          -- Results: Percent Average Deviation Error of Prediction from Actual\n             -- 11.84% for the IBL algorithm\n             -- 14.12% for the resulting linear regression equation\n \n 4. Relevant Information:\n    -- Description\n       This data set consists of three types of entities: (a) the\n       specification of an auto in terms of various characteristics, (b)\n       its assigned insurance risk rating, (c) its normalized losses in use\n       as compared to other cars.  The second rating corresponds to the\n       degree to which the auto is more risky than its price indicates.\n       Cars are initially assigned a risk factor symbol associated with its\n       price.   Then, if it is more risky (or less), this symbol is\n       adjusted by moving it up (or down) the scale.  Actuarians call this\n       process \"symboling\".  A value of +3 indicates that the auto is\n       risky, -3 that it is probably pretty safe.\n \n       The third factor is the relative average loss payment per insured\n       vehicle year.  This value is normalized for all autos within a\n       particular size classification (two-door small, station wagons,\n       sports/speciality, etc...), and represents the average loss per car\n       per year.\n \n    -- Note: Several of the attributes in the database could be used as a\n             \"class\" attribute.\n \n 5. Number of Instances: 205\n \n 6. Number of Attributes: 26 total\n    -- 15 continuous\n    -- 1 integer\n    -- 10 nominal\n \n 7. Attribute Information:     \n      Attribute:                Attribute Range:\n      ------------------        -----------------------------------------------\n   1. symboling:                -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback, convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n 8. Missing Attribute Values: (denoted by \"?\")\n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4%', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:18', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3644/dataset_2193_autoPrice.arff', 'true', 121, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:18'),
-(122, 1, 0, 'detroit', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n This is the data set called `DETROIT\' in the book `Subset selection in\n regression\' by Alan J. Miller published in the Chapman & Hall series of\n monographs on Statistics & Applied Probability, no. 40.   The data are\n unusual in that a subset of three predictors can be found which gives a\n very much better fit to the data than the subsets found from the Efroymson\n stepwise algorithm, or from forward selection or backward elimination.\n \n The original data were given in appendix A of `Regression analysis and its\n application: A data-oriented approach\' by Gunst & Mason, Statistics\n textbooks and monographs no. 24, Marcel Dekker.   It has caused problems\n because some copies of the Gunst & Mason book do not contain all of the data,\n and because Miller does not say which variables he used as predictors and\n which is the dependent variable.   (HOM was the dependent variable, and the\n predictors were FTP ... WE)\n \n The data were collected by J.C. Fisher and used in his paper: \"Homicide in\n Detroit: The Role of Firearms\", Criminology, vol.14, 387-400 (1976)\n \n \n The data are on the homicide rate in Detroit for the years 1961-1973.\n FTP    - Full-time police per 100,000 population\n UEMP   - %  unemployed in the population\n MAN    - number of manufacturing workers in thousands\n LIC    - Number of handgun licences per 100,000 population\n GR     - Number of handgun registrations per 100,000 population\n CLEAR  - %  homicides cleared by arrests\n WM     - Number of white males in the population\n NMAN   - Number of non-manufacturing workers in thousands\n GOV    - Number of government workers in thousands\n HE     - Average hourly earnings\n WE     - Average weekly earnings\n \n HOM    - Number of homicides per 100,000 of population\n ACC    - Death rate in accidents per 100,000 population\n ASR    - Number of assaults per 100,000 population\n \n N.B. Each case takes two lines.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:21', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3645/dataset_2194_detroit.arff', 'true', 122, 'ASR', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:21'),
-(123, 1, 0, 'quake', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:24', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3646/dataset_2195_quake.arff', 'true', 123, 'richter', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:24'),
-(124, 1, 0, 'cloud', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n These data are those collected in a cloud-seeding experiment in Tasmania\n between mid-1964 and January 1971.   Their analysis, using regression\n techniques and permutation tests, is discussed in:\n \n       Miller, A.J., Shaw, D.E., Veitch, L.G. & Smith, E.J. (1979).\n       `Analyzing the results of a cloud-seeding experiment in Tasmania\',\n       Communications in Statistics - Theory & Methods, vol.A8(10),\n       1017-1047.\n \n The rainfalls are period rainfalls in inches.   TE and TW are the east and\n west target areas respectively, while NC, SC and NWC are the corresponding\n rainfalls in the north, south and north-west control areas respectively.\n S = seeded, U = unseeded.\n\n Rain in eastern target region is being treated\n as the class attribute. (Attribute for rain\n in the western target region has been deleted.)', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:26', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3647/dataset_2196_cloud.arff', 'true', 124, 'TE', 'period', NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:26'),
-(125, 1, 0, 'longley', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n The infamous Longley data, \"An appraisal of least-squares programs from\n  the point of view of the user\", JASA, 62(1967) p819-841.\n\n Variables are: Number of people employed   (usually the y variable)\n                GNP implicit price deflator\n                GNP\n                Unemployed\n                Armed forces\n                Non-institutionalized population >=14 years of age\n                Year\n\n Employment is being treated as the class\n attribute.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3648/dataset_2197_longley.arff', 'true', 125, 'employed', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:28'),
-(126, 1, 0, 'diabetes_numeric', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis data set concerns the study of the factors affecting patterns of\n insulin-dependent diabetes mellitus in children.  The objective is to\n investigate the dependence of the level of serum C-peptide on the\n various other factors in order to understand the patterns of residual\n insulin secretion. The response measurement is the logarithm of\n C-peptide concentration (pmol/ml) at the diagnosis, and the predictor\n measurements age and base deficit, a measure of acidity.\n\n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original source: Book Generalized Additive Models (p.304) by Hastie &\n Tibshirani, Chapman & Hall.  \n Characteristics: 43 cases; 3 continuous variables', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:30', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3649/dataset_2198_diabetes_numeric.arff', 'true', 126, 'c_peptide', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-23 13:17:30'),
-(127, 1, 0, 'pharynx', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Case number deleted. \n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Name:  Pharynx (A clinical Trial in the Trt. of Carcinoma of the Oropharynx).\n SIZE:  195 observations, 13 variables.\n \n \n \n DESCRIPTIVE ABSTRACT:\n \n The .dat file gives the data for a part of a large clinical trial\n carried out by the Radiation Therapy Oncology Group in the United States. \n The full study included patients with squamous carcinoma of 15 sites in \n the mouth and throat, with 16 participating institutions, though only data \n on three sites in the oropharynx reported by the six largest institutions \n are considered here. Patients entering the study were randomly assigned to \n one of two treatment groups, radiation therapy alone or radiation therapy \n together with a chemotherapeutic agent.  One objective of the study was to \n compare the two treatment policies with respect to patient survival.\n \n \n \n SOURCE:  The Statistical Analysis of Failure Time Data, by JD Kalbfleisch\n          & RL Prentice, (1980),  Published by John Wiley & Sons \n \n \n \n VARIABLE DESCRIPTIONS:\n \n The data are in free format.  That is, at least one blank space separates \n each variable in the .dat file.  The variables are as follows:\n \n \n Case:         Case Number\n Inst:         Participating Institution\n sex:          1=male, 2=female\n Treatment:    1=standard, 2=test\n Grade:        1=well differentiated, 2=moderately differentiated, \n               3=poorly differentiated,  9=missing\n Age:          In years at time of diagnosis\n Condition:    1=no disability, 2=restricted work, 3=requires assistance with\n               self care, 4=bed confined,  9=missing\n Site:         1=faucial arch, 2=tonsillar fossa, 3=posterior pillar,\n               4=pharyngeal tongue, 5=posterior wall\n T staging:    1=primary tumor measuring 2 cm or less in largest diameter,\n               2=primary tumor measuring 2 cm to 4 cm in largest diameter with\n               minimal infiltration in depth, 3=primary tumor measuring more \n               than 4 cm, 4=massive invasive tumor\n N staging:    0=no clinical evidence of node metastases, 1=single positive\n               node 3 cm or less in diameter, not fixed, 2=single positive\n               node more than 3 cm in diameter, not fixed, 3=multiple\n               positive nodes or fixed positive nodes \n Entry Date:   Date of study entry: Day of year and year\n Status:       0=censored,  1=dead\n Time:         Survival time in days from day of diagnosis \n \n \n \n \n \n \n STORY BEHIND THE DATA:\n \n Approximately 30% of the survival times are censored owing primarily to \n patients surviving to the time of analysis. Some patients were lost\n to follow-up because the patient moved or transferred to an institution not\n participating in the study, though these cases were relatively rare. From \n a statistical point of view, an important feature of these data is the \n considerable lack of homogeneity between individuals being studied. \n Of course, as part of the study design, certain criteria for patient \n eligibility had to be met which eliminated extremes in the extent of disease, \n but still many factors are not controlled.\n \n This study included measurements of many covariates which would be expected \n to relate to survival experience. Six such variables are given in the data \n (sex, T staging, N staging, age, general condition, and grade).   The site \n of the primary tumor and possible differences between participating \n institutions require consideration as well.\n      \n The T,N staging classification gives a measure of the extent of the tumor at \n the primary site and at regional lymph nodes. T=1, refers to a small primary \n tumor, 2 centimeters or less in largest diameter, whereas T=4 is a massive \n tumor with extension to adjoining tissue. T=2 and T=3 refer to intermediate\n cases. N=0 refers to there being no clinical evidence of a lymph node \n metastasis and N=1, N=2, N=3 indicate, in increasing magnitude, the extent of \n existing lymph node involvement. Patients with classifications T=1,N=0; \n T=1,N=1;  T=2,N=0; or T=2,N=1, or with distant metastases were excluded \n from study.\n \n The variable general condition gives a measure of the functional capacity of \n the patient at the time of diagnosis (1 refers to no disability whereas\n 4 denotes bed confinement; 2 and 3 measure intermediate levels). The variable\n grade is a measure of the degree of differentiation of the tumor (the degree\n to which the tumor cell resembles the host cell) from 1 (well differentiated) \n to 3 (poorly differentiated)\n \n In addition to the primary question whether the combined treatment mode is\n preferable to the conventional radiation therapy, it is of considerable \n interest to determine the extent to which the several covariates relate to\n subsequent survival.  It is also imperative in answering the primary question \n to adjust the survivals for possible imbalance that may be present in the \n study with regard to the other covariates. Such problems are similar to those \n encountered in the classical theory of linear regression and the analysis of \n covariance.  Again, the need to accommodate censoring is an important \n distinguishing point. In many situations it is also important to develop \n nonparametric and robust procedures since there is frequently little empirical\n or theoretical work to support a particular family of failure time \n distributions.', 'ARFF', NULL, NULL, NULL, '2014-04-23 13:17:32', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3650/dataset_2199_pharynx.arff', 'true', 127, 'class', NULL, '\"Entry\"', NULL, 'public', NULL, NULL, 'set ignore feature', '2014-10-05 00:03:24'),
-(128, 1, 0, 'iris', '1', '1', '', 'ARFF', 'R.A. Fisher', NULL, '1936', '2014-04-06 23:23:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/61/dataset_61_iris.arff', 'true', 128, 'class', NULL, NULL, 'http://digital.library.adelaide.edu.au/dspace/handle/2440/15227', 'public', NULL, 'https://archive.ics.uci.edu/ml/datasets/Iris', NULL, '2014-04-06 23:23:39'),
-(129, 1, NULL, 'iris-challenge', '1', '1', '', 'ARFF', NULL, NULL, NULL, '2018-05-29 19:06:15', NULL, 'CC0', NULL, NULL, 'https://www.openml.org/data/download/19330175/iris-challenge.arff', 'true', 129, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2018-05-29 19:06:15'),
-(130, 1, 0, 'iris', '2', '1', '', 'ARFF', 'R.A. Fisher', NULL, '1936', '2014-04-06 23:23:39', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/61/dataset_61_iris.arff', 'true', 130, 'class', NULL, NULL, 'http://digital.library.adelaide.edu.au/dspace/handle/2440/15227', 'private', NULL, 'https://archive.ics.uci.edu/ml/datasets/Iris', NULL, '2014-04-06 23:23:39');
-
+(10, 1, 0, 'mfeat-pixel', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:48', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 'true', 10, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:48');
\ No newline at end of file
diff --git a/data/sql/dataset_tag.sql b/data/sql/dataset_tag.sql
index 65cd98a6b..db8a02107 100644
--- a/data/sql/dataset_tag.sql
+++ b/data/sql/dataset_tag.sql
@@ -8,94 +8,4 @@ INSERT INTO `dataset_tag` (`id`, `tag`, `uploader`, `date`) VALUES
 (7, 'study_14', 1, '2017-03-24 19:36:13'),
 (8, 'study_14', 1, '2017-03-24 19:36:13'),
 (9, 'study_14', 1, '2017-03-24 19:36:14'),
-(10, 'study_14', 1, '2017-03-24 19:36:14'),
-(11, 'study_14', 1, '2017-03-24 19:36:14'),
-(12, 'study_14', 1, '2017-03-24 19:36:15'),
-(13, 'study_14', 1, '2017-03-24 19:36:15'),
-(14, 'study_14', 1, '2017-03-24 19:36:15'),
-(15, 'study_14', 1, '2017-03-24 19:36:16'),
-(16, 'study_14', 1, '2017-03-24 19:36:16'),
-(17, 'study_14', 1, '2017-03-24 19:36:16'),
-(18, 'study_14', 1, '2017-03-24 19:36:16'),
-(19, 'study_14', 1, '2017-03-24 19:36:17'),
-(20, 'study_14', 1, '2017-03-24 19:36:17'),
-(21, 'study_14', 1, '2017-03-24 19:36:17'),
-(22, 'study_14', 1, '2017-03-24 19:36:18'),
-(23, 'study_14', 1, '2017-03-24 19:36:18'),
-(24, 'study_14', 1, '2017-03-24 19:36:18'),
-(25, 'study_14', 1, '2017-03-24 19:36:18'),
-(26, 'study_14', 1, '2017-03-24 19:36:19'),
-(27, 'study_14', 1, '2017-03-24 19:36:19'),
-(28, 'study_14', 1, '2017-03-24 19:36:19'),
-(29, 'study_14', 1, '2017-03-24 19:36:20'),
-(30, 'study_14', 1, '2017-03-24 19:36:20'),
-(31, 'study_14', 1, '2017-03-24 19:36:20'),
-(32, 'study_14', 1, '2017-03-24 19:36:21'),
-(33, 'study_14', 1, '2017-03-24 19:36:21'),
-(34, 'study_14', 1, '2017-03-24 19:36:21'),
-(35, 'study_14', 1, '2017-03-24 19:36:22'),
-(36, 'study_14', 1, '2017-03-24 19:36:22'),
-(37, 'study_14', 1, '2017-03-24 19:36:22'),
-(38, 'study_14', 1, '2017-03-24 19:36:22'),
-(39, 'study_14', 1, '2017-03-24 19:36:23'),
-(40, 'study_14', 1, '2017-03-24 19:36:23'),
-(41, 'study_14', 1, '2017-03-24 19:36:23'),
-(42, 'study_14', 1, '2017-03-24 19:36:24'),
-(43, 'study_14', 1, '2017-03-24 19:36:24'),
-(44, 'study_14', 1, '2017-03-24 19:36:24'),
-(45, 'study_14', 1, '2017-03-24 19:36:25'),
-(46, 'study_14', 1, '2017-03-24 19:36:25'),
-(47, 'study_14', 1, '2017-03-24 19:36:25'),
-(48, 'study_14', 1, '2017-03-24 19:36:26'),
-(49, 'study_14', 1, '2017-03-24 19:36:26'),
-(50, 'study_14', 1, '2017-03-24 19:36:26'),
-(51, 'study_14', 1, '2017-03-24 19:36:27'),
-(52, 'study_14', 1, '2017-03-24 19:36:27'),
-(53, 'study_14', 1, '2017-03-24 19:36:27'),
-(54, 'study_14', 1, '2017-03-24 19:36:28'),
-(55, 'study_14', 1, '2017-03-24 19:36:28'),
-(56, 'study_14', 1, '2017-03-24 19:36:28'),
-(57, 'study_14', 1, '2017-03-24 19:36:29'),
-(58, 'study_14', 1, '2017-03-24 19:36:29'),
-(59, 'study_14', 1, '2017-03-24 19:36:29'),
-(60, 'study_14', 1, '2017-03-24 19:36:30'),
-(61, 'study_14', 1, '2017-03-24 19:36:30'),
-(62, 'study_14', 1, '2017-03-24 19:36:30'),
-(63, 'study_14', 1, '2017-03-24 19:36:31'),
-(64, 'study_14', 1, '2017-03-24 19:36:31'),
-(65, 'study_14', 1, '2017-03-24 19:36:31'),
-(66, 'study_14', 1, '2017-03-24 19:36:31'),
-(67, 'study_14', 1, '2017-03-24 19:36:32'),
-(68, 'study_14', 1, '2017-03-24 19:36:32'),
-(69, 'study_14', 1, '2017-03-24 19:36:32'),
-(70, 'study_14', 1, '2017-03-24 19:36:33'),
-(71, 'study_14', 1, '2017-03-24 19:36:33'),
-(72, 'study_14', 1, '2017-03-24 19:36:33'),
-(73, 'study_14', 1, '2017-03-24 19:36:34'),
-(74, 'study_14', 1, '2017-03-24 19:36:34'),
-(75, 'study_14', 1, '2017-03-24 19:36:34'),
-(76, 'study_14', 1, '2017-03-24 19:36:35'),
-(77, 'study_14', 1, '2017-03-24 19:36:35'),
-(78, 'study_14', 1, '2017-03-24 19:36:35'),
-(79, 'study_14', 1, '2017-03-24 19:36:36'),
-(80, 'study_14', 1, '2017-03-24 19:36:36'),
-(81, 'study_14', 1, '2017-03-24 19:36:37'),
-(82, 'study_14', 1, '2017-03-24 19:36:37'),
-(83, 'study_14', 1, '2017-03-24 19:36:37'),
-(84, 'study_14', 1, '2017-03-24 19:36:38'),
-(85, 'study_14', 1, '2017-03-24 19:36:38'),
-(86, 'study_14', 1, '2017-03-24 19:36:38'),
-(87, 'study_14', 1, '2017-03-24 19:36:39'),
-(88, 'study_14', 1, '2017-03-24 19:36:39'),
-(89, 'study_14', 1, '2017-03-24 19:36:39'),
-(90, 'study_14', 1, '2017-03-24 19:36:40'),
-(91, 'study_14', 1, '2017-03-24 19:36:40'),
-(92, 'study_14', 1, '2017-03-24 19:36:41'),
-(93, 'study_14', 1, '2017-03-24 19:36:41'),
-(94, 'study_14', 1, '2017-03-24 19:36:41'),
-(95, 'study_14', 1, '2017-03-24 19:36:42'),
-(96, 'study_14', 1, '2017-03-24 19:36:42'),
-(97, 'study_14', 1, '2017-03-24 19:36:42'),
-(98, 'study_14', 1, '2017-03-24 19:36:43'),
-(99, 'study_14', 1, '2017-03-24 19:36:43'),
-(100, 'study_14', 1, '2017-03-24 19:36:43');
+(10, 'study_14', 1, '2017-03-24 19:36:14');
diff --git a/data/sql/file.sql b/data/sql/file.sql
index e240f88b8..cf01619f2 100644
--- a/data/sql/file.sql
+++ b/data/sql/file.sql
@@ -1,131 +1,11 @@
 INSERT INTO `file` (`id`, `creator`, `creation_date`, `filepath`, `filesize`, `filename_original`, `extension`, `mime_type`, `md5_hash`, `type`, `access_policy`) VALUES
-(1, 16, '2017-06-18 14:37:03', 'https://www.openml.org/data/download/1666876/phpFsFYVN', 84089, 'anneal.arff', 'arff', 'application/octet-stream', '4eaed8b6ec9d8211024b6c089b064761', 'url', 'public'),
-(2, 16, '2017-06-18 14:37:04', 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 489987, 'kr-vs-kp.arff', 'arff', 'text/plain;charset=UTF-8', 'ad6eb32b7492524d4382a40e23cdbb8e', 'url', 'public'),
-(3, 16, '2017-06-18 14:37:05', 'https://www.openml.org/data/download/6/dataset_6_letter.arff', 720419, 'letter.arff', 'arff', 'text/plain;charset=UTF-8', '9d8a79dccd72f429b67b88387e994db8', 'url', 'public'),
-(4, 16, '2017-06-18 14:37:06', 'https://www.openml.org/data/download/11/dataset_11_balance-scale.arff', 8724, 'balance-scale.arff', 'arff', 'text/plain;charset=UTF-8', '76938608d472f620c170cef9c8c1fa65', 'url', 'public'),
-(5, 16, '2017-06-18 14:37:07', 'https://www.openml.org/data/download/12/dataset_12_mfeat-factors.arff', 1475655, 'mfeat-factors.arff', 'arff', 'text/plain;charset=UTF-8', '2c55b6bb1ad6eaad86d6e6bc0a1d4e1b', 'url', 'public'),
-(6, 16, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/14/dataset_14_mfeat-fourier.arff', 1680757, 'mfeat-fourier.arff', 'arff', 'text/plain;charset=UTF-8', '498da46a7c832efe8914bf52abdfb87b', 'url', 'public'),
-(7, 16, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/52350/openml_phpJNxH0q', 19177, 'breast-w.arff', 'arff', 'application/octet-stream', '88633c065209e7a57323d3d7f2d00546', 'url', 'public'),
-(8, 16, '2017-06-18 14:37:09', 'https://www.openml.org/data/download/16/dataset_16_mfeat-karhunen.arff', 1480772, 'mfeat-karhunen.arff', 'arff', 'text/plain;charset=UTF-8', '76cd333768169bd895e87f3f90a07d2c', 'url', 'public'),
-(9, 16, '2017-06-18 14:37:10', 'https://www.openml.org/data/download/18/dataset_18_mfeat-morphological.arff', 149465, 'mfeat-morphological.arff', 'arff', 'text/plain;charset=UTF-8', '5600bd857927b893e2e4430240b2e74a', 'url', 'public'),
-(10, 16, '2017-06-18 14:37:11', 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 975410, 'mfeat-pixel.arff', 'arff', 'text/plain;charset=UTF-8', '0bdca54aadf13d4717810feb12a27c73', 'url', 'public'),
-(11, 16, '2017-06-18 14:37:12', 'https://www.openml.org/data/download/21/dataset_21_car.arff', 55488, 'car.arff', 'arff', 'text/plain;charset=UTF-8', 'ce8a51e4d55a480d91327d64bf09cffc', 'url', 'public'),
-(12, 16, '2017-06-18 14:37:13', 'https://www.openml.org/data/download/22/dataset_22_mfeat-zernike.arff', 1136136, 'mfeat-zernike.arff', 'arff', 'text/plain;charset=UTF-8', '590fe11f6c0eeacd456609f22c4eaf6d', 'url', 'public'),
-(13, 16, '2017-06-18 14:37:14', 'https://www.openml.org/data/download/23/dataset_23_cmc.arff', 33611, 'cmc.arff', 'arff', 'text/plain;charset=UTF-8', '3149646ecff276abac3e892d1556655f', 'url', 'public'),
-(14, 16, '2017-06-18 14:37:15', 'https://www.openml.org/data/download/24/dataset_24_mushroom.arff', 751225, 'mushroom.arff', 'arff', 'text/plain;charset=UTF-8', 'c8e898e453167e76d617eff67d198579', 'url', 'public'),
-(15, 16, '2017-06-18 14:37:16', 'https://www.openml.org/data/download/28/dataset_28_optdigits.arff', 832899, 'optdigits.arff', 'arff', 'text/plain;charset=UTF-8', 'd9d357ab8cfb6732853109e472d9cfce', 'url', 'public'),
-(16, 16, '2017-06-18 14:37:17', 'https://www.openml.org/data/download/29/dataset_29_credit-a.arff', 34347, 'credit-a.arff', 'arff', 'text/plain;charset=UTF-8', 'a948be968270a823fce432227c1aa36e', 'url', 'public'),
-(17, 16, '2017-06-18 14:37:18', 'https://www.openml.org/data/download/31/dataset_31_credit-g.arff', 162291, 'credit-g.arff', 'arff', 'text/plain;charset=UTF-8', '9a475053fed0c26ee95cd4525e50074c', 'url', 'public'),
-(18, 16, '2017-06-18 14:37:19', 'https://www.openml.org/data/download/32/dataset_32_pendigits.arff', 742221, 'pendigits.arff', 'arff', 'text/plain;charset=UTF-8', '6e472ab352f3f62cc4087dc78eda8d3d', 'url', 'public'),
-(19, 16, '2017-06-18 14:37:19', 'https://www.openml.org/data/download/36/dataset_36_segment.arff', 306054, 'segment.arff', 'arff', 'text/plain;charset=UTF-8', '037621812203bfd85c8dab1d7f16ebc6', 'url', 'public'),
-(20, 16, '2017-06-18 14:37:20', 'https://www.openml.org/data/download/37/dataset_37_diabetes.arff', 37419, 'diabetes.arff', 'arff', 'text/plain;charset=UTF-8', '3cbaa3e54586aa88cf6aacb4033e4470', 'url', 'public'),
-(21, 16, '2017-06-18 14:37:21', 'https://www.openml.org/data/download/38/dataset_38_sick.arff', 305843, 'sick.arff', 'arff', 'text/plain;charset=UTF-8', 'ee5cb4b7f41a5f44b0cd234f675ab492', 'url', 'public'),
-(22, 16, '2017-06-18 14:37:22', 'https://www.openml.org/data/download/42/dataset_42_soybean.arff', 203007, 'soybean.arff', 'arff', 'text/plain;charset=UTF-8', '39fa77494bb23a6bc09ca36ede1184a2', 'url', 'public'),
-(23, 16, '2017-06-18 14:37:23', 'https://www.openml.org/data/download/44/dataset_44_spambase.arff', 706994, 'spambase.arff', 'arff', 'text/plain;charset=UTF-8', 'd9ace01aeac3461e326a8e1b2d53fd84', 'url', 'public'),
-(24, 16, '2017-06-18 14:37:24', 'https://www.openml.org/data/download/46/dataset_46_splice.arff', 524138, 'splice.arff', 'arff', 'text/plain;charset=UTF-8', '21a60c8d1b14bbf0f146b4afeda39287', 'url', 'public'),
-(25, 16, '2017-06-18 14:37:25', 'https://www.openml.org/data/download/50/dataset_50_tic-tac-toe.arff', 29762, 'tic-tac-toe.arff', 'arff', 'text/plain;charset=UTF-8', '34b2992c41e5e23c42769817e96305ac', 'url', 'public'),
-(26, 16, '2017-06-18 14:37:26', 'https://www.openml.org/data/download/54/dataset_54_vehicle.arff', 63838, 'vehicle.arff', 'arff', 'text/plain;charset=UTF-8', 'fbba18157b188f309d772f9ca4e578f5', 'url', 'public'),
-(27, 16, '2017-06-18 14:37:27', 'https://www.openml.org/data/download/60/dataset_60_waveform-5000.arff', 1080193, 'waveform-5000.arff', 'arff', 'text/plain;charset=UTF-8', '996d42ae027122265c642efb845897ef', 'url', 'public'),
-(28, 16, '2017-06-18 14:37:28', 'https://www.openml.org/data/download/2419/electricity-normalized.arff', 3092013, 'electricity.arff', 'arff', 'text/plain;charset=UTF-8', '8ca97867d960ae029ae3a9ac2c923d34', 'url', 'public'),
-(29, 16, '2017-06-18 14:37:29', 'https://www.openml.org/data/download/3619/dataset_186_satimage.arff', 2218190, 'satimage.arff', 'arff', 'text/plain;charset=UTF-8', '72857195d1f6c4c2171c386594f870b2', 'url', 'public'),
-(30, 16, '2017-06-18 14:37:29', 'https://www.openml.org/data/download/3620/dataset_187_abalone.arff', 192330, 'abalone.arff', 'arff', 'text/plain;charset=UTF-8', 'a9555b1cfe39eb55df2c520f09172a96', 'url', 'public'),
-(31, 16, '2017-06-18 14:37:30', 'https://www.openml.org/data/download/3625/dataset_194_eucalyptus.arff', 78938, 'eucalyptus.arff', 'arff', 'text/plain;charset=UTF-8', 'adaf2938bd76922c010d3d32592bdb80', 'url', 'public'),
-(32, 16, '2017-06-18 14:37:31', 'https://www.openml.org/data/download/52405/phpB0xrNj', 31485310, 'isolet.arff', 'arff', 'application/octet-stream', '7c00e71486690813d4a60181bcbca7f6', 'url', 'public'),
-(33, 16, '2017-06-18 14:37:32', 'https://www.openml.org/data/download/52210/phpd8EoD9', 78659, 'vowel.arff', 'arff', 'application/octet-stream', '9c82432df30166fc62353fe4f8bb237a', 'url', 'public'),
-(34, 16, '2017-06-18 14:37:33', 'https://www.openml.org/data/download/1390080/phpuZu33P', 6211769, 'scene.arff', 'arff', 'application/octet-stream', '8ceccf81e895a6520a78bf6f756dde17', 'url', 'public'),
-(35, 16, '2017-06-18 14:37:34', 'https://www.openml.org/data/download/52236/phpAyyBys', 11090, 'monks-problems-1.arff', 'arff', 'application/octet-stream', '6cd008dccee6a34420c091dfe7cdb457', 'url', 'public'),
-(36, 16, '2017-06-18 14:37:35', 'https://www.openml.org/data/download/52237/php4fATLZ', 11720, 'monks-problems-2.arff', 'arff', 'application/octet-stream', '6fdd89666158107428442fb0f8847117', 'url', 'public'),
-(37, 16, '2017-06-18 14:37:36', 'https://www.openml.org/data/download/52238/phphZierv', 11062, 'monks-problems-3.arff', 'arff', 'application/octet-stream', '16e0d4dc4b7b0ad65c2ac1b0c0bebf57', 'url', 'public'),
-(38, 16, '2017-06-18 14:37:37', 'https://www.openml.org/data/download/52415/JapaneseVowels.arff', 1225033, 'JapaneseVowels.arff', 'arff', 'text/plain;charset=UTF-8', '8b59d000ae5310afb26ec66e97fc7724', 'url', 'public'),
-(39, 16, '2017-06-18 14:37:38', 'https://www.openml.org/data/download/52417/synthetic_control.arff', 299973, 'synthetic_control.arff', 'arff', 'text/plain;charset=UTF-8', '8762c53ac2c0a8f584915b06ed044611', 'url', 'public'),
-(40, 16, '2017-06-18 14:37:39', 'https://www.openml.org/data/download/52563/irish.arff', 39678, 'irish.arff', 'arff', 'text/plain;charset=UTF-8', 'b7f36a00bc8f9e2ab73d530b6f3ccb3d', 'url', 'public'),
-(41, 16, '2017-06-18 14:37:40', 'https://www.openml.org/data/download/52570/analcatdata_authorship.arff', 145060, 'analcatdata_authorship.arff', 'arff', 'text/plain;charset=UTF-8', 'bc8ba9aaf1615e1f0659bd4399a0f151', 'url', 'public'),
-(42, 16, '2017-06-18 14:37:40', 'https://www.openml.org/data/download/52581/analcatdata_dmft.arff', 23321, 'analcatdata_dmft.arff', 'arff', 'text/plain;charset=UTF-8', 'ad86226880052117a3e95479cf61faf6', 'url', 'public'),
-(43, 16, '2017-06-18 14:37:41', 'https://www.openml.org/data/download/52582/profb.arff', 27019, 'profb.arff', 'arff', 'text/plain;charset=UTF-8', '87e2345053544e0dc4352ee91add043c', 'url', 'public'),
-(44, 16, '2017-06-18 14:37:42', 'https://www.openml.org/data/download/52590/collins.arff', 62092, 'collins.arff', 'arff', 'text/plain;charset=UTF-8', '1956225e2fa96728d4d351f6147655e6', 'url', 'public'),
-(45, 16, '2017-06-18 14:37:43', 'https://www.openml.org/data/download/53919/sylva_agnostic.arff', 7363098, 'sylva_agnostic.arff', 'arff', 'text/plain;charset=UTF-8', 'f33e876c937d70387e5e0d2a4308bbf6', 'url', 'public'),
-(46, 16, '2017-06-18 14:37:44', 'https://www.openml.org/data/download/53921/gina_agnostic.arff', 8544666, 'gina_agnostic.arff', 'arff', 'text/plain;charset=UTF-8', 'e8b42f51e0a453e4b77c24f8e2569948', 'url', 'public'),
-(47, 16, '2017-06-18 14:37:46', 'https://www.openml.org/data/download/53926/ada_agnostic.arff', 488587, 'ada_agnostic.arff', 'arff', 'text/plain;charset=UTF-8', 'bdff0c6d2cfd9ce53a823464e136b3bd', 'url', 'public'),
-(48, 16, '2017-06-18 14:37:47', 'https://www.openml.org/data/download/53929/mozilla4.arff', 394305, 'mozilla4.arff', 'arff', 'text/plain;charset=UTF-8', '55aa6b33c54391f35bdd0f1c5a247c22', 'url', 'public'),
-(49, 16, '2017-06-18 14:37:48', 'https://www.openml.org/data/download/53932/pc4.arff', 184294, 'pc4.arff', 'arff', 'text/plain;charset=UTF-8', '907e332da09127c2b25620051ff7dbd3', 'url', 'public'),
-(50, 16, '2017-06-18 14:37:49', 'https://www.openml.org/data/download/53933/pc3.arff', 204104, 'pc3.arff', 'arff', 'text/plain;charset=UTF-8', 'a93c30f6c357d3ee74bccb247b10ef68', 'url', 'public'),
-(51, 16, '2017-06-18 14:37:50', 'https://www.openml.org/data/download/53936/jm1.arff', 861589, 'jm1.arff', 'arff', 'text/plain;charset=UTF-8', '0c19aefce52ccb955d0de57e4c0fe8bf', 'url', 'public'),
-(52, 16, '2017-06-18 14:37:51', 'https://www.openml.org/data/download/53946/kc2.arff', 55315, 'kc2.arff', 'arff', 'text/plain;charset=UTF-8', '99e1aabb687c9bc20225f2899746832f', 'url', 'public'),
-(53, 16, '2017-06-18 14:37:52', 'https://www.openml.org/data/download/53950/kc1.arff', 167423, 'kc1.arff', 'arff', 'text/plain;charset=UTF-8', 'bcc004464934075ecaf54d84e45c427a', 'url', 'public'),
-(54, 16, '2017-06-18 14:37:52', 'https://www.openml.org/data/download/53951/pc1.arff', 105904, 'pc1.arff', 'arff', 'text/plain;charset=UTF-8', '32f17b71c893473d8e0645d268796794', 'url', 'public'),
-(55, 16, '2017-06-18 14:37:54', 'https://www.openml.org/data/download/53995/KDDCup09_churn.arff', 34903843, 'KDDCup09_churn.arff', 'arff', 'text/plain;charset=UTF-8', '3cf8299181a0742893a10660e53f9d0a', 'url', 'public'),
-(56, 16, '2017-06-18 14:37:55', 'https://www.openml.org/data/download/53997/KDDCup09_upselling.arff', 34903723, 'KDDCup09_upselling.arff', 'arff', 'text/plain;charset=UTF-8', '7c517e830d9ff93e5e0437e3c58d9939', 'url', 'public'),
-(57, 16, '2017-06-18 14:37:56', 'https://www.openml.org/data/download/53999/musk.arff', 4684142, 'musk.arff', 'arff', 'text/plain;charset=UTF-8', '08a7f184b6cb2375d5b2b75bbeaaf611', 'url', 'public'),
-(58, 16, '2017-06-18 14:37:56', 'https://www.openml.org/data/download/54003/MagicTelescope.arff', 1580808, 'MagicTelescope.arff', 'arff', 'text/plain;charset=UTF-8', 'f6dee76810bd7211716fee7b2b0317cb', 'url', 'public'),
-(59, 16, '2017-06-18 14:37:57', 'https://www.openml.org/data/download/116567/phpCzcrGG', 20512636, 'Internet-Advertisements.arff', 'arff', 'application/octet-stream', '8407920c6389c88ab74dcaf5c7aa62a1', 'url', 'public'),
-(60, 16, '2017-06-18 14:37:58', 'https://www.openml.org/data/download/184157/phpfGCaQC', 2886080, 'Click_prediction_small.arff', 'arff', 'application/octet-stream', '2d0de4e34d32f9baf601ca3c917a9187', 'url', 'public'),
-(61, 16, '2017-06-18 14:37:59', 'https://www.openml.org/data/download/1586212/phpPQrHPH', 254223, 'artificial-characters.arff', 'arff', 'application/octet-stream', '9f872114cc6f0cf59ea4e3315b581a83', 'url', 'public'),
-(62, 16, '2017-06-18 14:38:00', 'https://www.openml.org/data/download/1586218/phpkIxskf', 3660226, 'bank-marketing.arff', 'arff', 'application/octet-stream', 'b29d2eb78e508569fd41172c97bda00e', 'url', 'public'),
-(63, 16, '2017-06-18 14:38:01', 'https://www.openml.org/data/download/1586223/php50jXam', 46233, 'banknote-authentication.arff', 'arff', 'application/octet-stream', 'baa2dc5b745775a943ebeb9c276401f8', 'url', 'public'),
-(64, 16, '2017-06-18 14:38:02', 'https://www.openml.org/data/download/1586225/php0iVrYT', 14143, 'blood-transfusion-service-center.arff', 'arff', 'application/octet-stream', 'c3242468edab8c2e7a907674122dc851', 'url', 'public'),
-(65, 16, '2017-06-18 14:38:03', 'https://www.openml.org/data/download/1586231/php9HX2u8', 211323, 'cardiotocography.arff', 'arff', 'application/octet-stream', 'b0cca84576bffc429099af497278763c', 'url', 'public'),
-(66, 16, '2017-06-18 14:38:04', 'https://www.openml.org/data/download/1586232/phpXeun7q', 90804, 'climate-model-simulation-crashes.arff', 'arff', 'application/octet-stream', 'c3d91dc521c59323cdfba6336a5d59a0', 'url', 'public'),
-(67, 16, '2017-06-18 14:38:05', 'https://www.openml.org/data/download/1586233/phpmcGu2X', 1871624, 'cnae-9.arff', 'arff', 'application/octet-stream', '76e93d20cebbeb462d01358a94a464e2', 'url', 'public'),
-(68, 16, '2017-06-18 14:38:06', 'https://www.openml.org/data/download/1587924/phplE7q6h', 1682156, 'eeg-eye-state.arff', 'arff', 'application/octet-stream', '32086b7bec4daaa9cbe5f19efa63368c', 'url', 'public'),
-(69, 16, '2017-06-18 14:38:07', 'https://www.openml.org/data/download/1587932/phpPbCMyg', 2685554, 'first-order-theorem-proving.arff', 'arff', 'application/octet-stream', '18085f8fa7a64f3870327ae4d9ff4123', 'url', 'public'),
-(70, 16, '2017-06-18 14:38:08', 'https://www.openml.org/data/download/1588715/phpbL6t4U', 17291112, 'gas-drift.arff', 'arff', 'application/octet-stream', 'e205264c230a3e15c7546c18ddceedaf', 'url', 'public'),
-(71, 16, '2017-06-18 14:38:09', 'https://www.openml.org/data/download/1589271/php88ZB4Q', 55364318, 'har.arff', 'arff', 'application/octet-stream', '7b1e71116a88aa1d88b43e697f7c67ee', 'url', 'public'),
-(72, 16, '2017-06-18 14:38:10', 'https://www.openml.org/data/download/1590101/php3isjYz', 838568, 'hill-valley.arff', 'arff', 'application/octet-stream', '347b591c7006712869fcae81a88dc0db', 'url', 'public'),
-(73, 16, '2017-06-18 14:38:11', 'https://www.openml.org/data/download/1590565/phpOJxGL9', 23485, 'ilpd.arff', 'arff', 'application/octet-stream', '155295549a459387fb71312caf1b8360', 'url', 'public'),
-(74, 16, '2017-06-18 14:38:12', 'https://www.openml.org/data/download/1590986/phpfLuQE4', 5217140, 'madelon.arff', 'arff', 'application/octet-stream', '9a4a0c7fa94e6f824962b4c6fbd1812c', 'url', 'public'),
-(75, 16, '2017-06-18 14:38:13', 'https://www.openml.org/data/download/1592278/phpDYCOet', 20200787, 'nomao.arff', 'arff', 'application/octet-stream', '8fc1ac73fbe5236892e166f9f24d7221', 'url', 'public'),
-(76, 16, '2017-06-18 14:38:14', 'https://www.openml.org/data/download/1592279/phpdReP6S', 901788, 'ozone-level-8hr.arff', 'arff', 'application/octet-stream', '6e6b7aeb382e30ae79149e4e5845cf2d', 'url', 'public'),
-(77, 16, '2017-06-18 14:38:15', 'https://www.openml.org/data/download/1592281/php8Mz7BG', 267467, 'phoneme.arff', 'arff', 'application/octet-stream', '902ff27649dbfb512f3fb01fc05c6b24', 'url', 'public'),
-(78, 16, '2017-06-18 14:38:16', 'https://www.openml.org/data/download/1592283/phpCsX3fx', 749631, 'one-hundred-plants-margin.arff', 'arff', 'application/octet-stream', '90d3ec061a62e527702faf31865c3c7e', 'url', 'public'),
-(79, 16, '2017-06-18 14:38:17', 'https://www.openml.org/data/download/1592284/php0FyS2T', 916577, 'one-hundred-plants-shape.arff', 'arff', 'application/octet-stream', 'ee55a0b769901a266e14b44588eaf31a', 'url', 'public'),
-(80, 16, '2017-06-18 14:38:18', 'https://www.openml.org/data/download/1592285/phpoOxxNn', 654948, 'one-hundred-plants-texture.arff', 'arff', 'application/octet-stream', 'cfe6cb4bd3cc847e38e61f7fd941307b', 'url', 'public'),
-(81, 16, '2017-06-18 14:38:19', 'https://www.openml.org/data/download/1592286/phpGUrE90', 154166, 'qsar-biodeg.arff', 'arff', 'application/octet-stream', 'a2c189cd65511103fa540d7186155c24', 'url', 'public'),
-(82, 16, '2017-06-18 14:38:20', 'https://www.openml.org/data/download/1592289/phpVeNa5j', 712551, 'wall-robot-navigation.arff', 'arff', 'application/octet-stream', '05d4fd7921f662a911dc41eafb5a5183', 'url', 'public'),
-(83, 16, '2017-06-18 14:38:21', 'https://www.openml.org/data/download/1592293/phptd5jYj', 825064, 'semeion.arff', 'arff', 'application/octet-stream', '0fb35a1d7db2f76c8587f9125abcf048', 'url', 'public'),
-(84, 16, '2017-06-18 14:38:22', 'https://www.openml.org/data/download/1592296/php9xWOpn', 298345, 'steel-plates-fault.arff', 'arff', 'application/octet-stream', '7917ace4246897272fc35e17131f1b38', 'url', 'public'),
-(85, 16, '2017-06-18 14:38:22', 'https://www.openml.org/data/download/1592297/phpvFYBg2', 1051629, 'tamilnadu-electricity.arff', 'arff', 'application/octet-stream', 'ac042ad58b470cb42e6355f1f13277b9', 'url', 'public'),
-(86, 16, '2017-06-18 14:38:23', 'https://www.openml.org/data/download/1592318/phpAmSP4g', 120617, 'wdbc.arff', 'arff', 'application/octet-stream', '7aa183d3657e364911ced0cbd6b272bd', 'url', 'public'),
-(87, 16, '2017-06-18 14:38:24', 'https://www.openml.org/data/download/1593707/phpHyLSNF', 2098230, 'micro-mass.arff', 'arff', 'application/octet-stream', '3eb0ad5e560b082f3f4bc2955ecdbec5', 'url', 'public'),
-(88, 16, '2017-06-18 14:38:25', 'https://www.openml.org/data/download/1675983/phpnThNfi', 246811, 'wilt.arff', 'arff', 'application/octet-stream', 'f56f95bac28580e0750b515885efbf49', 'url', 'public'),
-(89, 16, '2017-06-18 14:38:26', 'https://www.openml.org/data/download/1595261/phpMawTba', 5962724, 'adult.arff', 'arff', 'application/octet-stream', 'bb6510925e5d4b23d136715febb2cdf5', 'url', 'public'),
-(90, 16, '2017-06-18 14:38:27', 'https://www.openml.org/data/download/1681097/phpSSK7iA', 30223214, 'Bioresponse.arff', 'arff', 'text/xml;charset=UTF-8', 'f34816f743ec2f535dd92b8e5e01bb59', 'url', 'public'),
-(91, 16, '2017-06-18 14:38:28', 'https://www.openml.org/data/download/1681098/phpmPOD5A', 2134522, 'Amazon_employee_access.arff', 'arff', 'text/xml;charset=UTF-8', '5502ee73e563ddf5b699ffe0c633905b', 'url', 'public'),
-(92, 16, '2017-06-18 14:38:29', 'https://www.openml.org/data/download/1798106/phpV5QYya', 800920, 'PhishingWebsites.arff', 'arff', 'text/xml;charset=UTF-8', 'fa52215c3e8caba4b765afa33bc92657', 'url', 'public'),
-(93, 16, '2017-06-18 14:38:30', 'https://www.openml.org/data/download/1798765/phpYLeydd', 3506053, 'GesturePhaseSegmentationProcessed.arff', 'arff', 'text/xml;charset=UTF-8', '31a6e4f8cdae2b8141c1f754e0366007', 'url', 'public'),
-(94, 16, '2017-06-18 14:38:31', 'https://www.openml.org/data/download/1804243/phpO6CpB2', 1034699, 'MiceProtein.arff', 'arff', 'application/octet-stream', '3c479a6885bfa0438971388283a1ce32', 'url', 'public'),
-(95, 16, '2017-06-18 14:38:32', 'https://www.openml.org/data/download/1854224/phpAz9Len', 113973, 'cylinder-bands.arff', 'arff', 'application/octet-stream', '59ec3d17d1fb9fa79c3108a7a5a9bc5e', 'url', 'public'),
-(96, 16, '2017-06-18 14:38:33', 'https://www.openml.org/data/download/1910442/phpDAC5gS', 277989, 'cjs.arff', 'arff', 'application/octet-stream', 'e7dc5bf5b5c97173222f38db03e205fb', 'url', 'public'),
-(97, 16, '2017-06-18 14:38:34', 'https://www.openml.org/data/download/1910507/phpcFPMhq', 37842, 'dresses-sales.arff', 'arff', 'application/octet-stream', '996f7aacb2eab289cc1a968e263f5281', 'url', 'public'),
-(98, 16, '2017-06-18 14:38:35', 'https://www.openml.org/data/download/4535757/phpSj3fWL', 8274, 'LED-display-domain-7digit.arff', 'arff', 'text/xml;charset=UTF-8', '4509a210be57d02a8e468ad8c95ea277', 'url', 'public'),
-(99, 16, '2017-06-18 14:38:35', 'https://www.openml.org/data/download/4535764/phpBDgUyY', 1524277, 'texture.arff', 'arff', 'text/xml;charset=UTF-8', '47abf0594dbe574a3b04347e89e88a7a', 'url', 'public'),
-(100, 16, '2017-06-18 14:38:36', 'https://www.openml.org/data/download/4599631/phpjP8PU3', 52965, 'Australian.arff', 'arff', 'application/octet-stream', 'a910a2411e614ac874aacf83badc616f', 'url', 'public'),
-(101, 16, '2017-06-18 14:38:37', 'https://www.openml.org/data/download/52413/UNIX_user_data.arff', 1128106, 'UNIX_user_data.arff', 'arff', 'text/plain;charset=UTF-8', 'aa9fc5189e4d4d7430f2e55fcbae07e8', 'url', 'public'),
-(102, 16, '2017-06-18 14:38:38', 'https://www.openml.org/data/download/1594035/phpeLklTK', 25400, 'fourclass_scale.arff', 'arff', 'application/octet-stream', 'f323df25d3c1c7bfefacab580f5fbf81', 'url', 'public'),
-(103, 16, '2017-06-18 14:38:39', 'https://www.openml.org/data/download/3626/dataset_2175_kin8nm.arff', 1131204, 'kin8nm.arff', 'arff', 'text/plain;charset=UTF-8', 'b36414070701ab8bc985104c3ebd9b77', 'url', 'public'),
-(104, 16, '2017-06-18 14:38:40', 'https://www.openml.org/data/download/3627/dataset_2176_mbagrade.arff', 1015, 'mbagrade.arff', 'arff', 'text/plain;charset=UTF-8', 'e9395ccfd63dfcc9ca19eac8e285b10e', 'url', 'public'),
-(105, 16, '2017-06-18 14:38:40', 'https://www.openml.org/data/download/3628/dataset_2177_wisconsin.arff', 49308, 'wisconsin.arff', 'arff', 'text/plain;charset=UTF-8', '2d2241b4081fc0cb7f68b9a0a2104b05', 'url', 'public'),
-(106, 16, '2017-06-18 14:38:41', 'https://www.openml.org/data/download/3629/dataset_2178_vineyard.arff', 1163, 'vineyard.arff', 'arff', 'text/plain;charset=UTF-8', 'b261b616bccee824fb531e85007f2f90', 'url', 'public'),
-(107, 16, '2017-06-18 14:38:42', 'https://www.openml.org/data/download/3630/dataset_2179_bolts.arff', 4550, 'bolts.arff', 'arff', 'text/plain;charset=UTF-8', '9d119bd75be61765a1dd11feb8971c01', 'url', 'public'),
-(108, 16, '2017-06-18 14:38:43', 'https://www.openml.org/data/download/3631/dataset_2180_cleveland.arff', 21923, 'cleveland.arff', 'arff', 'text/plain;charset=UTF-8', 'aca21f966ac25ead1ac09fbb2c4e9dd6', 'url', 'public'),
-(109, 16, '2017-06-18 14:38:43', 'https://www.openml.org/data/download/3632/dataset_2181_auto_price.arff', 16564, 'auto_price.arff', 'arff', 'text/plain;charset=UTF-8', 'aa1f382817cccea56ebd5c4e2aaf871a', 'url', 'public'),
-(110, 16, '2017-06-18 14:38:44', 'https://www.openml.org/data/download/3633/dataset_2182_autoMpg.arff', 13325, 'autoMpg.arff', 'arff', 'text/plain;charset=UTF-8', 'fefd80d00774553658b04b29c97cb3ca', 'url', 'public'),
-(111, 16, '2017-06-18 14:38:45', 'https://www.openml.org/data/download/3634/dataset_2183_cpu_act.arff', 1058120, 'cpu_act.arff', 'arff', 'text/plain;charset=UTF-8', 'c48cbb08c17633f2db3a78633301c81b', 'url', 'public'),
-(112, 16, '2017-06-18 14:38:46', 'https://www.openml.org/data/download/52348/openml_phpmlDSTj', 331386, 'delta_elevators.arff', 'arff', 'application/octet-stream', '5092640d992fd31f2cf3f8d4fd08f8c9', 'url', 'public'),
-(113, 16, '2017-06-18 14:38:46', 'https://www.openml.org/data/download/3636/dataset_2185_fruitfly.arff', 8732, 'fruitfly.arff', 'arff', 'text/plain;charset=UTF-8', '7cc9ff041dfe3f02a5fe9f3afa899f2a', 'url', 'public'),
-(114, 16, '2017-06-18 14:38:47', 'https://www.openml.org/data/download/3637/dataset_2186_pbc.arff', 32954, 'pbc.arff', 'arff', 'text/plain;charset=UTF-8', '173be63cc24294e384f95f66913b5e18', 'url', 'public'),
-(115, 16, '2017-06-18 14:38:48', 'https://www.openml.org/data/download/3638/dataset_2187_pol.arff', 1726721, 'pol.arff', 'arff', 'text/plain;charset=UTF-8', 'd4f09062c182b5a5f911c6a59ed51fb0', 'url', 'public'),
-(116, 16, '2017-06-18 14:38:49', 'https://www.openml.org/data/download/3639/dataset_2188_autoHorse.arff', 30049, 'autoHorse.arff', 'arff', 'text/plain;charset=UTF-8', '5a561e70570035123f832423cf71ef0d', 'url', 'public'),
-(117, 16, '2017-06-18 14:38:49', 'https://www.openml.org/data/download/3640/dataset_2189_lowbwt.arff', 9058, 'lowbwt.arff', 'arff', 'text/plain;charset=UTF-8', 'f4a1a8578d9ac3ce0d003b4ed50dc9d2', 'url', 'public'),
-(118, 16, '2017-06-18 14:38:50', 'https://www.openml.org/data/download/3641/dataset_2190_cholesterol.arff', 22358, 'cholesterol.arff', 'arff', 'text/plain;charset=UTF-8', 'd058ba412634187a53ae376aaf45f925', 'url', 'public'),
-(119, 16, '2017-06-18 14:38:51', 'https://www.openml.org/data/download/3642/dataset_2191_sleep.arff', 3842, 'sleep.arff', 'arff', 'text/plain;charset=UTF-8', 'a45c303e3a998ef235987422baecb5ae', 'url', 'public'),
-(120, 16, '2017-06-18 14:38:52', 'https://www.openml.org/data/download/3643/dataset_2192_triazines.arff', 82105, 'triazines.arff', 'arff', 'text/plain;charset=UTF-8', '247b1ececf935a519a3f3e9669939795', 'url', 'public'),
-(121, 16, '2017-06-18 14:38:53', 'https://www.openml.org/data/download/3644/dataset_2193_autoPrice.arff', 16843, 'autoPrice.arff', 'arff', 'text/plain;charset=UTF-8', 'b1eb15c4ac30ab5a477ab48ec02ff63b', 'url', 'public'),
-(122, 16, '2017-06-18 14:38:53', 'https://www.openml.org/data/download/3645/dataset_2194_detroit.arff', 3626, 'detroit.arff', 'arff', 'text/plain;charset=UTF-8', '4c8aff8bbff7361387bf8ec32f9706d7', 'url', 'public'),
-(123, 16, '2017-06-18 14:38:54', 'https://www.openml.org/data/download/3646/dataset_2195_quake.arff', 44449, 'quake.arff', 'arff', 'text/plain;charset=UTF-8', '7ede4fd775db9eae5586b2f55c6d98c6', 'url', 'public'),
-(124, 16, '2017-06-18 14:38:55', 'https://www.openml.org/data/download/3647/dataset_2196_cloud.arff', 4564, 'cloud.arff', 'arff', 'text/plain;charset=UTF-8', '656f567557f538b4a1a2e271a8ae19fe', 'url', 'public'),
-(125, 16, '2017-06-18 14:38:56', 'https://www.openml.org/data/download/3648/dataset_2197_longley.arff', 1401, 'longley.arff', 'arff', 'text/plain;charset=UTF-8', 'a6d46686d199b6cb82dbdf0e34377b8a', 'url', 'public'),
-(126, 16, '2017-06-18 14:38:57', 'https://www.openml.org/data/download/3649/dataset_2198_diabetes_numeric.arff', 1460, 'diabetes_numeric.arff', 'arff', 'text/plain;charset=UTF-8', '39ee6f17581653e92c6d0b70ffac263a', 'url', 'public'),
-(127, 16, '2017-06-18 14:38:57', 'https://www.openml.org/data/download/3650/dataset_2199_pharynx.arff', 13471, 'pharynx.arff', 'arff', 'text/plain;charset=UTF-8', '0bdc5060f67578a2f715e541ee5d0138', 'url', 'public'),
-(128, 16, '2018-05-29 14:00:00', 'https://www.openml.org/data/download/61/iris.arff', 7487, 'iris.arff', 'arff', 'text/plain;charset=UTF-8', 'ad484452702105cbf3d30f8deaba39a9', 'url', 'public'),
-(129, 16, '2018-05-29 14:00:00', 'https://www.openml.org/data/download/19330175/iris-challenge.arff', 7103, 'iris-challenge.arff', 'arff', 'text/plain;charset=UTF-8', 'bce91bc6033c311070d198e78e7fc954', 'url', 'public'),
-(130, 16, '2018-05-29 14:00:00', 'https://www.openml.org/data/download/61/iris.arff', 7487, 'iris.arff', 'arff', 'text/plain;charset=UTF-8', 'ad484452702105cbf3d30f8deaba39a9', 'url', 'private');
+(1, 1, '2017-06-18 14:37:03', 'https://www.openml.org/data/download/1666876/phpFsFYVN', 84089, 'anneal.arff', 'arff', 'application/octet-stream', '4eaed8b6ec9d8211024b6c089b064761', 'url', 'public'),
+(2, 1, '2017-06-18 14:37:04', 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 489987, 'kr-vs-kp.arff', 'arff', 'text/plain;charset=UTF-8', 'ad6eb32b7492524d4382a40e23cdbb8e', 'url', 'public'),
+(3, 1, '2017-06-18 14:37:05', 'https://www.openml.org/data/download/6/dataset_6_letter.arff', 720419, 'letter.arff', 'arff', 'text/plain;charset=UTF-8', '9d8a79dccd72f429b67b88387e994db8', 'url', 'public'),
+(4, 1, '2017-06-18 14:37:06', 'https://www.openml.org/data/download/11/dataset_11_balance-scale.arff', 8724, 'balance-scale.arff', 'arff', 'text/plain;charset=UTF-8', '76938608d472f620c170cef9c8c1fa65', 'url', 'public'),
+(5, 1, '2017-06-18 14:37:07', 'https://www.openml.org/data/download/12/dataset_12_mfeat-factors.arff', 1475655, 'mfeat-factors.arff', 'arff', 'text/plain;charset=UTF-8', '2c55b6bb1ad6eaad86d6e6bc0a1d4e1b', 'url', 'public'),
+(6, 1, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/14/dataset_14_mfeat-fourier.arff', 1680757, 'mfeat-fourier.arff', 'arff', 'text/plain;charset=UTF-8', '498da46a7c832efe8914bf52abdfb87b', 'url', 'public'),
+(7, 1, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/52350/openml_phpJNxH0q', 19177, 'breast-w.arff', 'arff', 'application/octet-stream', '88633c065209e7a57323d3d7f2d00546', 'url', 'public'),
+(8, 1, '2017-06-18 14:37:09', 'https://www.openml.org/data/download/16/dataset_16_mfeat-karhunen.arff', 1480772, 'mfeat-karhunen.arff', 'arff', 'text/plain;charset=UTF-8', '76cd333768169bd895e87f3f90a07d2c', 'url', 'public'),
+(9, 1, '2017-06-18 14:37:10', 'https://www.openml.org/data/download/18/dataset_18_mfeat-morphological.arff', 149465, 'mfeat-morphological.arff', 'arff', 'text/plain;charset=UTF-8', '5600bd857927b893e2e4430240b2e74a', 'url', 'public'),
+(10, 1, '2017-06-18 14:37:11', 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 975410, 'mfeat-pixel.arff', 'arff', 'text/plain;charset=UTF-8', '0bdca54aadf13d4717810feb12a27c73', 'url', 'public');
diff --git a/data/sql/implementation_tag.sql b/data/sql/implementation_tag.sql
index c4c00ed78..424913996 100644
--- a/data/sql/implementation_tag.sql
+++ b/data/sql/implementation_tag.sql
@@ -1,15 +1,15 @@
 INSERT INTO `implementation_tag` (`id`, `tag`, `uploader`, `date`) VALUES
-(1, 'OpenmlWeka', 16, '2017-03-24 14:26:38'),
-(1, 'weka', 16, '2017-03-24 14:26:38'),
-(2, 'OpenmlWeka', 16, '2017-03-24 14:26:39'),
-(2, 'weka', 16, '2017-03-24 14:26:39'),
-(3, 'OpenmlWeka', 16, '2017-03-24 14:26:40'),
-(3, 'weka', 16, '2017-03-24 14:26:40'),
-(4, 'OpenmlWeka', 16, '2017-03-24 14:26:40'),
-(4, 'weka', 16, '2017-03-24 14:26:40'),
-(5, 'OpenmlWeka', 16, '2017-03-24 14:26:41'),
-(5, 'weka', 16, '2017-03-24 14:26:41'),
-(6, 'OpenmlWeka', 16, '2017-03-24 14:26:42'),
+(1, 'OpenmlWeka', 1, '2017-03-24 14:26:38'),
+(1, 'weka', 1, '2017-03-24 14:26:38'),
+(2, 'OpenmlWeka', 1, '2017-03-24 14:26:39'),
+(2, 'weka', 1, '2017-03-24 14:26:39'),
+(3, 'OpenmlWeka', 1, '2017-03-24 14:26:40'),
+(3, 'weka', 1, '2017-03-24 14:26:40'),
+(4, 'OpenmlWeka', 1, '2017-03-24 14:26:40'),
+(4, 'weka', 1, '2017-03-24 14:26:40'),
+(5, 'OpenmlWeka', 1, '2017-03-24 14:26:41'),
+(5, 'weka', 1, '2017-03-24 14:26:41'),
+(6, 'OpenmlWeka', 1, '2017-03-24 14:26:42'),
 (6, 'weka', 1, '2017-03-24 14:26:42'),
 (7, 'OpenmlWeka', 1, '2017-03-24 14:26:42'),
 (7, 'weka', 1, '2017-03-24 14:26:42'),
diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index 25c5fe9e7..00c17adcf 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -36,6 +36,7 @@ function __construct() {
     $this->load->model('Task_type_inout');
     $this->load->model('Estimation_procedure');
     $this->load->model('Run');
+    $this->load->model('Users');
     $this->load->Library('elasticSearch');
 
     $this->dir_suffix = 'dataset/cron/';
@@ -97,34 +98,36 @@ public function create_es_openml_index() {
 
   // Create admin user
   public function create_local_admin() {
-    echo "\r\nCreating local admin user...";
+    echo "\r\Fetching local admin user from user table...";
+
+    $session_hash = $this->Users->getColumnWhere('session_hash', 'id = ' . 1);
   
-    $username="admin";
-    $email="admin";
-    $session_hash='0123456789abcdef0123456789abcdef';
-    $additional_data = array(
-      'first_name' => "Admin",
-      'last_name'  => "",
-      'affiliation'=> "",
-      'country'    => "",
-      'bio'    	 => "",
-      'external_source' => null,
-      'external_id' => null,
-      'session_hash' => $session_hash
-    );
+    // $username="admin";
+    // $email="admin";
+    // $session_hash=md5(rand());
+    // $additional_data = array(
+    //   'first_name' => "Admin",
+    //   'last_name'  => "",
+    //   'affiliation'=> "",
+    //   'country'    => "",
+    //   'bio'    	 => "",
+    //   'external_source' => null,
+    //   'external_id' => null,
+    //   'session_hash' => $session_hash
+    // );
   
-    $password=substr(md5(uniqid()), 0, 12);    
-    echo "\r\nPassword:\t" .$password;
-    echo "\r\nAPI key:\t" .$session_hash;
+    // $password=substr(md5(uniqid()), 0, 12);    
+    // echo "\r\nPassword:\t" .$password;
+    // echo "\r\nAPI key:\t" .$session_hash;
 
-    echo "\r\nSaving key in shared config...";
+    echo "\r\nSaving key in shared config...config/api_key.txt";
     file_put_contents("/openmlconfig/api_key.txt",$session_hash);
    
-    $this->load->Library('ion_auth');
+    // $this->load->Library('ion_auth');
 
-    # Add to admin group
-    $group = array('1');
-    //$adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
+    // # Add to admin group
+    // $group = array('1');
+    // $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
     
     echo "\r\nDone! ";
     

From 5aefe929ef96ff83c1bb56ee3a4d82ee9dcefc57 Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Thu, 24 Dec 2020 13:33:29 +0100
Subject: [PATCH 09/15] add password and user name prints

---
 data/sql/users.sql             |  6 ++--
 openml_OS/controllers/Cron.php | 59 +++++++++++++++++++++-------------
 2 files changed, 40 insertions(+), 25 deletions(-)

diff --git a/data/sql/users.sql b/data/sql/users.sql
index 1c4ef64af..443d07323 100644
--- a/data/sql/users.sql
+++ b/data/sql/users.sql
@@ -1,3 +1,3 @@
-INSERT INTO `users` (`id`, `username`, `password`, `email`, `session_hash`, `gamification_visibility`) VALUES
-(1, 'admin@openml.org', MD5('admin'), 'admin@openml.org', '0123456789abcdef0123456789abcdef', 'show'),
-(2, 'user@openml.org', MD5('user'), 'user@openml.org', '00112233445566778899aabbccddeeff', 'show');
+INSERT INTO `users` (`id`, `username`, `password`, `email`, `active`,`session_hash`, `gamification_visibility`) VALUES
+(1, 'admin@openml.org', MD5('admin'), 'admin@openml.org', 1,'0123456789abcdef0123456789abcdef', 'show'),
+(2, 'user@openml.org', MD5('user'), 'user@openml.org', 1, '00112233445566778899aabbccddeeff', 'show');
diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index 00c17adcf..db6f819b6 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -99,13 +99,19 @@ public function create_es_openml_index() {
   // Create admin user
   public function create_local_admin() {
     echo "\r\Fetching local admin user from user table...";
-
-    $session_hash = $this->Users->getColumnWhere('session_hash', 'id = ' . 1);
-  
-    // $username="admin";
-    // $email="admin";
-    // $session_hash=md5(rand());
-    // $additional_data = array(
+ 
+    $record = $this->Users->getWhereSingle('id = ' . 1);  
+    $username= $record->username;
+    $session_hash=md5(rand());
+    $password = "admin";    
+    $this->Users->update(1, array('session_hash' => $session_hash));
+    echo "\r\nUser Name:\t" .$username; 
+    echo "\r\nPassword:\t" . $password;
+    echo "\r\nAPI key:\t" .$session_hash;
+    echo "\r\nSaving key in shared config...config/api_key.txt";
+    file_put_contents("/openmlconfig/api_key.txt",$session_hash);     
+    echo "\r\nDone! ";
+    // $user_data = array(
     //   'first_name' => "Admin",
     //   'last_name'  => "",
     //   'affiliation'=> "",
@@ -113,24 +119,33 @@ public function create_local_admin() {
     //   'bio'    	 => "",
     //   'external_source' => null,
     //   'external_id' => null,
-    //   'session_hash' => $session_hash
+    //   'session_hash' => $session_hash,
+    //   'created_on' => time(),
+    //   'active' => 1
     // );
-  
-    // $password=substr(md5(uniqid()), 0, 12);    
-    // echo "\r\nPassword:\t" .$password;
-    // echo "\r\nAPI key:\t" .$session_hash;
-
-    echo "\r\nSaving key in shared config...config/api_key.txt";
-    file_put_contents("/openmlconfig/api_key.txt",$session_hash);
-   
-    // $this->load->Library('ion_auth');
-
-    // # Add to admin group
-    // $group = array('1');
-    // $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
+    //$password=substr(md5(uniqid()), 0, 12);
+    // $this->load->Library('ion_auth'); 
+    // $hash = $this->ion_auth->hash_password($username, $password);
     
-    echo "\r\nDone! ";
+    // if ($hash === FALSE)
+    // {
+    //   $password = "admin";
+    // }
+    // else
+    // {
+    //   $data = [
+    //   'password' => $hash,
+    //   'remember_code' => NULL,
+    //   'forgotten_password_code' => NULL,
+    //   'forgotten_password_time' => NULL
+    // ];
+    //   $this->Users->update(1, $data);
+    // }
     
+    
+     // # Add to admin group
+    // $group = array('1');
+    // $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
   }
 
   

From 07f80c2f24198faf929440d180cbc96985ce89bf Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Thu, 24 Dec 2020 15:02:04 +0100
Subject: [PATCH 10/15] remove update

---
 openml_OS/controllers/Cron.php | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index db6f819b6..396cb1346 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -102,9 +102,10 @@ public function create_local_admin() {
  
     $record = $this->Users->getWhereSingle('id = ' . 1);  
     $username= $record->username;
-    $session_hash=md5(rand());
+    $session_hash = $record->session_hash;
+    // $session_hash=md5(rand());
     $password = "admin";    
-    $this->Users->update(1, array('session_hash' => $session_hash));
+    // $this->Users->update(1, array('session_hash' => $session_hash));
     echo "\r\nUser Name:\t" .$username; 
     echo "\r\nPassword:\t" . $password;
     echo "\r\nAPI key:\t" .$session_hash;

From d7c483f7f33d1c76ad83fb225c92373302d2269d Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Tue, 5 Jan 2021 00:01:49 +0100
Subject: [PATCH 11/15] add single dataset sample

---
 data/dataset_structured/0/11/dataset.arff | 3351 +++++++++++++++++++++
 data/sql/dataset.sql                      |   11 +-
 data/sql/dataset_tag.sql                  |   11 +-
 data/sql/file.sql                         |   11 +-
 4 files changed, 3354 insertions(+), 30 deletions(-)
 create mode 100644 data/dataset_structured/0/11/dataset.arff

diff --git a/data/dataset_structured/0/11/dataset.arff b/data/dataset_structured/0/11/dataset.arff
new file mode 100644
index 000000000..1738ca896
--- /dev/null
+++ b/data/dataset_structured/0/11/dataset.arff
@@ -0,0 +1,3351 @@
+% 1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7
+%    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square
+%    away from queening.  It is the King+Rook's side (white) to move.
+% 
+% 2. Sources:
+%     (a) Database originally generated and described by Alen Shapiro.
+%     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database
+%         was supplied to Holte by Peter Clark of the Turing Institute
+%         in Glasgow (pete@turing.ac.uk).
+%     (c) Date: 1 August 1989
+% 
+% 3. Past Usage:
+%      - Alen D. Shapiro (1983,1987), "Structured Induction in Expert Systems",
+%        Addison-Wesley.  This book is based on Shapiro's Ph.D. thesis (1983)
+%        at the University of Edinburgh entitled "The Role of Structured
+%        Induction in Expert Systems".
+%      - Stephen Muggleton (1987), "Structuring Knowledge by Asking Questions",
+%        pp.218-229 in "Progress in Machine Learning", edited by I. Bratko
+%        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.
+%      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),
+%        "Concept Learning and the Problem of Small Disjuncts",
+%        Proceedings of IJCAI.  Also available as technical report AI89-106,
+%        Computer Sciences Department, University of Texas at Austin,
+%        Austin, Texas 78712.
+% 
+% 4. Relevant Information:
+%       The dataset format is described below.  Note: the format of this
+%     database was modified on 2/26/90 to conform with the format of all
+%     the other databases in the UCI repository of machine learning databases.
+% 
+% 5. Number of Instances: 3196 total
+% 
+% 6. Number of Attributes: 36
+% 
+% 7. Attribute Summaries:
+%     Classes (2):  -- White-can-win ("won") and White-cannot-win ("nowin").
+%           I believe that White is deemed to be unable to win if the Black pawn
+%           can safely advance.
+%     Attributes: see Shapiro's book.
+% 
+% 8. Missing Attributes: --  none
+% 
+% 9. Class Distribution:
+%     In 1669 of the positions (52%), White can win.
+%     In 1527 of the positions (48%), White cannot win.
+% 
+% The format for instances in this database is a sequence of 37 attribute values.
+% Each instance is a board-descriptions for this chess endgame.  The first
+% 36 attributes describe the board.  The last (37th) attribute is the
+% classification: "win" or "nowin".  There are 0 missing values.
+% A typical board-description is
+% 
+% f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won
+% 
+% The names of the features do not appear in the board-descriptions.
+% Instead, each feature correponds to a particular position in the
+% feature-value list.  For example, the head of this list is the value
+% for the feature "bkblk".  The following is the list of features, in
+% the order in which their values appear in the feature-value list:
+% 
+% [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,
+%  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,
+%  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]
+% 
+% In the file, there is one instance (board position) per line.
+% 
+% 
+% Num Instances:     3196
+% Num Attributes:    37
+% Num Continuous:    0 (Int 0 / Real 0)
+% Num Discrete:      37
+% Missing values:    0 /  0.0%
+%
+%     name                      type enum ints real     missing    distinct  (1)
+%   1 'bkblk'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   2 'bknwy'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   3 'bkon8'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   4 'bkona'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   5 'bkspr'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   6 'bkxbq'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   7 'bkxcr'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   8 'bkxwp'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%   9 'blxwp'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  10 'bxqsq'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  11 'cntxt'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  12 'dsopp'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  13 'dwipd'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  14 'hdchk'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  15 'katri'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% 
+%  16 'mulch'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  17 'qxmsq'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  18 'r2ar8'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  19 'reskd'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  20 'reskr'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  21 'rimmx'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  22 'rkxwp'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  23 'rxmsq'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  24 'simpl'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  25 'skach'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  26 'skewr'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  27 'skrxp'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  28 'spcop'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  29 'stlmt'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  30 'thrsk'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  31 'wkcti'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  32 'wkna8'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  33 'wknck'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  34 'wkovl'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  35 'wkpos'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  36 'wtoeg'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%  37 'class'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% 
+%
+%
+@relation kr-vs-kp
+@attribute 'bkblk' {'t','f'}
+@attribute 'bknwy' {'t','f'}
+@attribute 'bkon8' {'t','f'}
+@attribute 'bkona' {'t','f'}
+@attribute 'bkspr' {'t','f'}
+@attribute 'bkxbq' {'t','f'}
+@attribute 'bkxcr' {'t','f'}
+@attribute 'bkxwp' {'t','f'}
+@attribute 'blxwp' {'t','f'}
+@attribute 'bxqsq' {'t','f'}
+@attribute 'cntxt' {'t','f'}
+@attribute 'dsopp' {'t','f'}
+@attribute 'dwipd' {'g','l'}
+@attribute 'hdchk' {'t','f'}
+@attribute 'katri' {'b','n','w'}
+@attribute 'mulch' {'t','f'}
+@attribute 'qxmsq' {'t','f'}
+@attribute 'r2ar8' {'t','f'}
+@attribute 'reskd' {'t','f'}
+@attribute 'reskr' {'t','f'}
+@attribute 'rimmx' {'t','f'}
+@attribute 'rkxwp' {'t','f'}
+@attribute 'rxmsq' {'t','f'}
+@attribute 'simpl' {'t','f'}
+@attribute 'skach' {'t','f'}
+@attribute 'skewr' {'t','f'}
+@attribute 'skrxp' {'t','f'}
+@attribute 'spcop' {'t','f'}
+@attribute 'stlmt' {'t','f'}
+@attribute 'thrsk' {'t','f'}
+@attribute 'wkcti' {'t','f'}
+@attribute 'wkna8' {'t','f'}
+@attribute 'wknck' {'t','f'}
+@attribute 'wkovl' {'t','f'}
+@attribute 'wkpos' {'t','f'}
+@attribute 'wtoeg' {'n','t','f'}
+@attribute 'class' {'won','nowin'}
+@data
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'t','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','t','f','l','f','w','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','f','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','t','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','f','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'t','f','f','f','t','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','t','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','t','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','t','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','t','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','t','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','b','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','t','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','t','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','t','t','f','f','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','t','t','f','f','f','f','f','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','g','f','b','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','f','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','g','f','n','f','t','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','t','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','t','f','t','t','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','t','f','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','g','f','b','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','g','f','n','f','t','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','t','t','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','w','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','t','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','t','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','t','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','t','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','t','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','t','t','t','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','w','t','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','t','t','f','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','t','t','f','f','f','f','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','f','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','t','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','t','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','t','t','f','f','f','f','f','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','t','f','f','f','f','f','f','f','f','f','f','l','f','b','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','f','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','t','t','f','f','f','t','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','w','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','t','t','f','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','t','f','t','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','b','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','t','f','t','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','t','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','t','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','f','t','f','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','f','t','t','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','w','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','t','f','t','t','t','t','f','t','f','l','f','b','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','t','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','b','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','g','f','n','f','t','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','g','f','n','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','t','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','w','f','t','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','f','t','t','t','f','t','t','l','f','n','f','f','t','f','t','f','f','f','t','f','t','t','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','f','f','f','t','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','t','t','f','t','f','f','t','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','t','t','f','t','f','t','t','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','t','t','f','t','f','f','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','t','t','f','f','f','t','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','t','t','f','t','f','t','t','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','t','f','f','t','f','t','t','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'t','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','w','f','f','f','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','w','f','t','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','w','f','t','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','w','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','w','f','t','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','n','f','f','t','f','t','f','t','t','f','f','t','f','f','f','f','f','f','f','t','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','n','won'
+'f','t','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','t','t','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','t','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','t','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','f','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','f','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','t','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','t','f','f','f','f','f','t','f','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','t','t','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','n','f','f','f','f','t','t','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','t','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','w','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','won'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','t','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','t','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','t','t','f','l','f','n','f','f','t','f','t','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','t','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','t','f','t','t','t','t','t','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','t','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','won'
+'f','f','t','f','f','f','t','f','f','t','f','t','l','f','n','f','f','t','f','t','t','f','f','t','f','t','f','f','f','f','f','t','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','t','l','f','w','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','t','f','f','t','t','t','t','f','t','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','t','t','t','l','f','n','f','t','t','f','f','t','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','t','t','f','t','t','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','t','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','t','t','t','t','t','t','t','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','t','f','t','l','f','n','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','t','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','t','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','n','t','f','f','f','t','t','t','f','t','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','n','t','f','f','f','t','t','t','f','t','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','f','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','n','t','f','f','f','t','t','t','f','f','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','n','t','f','f','f','t','t','t','f','f','f','t','f','f','f','t','t','f','t','f','f','n','won'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','w','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','w','t','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','w','t','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','t','f','f','f','f','t','f','f','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','t','f','f','f','f','t','f','f','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','t','f','f','f','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','w','t','f','f','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','t','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','t','t','f','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','t','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','t','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','t','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'t','f','f','f','t','t','f','f','t','f','t','f','l','f','w','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','t','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','won'
+'f','t','f','f','f','t','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','f','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','t','f','t','f','f','t','f','t','f','l','f','n','f','f','f','f','t','t','t','f','f','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','t','f','t','f','t','f','f','t','f','t','f','l','f','n','t','f','f','f','t','t','t','f','f','f','t','t','f','f','t','t','f','t','f','f','n','won'
+'f','f','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','w','f','f','f','f','t','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','t','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','t','f','t','f','f','t','f','t','f','l','f','w','f','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','f','f','t','f','f','t','f','t','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','t','f','t','f','t','f','f','t','f','t','f','l','f','w','t','f','f','f','t','t','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','won'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','won'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','t','f','f','f','t','f','t','f','t','n','won'
+'t','f','f','f','t','t','f','f','f','f','t','f','l','f','w','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','f','f','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','t','f','t','t','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','g','f','n','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','t','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','n','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','t','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','t','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','t','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','t','f','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','t','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','t','t','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','t','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','t','f','f','f','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','f','t','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','f','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','t','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','won'
+'f','t','f','f','f','f','f','f','t','f','f','t','l','f','n','t','f','f','f','f','t','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','won'
+'f','f','f','f','f','t','f','f','f','f','f','t','l','f','n','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','won'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','b','f','t','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','t','f','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','t','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','g','f','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','f','f','t','t','t','t','t','f','g','f','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','t','f','t','t','t','t','t','f','g','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','t','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','f','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','t','t','t','t','t','t','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','t','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'t','f','f','f','f','t','f','t','t','t','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','b','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','b','t','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','b','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','t','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','b','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','t','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','t','f','t','f','f','f','f','f','t','f','f','l','f','n','t','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','t','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','t','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','t','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','t','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','n','t','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','t','f','l','f','w','t','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','b','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','t','f','f','t','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','b','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','f','f','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','b','f','f','f','f','t','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','t','f','t','t','t','t','t','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','n','t','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','f','f','t','t','t','t','t','f','l','f','n','t','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'t','f','f','f','t','f','t','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','b','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','t','f','t','f','f','f','t','t','f','l','f','b','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','t','f','f','t','t','f','f','t','t','f','l','f','b','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','t','f','t','f','t','f','f','f','t','t','f','l','f','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','t','f','t','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','t','n','nowin'
+'f','t','f','t','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','t','t','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','t','t','t','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','t','n','nowin'
+'f','f','f','t','f','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'t','f','f','f','f','t','f','t','t','t','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','t','f','t','t','t','t','t','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','t','f','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','t','f','t','f','f','f','t','f','f','l','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','t','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','g','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','t','l','f','w','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','t','l','f','n','f','t','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','t','t','t','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','t','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','t','t','f','t','f','f','t','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','t','t','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','t','t','f','f','t','t','t','l','f','n','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','t','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','t','t','f','t','t','t','t','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','t','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','f','t','f','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','t','f','t','t','t','f','f','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','t','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','t','f','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','f','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','t','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','t','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','t','f','t','f','f','f','t','f','f','l','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','t','f','f','t','t','f','f','t','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','t','f','t','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','t','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','t','f','t','t','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','f','t','t','t','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','t','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','t','f','t','l','f','n','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','g','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','f','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','t','g','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','t','f','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','f','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','g','f','b','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','t','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','f','t','f','f','f','f','f','f','f','t','f','g','t','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','t','t','f','f','f','f','f','f','f','t','f','g','t','b','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','t','f','f','f','f','f','f','t','f','g','t','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','t','f','t','f','f','f','f','f','f','t','f','g','t','b','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','t','f','f','f','f','f','f','f','t','f','g','t','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','t','t','f','f','f','f','f','f','f','t','f','g','t','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','f','f','t','f','f','f','f','f','f','t','f','g','t','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'t','t','f','t','f','f','f','f','f','f','t','f','g','t','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','g','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','t','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','t','f','t','t','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','b','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','b','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','b','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','t','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','t','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','t','f','f','f','f','f','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'t','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','t','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','t','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','t','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','t','t','f','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','f','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','t','f','f','f','f','t','f','t','f','f','f','t','t','t','f','f','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','t','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'t','t','f','f','f','f','f','f','f','f','f','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','t','f','f','f','f','f','f','f','f','f','f','l','f','b','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','t','f','f','f','n','nowin'
+'t','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','f','f','f','t','f','f','f','t','f','f','f','l','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','f','t','f','t','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','t','t','f','f','f','l','f','w','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','l','f','n','f','f','t','f','t','f','t','t','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','l','f','n','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','t','t','f','f','f','n','nowin'
+'t','t','t','f','f','f','f','f','f','f','f','f','l','t','b','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','t','t','f','f','f','f','f','f','f','f','f','l','t','n','t','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','t','f','t','f','f','f','f','f','f','f','f','l','t','n','t','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','t','f','t','f','f','f','f','f','f','f','f','l','t','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'t','f','f','f','f','f','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','t','f','t','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','t','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','t','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','b','t','f','t','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','b','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','f','f','t','t','t','f','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','n','t','f','t','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','t','f','l','f','n','t','f','f','t','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','t','f','f','t','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','t','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','t','t','t','f','f','f','t','f','l','f','n','f','f','f','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','t','f','t','f','l','f','n','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','f','t','f','f','n','nowin'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','t','f','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','f','t','f','t','f','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','t','f','f','t','t','f','f','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','t','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','t','f','t','f','t','f','f','f','f','f','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','f','f','t','f','t','f','t','f','f','f','t','t','f','f','t','t','f','t','t','f','n','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','t','f','t','t','t','t','f','t','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','t','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','f','t','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','t','t','t','t','t','t','f','t','f','l','f','w','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','n','nowin'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'t','f','f','f','f','t','f','f','f','f','t','f','l','f','n','f','t','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','t','t','f','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','g','f','n','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','t','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','g','f','n','f','f','t','f','f','f','f','t','t','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','g','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','f','t','f','f','f','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','f','f','f','f','t','t','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','t','f','t','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','n','f','f','t','f','t','f','t','f','t','f','t','t','f','f','t','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','f','t','f','f','f','t','t','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','t','f','t','t','l','f','n','f','f','t','f','t','f','t','t','f','f','t','f','f','f','f','t','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','f','f','f','t','t','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','t','t','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','t','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','n','f','f','t','f','t','f','t','t','f','f','t','f','f','f','f','f','f','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','f','f','t','t','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','l','f','n','f','f','t','f','t','f','t','t','f','f','t','f','f','f','f','t','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','t','t','l','f','n','f','f','f','f','t','f','f','t','f','f','t','f','f','f','f','f','f','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','t','l','f','n','f','f','f','f','t','f','f','t','t','f','t','f','f','f','f','t','t','f','f','f','n','nowin'
+'t','f','f','t','f','f','t','t','t','f','f','t','l','t','n','t','f','t','f','t','f','f','f','t','f','t','f','f','f','f','f','t','f','f','f','n','nowin'
+'f','f','f','f','f','t','t','f','t','f','t','t','l','f','n','f','f','t','f','t','f','t','f','t','f','t','f','f','f','f','t','f','t','t','f','n','nowin'
+'f','f','f','f','f','t','t','f','f','f','t','t','l','f','n','f','f','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','t','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','t','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','t','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','t','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','t','f','f','f','f','f','f','f','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','g','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','t','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','f','t','f','t','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','n','f','t','t','f','f','f','f','t','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','t','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','t','f','f','f','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','t','f','f','f','f','f','f','f','t','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','t','f','t','t','t','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','t','f','f','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','f','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','f','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','t','f','t','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','t','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','t','t','t','t','f','f','f','l','f','n','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','f','t','f','f','f','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','t','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','g','f','n','f','f','t','f','f','f','t','f','f','f','f','t','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','t','t','f','f','t','f','f','t','f','t','t','nowin'
+'f','f','f','f','t','f','f','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','t','f','f','f','f','f','f','f','f','f','t','l','f','n','t','f','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','t','l','f','n','f','f','f','f','f','f','f','f','f','f','f','t','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','t','t','t','nowin'
+'f','f','f','f','f','t','f','f','t','f','f','t','l','f','n','f','f','f','f','f','f','t','f','f','f','f','f','f','f','f','f','f','t','f','t','t','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','g','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','g','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','f','g','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','f','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','l','f','w','f','f','t','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','f','f','f','f','f','f','l','f','w','f','f','f','f','t','f','f','f','t','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+'f','f','f','t','f','f','f','f','f','f','f','f','l','f','w','f','f','f','t','t','f','f','f','t','f','t','f','f','t','t','f','t','f','f','f','n','nowin'
+'f','f','f','f','t','f','t','f','t','f','f','f','l','f','n','f','f','f','f','f','f','t','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','f','f','f','f','f','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','f','f','f','f','f','f','l','f','w','f','f','f','f','t','f','f','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','t','f','f','f','l','f','w','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','t','t','f','f','f','l','f','w','f','f','t','f','f','f','t','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','f','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','f','l','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','t','g','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','g','f','b','f','f','t','f','f','f','f','t','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','g','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','g','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','t','f','t','f','t','f','f','t','l','f','b','f','f','t','f','f','f','t','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','t','f','t','f','t','f','f','t','l','f','n','f','f','t','f','f','f','t','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','f','t','f','t','f','t','f','f','t','l','f','n','f','f','f','f','f','f','t','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','t','f','t','f','t','f','f','t','l','f','b','f','f','f','f','f','f','t','f','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','t','t','t','f','f','t','l','f','w','f','f','t','f','f','f','f','f','f','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+'f','f','f','t','f','f','t','t','t','f','f','t','l','f','w','f','f','t','f','f','f','t','f','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'f','f','f','t','f','f','t','t','t','f','f','t','l','f','w','f','f','f','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+'f','f','f','f','f','f','t','f','f','f','f','t','l','f','n','f','f','t','f','t','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','l','f','b','f','f','t','f','f','f','f','t','t','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','l','f','b','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','f','f','f','f','t','f','f','f','f','t','l','f','n','f','f','t','f','f','f','f','t','f','f','t','f','f','t','f','f','t','f','t','f','n','nowin'
+'t','f','t','f','f','f','t','f','f','f','f','t','l','t','b','t','f','t','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+'t','f','t','f','f','f','t','f','f','f','f','t','l','t','n','t','f','t','f','f','f','f','f','t','f','t','f','f','t','f','f','t','f','f','f','n','nowin'
+%
+%
+%
diff --git a/data/sql/dataset.sql b/data/sql/dataset.sql
index 654edd1cd..cbce168c1 100644
--- a/data/sql/dataset.sql
+++ b/data/sql/dataset.sql
@@ -1,11 +1,2 @@
 INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_label`, `description`, `format`, `creator`, `contributor`, `collection_date`, `upload_date`, `language`, `licence`, `citation`, `collection`, `url`, `isOriginal`, `file_id`, `default_target_attribute`, `row_id_attribute`, `ignore_attribute`, `paper_url`, `visibility`, `original_data_id`, `original_data_url`, `update_comment`, `last_update`) VALUES
-(1, 1, 0, 'anneal', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Annealing Data\n \n 2. Source Information: donated by David Sterling and Wray Buntine.\n \n 3. Past Usage: unknown\n \n 4. Relevant Information:\n    -- Explanation: I suspect this was left by Ross Quinlan in 1987 at the\n       4th Machine Learning Workshop.  I\'d have to check with Jeff Schlimmer\n       to double check this.\n \n 5. Number of Instances: 798\n \n 6. Number of Attributes: 38\n    -- 6 continuously-valued\n    -- 3 integer-valued\n    -- 29 nominal-valued\n \n 7. Attribute Information:\n     1. family:          --,GB,GK,GS,TN,ZA,ZF,ZH,ZM,ZS\n     2. product-type:    C, H, G\n     3. steel:           -,R,A,U,K,M,S,W,V\n     4. carbon:          continuous\n     5. hardness:        continuous\n     6. temper_rolling:  -,T\n     7. condition:       -,S,A,X\n     8. formability:     -,1,2,3,4,5\n     9. strength:        continuous\n    10. non-ageing:      -,N\n    11. surface-finish:  P,M,-\n    12. surface-quality: -,D,E,F,G\n    13. enamelability:   -,1,2,3,4,5\n    14. bc:              Y,-\n    15. bf:              Y,-\n    16. bt:              Y,-\n    17. bw/me:           B,M,-\n    18. bl:              Y,-\n    19. m:               Y,-\n    20. chrom:           C,-\n    21. phos:            P,-\n    22. cbond:           Y,-\n    23. marvi:           Y,-\n    24. exptl:           Y,-\n    25. ferro:           Y,-\n    26. corr:            Y,-\n    27. blue/bright/varn/clean:          B,R,V,C,-\n    28. lustre:          Y,-\n    29. jurofm:          Y,-\n    30. s:               Y,-\n    31. p:               Y,-\n    32. shape:           COIL, SHEET\n    33. thick:           continuous\n    34. width:           continuous\n    35. len:             continuous\n    36. oil:             -,Y,N\n    37. bore:            0000,0500,0600,0760\n    38. packing: -,1,2,3\n    classes:        1,2,3,4,5,U\n  \n    -- The \'-\' values are actually \'not_applicable\' values rather than\n       \'missing_values\' (and so can be treated as legal discrete\n       values rather than as showing the absence of a discrete value).\n \n 8. Missing Attribute Values: Signified with \"?\"\n    Attribute:  Number of instances missing its value:\n    1           0\n    2           0\n    3           70\n    4           0\n    5           0\n    6           675\n    7           271\n    8           283\n    9           0\n   10           703\n   11           790\n   12           217\n   13           785\n   14           797\n   15           680\n   16           736\n   17           609\n   18           662\n   19           798\n   20           775\n   21           791\n   22           730\n   23           798\n   24           796\n   25           772\n   26           798\n   27           793\n   28           753\n   29           798\n   30           798\n   31           798\n   32           0\n   33           0\n   34           0\n   35           0\n   36           740\n   37           0\n   38           789\n   39           0\n \n 9. Distribution of Classes\n      Class Name:   Number of Instances:\n      1               8\n      2              88\n      3             608\n      4               0\n      5              60\n      U              34\n                    ---\n                    798', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:24', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/1666876/phpFsFYVN', 'true', 1, 'class', NULL, NULL, NULL, 'public', NULL, NULL, 'Restoring dataset file', '2015-09-02 00:49:04'),
-(2, 1, 0, 'kr-vs-kp', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7\n    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square\n    away from queening.  It is the King+Rook\'s side (white) to move.\n \n 2. Sources:\n     (a) Database originally generated and described by Alen Shapiro.\n     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database\n         was supplied to Holte by Peter Clark of the Turing Institute\n         in Glasgow (pete@turing.ac.uk).\n     (c) Date: 1 August 1989\n \n 3. Past Usage:\n      - Alen D. Shapiro (1983,1987), \"Structured Induction in Expert Systems\",\n        Addison-Wesley.  This book is based on Shapiro\'s Ph.D. thesis (1983)\n        at the University of Edinburgh entitled \"The Role of Structured\n        Induction in Expert Systems\".\n      - Stephen Muggleton (1987), \"Structuring Knowledge by Asking Questions\",\n        pp.218-229 in \"Progress in Machine Learning\", edited by I. Bratko\n        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.\n      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),\n        \"Concept Learning and the Problem of Small Disjuncts\",\n        Proceedings of IJCAI.  Also available as technical report AI89-106,\n        Computer Sciences Department, University of Texas at Austin,\n        Austin, Texas 78712.\n \n 4. Relevant Information:\n       The dataset format is described below.  Note: the format of this\n     database was modified on 2/26/90 to conform with the format of all\n     the other databases in the UCI repository of machine learning databases.\n \n 5. Number of Instances: 3196 total\n \n 6. Number of Attributes: 36\n \n 7. Attribute Summaries:\n     Classes (2):  -- White-can-win (\"won\") and White-cannot-win (\"nowin\").\n           I believe that White is deemed to be unable to win if the Black pawn\n           can safely advance.\n     Attributes: see Shapiro\'s book.\n \n 8. Missing Attributes: --  none\n \n 9. Class Distribution:\n     In 1669 of the positions (52%), White can win.\n     In 1527 of the positions (48%), White cannot win.\n \n The format for instances in this database is a sequence of 37 attribute values.\n Each instance is a board-descriptions for this chess endgame.  The first\n 36 attributes describe the board.  The last (37th) attribute is the\n classification: \"win\" or \"nowin\".  There are 0 missing values.\n A typical board-description is\n \n f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won\n \n The names of the features do not appear in the board-descriptions.\n Instead, each feature correponds to a particular position in the\n feature-value list.  For example, the head of this list is the value\n for the feature \"bkblk\".  The following is the list of features, in\n the order in which their values appear in the feature-value list:\n \n [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,\n  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,\n  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]\n \n In the file, there is one instance (board position) per line.\n \n \n Num Instances:     3196\n Num Attributes:    37\n Num Continuous:    0 (Int 0 / Real 0)\n Num Discrete:      37\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'bkblk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   2 \'bknwy\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   3 \'bkon8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'bkona\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'bkspr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'bkxbq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'bkxcr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'bkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'blxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'bxqsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'cntxt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'dsopp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'dwipd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'hdchk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'katri\'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n  16 \'mulch\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'qxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'r2ar8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  19 \'reskd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'reskr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  21 \'rimmx\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'rkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  23 \'rxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'simpl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  25 \'skach\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'skewr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  27 \'skrxp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  28 \'spcop\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  29 \'stlmt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  30 \'thrsk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  31 \'wkcti\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  32 \'wkna8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  33 \'wknck\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  34 \'wkovl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  35 \'wkpos\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  36 \'wtoeg\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  37 \'class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 'true', 2, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:28'),
-(3, 1, 0, 'letter', '1', '1', '**Author**: David J. Slate  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Letter+Recognition) - 01-01-1991  \r\n**Please cite**: P. W. Frey and D. J. Slate. \"Letter Recognition Using Holland-style Adaptive Classifiers\". Machine Learning 6(2), 1991  \r\n\r\n1. TITLE: \r\n  Letter Image Recognition Data \r\n \r\n    The objective is to identify each of a large number of black-and-white\r\n    rectangular pixel displays as one of the 26 capital letters in the English\r\n    alphabet.  The character images were based on 20 different fonts and each\r\n    letter within these 20 fonts was randomly distorted to produce a file of\r\n    20,000 unique stimuli.  Each stimulus was converted into 16 primitive\r\n    numerical attributes (statistical moments and edge counts) which were then\r\n    scaled to fit into a range of integer values from 0 through 15.  We\r\n    typically train on the first 16000 items and then use the resulting model\r\n    to predict the letter category for the remaining 4000.  See the article\r\n    cited above for more details.', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:41', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/6/dataset_6_letter.arff', 'true', 3, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:41'),
-(4, 1, 0, 'balance-scale', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Balance Scale Weight & Distance Database\n\n2. Source Information: \n    (a) Source: Generated to model psychological experiments reported\n		by Siegler, R. S. (1976).  Three Aspects of Cognitive\n		Development.  Cognitive Psychology, 8, 481-520.\n    (b) Donor: Tim Hume (hume@ics.uci.edu)\n    (c) Date: 22 April 1994\n\n3. Past Usage: (possibly different formats of this data)\n   - Publications\n	1. Klahr, D., & Siegler, R.S. (1978).  The Representation of\n	   Children\'s Knowledge.  In H. W. Reese & L. P. Lipsitt (Eds.),\n	   Advances in Child Development and Behavior, pp. 61-116.  New\n	   York: Academic Press \n	2. Langley,P. (1987).  A General Theory of Discrimination\n	   Learning.  In D. Klahr, P. Langley, & R. Neches (Eds.),\n	   Production System Models of Learning and Development, pp.\n	   99-161. Cambridge, MA: MIT Press\n	3. Newell, A. (1990).  Unified Theories of Cognition.\n	   Cambridge, MA: Harvard University Press\n	4. McClelland, J.L. (1988).  Parallel Distibuted Processing:\n	   Implications for Cognition and Development.  Technical\n	   Report AIP-47, Department of Psychology, Carnegie-Mellon\n	   University \n	5. Shultz, T., Mareschal, D., & Schmidt, W. (1994).  Modeling\n	   Cognitive Development on Balance Scale Phenomena. Machine\n	   Learning, Vol. 16, pp. 59-88.\n\n4. Relevant Information: \n	This data set was generated to model psychological\n	experimental results.  Each example is classified as having the\n	balance scale tip to the right, tip to the left, or be\n	balanced.  The attributes are the left weight, the left\n	distance, the right weight, and the right distance.  The\n	correct way to find the class is the greater of \n	(left-distance * left-weight) and (right-distance *\n	right-weight).  If they are equal, it is balanced.\n\n5. Number of Instances: 625 (49 balanced, 288 left, 288 right)\n\n6. Number of Attributes: 4 (numeric) + class name = 5\n\n7. Attribute Information:\n	1. Class Name: 3 (L, B, R)\n	2. Left-Weight: 5 (1, 2, 3, 4, 5)\n	3. Left-Distance: 5 (1, 2, 3, 4, 5)\n	4. Right-Weight: 5 (1, 2, 3, 4, 5)\n	5. Right-Distance: 5 (1, 2, 3, 4, 5)\n\n8. Missing Attribute Values: \n	none\n\n9. Class Distribution: \n   1. 46.08 percent are L\n   2. 07.84 percent are B\n   3. 46.08 percent are R', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:55', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/11/dataset_11_balance-scale.arff', 'true', 4, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:55'),
-(5, 1, 0, 'mfeat-factors', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:04', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/12/dataset_12_mfeat-factors.arff', 'true', 5, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:04'),
-(6, 1, 0, 'mfeat-fourier', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:17', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/14/dataset_14_mfeat-fourier.arff', 'true', 6, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:17'),
-(7, 1, 0, 'breast-w', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:20', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/52350/openml_phpJNxH0q', 'true', 7, 'Class', NULL, NULL, NULL, 'public', NULL, NULL, 'added special attributes', '2014-09-21 23:04:47'),
-(8, 1, 0, 'mfeat-karhunen', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:30', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/16/dataset_16_mfeat-karhunen.arff', 'true', 8, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:30'),
-(9, 1, 0, 'mfeat-morphological', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (0 - 9)\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class \'0\', followed by sets of 200 patterns\n for each of the classes \'1\' - \'9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:37', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/18/dataset_18_mfeat-morphological.arff', 'true', 9, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:37'),
-(10, 1, 0, 'mfeat-pixel', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:20:48', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 'true', 10, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:20:48');
\ No newline at end of file
+(1, 1, 0, 'kr-vs-kp', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7\n    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square\n    away from queening.  It is the King+Rook\'s side (white) to move.\n \n 2. Sources:\n     (a) Database originally generated and described by Alen Shapiro.\n     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database\n         was supplied to Holte by Peter Clark of the Turing Institute\n         in Glasgow (pete@turing.ac.uk).\n     (c) Date: 1 August 1989\n \n 3. Past Usage:\n      - Alen D. Shapiro (1983,1987), \"Structured Induction in Expert Systems\",\n        Addison-Wesley.  This book is based on Shapiro\'s Ph.D. thesis (1983)\n        at the University of Edinburgh entitled \"The Role of Structured\n        Induction in Expert Systems\".\n      - Stephen Muggleton (1987), \"Structuring Knowledge by Asking Questions\",\n        pp.218-229 in \"Progress in Machine Learning\", edited by I. Bratko\n        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.\n      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),\n        \"Concept Learning and the Problem of Small Disjuncts\",\n        Proceedings of IJCAI.  Also available as technical report AI89-106,\n        Computer Sciences Department, University of Texas at Austin,\n        Austin, Texas 78712.\n \n 4. Relevant Information:\n       The dataset format is described below.  Note: the format of this\n     database was modified on 2/26/90 to conform with the format of all\n     the other databases in the UCI repository of machine learning databases.\n \n 5. Number of Instances: 3196 total\n \n 6. Number of Attributes: 36\n \n 7. Attribute Summaries:\n     Classes (2):  -- White-can-win (\"won\") and White-cannot-win (\"nowin\").\n           I believe that White is deemed to be unable to win if the Black pawn\n           can safely advance.\n     Attributes: see Shapiro\'s book.\n \n 8. Missing Attributes: --  none\n \n 9. Class Distribution:\n     In 1669 of the positions (52%), White can win.\n     In 1527 of the positions (48%), White cannot win.\n \n The format for instances in this database is a sequence of 37 attribute values.\n Each instance is a board-descriptions for this chess endgame.  The first\n 36 attributes describe the board.  The last (37th) attribute is the\n classification: \"win\" or \"nowin\".  There are 0 missing values.\n A typical board-description is\n \n f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won\n \n The names of the features do not appear in the board-descriptions.\n Instead, each feature correponds to a particular position in the\n feature-value list.  For example, the head of this list is the value\n for the feature \"bkblk\".  The following is the list of features, in\n the order in which their values appear in the feature-value list:\n \n [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,\n  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,\n  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]\n \n In the file, there is one instance (board position) per line.\n \n \n Num Instances:     3196\n Num Attributes:    37\n Num Continuous:    0 (Int 0 / Real 0)\n Num Discrete:      37\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'bkblk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   2 \'bknwy\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   3 \'bkon8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'bkona\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'bkspr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'bkxbq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'bkxcr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'bkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'blxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'bxqsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'cntxt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'dsopp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'dwipd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'hdchk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'katri\'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n  16 \'mulch\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'qxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'r2ar8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  19 \'reskd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'reskr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  21 \'rimmx\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'rkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  23 \'rxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'simpl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  25 \'skach\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'skewr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  27 \'skrxp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  28 \'spcop\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  29 \'stlmt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  30 \'thrsk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  31 \'wkcti\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  32 \'wkna8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  33 \'wknck\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  34 \'wkovl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  35 \'wkpos\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  36 \'wtoeg\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  37 \'class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 'true', 1, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:28');
diff --git a/data/sql/dataset_tag.sql b/data/sql/dataset_tag.sql
index db8a02107..5908d334f 100644
--- a/data/sql/dataset_tag.sql
+++ b/data/sql/dataset_tag.sql
@@ -1,11 +1,2 @@
 INSERT INTO `dataset_tag` (`id`, `tag`, `uploader`, `date`) VALUES
-(1, 'study_14', 1, '2017-03-24 19:36:11'),
-(2, 'study_14', 1, '2017-03-24 19:36:12'),
-(3, 'study_14', 1, '2017-03-24 19:36:12'),
-(4, 'study_14', 1, '2017-03-24 19:36:12'),
-(5, 'study_14', 1, '2017-03-24 19:36:12'),
-(6, 'study_14', 1, '2017-03-24 19:36:13'),
-(7, 'study_14', 1, '2017-03-24 19:36:13'),
-(8, 'study_14', 1, '2017-03-24 19:36:13'),
-(9, 'study_14', 1, '2017-03-24 19:36:14'),
-(10, 'study_14', 1, '2017-03-24 19:36:14');
+(1, 'study_14', 1, '2017-03-24 19:36:11');
diff --git a/data/sql/file.sql b/data/sql/file.sql
index cf01619f2..e2e43cba0 100644
--- a/data/sql/file.sql
+++ b/data/sql/file.sql
@@ -1,11 +1,2 @@
 INSERT INTO `file` (`id`, `creator`, `creation_date`, `filepath`, `filesize`, `filename_original`, `extension`, `mime_type`, `md5_hash`, `type`, `access_policy`) VALUES
-(1, 1, '2017-06-18 14:37:03', 'https://www.openml.org/data/download/1666876/phpFsFYVN', 84089, 'anneal.arff', 'arff', 'application/octet-stream', '4eaed8b6ec9d8211024b6c089b064761', 'url', 'public'),
-(2, 1, '2017-06-18 14:37:04', 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 489987, 'kr-vs-kp.arff', 'arff', 'text/plain;charset=UTF-8', 'ad6eb32b7492524d4382a40e23cdbb8e', 'url', 'public'),
-(3, 1, '2017-06-18 14:37:05', 'https://www.openml.org/data/download/6/dataset_6_letter.arff', 720419, 'letter.arff', 'arff', 'text/plain;charset=UTF-8', '9d8a79dccd72f429b67b88387e994db8', 'url', 'public'),
-(4, 1, '2017-06-18 14:37:06', 'https://www.openml.org/data/download/11/dataset_11_balance-scale.arff', 8724, 'balance-scale.arff', 'arff', 'text/plain;charset=UTF-8', '76938608d472f620c170cef9c8c1fa65', 'url', 'public'),
-(5, 1, '2017-06-18 14:37:07', 'https://www.openml.org/data/download/12/dataset_12_mfeat-factors.arff', 1475655, 'mfeat-factors.arff', 'arff', 'text/plain;charset=UTF-8', '2c55b6bb1ad6eaad86d6e6bc0a1d4e1b', 'url', 'public'),
-(6, 1, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/14/dataset_14_mfeat-fourier.arff', 1680757, 'mfeat-fourier.arff', 'arff', 'text/plain;charset=UTF-8', '498da46a7c832efe8914bf52abdfb87b', 'url', 'public'),
-(7, 1, '2017-06-18 14:37:08', 'https://www.openml.org/data/download/52350/openml_phpJNxH0q', 19177, 'breast-w.arff', 'arff', 'application/octet-stream', '88633c065209e7a57323d3d7f2d00546', 'url', 'public'),
-(8, 1, '2017-06-18 14:37:09', 'https://www.openml.org/data/download/16/dataset_16_mfeat-karhunen.arff', 1480772, 'mfeat-karhunen.arff', 'arff', 'text/plain;charset=UTF-8', '76cd333768169bd895e87f3f90a07d2c', 'url', 'public'),
-(9, 1, '2017-06-18 14:37:10', 'https://www.openml.org/data/download/18/dataset_18_mfeat-morphological.arff', 149465, 'mfeat-morphological.arff', 'arff', 'text/plain;charset=UTF-8', '5600bd857927b893e2e4430240b2e74a', 'url', 'public'),
-(10, 1, '2017-06-18 14:37:11', 'https://www.openml.org/data/download/20/dataset_20_mfeat-pixel.arff', 975410, 'mfeat-pixel.arff', 'arff', 'text/plain;charset=UTF-8', '0bdca54aadf13d4717810feb12a27c73', 'url', 'public');
+(1, 1, '2021-01-04 22:01:06', 'dataset_structured//0/11//dataset.arff', 493338, 'dataset.arff', 'arff', 'text/xml', '6f6215ebe8c370b63991015b02afcd1f', 'dataset', 'public');

From b2cd923bf4aee17ae26ede115da47ff99bc586f3 Mon Sep 17 00:00:00 2001
From: sahithyaravi1493 <sahithyaravi1493@gmail.com>
Date: Fri, 8 Jan 2021 13:59:06 +0100
Subject: [PATCH 12/15] fix write model

---
 openml_OS/models/Dataset_topic.php | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/openml_OS/models/Dataset_topic.php b/openml_OS/models/Dataset_topic.php
index 1771e201d..2e523f838 100644
--- a/openml_OS/models/Dataset_topic.php
+++ b/openml_OS/models/Dataset_topic.php
@@ -1,5 +1,5 @@
 <?php
-class Dataset_topic extends Database_write {
+class Dataset_topic extends MY_Database_Write_Model {
 
   function __construct() {
     parent::__construct();

From 2074f642a7e7e6c29b6cd282e6e300d385f3db44 Mon Sep 17 00:00:00 2001
From: Joost van der Gaag <joost@smartfarmsensing.com>
Date: Fri, 8 Jul 2022 12:09:23 +0200
Subject: [PATCH 13/15] Updated SQLand Env Config

Updated SQL insertion files for demo data
Added BASE_CONFIG file for using with environment variables (Docker compose/ AWS) instead of hard coded configurations
---
 data/sql/dataset.sql                       |   4 +-
 data/sql/dataset_description.sql           | 131 +------------
 openml_OS/config/BASE_CONFIG-BLANK_ENV.php | 205 +++++++++++++++++++++
 openml_OS/controllers/Cron.php             |   4 +-
 4 files changed, 210 insertions(+), 134 deletions(-)
 create mode 100644 openml_OS/config/BASE_CONFIG-BLANK_ENV.php

diff --git a/data/sql/dataset.sql b/data/sql/dataset.sql
index cbce168c1..161c0e24e 100644
--- a/data/sql/dataset.sql
+++ b/data/sql/dataset.sql
@@ -1,2 +1,2 @@
-INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_label`, `description`, `format`, `creator`, `contributor`, `collection_date`, `upload_date`, `language`, `licence`, `citation`, `collection`, `url`, `isOriginal`, `file_id`, `default_target_attribute`, `row_id_attribute`, `ignore_attribute`, `paper_url`, `visibility`, `original_data_id`, `original_data_url`, `update_comment`, `last_update`) VALUES
-(1, 1, 0, 'kr-vs-kp', '1', '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7\n    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square\n    away from queening.  It is the King+Rook\'s side (white) to move.\n \n 2. Sources:\n     (a) Database originally generated and described by Alen Shapiro.\n     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database\n         was supplied to Holte by Peter Clark of the Turing Institute\n         in Glasgow (pete@turing.ac.uk).\n     (c) Date: 1 August 1989\n \n 3. Past Usage:\n      - Alen D. Shapiro (1983,1987), \"Structured Induction in Expert Systems\",\n        Addison-Wesley.  This book is based on Shapiro\'s Ph.D. thesis (1983)\n        at the University of Edinburgh entitled \"The Role of Structured\n        Induction in Expert Systems\".\n      - Stephen Muggleton (1987), \"Structuring Knowledge by Asking Questions\",\n        pp.218-229 in \"Progress in Machine Learning\", edited by I. Bratko\n        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.\n      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),\n        \"Concept Learning and the Problem of Small Disjuncts\",\n        Proceedings of IJCAI.  Also available as technical report AI89-106,\n        Computer Sciences Department, University of Texas at Austin,\n        Austin, Texas 78712.\n \n 4. Relevant Information:\n       The dataset format is described below.  Note: the format of this\n     database was modified on 2/26/90 to conform with the format of all\n     the other databases in the UCI repository of machine learning databases.\n \n 5. Number of Instances: 3196 total\n \n 6. Number of Attributes: 36\n \n 7. Attribute Summaries:\n     Classes (2):  -- White-can-win (\"won\") and White-cannot-win (\"nowin\").\n           I believe that White is deemed to be unable to win if the Black pawn\n           can safely advance.\n     Attributes: see Shapiro\'s book.\n \n 8. Missing Attributes: --  none\n \n 9. Class Distribution:\n     In 1669 of the positions (52%), White can win.\n     In 1527 of the positions (48%), White cannot win.\n \n The format for instances in this database is a sequence of 37 attribute values.\n Each instance is a board-descriptions for this chess endgame.  The first\n 36 attributes describe the board.  The last (37th) attribute is the\n classification: \"win\" or \"nowin\".  There are 0 missing values.\n A typical board-description is\n \n f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won\n \n The names of the features do not appear in the board-descriptions.\n Instead, each feature correponds to a particular position in the\n feature-value list.  For example, the head of this list is the value\n for the feature \"bkblk\".  The following is the list of features, in\n the order in which their values appear in the feature-value list:\n \n [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,\n  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,\n  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]\n \n In the file, there is one instance (board position) per line.\n \n \n Num Instances:     3196\n Num Attributes:    37\n Num Continuous:    0 (Int 0 / Real 0)\n Num Discrete:      37\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'bkblk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   2 \'bknwy\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   3 \'bkon8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'bkona\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'bkspr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'bkxbq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'bkxcr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'bkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'blxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'bxqsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'cntxt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'dsopp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'dwipd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'hdchk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'katri\'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n  16 \'mulch\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'qxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'r2ar8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  19 \'reskd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'reskr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  21 \'rimmx\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'rkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  23 \'rxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'simpl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  25 \'skach\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'skewr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  27 \'skrxp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  28 \'spcop\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  29 \'stlmt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  30 \'thrsk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  31 \'wkcti\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  32 \'wkna8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  33 \'wknck\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  34 \'wkovl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  35 \'wkpos\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  36 \'wtoeg\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  37 \'class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 'true', 1, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:28');
+INSERT INTO `dataset` (`did`, `uploader`, `source`, `name`, `version`, `version_label`, `format`, `creator`, `contributor`, `collection_date`, `upload_date`, `language`, `licence`, `citation`, `collection`, `url`, `isOriginal`, `file_id`, `default_target_attribute`, `row_id_attribute`, `ignore_attribute`, `paper_url`, `visibility`, `original_data_id`, `original_data_url`, `update_comment`, `last_update`) VALUES
+(1, 1, 0, 'kr-vs-kp', '1', '1', 'ARFF', NULL, NULL, NULL, '2014-04-06 23:19:28', NULL, 'Public', NULL, NULL, 'https://www.openml.org/data/download/3/dataset_3_kr-vs-kp.arff', 'true', 1, 'class', NULL, NULL, NULL, 'public', NULL, NULL, NULL, '2014-04-06 23:19:28');
diff --git a/data/sql/dataset_description.sql b/data/sql/dataset_description.sql
index c34a3f066..f40f39e0d 100644
--- a/data/sql/dataset_description.sql
+++ b/data/sql/dataset_description.sql
@@ -1,132 +1,3 @@
 INSERT INTO `dataset_description` (`did`, `version`, `description`, `uploader`) VALUES
-(1, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Annealing Data\n \n 2. Source Information: donated by David Sterling and Wray Buntine.\n \n 3. Past Usage: unknown\n \n 4. Relevant Information:\n    -- Explanation: I suspect this was left by Ross Quinlan in 1987 at the\n       4th Machine Learning Workshop.  I\'d have to check with Jeff Schlimmer\n       to double check this.\n \n 5. Number of Instances: 798\n \n 6. Number of Attributes: 38\n    -- 6 continuously-valued\n    -- 3 integer-valued\n    -- 29 nominal-valued\n \n 7. Attribute Information:\n     1. family:          --,GB,GK,GS,TN,ZA,ZF,ZH,ZM,ZS\n     2. product-type:    C, H, G\n     3. steel:           -,R,A,U,K,M,S,W,V\n     4. carbon:          continuous\n     5. hardness:        continuous\n     6. temper_rolling:  -,T\n     7. condition:       -,S,A,X\n     8. formability:     -,1,2,3,4,5\n     9. strength:        continuous\n    10. non-ageing:      -,N\n    11. surface-finish:  P,M,-\n    12. surface-quality: -,D,E,F,G\n    13. enamelability:   -,1,2,3,4,5\n    14. bc:              Y,-\n    15. bf:              Y,-\n    16. bt:              Y,-\n    17. bw/me:           B,M,-\n    18. bl:              Y,-\n    19. m:               Y,-\n    20. chrom:           C,-\n    21. phos:            P,-\n    22. cbond:           Y,-\n    23. marvi:           Y,-\n    24. exptl:           Y,-\n    25. ferro:           Y,-\n    26. corr:            Y,-\n    27. blue/bright/varn/clean:          B,R,V,C,-\n    28. lustre:          Y,-\n    29. jurofm:          Y,-\n    30. s:               Y,-\n    31. p:               Y,-\n    32. shape:           COIL, SHEET\n    33. thick:           continuous\n    34. width:           continuous\n    35. len:             continuous\n    36. oil:             -,Y,N\n    37. bore:            0000,0500,0600,0760\n    38. packing: -,1,2,3\n    classes:        1,2,3,4,5,U\n  \n    -- The \'-\' values are actually \'not_applicable\' values rather than\n       \'missing_values\' (and so can be treated as legal discrete\n       values rather than as showing the absence of a discrete value).\n \n 8. Missing Attribute Values: Signified with \"?\"\n    Attribute:  Number of instances missing its value:\n    1           0\n    2           0\n    3           70\n    4           0\n    5           0\n    6           675\n    7           271\n    8           283\n    9           0\n   10           703\n   11           790\n   12           217\n   13           785\n   14           797\n   15           680\n   16           736\n   17           609\n   18           662\n   19           798\n   20           775\n   21           791\n   22           730\n   23           798\n   24           796\n   25           772\n   26           798\n   27           793\n   28           753\n   29           798\n   30           798\n   31           798\n   32           0\n   33           0\n   34           0\n   35           0\n   36           740\n   37           0\n   38           789\n   39           0\n \n 9. Distribution of Classes\n      Class Name:   Number of Instances:\n      1               8\n      2              88\n      3             608\n      4               0\n      5              60\n      U              34\n                    ---\n                    798', 1),
-(2, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7\n    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square\n    away from queening.  It is the King+Rook\'s side (white) to move.\n \n 2. Sources:\n     (a) Database originally generated and described by Alen Shapiro.\n     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database\n         was supplied to Holte by Peter Clark of the Turing Institute\n         in Glasgow (pete@turing.ac.uk).\n     (c) Date: 1 August 1989\n \n 3. Past Usage:\n      - Alen D. Shapiro (1983,1987), \"Structured Induction in Expert Systems\",\n        Addison-Wesley.  This book is based on Shapiro\'s Ph.D. thesis (1983)\n        at the University of Edinburgh entitled \"The Role of Structured\n        Induction in Expert Systems\".\n      - Stephen Muggleton (1987), \"Structuring Knowledge by Asking Questions\",\n        pp.218-229 in \"Progress in Machine Learning\", edited by I. Bratko\n        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.\n      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),\n        \"Concept Learning and the Problem of Small Disjuncts\",\n        Proceedings of IJCAI.  Also available as technical report AI89-106,\n        Computer Sciences Department, University of Texas at Austin,\n        Austin, Texas 78712.\n \n 4. Relevant Information:\n       The dataset format is described below.  Note: the format of this\n     database was modified on 2/26/90 to conform with the format of all\n     the other databases in the UCI repository of machine learning databases.\n \n 5. Number of Instances: 3196 total\n \n 6. Number of Attributes: 36\n \n 7. Attribute Summaries:\n     Classes (2):  -- White-can-win (\"won\") and White-cannot-win (\"nowin\").\n           I believe that White is deemed to be unable to win if the Black pawn\n           can safely advance.\n     Attributes: see Shapiro\'s book.\n \n 8. Missing Attributes: --  none\n \n 9. Class Distribution:\n     In 1669 of the positions (52%), White can win.\n     In 1527 of the positions (48%), White cannot win.\n \n The format for instances in this database is a sequence of 37 attribute values.\n Each instance is a board-descriptions for this chess endgame.  The first\n 36 attributes describe the board.  The last (37th) attribute is the\n classification: \"win\" or \"nowin\".  There are 0 missing values.\n A typical board-description is\n \n f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won\n \n The names of the features do not appear in the board-descriptions.\n Instead, each feature correponds to a particular position in the\n feature-value list.  For example, the head of this list is the value\n for the feature \"bkblk\".  The following is the list of features, in\n the order in which their values appear in the feature-value list:\n \n [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,\n  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,\n  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]\n \n In the file, there is one instance (board position) per line.\n \n \n Num Instances:     3196\n Num Attributes:    37\n Num Continuous:    0 (Int 0 / Real 0)\n Num Discrete:      37\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'bkblk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   2 \'bknwy\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   3 \'bkon8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'bkona\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'bkspr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'bkxbq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'bkxcr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'bkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'blxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'bxqsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'cntxt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'dsopp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'dwipd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'hdchk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'katri\'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n  16 \'mulch\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'qxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'r2ar8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  19 \'reskd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'reskr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  21 \'rimmx\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'rkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  23 \'rxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'simpl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  25 \'skach\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'skewr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  27 \'skrxp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  28 \'spcop\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  29 \'stlmt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  30 \'thrsk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  31 \'wkcti\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  32 \'wkna8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  33 \'wknck\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  34 \'wkovl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  35 \'wkpos\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  36 \'wtoeg\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  37 \'class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 1),
-(3, '1', '**Author**: David J. Slate  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Letter+Recognition) - 01-01-1991  \r\n**Please cite**: P. W. Frey and D. J. Slate. \"Letter Recognition Using Holland-style Adaptive Classifiers\". Machine Learning 6(2), 1991  \r\n\r\n1. TITLE: \r\n  Letter Image Recognition Data \r\n \r\n    The objective is to identify each of a large number of black-and-white\r\n    rectangular pixel displays as one of the 26 capital letters in the English\r\n    alphabet.  The character images were based on 20 different fonts and each\r\n    letter within these 20 fonts was randomly distorted to produce a file of\r\n    20,000 unique stimuli.  Each stimulus was converted into 16 primitive\r\n    numerical attributes (statistical moments and edge counts) which were then\r\n    scaled to fit into a range of integer values from 0 through 15.  We\r\n    typically train on the first 16000 items and then use the resulting model\r\n    to predict the letter category for the remaining 4000.  See the article\r\n    cited above for more details.', 1),
-(4, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Balance Scale Weight & Distance Database\n\n2. Source Information: \n    (a) Source: Generated to model psychological experiments reported\n		by Siegler, R. S. (1976).  Three Aspects of Cognitive\n		Development.  Cognitive Psychology, 8, 481-520.\n    (b) Donor: Tim Hume (hume@ics.uci.edu)\n    (c) Date: 22 April 1994\n\n3. Past Usage: (possibly different formats of this data)\n   - Publications\n	1. Klahr, D., & Siegler, R.S. (1978).  The Representation of\n	   Children\'s Knowledge.  In H. W. Reese & L. P. Lipsitt (Eds.),\n	   Advances in Child Development and Behavior, pp. 61-116.  New\n	   York: Academic Press \n	2. Langley,P. (1987).  A General Theory of Discrimination\n	   Learning.  In D. Klahr, P. Langley, & R. Neches (Eds.),\n	   Production System Models of Learning and Development, pp.\n	   99-161. Cambridge, MA: MIT Press\n	3. Newell, A. (1990).  Unified Theories of Cognition.\n	   Cambridge, MA: Harvard University Press\n	4. McClelland, J.L. (1988).  Parallel Distibuted Processing:\n	   Implications for Cognition and Development.  Technical\n	   Report AIP-47, Department of Psychology, Carnegie-Mellon\n	   University \n	5. Shultz, T., Mareschal, D., & Schmidt, W. (1994).  Modeling\n	   Cognitive Development on Balance Scale Phenomena. Machine\n	   Learning, Vol. 16, pp. 59-88.\n\n4. Relevant Information: \n	This data set was generated to model psychological\n	experimental results.  Each example is classified as having the\n	balance scale tip to the right, tip to the left, or be\n	balanced.  The attributes are the left weight, the left\n	distance, the right weight, and the right distance.  The\n	correct way to find the class is the greater of \n	(left-distance * left-weight) and (right-distance *\n	right-weight).  If they are equal, it is balanced.\n\n5. Number of Instances: 625 (49 balanced, 288 left, 288 right)\n\n6. Number of Attributes: 4 (numeric) + class name = 5\n\n7. Attribute Information:\n	1. Class Name: 3 (L, B, R)\n	2. Left-Weight: 5 (1, 2, 3, 4, 5)\n	3. Left-Distance: 5 (1, 2, 3, 4, 5)\n	4. Right-Weight: 5 (1, 2, 3, 4, 5)\n	5. Right-Distance: 5 (1, 2, 3, 4, 5)\n\n8. Missing Attribute Values: \n	none\n\n9. Class Distribution: \n   1. 46.08 percent are L\n   2. 07.84 percent are B\n   3. 46.08 percent are R', 1),
-(5, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(6, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(7, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n', 1),
-(8, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(9, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (0 - 9)\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class \'0\', followed by sets of 200 patterns\n for each of the classes \'1\' - \'9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(10, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(11, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Car Evaluation Database\n \n 2. Sources:\n    (a) Creator: Marko Bohanec\n    (b) Donors: Marko Bohanec   (marko.bohanec@ijs.si)\n                Blaz Zupan      (blaz.zupan@ijs.si)\n    (c) Date: June, 1997\n \n 3. Past Usage:\n \n    The hierarchical decision model, from which this dataset is\n    derived, was first presented in \n \n    M. Bohanec and V. Rajkovic: Knowledge acquisition and explanation for\n    multi-attribute decision making. In 8th Intl Workshop on Expert\n    Systems and their Applications, Avignon, France. pages 59-78, 1988.\n \n    Within machine-learning, this dataset was used for the evaluation\n    of HINT (Hierarchy INduction Tool), which was proved to be able to\n    completely reconstruct the original hierarchical model. This,\n    together with a comparison with C4.5, is presented in\n \n    B. Zupan, M. Bohanec, I. Bratko, J. Demsar: Machine learning by\n    function decomposition. ICML-97, Nashville, TN. 1997 (to appear)\n \n 4. Relevant Information Paragraph:\n \n    Car Evaluation Database was derived from a simple hierarchical\n    decision model originally developed for the demonstration of DEX\n    (M. Bohanec, V. Rajkovic: Expert system for decision\n    making. Sistemica 1(1), pp. 145-157, 1990.). The model evaluates\n    cars according to the following concept structure:\n \n    CAR                      car acceptability\n    . PRICE                  overall price\n    . . buying               buying price\n    . . maint                price of the maintenance\n    . TECH                   technical characteristics\n    . . COMFORT              comfort\n    . . . doors              number of doors\n    . . . persons            capacity in terms of persons to carry\n    . . . lug_boot           the size of luggage boot\n    . . safety               estimated safety of the car\n \n    Input attributes are printed in lowercase. Besides the target\n    concept (CAR), the model includes three intermediate concepts:\n    PRICE, TECH, COMFORT. Every concept is in the original model\n    related to its lower level descendants by a set of examples (for\n    these examples sets see http://www-ai.ijs.si/BlazZupan/car.html).\n \n    The Car Evaluation Database contains examples with the structural\n    information removed, i.e., directly relates CAR to the six input\n    attributes: buying, maint, doors, persons, lug_boot, safety.\n \n    Because of known underlying concept structure, this database may be\n    particularly useful for testing constructive induction and\n    structure discovery methods.\n \n 5. Number of Instances: 1728\n    (instances completely cover the attribute space)\n \n 6. Number of Attributes: 6\n \n 7. Attribute Values:\n \n    buying       v-high, high, med, low\n    maint        v-high, high, med, low\n    doors        2, 3, 4, 5-more\n    persons      2, 4, more\n    lug_boot     small, med, big\n    safety       low, med, high\n \n 8. Missing Attribute Values: none\n \n 9. Class Distribution (number of instances per class)\n \n    class      N          N[%]\n    -----------------------------\n    unacc     1210     (70.023 %) \n    acc        384     (22.222 %) \n    good        69     ( 3.993 %) \n    v-good      65     ( 3.762 %) \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(12, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe multi-feature digit dataset\n -------------------------------\n \n Oowned and donated by:\n ----------------------\n \n Robert P.W. Duin\n Department of Applied Physics \n Delft University of Technology \n P.O. Box 5046, 2600 GA Delft\n The Netherlands\n \n email: duin@ph.tn.tudelft.nl\n http : //www.ph.tn.tudelft.nl/~duin\n tel +31 15 2786143\n \n Usage\n -----\n A slightly different version of the database is used in\n \n M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten\n      digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4,\n      1998, 381-386.\n \n M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined\n      Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR\'98,\n      Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20),\n \n The database as it is is used in:\n \n A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review,\n      in preparation\n \n Description\n -----------\n \n This dataset consists of features of handwritten numerals (`0\'--`9\')\n extracted from a collection of Dutch utility maps. 200 patterns per\n class (for a total of 2,000 patterns) have been digitized in  binary\n images. These digits are represented in terms of the following six\n feature sets (files): \n \n 1. mfeat-fou: 76 Fourier coefficients of the character shapes; \n 2. mfeat-fac: 216 profile correlations; \n 3. mfeat-kar: 64 Karhunen-Love coefficients; \n 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; \n 5. mfeat-zer: 47 Zernike moments; \n 6. mfeat-mor: 6 morphological features. \n \n In each file the 2000 patterns are stored in ASCI on 2000 lines. The\n first 200 patterns are of class `0\', followed by sets of 200 patterns\n for each of the classes `1\' - `9\'. Corresponding patterns in different\n feature sets (files) correspond to the same original character.\n \n The source image dataset is lost. Using the pixel-dataset (mfeat-pix)\n sampled versions of the original images may be obtained (15 x 16 pixels).\n \n Total number of instances:\n --------------------------\n 2000 (200 instances per class)\n \n Total number of attributes:\n ---------------------------\n 649 (distributed over 6 datasets,see above)\n \n no missing attributes\n \n Total number of classes:\n ------------------------\n 10\n \n Format:\n ------\n 6 files, see above.\n Each file contains 2000 lines, one for each instance.\n Attributes are SPACE separated and can be loaded by Matlab as\n > load filename\n No missing attributes. Some are integer, others are real.\n \n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(13, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Contraceptive Method Choice\n \n 2. Sources:\n    (a) Origin:  This dataset is a subset of the 1987 National Indonesia\n                 Contraceptive Prevalence Survey\n    (b) Creator: Tjen-Sien Lim (limt@stat.wisc.edu)\n    (c) Donor:   Tjen-Sien Lim (limt@stat.wisc.edu)\n    (c) Date:    June 7, 1997\n \n 3. Past Usage:\n    Lim, T.-S., Loh, W.-Y. & Shih, Y.-S. (1999). A Comparison of\n    Prediction Accuracy, Complexity, and Training Time of Thirty-three\n    Old and New Classification Algorithms. Machine Learning. Forthcoming.\n    (ftp://ftp.stat.wisc.edu/pub/loh/treeprogs/quest1.7/mach1317.pdf or\n    (http://www.stat.wisc.edu/~limt/mach1317.pdf)\n \n 4. Relevant Information:\n    This dataset is a subset of the 1987 National Indonesia Contraceptive\n    Prevalence Survey. The samples are married women who were either not \n    pregnant or do not know if they were at the time of interview. The \n    problem is to predict the current contraceptive method choice \n    (no use, long-term methods, or short-term methods) of a woman based \n    on her demographic and socio-economic characteristics.\n \n 5. Number of Instances: 1473\n \n 6. Number of Attributes: 10 (including the class attribute)\n \n 7. Attribute Information:\n \n    1. Wife\'s age                     (numerical)\n    2. Wife\'s education               (categorical)      1=low, 2, 3, 4=high\n    3. Husband\'s education            (categorical)      1=low, 2, 3, 4=high\n    4. Number of children ever born   (numerical)\n    5. Wife\'s religion                (binary)           0=Non-Islam, 1=Islam\n    6. Wife\'s now working?            (binary)           0=Yes, 1=No\n    7. Husband\'s occupation           (categorical)      1, 2, 3, 4\n    8. Standard-of-living index       (categorical)      1=low, 2, 3, 4=high\n    9. Media exposure                 (binary)           0=Good, 1=Not good\n    10. Contraceptive method used     (class attribute)  1=No-use \n                                                         2=Long-term\n                                                         3=Short-term\n \n 8. Missing Attribute Values: None\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(14, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Mushroom Database\n \n 2. Sources: \n     (a) Mushroom records drawn from The Audubon Society Field Guide to North\n         American Mushrooms (1981). G. H. Lincoff (Pres.), New York: Alfred\n         A. Knopf\n     (b) Donor: Jeff Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n     (c) Date: 27 April 1987\n \n 3. Past Usage:\n     1. Schlimmer,J.S. (1987). Concept Acquisition Through Representational\n        Adjustment (Technical Report 87-19).  Doctoral disseration, Department\n        of Information and Computer Science, University of California, Irvine.\n        --- STAGGER: asymptoted to 95% classification accuracy after reviewing\n            1000 instances.\n     2. Iba,W., Wogulis,J., & Langley,P. (1988).  Trading off Simplicity\n        and Coverage in Incremental Concept Learning. In Proceedings of \n        the 5th International Conference on Machine Learning, 73-79.\n        Ann Arbor, Michigan: Morgan Kaufmann.  \n        -- approximately the same results with their HILLARY algorithm    \n     3. In the following references a set of rules (given below) were\n         learned for this data set which may serve as a point of\n         comparison for other researchers.\n \n         Duch W, Adamczak R, Grabczewski K (1996) Extraction of logical rules\n         from training data using backpropagation networks, in: Proc. of the\n         The 1st Online Workshop on Soft Computing, 19-30.Aug.1996, pp. 25-30,\n         available on-line at: http://www.bioele.nuee.nagoya-u.ac.jp/wsc1/\n \n         Duch W, Adamczak R, Grabczewski K, Ishikawa M, Ueda H, Extraction of\n         crisp logical rules using constrained backpropagation networks -\n         comparison of two new approaches, in: Proc. of the European Symposium\n         on Artificial Neural Networks (ESANN\'97), Bruge, Belgium 16-18.4.1997,\n         pp. xx-xx\n \n         Wlodzislaw Duch, Department of Computer Methods, Nicholas Copernicus\n         University, 87-100 Torun, Grudziadzka 5, Poland\n         e-mail: duch@phys.uni.torun.pl\n         WWW     http://www.phys.uni.torun.pl/kmk/\n         \n         Date: Mon, 17 Feb 1997 13:47:40 +0100\n         From: Wlodzislaw Duch <duch@phys.uni.torun.pl>\n         Organization: Dept. of Computer Methods, UMK\n \n         I have attached a file containing logical rules for mushrooms.\n         It should be helpful for other people since only in the last year I\n         have seen about 10 papers analyzing this dataset and obtaining quite\n         complex rules. We will try to contribute other results later.\n \n         With best regards, Wlodek Duch\n         ________________________________________________________________\n \n         Logical rules for the mushroom data sets.\n \n         Logical rules given below seem to be the simplest possible for the\n         mushroom dataset and therefore should be treated as benchmark results.\n \n         Disjunctive rules for poisonous mushrooms, from most general\n         to most specific:\n \n         P_1) odor=NOT(almond.OR.anise.OR.none)\n              120 poisonous cases missed, 98.52% accuracy\n \n         P_2) spore-print-color=green\n              48 cases missed, 99.41% accuracy\n          \n         P_3) odor=none.AND.stalk-surface-below-ring=scaly.AND.\n                   (stalk-color-above-ring=NOT.brown) \n              8 cases missed, 99.90% accuracy\n          \n         P_4) habitat=leaves.AND.cap-color=white\n                  100% accuracy     \n \n         Rule P_4) may also be\n \n         P_4\') population=clustered.AND.cap_color=white\n \n         These rule involve 6 attributes (out of 22). Rules for edible\n         mushrooms are obtained as negation of the rules given above, for\n         example the rule:\n \n         odor=(almond.OR.anise.OR.none).AND.spore-print-color=NOT.green\n \n         gives 48 errors, or 99.41% accuracy on the whole dataset.\n \n         Several slightly more complex variations on these rules exist,\n         involving other attributes, such as gill_size, gill_spacing,\n         stalk_surface_above_ring, but the rules given above are the simplest\n         we have found.\n \n \n 4. Relevant Information:\n     This data set includes descriptions of hypothetical samples\n     corresponding to 23 species of gilled mushrooms in the Agaricus and\n     Lepiota Family (pp. 500-525).  Each species is identified as\n     definitely edible, definitely poisonous, or of unknown edibility and\n     not recommended.  This latter class was combined with the poisonous\n     one.  The Guide clearly states that there is no simple rule for\n     determining the edibility of a mushroom; no rule like ``leaflets\n     three, let it be\'\' for Poisonous Oak and Ivy.\n \n 5. Number of Instances: 8124\n \n 6. Number of Attributes: 22 (all nominally valued)\n \n 7. Attribute Information: (classes: edible=e, poisonous=p)\n      1. cap-shape:                bell=b,conical=c,convex=x,flat=f,\n                                   knobbed=k,sunken=s\n      2. cap-surface:              fibrous=f,grooves=g,scaly=y,smooth=s\n      3. cap-color:                brown=n,buff=b,cinnamon=c,gray=g,green=r,\n                                   pink=p,purple=u,red=e,white=w,yellow=y\n      4. bruises?:                 bruises=t,no=f\n      5. odor:                     almond=a,anise=l,creosote=c,fishy=y,foul=f,\n                                   musty=m,none=n,pungent=p,spicy=s\n      6. gill-attachment:          attached=a,descending=d,free=f,notched=n\n      7. gill-spacing:             close=c,crowded=w,distant=d\n      8. gill-size:                broad=b,narrow=n\n      9. gill-color:               black=k,brown=n,buff=b,chocolate=h,gray=g,\n                                   green=r,orange=o,pink=p,purple=u,red=e,\n                                   white=w,yellow=y\n     10. stalk-shape:              enlarging=e,tapering=t\n     11. stalk-root:               bulbous=b,club=c,cup=u,equal=e,\n                                   rhizomorphs=z,rooted=r,missing=?\n     12. stalk-surface-above-ring: ibrous=f,scaly=y,silky=k,smooth=s\n     13. stalk-surface-below-ring: ibrous=f,scaly=y,silky=k,smooth=s\n     14. stalk-color-above-ring:   brown=n,buff=b,cinnamon=c,gray=g,orange=o,\n                                   pink=p,red=e,white=w,yellow=y\n     15. stalk-color-below-ring:   brown=n,buff=b,cinnamon=c,gray=g,orange=o,\n                                   pink=p,red=e,white=w,yellow=y\n     16. veil-type:                partial=p,universal=u\n     17. veil-color:               brown=n,orange=o,white=w,yellow=y\n     18. ring-number:              none=n,one=o,two=t\n     19. ring-type:                cobwebby=c,evanescent=e,flaring=f,large=l,\n                                   none=n,pendant=p,sheathing=s,zone=z\n     20. spore-print-color:        black=k,brown=n,buff=b,chocolate=h,green=r,\n                                   orange=o,purple=u,white=w,yellow=y\n     21. population:               abundant=a,clustered=c,numerous=n,\n                                   scattered=s,several=v,solitary=y\n     22. habitat:                  grasses=g,leaves=l,meadows=m,paths=p,\n                                   urban=u,waste=w,woods=d\n \n 8. Missing Attribute Values: 2480 of them (denoted by \"?\"), all for\n    attribute #11.\n \n 9. Class Distribution: \n     --    edible: 4208 (51.8%)\n     -- poisonous: 3916 (48.2%)\n     --     total: 8124 instances', 1),
-(15, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Optical Recognition of Handwritten Digits\n \n 2. Source:\n 	E. Alpaydin, C. Kaynak\n 	Department of Computer Engineering\n 	Bogazici University, 80815 Istanbul Turkey\n 	alpaydin@boun.edu.tr\n 	July 1998\n \n 3. Past Usage:\n 	C. Kaynak (1995) Methods of Combining Multiple Classifiers and Their\n 	Applications to Handwritten Digit Recognition, \n 	MSc Thesis, Institute of Graduate Studies in Science and \n 	Engineering, Bogazici University.\n \n 	E. Alpaydin, C. Kaynak (1998) Cascading Classifiers, Kybernetika,\n 	to appear. ftp://ftp.icsi.berkeley.edu/pub/ai/ethem/kyb.ps.Z\n \n 4. Relevant Information:\n 	We used preprocessing programs made available by NIST to extract\n 	normalized bitmaps of handwritten digits from a preprinted form. From\n 	a total of 43 people, 30 contributed to the training set and different\n 	13 to the test set. 32x32 bitmaps are divided into nonoverlapping \n 	blocks of 4x4 and the number of on pixels are counted in each block.\n 	This generates an input matrix of 8x8 where each element is an \n 	integer in the range 0..16. This reduces dimensionality and gives \n 	invariance to small distortions.\n \n 	For info on NIST preprocessing routines, see \n 	M. D. Garris, J. L. Blue, G. T. Candela, D. L. Dimmick, J. Geist, \n 	P. J. Grother, S. A. Janet, and C. L. Wilson, NIST Form-Based \n 	Handprint Recognition System, NISTIR 5469, 1994.\n \n 5. Number of Instances\n 	optdigits.tra	Training	3823\n 	optdigits.tes	Testing		1797\n 	\n 	The way we used the dataset was to use half of training for \n 	actual training, one-fourth for validation and one-fourth\n 	for writer-dependent testing. The test set was used for \n 	writer-independent testing and is the actual quality measure.\n \n 6. Number of Attributes\n 	64 input+1 class attribute\n \n 7. For Each Attribute:\n 	All input attributes are integers in the range 0..16.\n 	The last attribute is the class code 0..9\n \n 8. Missing Attribute Values\n 	None\n \n 9. Class Distribution\n 	Class:	No of examples in training set\n 	0:  376\n 	1:  389\n 	2:  380\n 	3:  389\n 	4:  387\n 	5:  376\n 	6:  377\n 	7:  387\n 	8:  380\n 	9:  382\n \n 	Class: No of examples in testing set\n 	0:  178\n 	1:  182\n 	2:  177\n 	3:  183\n 	4:  181\n 	5:  182\n 	6:  181\n 	7:  179\n 	8:  174\n 	9:  180\n \n Accuracy on the testing set with k-nn \n using Euclidean distance as the metric\n \n  k =  1   : 98.00\n  k =  2   : 97.38\n  k =  3   : 97.83\n  k =  4   : 97.61\n  k =  5   : 97.89\n  k =  6   : 97.77\n  k =  7   : 97.66\n  k =  8   : 97.66\n  k =  9   : 97.72\n  k = 10   : 97.55\n  k = 11   : 97.89', 1),
-(16, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Credit Approval\n \n 2. Sources: \n     (confidential)\n     Submitted by quinlan@cs.su.oz.au\n \n 3.  Past Usage:\n \n     See Quinlan,\n     * \"Simplifying decision trees\", Int J Man-Machine Studies 27,\n       Dec 1987, pp. 221-234.\n     * \"C4.5: Programs for Machine Learning\", Morgan Kaufmann, Oct 1992\n   \n 4.  Relevant Information:\n \n     This file concerns credit card applications.  All attribute names\n     and values have been changed to meaningless symbols to protect\n     confidentiality of the data.\n   \n     This dataset is interesting because there is a good mix of\n     attributes -- continuous, nominal with small numbers of\n     values, and nominal with larger numbers of values.  There\n     are also a few missing values.\n   \n 5.  Number of Instances: 690\n \n 6.  Number of Attributes: 15 + class attribute\n \n 7.  Attribute Information:\n \n     A1:	b, a.\n     A2:	continuous.\n     A3:	continuous.\n     A4:	u, y, l, t.\n     A5:	g, p, gg.\n     A6:	c, d, cc, i, j, k, m, r, q, w, x, e, aa, ff.\n     A7:	v, h, bb, j, n, z, dd, ff, o.\n     A8:	continuous.\n     A9:	t, f.\n     A10:	t, f.\n     A11:	continuous.\n     A12:	t, f.\n     A13:	g, p, s.\n     A14:	continuous.\n     A15:	continuous.\n     A16: +,-         (class attribute)\n \n 8.  Missing Attribute Values:\n     37 cases (5%) have one or more missing values.  The missing\n     values from particular attributes are:\n \n     A1:  12\n     A2:  12\n     A4:   6\n     A5:   6\n     A6:   9\n     A7:   9\n     A14: 13\n \n 9.  Class Distribution\n   \n     +: 307 (44.5%)\n     -: 383 (55.5%)', 1),
-(17, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDescription of the German credit dataset.\n \n 1. Title: German Credit data\n \n 2. Source Information\n \n Professor Dr. Hans Hofmann  \n Institut f\"ur Statistik und \"Okonometrie  \n Universit\"at Hamburg  \n FB Wirtschaftswissenschaften  \n Von-Melle-Park 5    \n 2000 Hamburg 13 \n \n 3. Number of Instances:  1000\n \n Two datasets are provided.  the original dataset, in the form provided\n by Prof. Hofmann, contains categorical/symbolic attributes and\n is in the file \"german.data\".   \n  \n For algorithms that need numerical attributes, Strathclyde University \n produced the file \"german.data-numeric\".  This file has been edited \n and several indicator variables added to make it suitable for \n algorithms which cannot cope with categorical variables.   Several\n attributes that are ordered categorical (such as attribute 17) have\n been coded as integer.    This was the form used by StatLog.\n \n \n 6. Number of Attributes german: 20 (7 numerical, 13 categorical)\n    Number of Attributes german.numer: 24 (24 numerical)\n \n \n 7.  Attribute description for german\n \n Attribute 1:  (qualitative)\n 	       Status of existing checking account\n                A11 :      ... <    0 DM\n 	       A12 : 0 <= ... <  200 DM\n 	       A13 :      ... >= 200 DM /\n 		     salary assignments for at least 1 year\n                A14 : no checking account\n \n Attribute 2:  (numerical)\n 	      Duration in month\n \n Attribute 3:  (qualitative)\n 	      Credit history\n 	      A30 : no credits taken/\n 		    all credits paid back duly\n               A31 : all credits at this bank paid back duly\n 	      A32 : existing credits paid back duly till now\n               A33 : delay in paying off in the past\n 	      A34 : critical account/\n 		    other credits existing (not at this bank)\n \n Attribute 4:  (qualitative)\n 	      Purpose\n 	      A40 : car (new)\n 	      A41 : car (used)\n 	      A42 : furniture/equipment\n 	      A43 : radio/television\n 	      A44 : domestic appliances\n 	      A45 : repairs\n 	      A46 : education\n 	      A47 : (vacation - does not exist?)\n 	      A48 : retraining\n 	      A49 : business\n 	      A410 : others\n \n Attribute 5:  (numerical)\n 	      Credit amount\n \n Attibute 6:  (qualitative)\n 	      Savings account/bonds\n 	      A61 :          ... <  100 DM\n 	      A62 :   100 <= ... <  500 DM\n 	      A63 :   500 <= ... < 1000 DM\n 	      A64 :          .. >= 1000 DM\n               A65 :   unknown/ no savings account\n \n Attribute 7:  (qualitative)\n 	      Present employment since\n 	      A71 : unemployed\n 	      A72 :       ... < 1 year\n 	      A73 : 1  <= ... < 4 years  \n 	      A74 : 4  <= ... < 7 years\n 	      A75 :       .. >= 7 years\n \n Attribute 8:  (numerical)\n 	      Installment rate in percentage of disposable income\n \n Attribute 9:  (qualitative)\n 	      Personal status and sex\n 	      A91 : male   : divorced/separated\n 	      A92 : female : divorced/separated/married\n               A93 : male   : single\n 	      A94 : male   : married/widowed\n 	      A95 : female : single\n \n Attribute 10: (qualitative)\n 	      Other debtors / guarantors\n 	      A101 : none\n 	      A102 : co-applicant\n 	      A103 : guarantor\n \n Attribute 11: (numerical)\n 	      Present residence since\n \n Attribute 12: (qualitative)\n 	      Property\n 	      A121 : real estate\n 	      A122 : if not A121 : building society savings agreement/\n 				   life insurance\n               A123 : if not A121/A122 : car or other, not in attribute 6\n 	      A124 : unknown / no property\n \n Attribute 13: (numerical)\n 	      Age in years\n \n Attribute 14: (qualitative)\n 	      Other installment plans \n 	      A141 : bank\n 	      A142 : stores\n 	      A143 : none\n \n Attribute 15: (qualitative)\n 	      Housing\n 	      A151 : rent\n 	      A152 : own\n 	      A153 : for free\n \n Attribute 16: (numerical)\n               Number of existing credits at this bank\n \n Attribute 17: (qualitative)\n 	      Job\n 	      A171 : unemployed/ unskilled  - non-resident\n 	      A172 : unskilled - resident\n 	      A173 : skilled employee / official\n 	      A174 : management/ self-employed/\n 		     highly qualified employee/ officer\n \n Attribute 18: (numerical)\n 	      Number of people being liable to provide maintenance for\n \n Attribute 19: (qualitative)\n 	      Telephone\n 	      A191 : none\n 	      A192 : yes, registered under the customers name\n \n Attribute 20: (qualitative)\n 	      foreign worker\n 	      A201 : yes\n 	      A202 : no\n \n \n \n 8.  Cost Matrix\n \n This dataset requires use of a cost matrix (see below)\n \n \n       1        2\n ----------------------------\n   1   0        1\n -----------------------\n   2   5        0\n \n (1 = Good,  2 = Bad)\n \n the rows represent the actual classification and the columns\n the predicted classification.\n \n It is worse to class a customer as good when they are bad (5), \n than it is to class a customer as bad when they are good (1).\n \n\n\n\n\n Relabeled values in attribute checking_status\n    From: A11                     To: \'<0\'                \n    From: A12                     To: \'0<=X<200\'          \n    From: A13                     To: \'>=200\'             \n    From: A14                     To: \'no checking\'       \n\n\n Relabeled values in attribute credit_history\n    From: A30                     To: \'no credits/all paid\'\n    From: A31                     To: \'all paid\'          \n    From: A32                     To: \'existing paid\'     \n    From: A33                     To: \'delayed previously\'\n    From: A34                     To: \'critical/other existing credit\'\n\n\n Relabeled values in attribute purpose\n    From: A40                     To: \'new car\'           \n    From: A41                     To: \'used car\'          \n    From: A42                     To: furniture/equipment \n    From: A43                     To: radio/tv            \n    From: A44                     To: \'domestic appliance\'\n    From: A45                     To: repairs             \n    From: A46                     To: education           \n    From: A47                     To: vacation            \n    From: A48                     To: retraining          \n    From: A49                     To: business            \n    From: A410                    To: other               \n\n\n Relabeled values in attribute savings_status\n    From: A61                     To: \'<100\'              \n    From: A62                     To: \'100<=X<500\'        \n    From: A63                     To: \'500<=X<1000\'       \n    From: A64                     To: \'>=1000\'            \n    From: A65                     To: \'no known savings\'  \n\n\n Relabeled values in attribute employment\n    From: A71                     To: unemployed          \n    From: A72                     To: \'<1\'                \n    From: A73                     To: \'1<=X<4\'            \n    From: A74                     To: \'4<=X<7\'            \n    From: A75                     To: \'>=7\'               \n\n\n Relabeled values in attribute personal_status\n    From: A91                     To: \'male div/sep\'      \n    From: A92                     To: \'female div/dep/mar\'\n    From: A93                     To: \'male single\'       \n    From: A94                     To: \'male mar/wid\'      \n    From: A95                     To: \'female single\'     \n\n\n Relabeled values in attribute other_parties\n    From: A101                    To: none                \n    From: A102                    To: \'co applicant\'      \n    From: A103                    To: guarantor           \n\n\n Relabeled values in attribute property_magnitude\n    From: A121                    To: \'real estate\'       \n    From: A122                    To: \'life insurance\'    \n    From: A123                    To: car                 \n    From: A124                    To: \'no known property\' \n\n\n Relabeled values in attribute other_payment_plans\n    From: A141                    To: bank                \n    From: A142                    To: stores              \n    From: A143                    To: none                \n\n\n Relabeled values in attribute housing\n    From: A151                    To: rent                \n    From: A152                    To: own                 \n    From: A153                    To: \'for free\'          \n\n\n Relabeled values in attribute job\n    From: A171                    To: \'unemp/unskilled non res\'\n    From: A172                    To: \'unskilled resident\'\n    From: A173                    To: skilled             \n    From: A174                    To: \'high qualif/self emp/mgmt\'\n\n\n Relabeled values in attribute own_telephone\n    From: A191                    To: none                \n    From: A192                    To: yes                 \n\n\n Relabeled values in attribute foreign_worker\n    From: A201                    To: yes                 \n    From: A202                    To: no                  \n\n\n Relabeled values in attribute class\n    From: 1                       To: good                \n    From: 2                       To: bad', 1),
-(18, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Pen-Based Recognition of Handwritten Digits\n \n 2. Source:\n 	E. Alpaydin, F. Alimoglu\n 	Department of Computer Engineering\n 	Bogazici University, 80815 Istanbul Turkey\n 	alpaydin@boun.edu.tr\n 	July 1998\n \n 3. Past Usage:\n 	F. Alimoglu (1996) Combining Multiple Classifiers for Pen-Based\n 	Handwritten Digit Recognition, \n 	MSc Thesis, Institute of Graduate Studies in Science and \n 	Engineering, Bogazici University.\n 	http://www.cmpe.boun.edu.tr/~alimoglu/alimoglu.ps.gz\n \n 	F. Alimoglu, E. Alpaydin, \"Methods of Combining Multiple Classifiers \n 	Based on Different Representations for Pen-based Handwriting\n 	Recognition,\" Proceedings of the Fifth Turkish Artificial \n 	Intelligence and Artificial Neural Networks Symposium (TAINN 96), \n 	June 1996, Istanbul, Turkey.\n 	http://www.cmpe.boun.edu.tr/~alimoglu/tainn96.ps.gz\n \n 	\n 4. Relevant Information:\n \n 	We create a digit database by collecting 250 samples from 44 writers.\n 	The samples written by 30 writers are used for training,\n 	cross-validation and writer dependent testing, and the digits \n 	written by the other 14 are used for writer independent testing. This\n 	database is also available in the UNIPEN format.\n \n 	We use a WACOM PL-100V pressure sensitive tablet with an integrated \n 	LCD display and a cordless stylus. The input and display areas are\n 	located in the same place. Attached to the serial port of an Intel \n 	486 based PC, it allows us to collect handwriting samples. The tablet\n 	sends $x$ and $y$ tablet coordinates and pressure level values of the\n 	pen at fixed time intervals (sampling rate) of 100 miliseconds. \n \n 	These writers are asked to write 250 digits in random order inside \n 	boxes of 500 by 500 tablet pixel resolution.  Subject are monitored \n 	only during the first entry screens. Each screen contains five boxes\n 	with the digits to be written displayed above. Subjects are told to\n 	write only inside these boxes.  If they make a mistake or are unhappy\n 	with their writing, they are instructed to clear the content of a box \n 	by using an on-screen button. The first ten digits are ignored \n 	because most writers are not familiar with this type of input devices,\n 	but subjects are not aware of this. \n \n 	In our study, we use only ($x, y$) coordinate information. The stylus\n 	pressure level values are ignored. First we apply normalization to \n 	make our representation invariant to translations and scale \n 	distortions. The raw data that we capture from the tablet consist of\n 	integer values between 0 and 500 (tablet input box resolution). The \n 	new coordinates are such that the coordinate which has the maximum \n 	range varies between 0 and 100. Usually $x$ stays in this range, since\n 	most characters are taller than they are wide.  \n \n 	In order to train and test our classifiers, we need to represent \n 	digits as constant length feature vectors. A commonly used technique\n 	leading to good results is resampling the ( x_t, y_t) points. \n 	Temporal resampling (points regularly spaced in time) or spatial\n 	resampling (points regularly spaced in arc length) can be used here. \n 	Raw point data are already regularly spaced in time but the distance\n 	between them is variable. Previous research showed that spatial\n 	resampling to obtain a constant number of regularly spaced points \n 	on the trajectory yields much better performance, because it provides \n 	a better alignment between points. Our resampling algorithm uses \n 	simple linear interpolation between pairs of points. The resampled\n 	digits are represented as a sequence of T points ( x_t, y_t )_{t=1}^T,\n 	regularly spaced in arc length, as opposed to the input sequence, \n 	which is regularly spaced in time.\n \n 	So, the input vector size is 2*T, two times the number of points\n 	resampled. We considered spatial resampling to T=8,12,16 points in our\n 	experiments and found that T=8 gave the best trade-off between \n 	accuracy and complexity.\n \n \n 5. Number of Instances\n 	pendigits.tra	Training	7494\n 	pendigits.tes	Testing		3498\n 	\n 	The way we used the dataset was to use first half of training for \n 	actual training, one-fourth for validation and one-fourth\n 	for writer-dependent testing. The test set was used for \n 	writer-independent testing and is the actual quality measure.\n \n 6. Number of Attributes\n 	16 input+1 class attribute\n \n 7. For Each Attribute:\n 	All input attributes are integers in the range 0..100.\n 	The last attribute is the class code 0..9\n \n 8. Missing Attribute Values\n 	None\n \n 9. Class Distribution\n 	Class: No of examples in training set\n 	0:  780\n 	1:  779\n 	2:  780\n 	3:  719\n 	4:  780\n 	5:  720\n 	6:  720\n 	7:  778\n 	8:  719\n 	9:  719\n 	Class: No of examples in testing set\n 	0:  363\n 	1:  364\n 	2:  364\n 	3:  336\n 	4:  364\n 	5:  335\n 	6:  336\n 	7:  364\n 	8:  336\n 	9:  336\n \n Accuracy on the testing set with k-nn \n using Euclidean distance as the metric\n \n  k =  1 : 97.74\n  k =  2 : 97.37\n  k =  3 : 97.80\n  k =  4 : 97.66\n  k =  5 : 97.60\n  k =  6 : 97.57\n  k =  7 : 97.54\n  k =  8 : 97.54\n  k =  9 : 97.46\n  k = 10 : 97.48\n  k = 11 : 97.34', 1),
-(19, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Image Segmentation data\n \n 2. Source Information\n    -- Creators: Vision Group, University of Massachusetts\n    -- Donor: Vision Group (Carla Brodley, brodley@cs.umass.edu)\n    -- Date: November, 1990\n  \n 3. Past Usage: None yet published\n \n 4. Relevant Information:\n \n    The instances were drawn randomly from a database of 7 outdoor \n    images.  The images were handsegmented to create a classification\n    for every pixel.  \n \n    Each instance is a 3x3 region.\n \n 5. Number of Instances: Training data: 210  Test data: 2100\n \n 6. Number of Attributes: 19 continuous attributes\n \n 7. Attribute Information:\n \n     1.  region-centroid-col:  the column of the center pixel of the region.\n     2.  region-centroid-row:  the row of the center pixel of the region.\n     3.  region-pixel-count:  the number of pixels in a region = 9.\n     4.  short-line-density-5:  the results of a line extractoin algorithm that \n          counts how many lines of length 5 (any orientation) with\n          low contrast, less than or equal to 5, go through the region.\n     5.  short-line-density-2:  same as short-line-density-5 but counts lines\n          of high contrast, greater than 5.\n     6.  vedge-mean:  measure the contrast of horizontally\n          adjacent pixels in the region.  There are 6, the mean and \n          standard deviation are given.  This attribute is used as\n         a vertical edge detector.\n     7.  vegde-sd:  (see 6)\n     8.  hedge-mean:  measures the contrast of vertically adjacent\n           pixels. Used for horizontal line detection. \n     9.  hedge-sd: (see 8).\n     10. intensity-mean:  the average over the region of (R + G + B)/3\n     11. rawred-mean: the average over the region of the R value.\n     12. rawblue-mean: the average over the region of the B value.\n     13. rawgreen-mean: the average over the region of the G value.\n     14. exred-mean: measure the excess red:  (2R - (G + B))\n     15. exblue-mean: measure the excess blue:  (2B - (G + R))\n     16. exgreen-mean: measure the excess green:  (2G - (R + B))\n     17. value-mean:  3-d nonlinear transformation\n          of RGB. (Algorithm can be found in Foley and VanDam, Fundamentals\n          of Interactive Computer Graphics)\n     18. saturatoin-mean:  (see 17)\n     19. hue-mean:  (see 17)\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: \n \n    Classes:  brickface, sky, foliage, cement, window, path, grass.\n \n    30 instances per class for training data.\n    300 instances per class for test data.\n \n\n\n\n\n Relabeled values in attribute class\n    From: 1                       To: brickface           \n    From: 2                       To: sky                 \n    From: 3                       To: foliage             \n    From: 4                       To: cement              \n    From: 5                       To: window              \n    From: 6                       To: path                \n    From: 7                       To: grass', 1),
-(20, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Pima Indians Diabetes Database\n \n 2. Sources:\n    (a) Original owners: National Institute of Diabetes and Digestive and\n                         Kidney Diseases\n    (b) Donor of database: Vincent Sigillito (vgs@aplcen.apl.jhu.edu)\n                           Research Center, RMI Group Leader\n                           Applied Physics Laboratory\n                           The Johns Hopkins University\n                           Johns Hopkins Road\n                           Laurel, MD 20707\n                           (301) 953-6231\n    (c) Date received: 9 May 1990\n \n 3. Past Usage:\n     1. Smith,~J.~W., Everhart,~J.~E., Dickson,~W.~C., Knowler,~W.~C., &\n        Johannes,~R.~S. (1988). Using the ADAP learning algorithm to forecast\n        the onset of diabetes mellitus.  In {it Proceedings of the Symposium\n        on Computer Applications and Medical Care} (pp. 261--265).  IEEE\n        Computer Society Press.\n \n        The diagnostic, binary-valued variable investigated is whether the\n        patient shows signs of diabetes according to World Health Organization\n        criteria (i.e., if the 2 hour post-load plasma glucose was at least \n        200 mg/dl at any survey  examination or if found during routine medical\n        care).   The population lives near Phoenix, Arizona, USA.\n \n        Results: Their ADAP algorithm makes a real-valued prediction between\n        0 and 1.  This was transformed into a binary decision using a cutoff of \n        0.448.  Using 576 training instances, the sensitivity and specificity\n        of their algorithm was 76% on the remaining 192 instances.\n \n 4. Relevant Information:\n       Several constraints were placed on the selection of these instances from\n       a larger database.  In particular, all patients here are females at\n       least 21 years old of Pima Indian heritage.  ADAP is an adaptive learning\n       routine that generates and executes digital analogs of perceptron-like\n       devices.  It is a unique algorithm; see the paper for details.\n \n 5. Number of Instances: 768\n \n 6. Number of Attributes: 8 plus class \n \n 7. For Each Attribute: (all numeric-valued)\n    1. Number of times pregnant\n    2. Plasma glucose concentration a 2 hours in an oral glucose tolerance test\n    3. Diastolic blood pressure (mm Hg)\n    4. Triceps skin fold thickness (mm)\n    5. 2-Hour serum insulin (mu U/ml)\n    6. Body mass index (weight in kg/(height in m)^2)\n    7. Diabetes pedigree function\n    8. Age (years)\n    9. Class variable (0 or 1)\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: (class value 1 is interpreted as \"tested positive for\n    diabetes\")\n \n    Class Value  Number of instances\n    0            500\n    1            268\n \n 10. Brief statistical analysis:\n \n     Attribute number:    Mean:   Standard Deviation:\n     1.                     3.8     3.4\n     2.                   120.9    32.0\n     3.                    69.1    19.4\n     4.                    20.5    16.0\n     5.                    79.8   115.2\n     6.                    32.0     7.9\n     7.                     0.5     0.3\n     8.                    33.2    11.8\n \n \n\n\n\n\n Relabeled values in attribute \'class\'\n    From: 0                       To: tested_negative     \n    From: 1                       To: tested_positive', 1),
-(21, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n;\n ; Thyroid disease records supplied by the Garavan Institute and J. Ross\n ; Quinlan, New South Wales Institute, Syndney, Australia.\n ;\n ; 1987.\n ;\n \n sick, negative.			|  classes\n \n age:				continuous.\n sex:				M, F.\n on thyroxine:			f, t.\n query on thyroxine:		f, t.\n on antithyroid medication:	f, t.\n sick:				f, t.\n pregnant:			f, t.\n thyroid surgery:		f, t.\n I131 treatment:			f, t.\n query hypothyroid:		f, t.\n query hyperthyroid:		f, t.\n lithium:			f, t.\n goitre:				f, t.\n tumor:				f, t.\n hypopituitary:			f, t.\n psych:				f, t.\n TSH measured:			f, t.\n TSH:				continuous.\n T3 measured:			f, t.\n T3:				continuous.\n TT4 measured:			f, t.\n TT4:				continuous.\n T4U measured:			f, t.\n T4U:				continuous.\n FTI measured:			f, t.\n FTI:				continuous.\n TBG measured:			f, t.\n TBG:				continuous.\n referral source:		WEST, STMW, SVHC, SVI, SVHD, other.\n \n\n Num Instances:     3772\n Num Attributes:    30\n Num Continuous:    7 (Int 1 / Real 6)\n Num Discrete:      23\n Missing values:    6064 /  5.4%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'age\'                     Int    0% 100%   0%     1 /  0%    93 /  2%   0% \n   2 \'sex\'                     Enum  96%   0%   0%   150 /  4%     2 /  0%   0% \n   3 \'on thyroxine\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'query on thyroxine\'      Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'on antithyroid medicati  Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'sick\'                    Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'pregnant\'                Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'thyroid surgery\'         Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'I131 treatment\'          Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'query hypothyroid\'       Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'query hyperthyroid\'      Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'lithium\'                 Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'goitre\'                  Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'tumor\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'hypopituitary\'           Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  16 \'psych\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'TSH measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'TSH\'                     Real   0%  11%  79%   369 / 10%   287 /  8%   2% \n  19 \'T3 measured\'             Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'T3\'                      Real   0%   9%  71%   769 / 20%    69 /  2%   0% \n  21 \'TT4 measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'TT4\'                     Real   0%  94%   0%   231 /  6%   241 /  6%   1% \n  23 \'T4U measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'T4U\'                     Real   0%   2%  87%   387 / 10%   146 /  4%   1% \n  25 \'FTI measured\'            Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'FTI\'                     Real   0%  90%   0%   385 / 10%   234 /  6%   2% \n  27 \'TBG measured\'            Enum 100%   0%   0%     0 /  0%     1 /  0%   0% \n  28 \'TBG\'                     Real   0%   0%   0%  3772 /100%     0 /  0%   0% \n  29 \'referral source\'         Enum 100%   0%   0%     0 /  0%     5 /  0%   0% \n  30 \'Class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 1),
-(22, '1', '**Author**: R.S. Michalski and R.L. Chilausky (Donors: Ming Tan & Jeff Schlimmer)  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Soybean+(Large)) - 1988  \r\n**Please cite**: R.S. Michalski and R.L. Chilausky \"Learning by Being Told and Learning from Examples: An Experimental Comparison of the Two Methods of Knowledge Acquisition in the Context of Developing an Expert System for Soybean Disease Diagnosis\", International Journal of Policy Analysis and Information Systems, Vol. 4, No. 2, 1980.  \r\n\r\n**Large Soybean Database**  \r\nThis is the large soybean database from the UCI repository, with its training and test database combined into a single file. There are 19 classes, only the first 15 of which have been used in prior work.  The folklore seems to be that the last four classes are unjustified by the data since they have so few examples. There are 35 categorical attributes, some nominal and some ordered.  The value \'dna\' means does not apply.  The values for attributes are encoded numerically, with the first value encoded as \'0\', the second as \'1\', and so forth.  An unknown values is encoded as \'?\'.', 1),
-(23, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title:  SPAM E-mail Database\n \n 2. Sources:\n    (a) Creators: Mark Hopkins, Erik Reeber, George Forman, Jaap Suermondt\n         Hewlett-Packard Labs, 1501 Page Mill Rd., Palo Alto, CA 94304\n    (b) Donor: George Forman (gforman at nospam hpl.hp.com)  650-857-7835\n    (c) Generated: June-July 1999\n \n 3. Past Usage:\n    (a) Hewlett-Packard Internal-only Technical Report. External forthcoming.\n    (b) Determine whether a given email is spam or not.\n    (c) ~7% misclassification error.\n        False positives (marking good mail as spam) are very undesirable.\n        If we insist on zero false positives in the training/testing set,\n        20-25% of the spam passed through the filter.\n \n 4. Relevant Information:\n         The \"spam\" concept is diverse: advertisements for products/web\n         sites, make money fast schemes, chain letters, pornography...\n 	Our collection of spam e-mails came from our postmaster and \n 	individuals who had filed spam.  Our collection of non-spam \n 	e-mails came from filed work and personal e-mails, and hence\n 	the word \'george\' and the area code \'650\' are indicators of \n 	non-spam.  These are useful when constructing a personalized \n 	spam filter.  One would either have to blind such non-spam \n 	indicators or get a very wide collection of non-spam to \n 	generate a general purpose spam filter.\n \n         For background on spam:\n         Cranor, Lorrie F., LaMacchia, Brian A.  Spam! \n         Communications of the ACM, 41(8):74-83, 1998.\n \n 5. Number of Instances: 4601 (1813 Spam = 39.4%)\n \n 6. Number of Attributes: 58 (57 continuous, 1 nominal class label)\n \n 7. Attribute Information:\n The last column of \'spambase.data\' denotes whether the e-mail was \n considered spam (1) or not (0), i.e. unsolicited commercial e-mail.  \n Most of the attributes indicate whether a particular word or\n character was frequently occuring in the e-mail.  The run-length\n attributes (55-57) measure the length of sequences of consecutive \n capital letters.  For the statistical measures of each attribute, \n see the end of this file.  Here are the definitions of the attributes:\n \n 48 continuous real [0,100] attributes of type word_freq_WORD \n = percentage of words in the e-mail that match WORD,\n i.e. 100 * (number of times the WORD appears in the e-mail) / \n total number of words in e-mail.  A \"word\" in this case is any \n string of alphanumeric characters bounded by non-alphanumeric \n characters or end-of-string.\n \n 6 continuous real [0,100] attributes of type char_freq_CHAR\n = percentage of characters in the e-mail that match CHAR,\n i.e. 100 * (number of CHAR occurences) / total characters in e-mail\n \n 1 continuous real [1,...] attribute of type capital_run_length_average\n = average length of uninterrupted sequences of capital letters\n \n 1 continuous integer [1,...] attribute of type capital_run_length_longest\n = length of longest uninterrupted sequence of capital letters\n \n 1 continuous integer [1,...] attribute of type capital_run_length_total\n = sum of length of uninterrupted sequences of capital letters\n = total number of capital letters in the e-mail\n \n 1 nominal {0,1} class attribute of type spam\n = denotes whether the e-mail was considered spam (1) or not (0), \n i.e. unsolicited commercial e-mail.  \n \n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution:\n 	Spam	  1813  (39.4%)\n 	Non-Spam  2788  (60.6%)\n \n \n Attribute Statistics:\n    Min: Max:   Average:  Std.Dev: Coeff.Var_%: \n 1  0    4.54   0.10455   0.30536  292          \n 2  0    14.28  0.21301   1.2906   606          \n 3  0    5.1    0.28066   0.50414  180          \n 4  0    42.81  0.065425  1.3952   2130         \n 5  0    10     0.31222   0.67251  215          \n 6  0    5.88   0.095901  0.27382  286          \n 7  0    7.27   0.11421   0.39144  343          \n 8  0    11.11  0.10529   0.40107  381          \n 9  0    5.26   0.090067  0.27862  309          \n 10 0    18.18  0.23941   0.64476  269          \n 11 0    2.61   0.059824  0.20154  337          \n 12 0    9.67   0.5417    0.8617   159          \n 13 0    5.55   0.09393   0.30104  320          \n 14 0    10     0.058626  0.33518  572          \n 15 0    4.41   0.049205  0.25884  526          \n 16 0    20     0.24885   0.82579  332          \n 17 0    7.14   0.14259   0.44406  311          \n 18 0    9.09   0.18474   0.53112  287          \n 19 0    18.75  1.6621    1.7755   107          \n 20 0    18.18  0.085577  0.50977  596          \n 21 0    11.11  0.80976   1.2008   148          \n 22 0    17.1   0.1212    1.0258   846          \n 23 0    5.45   0.10165   0.35029  345          \n 24 0    12.5   0.094269  0.44264  470          \n 25 0    20.83  0.5495    1.6713   304          \n 26 0    16.66  0.26538   0.88696  334          \n 27 0    33.33  0.7673    3.3673   439          \n 28 0    9.09   0.12484   0.53858  431          \n 29 0    14.28  0.098915  0.59333  600          \n 30 0    5.88   0.10285   0.45668  444          \n 31 0    12.5   0.064753  0.40339  623          \n 32 0    4.76   0.047048  0.32856  698          \n 33 0    18.18  0.097229  0.55591  572          \n 34 0    4.76   0.047835  0.32945  689          \n 35 0    20     0.10541   0.53226  505          \n 36 0    7.69   0.097477  0.40262  413          \n 37 0    6.89   0.13695   0.42345  309          \n 38 0    8.33   0.013201  0.22065  1670         \n 39 0    11.11  0.078629  0.43467  553          \n 40 0    4.76   0.064834  0.34992  540          \n 41 0    7.14   0.043667  0.3612   827          \n 42 0    14.28  0.13234   0.76682  579          \n 43 0    3.57   0.046099  0.22381  486          \n 44 0    20     0.079196  0.62198  785          \n 45 0    21.42  0.30122   1.0117   336          \n 46 0    22.05  0.17982   0.91112  507          \n 47 0    2.17   0.0054445 0.076274 1400         \n 48 0    10     0.031869  0.28573  897          \n 49 0    4.385  0.038575  0.24347  631          \n 50 0    9.752  0.13903   0.27036  194          \n 51 0    4.081  0.016976  0.10939  644          \n 52 0    32.478 0.26907   0.81567  303          \n 53 0    6.003  0.075811  0.24588  324          \n 54 0    19.829 0.044238  0.42934  971          \n 55 1    1102.5 5.1915    31.729   611          \n 56 1    9989   52.173    194.89   374          \n 57 1    15841  283.29    606.35   214          \n 58 0    1      0.39404   0.4887   124          \n \n \n This file: \'spambase.DOCUMENTATION\' at the UCI Machine Learning Repository\n http://www.ics.uci.edu/~mlearn/MLRepository.html\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(24, '1', '**Author**: Genbank  \nDonor: G. Towell, M. Noordewier, and J. Shavlik  \n**Source**: [Genbank 64.1](genbank.bio.net) - 1/1/92  \n**Please cite**:   \n\nPrimate splice-junction gene sequences (DNA) with associated imperfect domain theory. All examples taken from Genbank 64.1. Categories \"ei\" and \"ie\" include every \"split-gene\" for primates in Genbank 64.1. Non-splice examples taken from sequences known not to include a splicing site.\n \nProblem Description:  \nSplice junctions are points on a DNA sequence at which \'superfluous\' DNA is removed during the process of protein creation in higher organisms. The problem posed in this dataset is to recognize, given a sequence of DNA, the boundaries between exons (the parts of the DNA sequence retained after splicing) and introns (the parts of the DNA sequence that are spliced out). This problem consists of two subtasks: recognizing exon/intron boundaries (referred to as EI sites), and recognizing intron/exon boundaries (IE sites). (In the biological community, IE borders are referred to a \'\'acceptors\'\' while EI borders are referred to as \'\'donors\'\'.)\n \nThis dataset has been developed to help evaluate a \"hybrid\" learning algorithm (KBANN) that uses examples to inductively refine preexisting knowledge.\n        \nAttributes:  \n>\n              1   One of {n ei ie}, indicating the class.\n              2   The instance name.\n           3-62   The remaining 60 fields are the sequence, starting at \n                  position -30 and ending at position +30. Each of\n                  these fields is almost always filled by one of \n                  {a, g, t, c}. Other characters indicate ambiguity among\n                  the standard characters according to the following table:\n    character: meaning\n        D: A or G or T\n        N: A or G or C or T\n        S: C or G\n        R: A or G\n\nNotes:  \n* Instance_name is an identifier and should be ignored for modelling', 1),
-(25, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Tic-Tac-Toe Endgame database\n \n 2. Source Information\n    -- Creator: David W. Aha (aha@cs.jhu.edu)\n    -- Donor: David W. Aha (aha@cs.jhu.edu)\n    -- Date: 19 August 1991\n  \n 3. Known Past Usage: \n    1. Matheus,~C.~J., & Rendell,~L.~A. (1989).  Constructive\n       induction on decision trees.  In {it Proceedings of the\n       Eleventh International Joint Conference on Artificial Intelligence} \n       (pp. 645--650).  Detroit, MI: Morgan Kaufmann.\n       -- CITRE was applied to 100-instance training and 200-instance test\n          sets.  In a study using various amounts of domain-specific\n          knowledge, its highest average accuracy was 76.7% (using the\n          final decision tree created for testing).\n \n    2. Matheus,~C.~J. (1990). Adding domain knowledge to SBL through\n       feature construction.  In {it Proceedings of the Eighth National\n       Conference on Artificial Intelligence} (pp. 803--808). \n       Boston, MA: AAAI Press.\n       -- Similar experiments with CITRE, includes learning curves up\n          to 500-instance training sets but used _all_ instances in the\n          database for testing.  Accuracies reached above 90%, but specific\n          values are not given (see Chris\'s dissertation for more details).\n \n    3. Aha,~D.~W. (1991). Incremental constructive induction: An instance-based\n       approach.  In {it Proceedings of the Eighth International Workshop\n       on Machine Learning} (pp. 117--121).  Evanston, ILL: Morgan Kaufmann.\n       -- Used 70% for training, 30% of the instances for testing, evaluated\n          over 10 trials.  Results reported for six algorithms:\n          -- NewID:   84.0%\n          -- CN2:     98.1%  \n          -- MBRtalk: 88.4%\n          -- IB1:     98.1% \n          -- IB3:     82.0%\n          -- IB3-CI:  99.1%\n       -- Results also reported when adding an additional 10 irrelevant \n          ternary-valued attributes; similar _relative_ results except that\n          IB1\'s performance degraded more quickly than the others.\n \n 4. Relevant Information:\n \n    This database encodes the complete set of possible board configurations\n    at the end of tic-tac-toe games, where \"x\" is assumed to have played\n    first.  The target concept is \"win for x\" (i.e., true when \"x\" has one\n    of 8 possible ways to create a \"three-in-a-row\").  \n \n    Interestingly, this raw database gives a stripped-down decision tree\n    algorithm (e.g., ID3) fits.  However, the rule-based CN2 algorithm, the\n    simple IB1 instance-based learning algorithm, and the CITRE \n    feature-constructing decision tree algorithm perform well on it.\n \n 5. Number of Instances: 958 (legal tic-tac-toe endgame boards)\n \n 6. Number of Attributes: 9, each corresponding to one tic-tac-toe square\n \n 7. Attribute Information: (x=player x has taken, o=player o has taken, b=blank)\n \n     1. top-left-square: {x,o,b}\n     2. top-middle-square: {x,o,b}\n     3. top-right-square: {x,o,b}\n     4. middle-left-square: {x,o,b}\n     5. middle-middle-square: {x,o,b}\n     6. middle-right-square: {x,o,b}\n     7. bottom-left-square: {x,o,b}\n     8. bottom-middle-square: {x,o,b}\n     9. bottom-right-square: {x,o,b}\n    10. Class: {positive,negative}\n \n 8. Missing Attribute Values: None\n \n 9. Class Distribution: About 65.3% are positive (i.e., wins for \"x\")\n\n Information about the dataset\n CLASSTYPE: nominal\n CLASSINDEX: last', 1),
-(26, '1', '**Author**:  Peter Mowforth  \n**Source**: UCI -   \n**Please cite**: Siebert,JP. Turing Institute Research Memorandum TIRM-87-018 \"Vehicle Recognition Using Rule Based Methods\" (March 1987)  \n\n NAME\n         vehicle silhouettes\n \n PURPOSE\n         to classify a given silhouette as one of four types of vehicle,\n         using  a set of features extracted from the silhouette. The\n         vehicle may be viewed from one of many different angles.  \n \n PROBLEM TYPE\n         classification\n         \n SOURCE\n         Drs.Pete Mowforth and Barry Shepherd\n         Turing Institute\n         George House\n         36 North Hanover St.\n         Glasgow\n         G1 2AD\n \n CONTACT\n         Alistair Sutherland\n         Statistics Dept.\n         Strathclyde University\n         Livingstone Tower\n         26 Richmond St.\n         GLASGOW G1 1XH\n         Great Britain\n         \n         Tel: 041 552 4400 x3033\n         \n         Fax: 041 552 4711 \n         \n         e-mail: alistair@uk.ac.strathclyde.stams\n \n HISTORY\n         This data was originally gathered at the TI in 1986-87 by\n         JP Siebert. It was partially financed by Barr and Stroud Ltd.\n         The original purpose was to find a method of distinguishing\n         3D objects within a 2D image by application of an ensemble of\n         shape feature extractors to the 2D silhouettes of the objects.\n         Measures of shape features extracted from example silhouettes\n         of objects to be discriminated were used to generate a class-\n         ification rule tree by means of computer induction.\n          This object recognition strategy was successfully used to \n         discriminate between silhouettes of model cars, vans and buses\n         viewed from constrained elevation but all angles of rotation.\n          The rule tree classification performance compared favourably\n         to MDC (Minimum Distance Classifier) and k-NN (k-Nearest Neigh-\n         bour) statistical classifiers in terms of both error rate and\n         computational efficiency. An investigation of these rule trees\n         generated by example indicated that the tree structure was \n         heavily influenced by the orientation of the objects, and grouped\n         similar object views into single decisions.\n \n DESCRIPTION\n          The features were extracted from the silhouettes by the HIPS\n         (Hierarchical Image Processing System) extension BINATTS, which \n         extracts a combination of scale independent features utilising\n         both classical moments based measures such as scaled variance,\n         skewness and kurtosis about the major/minor axes and heuristic\n         measures such as hollows, circularity, rectangularity and\n         compactness.\n          Four \"Corgie\" model vehicles were used for the experiment:\n         a double decker bus, Cheverolet van, Saab 9000 and an Opel Manta 400.\n         This particular combination of vehicles was chosen with the \n         expectation that the bus, van and either one of the cars would\n         be readily distinguishable, but it would be more difficult to\n         distinguish between the cars.\n          The images were acquired by a camera looking downwards at the\n         model vehicle from a fixed angle of elevation (34.2 degrees\n         to the horizontal). The vehicles were placed on a diffuse\n         backlit surface (lightbox). The vehicles were painted matte black\n         to minimise highlights. The images were captured using a CRS4000\n         framestore connected to a vax 750. All images were captured with\n         a spatial resolution of 128x128 pixels quantised to 64 greylevels.\n         These images were thresholded to produce binary vehicle silhouettes,\n         negated (to comply with the processing requirements of BINATTS) and\n         thereafter subjected to shrink-expand-expand-shrink HIPS modules to\n         remove \"salt and pepper\" image noise.\n          The vehicles were rotated and their angle of orientation was measured\n         using a radial graticule beneath the vehicle. 0 and 180 degrees\n         corresponded to \"head on\" and \"rear\" views respectively while 90 and\n         270 corresponded to profiles in opposite directions. Two sets of\n         60 images, each set covering a full 360 degree rotation, were captured\n         for each vehicle. The vehicle was rotated by a fixed angle between \n         images. These datasets are known as e2 and e3 respectively.\n          A further two sets of images, e4 and e5, were captured with the camera \n         at elevations of 37.5 degs and 30.8 degs respectively. These sets\n         also contain 60 images per vehicle apart from e4.van which contains\n         only 46 owing to the difficulty of containing the van in the image\n         at some orientations.\n \n ATTRIBUTES\n         \n         COMPACTNESS     (average perim)**2/area\n         \n         CIRCULARITY     (average radius)**2/area\n         \n         DISTANCE CIRCULARITY    area/(av.distance from border)**2\n         \n         RADIUS RATIO    (max.rad-min.rad)/av.radius\n         \n         PR.AXIS ASPECT RATIO    (minor axis)/(major axis)\n         \n         MAX.LENGTH ASPECT RATIO (length perp. max length)/(max length)\n         \n         SCATTER RATIO   (inertia about minor axis)/(inertia about major axis)\n         \n         ELONGATEDNESS           area/(shrink width)**2\n         \n         PR.AXIS RECTANGULARITY  area/(pr.axis length*pr.axis width)\n         \n         MAX.LENGTH RECTANGULARITY area/(max.length*length perp. to this)\n         \n         SCALED VARIANCE         (2nd order moment about minor axis)/area\n         ALONG MAJOR AXIS\n         \n         SCALED VARIANCE         (2nd order moment about major axis)/area\n         ALONG MINOR AXIS \n         \n         SCALED RADIUS OF GYRATION       (mavar+mivar)/area\n         \n         SKEWNESS ABOUT  (3rd order moment about major axis)/sigma_min**3\n         MAJOR AXIS\n         \n         SKEWNESS ABOUT  (3rd order moment about minor axis)/sigma_maj**3\n         MINOR AXIS\n                 \n         KURTOSIS ABOUT  (4th order moment about major axis)/sigma_min**4\n         MINOR AXIS  \n                 \n         KURTOSIS ABOUT  (4th order moment about minor axis)/sigma_maj**4\n         MAJOR AXIS\n         \n         HOLLOWS RATIO   (area of hollows)/(area of bounding polygon)\n         \n          Where sigma_maj**2 is the variance along the major axis and\n         sigma_min**2 is the variance along the minor axis, and\n         \n         area of hollows= area of bounding poly-area of object \n         \n          The area of the bounding polygon is found as a side result of\n         the computation to find the maximum length. Each individual\n         length computation yields a pair of calipers to the object\n         orientated at every 5 degrees. The object is propagated into\n         an image containing the union of these calipers to obtain an\n         image of the bounding polygon. \n         \n NUMBER OF CLASSES\n \n         4       OPEL, SAAB, BUS, VAN\n \n NUMBER OF EXAMPLES\n \n                 Total no. = 946\n                 \n                 No. in each class\n                 \n                   opel 240\n                   saab 240\n                   bus  240\n                   van  226\n                 \n                 \n                 100 examples are being kept by Strathclyde for validation.\n                 So StatLog partners will receive 846 examples.\n \n NUMBER OF ATTRIBUTES\n \n                 No. of atts. = 18', 1),
-(27, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Waveform Database Generator (written in C)\n  \n 2. Source:\n    (a) Breiman,L., Friedman,J.H., Olshen,R.A., & Stone,C.J. (1984). \n        Classification and Regression Trees.  Wadsworth International\n        Group: Belmont, California.  (see pages 43-49).\n    (b) Donor: David Aha \n    (c) Date: 11/10/1988\n \n 3. Past Usage:\n      1. CART book (above):\n         -- Optimal Bayes classification rate: 86% accuracy\n         -- CART decision tree algorithm: 72%\n         -- Nearest Neighbor Algorithm: 78%\n            --	300 training and 5000 test instances\n \n 4. Relevant Information:\n      -- 3 classes of waves\n      -- 21 attributes, all of which include noise\n      -- See the book for details (49-55, 169)\n      -- waveform.data.Z contains 5000 instances\n \n 5. Number of Instances: chosen by user\n \n 6. Number of Attributes:\n     -- 21 attributes with continuous values between 0 and 6\n \n 7. Attribute Information:\n     -- Each class is generated from a combination of 2 of 3 \"base\" waves\n     -- Each instance is generated f added noise (mean 0, variance 1) in \n        each attribute\n      -- See the book for details (49-55, 169)\n     \n 8. Missing Attribute Values: none\n \n 9. Class Distribution: 33% for each of 3 classes', 1),
-(28, '1', '**Author**: M. Harries, J. Gama, A. Bifet  \n**Source**: [Gama](http://www.inescporto.pt/~jgama/ales/ales_5.html) - 2009  \n**Please cite**:   \n\n**Electricity** is a widely used dataset described by M. Harries and analysed by Gama. This data was collected from the Australian New South Wales Electricity Market. In this market, prices are not fixed and are affected by demand and supply of the market. They are set every five minutes. The ELEC dataset contains 45, 312 instances. The class label identifies the change of the price relative to a moving average of the last 24 hours.', 1),
-(29, '1', '**Author**: Ashwin Srinivasan, Department of Statistics and Data Modeling, University of Strathclyde \r\n**Source**: https://archive.ics.uci.edu/ml/datasets/Statlog+(Landsat+Satellite)\r\n**Please cite**:   \r\n\r\nThe database consists of the multi-spectral values of pixels in 3x3 neighbourhoods in a satellite image, and the classification associated with the central pixel in each neighbourhood. The aim is to predict this classification, given the multi-spectral values. In the sample database, the class of a pixel is coded as a number.', 1),
-(30, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title of Database: Abalone data\n \n 2. Sources:\n \n    (a) Original owners of database:\n 	Marine Resources Division\n 	Marine Research Laboratories - Taroona\n 	Department of Primary Industry and Fisheries, Tasmania\n 	GPO Box 619F, Hobart, Tasmania 7001, Australia\n 	(contact: Warwick Nash +61 02 277277, wnash@dpi.tas.gov.au)\n \n    (b) Donor of database:\n 	Sam Waugh (Sam.Waugh@cs.utas.edu.au)\n 	Department of Computer Science, University of Tasmania\n 	GPO Box 252C, Hobart, Tasmania 7001, Australia\n \n    (c) Date received: December 1995\n \n \n 3. Past Usage:\n \n    Sam Waugh (1995) \"Extending and benchmarking Cascade-Correlation\", PhD\n    thesis, Computer Science Department, University of Tasmania.\n \n    -- Test set performance (final 1044 examples, first 3133 used for training):\n 	24.86% Cascade-Correlation (no hidden nodes)\n 	26.25% Cascade-Correlation (5 hidden nodes)\n 	21.5%  C4.5\n 	 0.0%  Linear Discriminate Analysis\n 	 3.57% k=5 Nearest Neighbour\n       (Problem encoded as a classification task)\n \n    -- Data set samples are highly overlapped.  Further information is required\n 	to separate completely using affine combinations.  Other restrictions\n 	to data set examined.\n \n    David Clark, Zoltan Schreter, Anthony Adams \"A Quantitative Comparison of\n    Dystal and Backpropagation\", submitted to the Australian Conference on\n    Neural Networks (ACNN\'96). Data set treated as a 3-category classification\n    problem (grouping ring classes 1-8, 9 and 10, and 11 on).\n \n    -- Test set performance (3133 training, 1044 testing as above):\n 	64%    Backprop\n 	55%    Dystal\n    -- Previous work (Waugh, 1995) on same data set:\n 	61.40% Cascade-Correlation (no hidden nodes)\n 	65.61% Cascade-Correlation (5 hidden nodes)\n 	59.2%  C4.5\n 	32.57% Linear Discriminate Analysis\n 	62.46% k=5 Nearest Neighbour\n \n \n 4. Relevant Information Paragraph:\n \n    Predicting the age of abalone from physical measurements.  The age of\n    abalone is determined by cutting the shell through the cone, staining it,\n    and counting the number of rings through a microscope -- a boring and\n    time-consuming task.  Other measurements, which are easier to obtain, are\n    used to predict the age.  Further information, such as weather patterns\n    and location (hence food availability) may be required to solve the problem.\n \n    From the original data examples with missing values were removed (the\n    majority having the predicted value missing), and the ranges of the\n    continuous values have been scaled for use with an ANN (by dividing by 200).\n \n    Data comes from an original (non-machine-learning) study:\n \n 	Warwick J Nash, Tracy L Sellers, Simon R Talbot, Andrew J Cawthorn and\n 	Wes B Ford (1994) \"The Population Biology of Abalone (_Haliotis_\n 	species) in Tasmania. I. Blacklip Abalone (_H. rubra_) from the North\n 	Coast and Islands of Bass Strait\", Sea Fisheries Division, Technical\n 	Report No. 48 (ISSN 1034-3288)\n \n \n 5. Number of Instances: 4177\n \n \n 6. Number of Attributes: 8\n \n \n 7. Attribute information:\n \n    Given is the attribute name, attribute type, the measurement unit and a\n    brief description.  The number of rings is the value to predict: either\n    as a continuous value or as a classification problem.\n \n 	Name		Data Type	Meas.	Description\n 	----		---------	-----	-----------\n 	Sex		nominal			M, F, and I (infant)\n 	Length		continuous	mm	Longest shell measurement\n 	Diameter	continuous	mm	perpendicular to length\n 	Height		continuous	mm	with meat in shell\n 	Whole weight	continuous	grams	whole abalone\n 	Shucked weight	continuous	grams	weight of meat\n 	Viscera weight	continuous	grams	gut weight (after bleeding)\n 	Shell weight	continuous	grams	after being dried\n 	Rings		integer			+1.5 gives the age in years\n \n    Statistics for numeric domains:\n \n 		Length	Diam	Height	Whole	Shucked	Viscera	Shell	Rings\n 	Min	0.075	0.055	0.000	0.002	0.001	0.001	0.002	    1\n 	Max	0.815	0.650	1.130	2.826	1.488	0.760	1.005	   29\n 	Mean	0.524	0.408	0.140	0.829	0.359	0.181	0.239	9.934\n 	SD	0.120	0.099	0.042	0.490	0.222	0.110	0.139	3.224\n 	Correl	0.557	0.575	0.557	0.540	0.421	0.504	0.628	  1.0\n \n \n 8. Missing Attribute Values: None\n \n \n 9. Class Distribution:\n \n 	Class	Examples\n 	-----	--------\n 	1	1\n 	2	1\n 	3	15\n 	4	57\n 	5	115\n 	6	259\n 	7	391\n 	8	568\n 	9	689\n 	10	634\n 	11	487\n 	12	267\n 	13	203\n 	14	126\n 	15	103\n 	16	67\n 	17	58\n 	18	42\n 	19	32\n 	20	26\n 	21	14\n 	22	6\n 	23	9\n 	24	2\n 	25	1\n 	26	1\n 	27	2\n 	29	1\n 	-----	----\n 	Total	4177\n \n Num Instances:     4177\n Num Attributes:    9\n Num Continuous:    8 (Int 1 / Real 7)\n Num Discrete:      1\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'Sex\'                     Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n   2 \'Length\'                  Real   0%   0% 100%     0 /  0%   134 /  3%   0% \n   3 \'Diameter\'                Real   0%   0% 100%     0 /  0%   111 /  3%   0% \n   4 \'Height\'                  Real   0%   0% 100%     0 /  0%    51 /  1%   0% \n   5 \'Whole weight\'            Real   0%   0% 100%     0 /  0%  2429 / 58%  31% \n   6 \'Shucked weight\'          Real   0%   0% 100%     0 /  0%  1515 / 36%  10% \n   7 \'Viscera weight\'          Real   0%   0% 100%     0 /  0%   880 / 21%   3% \n   8 \'Shell weight\'            Real   0%   0% 100%     0 /  0%   926 / 22%   8% \n   9 \'Class_Rings\'             Int    0% 100%   0%     0 /  0%    28 /  1%   0%', 1),
-(31, '1', '**Author**: Bruce Bulloch  \n**Source**: [WEKA Dataset Collection](http://www.cs.waikato.ac.nz/ml/weka/datasets.html)   \n**Please cite**:   \n\n**Eucalyptus Soil Conservation**  \nThe objective was to determine which seedlots in a species are best for soil conservation in seasonally dry hill country. Determination is found by measurement of height, diameter by height, survival, and other contributing factors. \n \nIt is important to note that eucalypt trial methods changed over time; earlier trials included mostly 15 - 30cm tall seedling grown in peat plots and the later trials have included mostly three replications of eight trees grown. This change may contribute to less significant results.\n\nExperimental data recording procedures which require noting include:\n - instances with no data recorded due to experimental recording procedures\n   require that the absence of a species from one replicate at a site was\n   treated as a missing value, but if absent from two or more replicates at a\n   site the species was excluded from the site\'s analyses.\n - missing data for survival, vigour, insect resistance, stem form, crown form\n   and utility especially for the data recorded at the Morea Station; this \n   could indicate the death of species in these areas or a lack in collection\n   of data.  \n\n Attribute Information:  \n  1.  Abbrev - site abbreviation - enumerated\n  2.  Rep - site rep - integer\n  3.  Locality - site locality in the North Island - enumerated\n  4.  Map_Ref - map location in the North Island - enumerated\n  5.  Latitude - latitude approximation - enumerated\n  6.  Altitude - altitude approximation - integer\n  7.  Rainfall - rainfall (mm pa) - integer\n  8.  Frosts - frosts (deg. c) - integer\n  9.  Year - year of planting - integer\n  10. Sp - species code - enumerated\n  11. PMCno - seedlot number - integer\n  12. DBH - best diameter base height (cm) - real\n  13. Ht - height (m) - real\n  14. Surv - survival - integer\n  15. Vig - vigour - real\n  16. Ins_res - insect resistance - real\n  17. Stem_Fm - stem form - real\n  18. Crown_Fm - crown form - real\n  19. Brnch_Fm - branch form - real\n  Class:\n  20. Utility - utility rating - enumerated\n\n Class Distribution:\n  none    - 180\n  low     - 107\n  average - 130\n  good    - 214\n  best    - 105\n\n Contact: Bruce Bulloch, 128 Cook Street, Palmerston North, New Zealand', 1),
-(32, '1', '**Author**: Ron Cole and Mark Fanty (cole@cse.ogi.edu, fanty@cse.ogi.edu)  \n**Donor**: Tom Dietterich (tgd@cs.orst.edu)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/ISOLET) - 1994  \n**Please cite**:   \n\n**ISOLET (Isolated Letter Speech Recognition)**  \nThis data set was generated as follows. 150 subjects spoke the name of each letter of the alphabet twice. Hence, we have 52 training examples from each speaker. The speakers are grouped into sets of 30 speakers each, 4 groups can serve as trainings set, the last group as the test set.\n\nYou will note that 3 examples are missing.  I believe they were dropped due to difficulties in recording.\n\nI believe this is a good domain for a noisy, perceptual task.  It is also a very good domain for testing the scaling abilities of algorithms. For example, C4.5 on this domain is slower than backpropagation!\n\nPast Usage:\n* Fanty, M., Cole, R. (1991).  Spoken letter recognition.  In Lippman, R. P., Moody, J., and Touretzky, D. S. (Eds). Advances in Neural Information Processing Systems 3.  San Mateo, CA: Morgan Kaufmann.\nGoal: Predict which letter-name was spoken, a simple classification task.  95.9% correct classification using the OPT backpropagation implementation.  Training on isolet1+2+3+4, testing on isolet5.  Network architecture: 56 hidden units, 26 output units (one-per-class).\n\n* Dietterich, T. G., Bakiri, G. (1991)  Error-correcting output codes: A general method for improving multiclass inductive learning programs.  Proceedings of the Ninth National Conference on Artificial Intelligence (AAAI-91), Anaheim, CA: AAAI Press.\nGoal: same as above. 95.83% correct using OPT backpropagation.  (Architecture: 78 hidden units, 26 output units, one-per-class). 96.73% correct using a 30-bit error-correcting output code with OPT (Architecture: 156 hidden units, 30 output units).\n\n**Attributes**  \nAll attributes are continuous, real-valued attributes scaled into the range -1.0 to 1.0. The features are described in the paper by Cole and Fanty cited above. The features include spectral coefficients; contour features, sonorant features, pre-sonorant features, and post-sonorant features.  Exact order of appearance of the features is not known.', 94),
-(33, '1', '**Author**: Peter Turney (peter@ai.iit.nrc.ca)   \n**Source**: [UCI](https://archive.ics.uci.edu/ml/machine-learning-databases/undocumented/connectionist-bench/vowel/) - date unknown  \n**Please cite**:   \n\n**Vowel Recognition Data**  \nIn my work on context-sensitive learning, I used the \"Deterding Vowel Recognition Data\", but I found it necessary to reformulate the data. Implicit in the original data is contextual information on the speaker\'s gender and identity. For my work, it was necessary to make this information explicit. This dataset adds the speaker\'s sex and identity as new features. \n\nNotes:  \n* This is version 2. Version 1 is hidden because it includes a feature dividing the data in train and test set. In OpenML this information is explicitly available in the corresponding task.', 2),
-(34, '1', '**Author**:   \n  \n**Source**: Unknown -   \n**Please cite**:   \n\nScene recognition dataset\n\nSource:\nMatthew R. Boutell, Jiebo Luo, Xipeng Shen, and Christopher M. Brown.\nLearning multi-label scene classification.\nPattern Recognition, 37(9):1757-1771, 2004. \n\n1: Description.\nScenes data set contains characteristics about images and thier classes. One image can belong to one or more classes.\n\n2: Type: Multi label \n3: Origin: Real world\n4: Instances: 2407\n5: Features: 294\n6: Labels: 6 \n7: Missing values: No', 94),
-(35, '1', '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 1\n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks. \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA  E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992\n\n4. Relevant Information:  The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991.  One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users.  There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.\n\nAttribute information: \n1. class: 0, 1  \n2. a1: 1, 2, 3 \n3. a2: 1, 2, 3 \n4. a3: 1, 2 \n5. a4: 1, 2, 3 \n6. a5: 1, 2, 3, 4 \n7. a6: 1, 2 8. \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 2),
-(36, '1', '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) - October 1992  \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 2  \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992  \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.  \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2 \n\n8.   Target Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 2),
-(37, '1', '**Author**: Sebastian Thrun  \n**Source**: [original](https://archive.ics.uci.edu/ml/datasets/MONK\'s+Problems) -   \n**Please cite**:   \n\nThe Monk\'s Problems: Problem 3   \n\nThis is a merged version of the separate train and test set which are usually distributed. On OpenML this train-test split can be found as one of the possible tasks.   \n\nSources:  \n(a) Donor: Sebastian Thrun School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA E-mail: thrun@cs.cmu.edu  \n(b) Date: October 1992   \n\nRelevant Information: The MONK\'s problem were the basis of a first international comparison of learning algorithms. The result of this comparison is summarized in \"The MONK\'s Problems - A Performance Comparison of Different Learning algorithms\" by S.B. Thrun, J. Bala, E. Bloedorn, I. Bratko, B. Cestnik, J. Cheng, K. De Jong, S. Dzeroski, S.E. Fahlman, D. Fisher, R. Hamann, K. Kaufman, S. Keller, I. Kononenko, J. Kreuziger, R.S. Michalski, T. Mitchell, P. Pachowicz, Y. Reich H. Vafaie, W. Van de Welde, W. Wenzel, J. Wnek, and J. Zhang has been published as Technical Report CS-CMU-91-197, Carnegie Mellon University in Dec. 1991. One significant characteristic of this comparison is that it was performed by a collection of researchers, each of whom was an advocate of the technique they tested (often they were the creators of the various methods). In this sense, the results are less biased than in comparisons performed by a single person advocating a specific learning method, and more accurately reflect the generalization behavior of the learning techniques as applied by knowledgeable users. There are three MONK\'s problems. The domains for all MONK\'s problems are the same (described below). One of the MONK\'s problems has noise added. For each problem, the domain has been partitioned into a train and test set.   \n\nAttribute information:  \n1. class: 0, 1  \n2. a1: 1, 2, 3  \n3. a2: 1, 2, 3  \n4. a3: 1, 2  \n5. a4: 1, 2, 3  \n6. a5: 1, 2, 3, 4  \n7. a6: 1, 2  \n\nTarget Concepts associated to the MONK\'s problem:  \nMONK-1: (a1 = a2) or (a5 = 1)  \nMONK-2: EXACTLY TWO of {a1 = 1, a2 = 1, a3 = 1, a4 = 1, a5 = 1, a6 = 1}  \nMONK-3: (a5 = 3 and a4 = 1) or (a5 /= 4 and a2 /= 3) (5% class noise added to the training set)', 2),
-(38, '1', '**Author**: Mineichi Kudo, Jun Toyama, Masaru Shimbo ({mine,jun,shimbo}@main.eng.hokudai.ac.jp)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Japanese+Vowels) - 2000  \n**Please cite**:   \n\n**Japanese vowels**  \nThis dataset records 640 time series of 12 LPC cepstrum coefficients taken from nine male speakers.\n\nThe data was collected for examining our newly developed classifier for multidimensional curves (multidimensional time series). Nine male speakers uttered two Japanese vowels /ae/ successively. For each utterance, with the analysis parameters described below, we applied 12-degree linear prediction analysis to it to obtain a discrete-time series with 12 LPC cepstrum coefficients. This means that one utterance by a speaker forms a time series whose length is in the range 7-29 and each point of a time series is of 12 features (12 coefficients).\n\nSimilar data are available for different utterances /ei/, /iu/, /uo/, /oa/ in addition to /ae/. Please contact the donor if you are interested in using this data.\n\nThe number of the time series is 640 in total. We used one set of 270 time series for training and the other set of 370 time series for testing.\n\nAnalysis parameters:  \n* Sampling rate : 10kHz\n* Frame length : 25.6 ms\n* Shift length : 6.4ms\n* Degree of LPC coefficients : 12\n\nEach line represents 12 LPC coefficients in the increasing order separated by spaces. This corresponds to one analysis\nframe. Lines are organized into blocks, which are a set of 7-29 lines separated by blank lines and corresponds to a single speech utterance of /ae/ with 7-29 frames.\n\nEach speaker is a set of consecutive blocks. In ae.train there are 30 blocks for each speaker. Blocks 1-30 represent speaker 1, blocks 31-60 represent speaker 2, and so on up to speaker 9. In ae.test, speakers 1 to 9 have the corresponding number of blocks: 31 35 88 44 29 24 40 50 29. Thus, blocks 1-31 represent speaker 1 (31 utterances of /ae/), blocks 32-66 represent speaker 2 (35 utterances of /ae/), and so on.\n\n**Past Usage**\n\nM. Kudo, J. Toyama and M. Shimbo. (1999). \"Multidimensional Curve Classification Using Passing-Through Regions\". Pattern Recognition Letters, Vol. 20, No. 11--13, pages 1103--1111.\n\nIf you publish any work using the dataset, please inform the donor. Use for commercial purposes requires donor permission.\n\nReferences  \n\n1. http://ips9.main.eng.hokudai.ac.jp/index_e.html\n2. mailto:mine@main.eng.hokudai.ac.jp\n3. mailto:jun@main.eng.hokudai.ac.jp\n4. mailto:shimbo@main.eng.hokudai.ac.jp\n5. http://kdd.ics.uci.edu/\n6. http://www.ics.uci.edu/\n7. http://www.uci.edu/', 2),
-(39, '1', '**Author**: Dr Robert Alcock (rob@skyblue.csd.auth.gr)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Synthetic+Control+Chart+Time+Series) - 1999  \n**Please cite**:   \n\n**Synthetic Control Chart Time Series**  \nThis data consists of synthetically generated control charts. This dataset contains 600 examples of control charts synthetically generated by the process in Alcock and Manolopoulos (1999). There are six different classes of control charts:\n\n1. Normal\n2. Cyclic\n3. Increasing trend\n4. Decreasing trend\n5. Upward shift\n6. Downward shift\n\n**Past Usage**  \nAlcock R.J. and Manolopoulos Y. Time-Series Similarity Queries Employing a Feature-Based Approach. 7th Hellenic Conference on\nInformatics. August 27-29. Ioannina,Greece 1999.\n\n**References and Further Information**  \nD.T. Pham and A.B. Chan \"Control Chart Pattern Recognition using a New Type of Self Organizing Neural Network\" Proc. Instn, Mech, Engrs. Vol 212, No 1, pp 115-127 1998.\n\nReferences  \n\n1. http://skyblue.csd.auth.gr/~rob/\n2. mailto:rob@skyblue.csd.auth.gr\n3. http://kdd.ics.uci.edu/\n4. http://www.ics.uci.edu/\n5. http://www.uci.edu/', 2),
-(40, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nIrish Educational Transitions Data\n\nBelow are shown data on educational transitions for a sample of 500\nIrish schoolchildren aged 11 in 1967. The data were collected by\nGreaney and Kelleghan (1984), and reanalyzed by Raftery and Hout (1985, 1993).\n\nThe data were also used, in a simplified form, as an example to illustrate\nBayesian model selection methods by Raftery (1988) and Kass and Raftery (1993).\nIn that simplified form, primary terminal leavers and cases with any missing\ndata were removed, leaving 441 cases. The Leaving Certificate variable\nwas used as the dependent variable in a logistic regression analysis.\n\nThe variables shown are as follows:\n\n1. Sex: 1=male; 2=female.\n\n2. DVRT (Drumcondra Verbal Reasoning Test Score).\n\n3. Educational level attained:\n1		Primary terminal leaver\n2	 	Junior cycle incomplete: vocational school\n3		Junior cycle incomplete: secondary school\n4		Junior cycle terminal leaver: vocational school\n5		Junior cycle terminal leaver: secondary school\n6		Senior cycle incomplete: vocational school\n7		Senior cycle incomplete: secondary school\n8		Senior cycle terminal leaver: vocational school\n9		Senior cycle terminal leaver: secondary school\n10		3rd level incomplete\n11		3rd level complete\n\n4. Leaving Certificate. 1 if Leaving Certificate not taken; 2 if taken.\n\n5. Prestige score for father\'s occupation\n(calculated by Raftery and Hout, 1985). 0 if missing.\n\n6. Type of school: 1=secondary; 2=vocational; 9=primary terminal leaver.\n\nREFERENCES\n\nGreaney, V. and Kelleghan, T. (1984). Equality of Opportunity in Irish\nSchools. Dublin: Educational Company.\n\nKass, R.E. and Raftery, A.E. (1993). Bayes factors and model uncertainty.\nTechnical Report no. 254, Department of Statistics, University of Washington.\nRevised version to appear in Journal of the American Statistical\nAssociation.\n\nRaftery, A.E. (1988). Approximate Bayes factors for generalized linear models.\nTechnical Report no. 121, Department of Statistics, University of Washington.\n\nRaftery, A.E. and Hout, M. (1985). Does Irish education approach the\nmeritocratic ideal? A logistic analysis.\nEconomic and Social Review, 16, 115-140.\n\nRaftery, A.E. and Hout, M. (1993). Maximally maintained inequality:\nExpansion, reform and opportunity in Irish schools.\nSociology of Education, 66, 41-62.\n\n\nOWNERSHIP STATEMENT\n\nThis data belongs to Vincent Greaney and Thomas Kelleghan,\nEducational Research Centre, St. Patrick\'s College, Drumcondra,\nDublin 9, Ireland, who retain the copyright.\n\nIn the form given here, it may be used solely as an example for research\non the development of statistical methods. For any other use of the data,\npermission must be obtained from the owners.  Subject to this statement,\npermission is hereby given to StatLib to distribute this data freely.\n\nSubmitted by Adrian Raftery (raftery@stat.washington.edu).\n\nCopyright 1984 Vincent Greaney and Thomas Kelleghan.\n\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: 4', 2),
-(41, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nanalcatdata    A collection of data sets used in the book \"Analyzing Categorical Data,\"\nby Jeffrey S. Simonoff, Springer-Verlag, New York, 2003. The submission\nconsists of a zip file containing two versions of each of 84 data sets,\nplus this README file. Each data set is given in comma-delimited ASCII\n(.csv) form, and Microsoft Excel (.xls) form.\n\nNOTICE: These data sets may be used freely for scientific, educational and/or\nnoncommercial purposes, provided suitable acknowledgment is given (by citing\nthe above-named reference).\n\nFurther details concerning the book, including information on statistical software\n(including sample S-PLUS/R and SAS code), are available at the web site\n\nhttp://www.stern.nyu.edu/~jsimonof/AnalCatData\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last\n\n\nNote: Quotes, Single-Quotes and Backslashes were removed, Blanks replaced\nwith Underscores', 2),
-(42, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nanalcatdata    A collection of data sets used in the book \"Analyzing Categorical Data,\"\nby Jeffrey S. Simonoff, Springer-Verlag, New York, 2003. The submission\nconsists of a zip file containing two versions of each of 84 data sets,\nplus this README file. Each data set is given in comma-delimited ASCII\n(.csv) form, and Microsoft Excel (.xls) form.\n\nNOTICE: These data sets may be used freely for scientific, educational and/or\nnoncommercial purposes, provided suitable acknowledgment is given (by citing\nthe above-named reference).\n\nFurther details concerning the book, including information on statistical software\n(including sample S-PLUS/R and SAS code), are available at the web site\n\nhttp://www.stern.nyu.edu/~jsimonof/AnalCatData\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last\n\n\nNote: Quotes, Single-Quotes and Backslashes were removed, Blanks replaced\nwith Underscores', 2),
-(43, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nPRO FOOTBALL SCORES  (raw data appears after the description below)\n\nHow well do the oddsmakers of Las Vegas predict the outcome of\nprofessional football games?  Is there really a home field advantage - if\nso how large is it?  Are teams that play the Monday Night game at a\ndisadvantage when they play again the following Sunday?  Do teams benefit\nfrom having a \"bye\" week off in the current schedule?  These questions and\na host of others can be investigated using this data set.\n\nHal Stern from the Statistics Department at Harvard University has\nmade available his compilation of scores for all National Football League\ngames from the 1989, 1990, and 1991 seasons.  Dr. Stern used these data as\npart of his presentation \"Who\'s Number One?\" in the special \"Best of\nBoston\" session at the 1992 Joint Statistics Meetings.\n\nSeveral variables in the data are keyed to the oddsmakers \"point\nspread\" for each game.  The point spread is a value assigned before each\ngame to serve as a handicap for whichever is perceived to be the better\nteam.  Thus, to win against the point spread, the \"favorite\" team must beat\nthe \"underdog\" team by more points than the spread.  The underdog \"wins\"\nagainst the spread if it wins the game outright or manages to lose by fewer\npoints than the spread.  In theory, the point spread should represent the\n\"expert\" prediction as to the game\'s outcome.  In practice, it more usually\ndenotes a point at which an equal amount of money will be wagered both for\nand against the favored team.\n\nRaw data below contains 672 cases (all 224 regular season games in\neach season and informatino on the following 9 varialbes:     .\n\nHome/Away       = Favored team is at home (1) or away (0)\nFavorite Points = Points scored by the favored team\nUnderdog Points = Points scored by the underdog team\nPointspread     = Oddsmaker\'s points to handicap the favored team\nFavorite Name   = Code for favored team\'s name\nUnderdog name   = Code for underdog\'s name\nYear            = 89, 90, or 91\nWeek            = 1, 2, ... 17\nSpecial         = Mon.night (M), Sat. (S), Thur. (H), Sun. night (N)\not - denotes an overtime game\n\n\nData were submitted by: Robin Lock  (rlock@stlawu.bitnet)\nMathematics Department, St. Lawrence University\n\nData were compiled by: Hal Stern, Dept. of Statistics, Harvard University\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: 1', 2),
-(44, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nThe following are data used in an analysis of the Brown and Frown corpora for my doctoral dissertation titled ``Variations in Written English: Characterizing Authors\' Rhetorical Language Choices Across Corpora of Published Texts\" (Completed at Carnegie Mellon Univ, 2003).  The source of the corpora was the ICAME CD-ROM  (get info at <http://www.hit.uib.no/icame/cd>).\n\nThe data were generated from the texts using tagging and visualization software, Docuscope.\n\nThe first row is the variable names. The genre of each text (assigned by the Brown corpus compilers) is in \'Genre\' column and the corpus is listed in the \'corpus\' column with 1=Brown and 2=Frown corpus.\n\nThe dataset may be freely used and distributed for non-commercial purposes.\n\nJeff Collins <jeff.collins@acm.org> 11 July 2003\n\n\n\nInformation about the dataset\nCLASSTYPE: nominal\nCLASSINDEX: last', 2),
-(45, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\nModified by TunedIT (converted to ARFF format)\n\n\nSYLVA is the ecology database\n\n\nThe task of SYLVA is to classify forest cover types. The forest cover type for 30 x 30 meter cells is obtained from US Forest Service (USFS) Region 2 Resource Information System (RIS) data. We brought it back to a two-class classification problem (classifying Ponderosa pine vs. everything else). The \"agnostic learning track\" data consists in 216 input variables. Each pattern is composed of 4 records: 2 true records matching the target and 2 records picked at random. Thus 1/2 of the features are distracters.\n\nData type: non-sparse\nNumber of features: 216\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	  805	12281	13086	238271607.00\nValid	   81	 1228	 1309	23817234.00\n\n\nThis dataset contains samples from both training and validation datasets.', 2),
-(46, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\n\nModified by TunedIT (converted to ARFF format)\n\n\nGINA is digit recognition database\n\nThe task of GINA is handwritten digit recognition. For the \"agnostic learning track\" we chose the problem of separating two-digit odd numbers from two-digit even numbers. Only the unit digit is informative for that task, therefore at least 1/2 of the features are distracters. Additionally, the pixels that are almost always blank were removed and the pixel order was randomized to hide the feature identity.  This is a two class classification problem with sparse continuous input variables, in which each class is composed of several clusters. It is a problem with heterogeneous classes.\n\nData type: non-sparse\nNumber of features: 970\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	 1550	 1603	 3153	164947945.00\nValid	  155	  160	  315	16688946.00\n\n\nThis dataset contains samples from both training and validation datasets.', 2),
-(47, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from the Agnostic Learning vs. Prior Knowledge Challenge (http://www.agnostic.inf.ethz.ch)\n\nDataset from: http://www.agnostic.inf.ethz.ch/datasets.php\n\n\nModified by TunedIT (converted to ARFF format)\n\n\nADA is the marketing database\n\nThe task of ADA is to discover high revenue people from census data. This is a two-class classification problem. The raw data from the census bureau is known as the Adult database in the UCI machine-learning repository. The 14 original attributes (features) include age, workclass,  education,\nmarital status, occupation, native country, etc. It contains continuous, binary and categorical features. This dataset is from the \"agnostic learning track\", i.e. has access to a preprocessed numeric representation eliminating categorical variables, but the identity of the features is not revealed.\n\n\n\nData type: non-sparse\nNumber of features: 48\nNumber of examples and check-sums:\nPos_ex	Neg_ex	Tot_ex	Check_sum\nTrain	 1029	 3118	 4147	6798109.00\nValid	  103	  312	  415	681151.00\n\n\nThis dataset contains samples from both training and validation datasets.', 2),
-(48, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promisedata.org/repository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n(c) 2007  A. Gunes Koru\nContact: gkoru AT umbc DOT edu Phone: +1 (410) 455 8843\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\n\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n\n\n1. Title: Recurrent event (defect fix) and size data for Mozilla Classes\nThis one includes a binary attribute (event) to show defect fix.\nThe data is at the \"observation\" level. Each modification made to\na C++ class was entered as an observation. A newly added class\ncreated an observation. The observation period was between\nMay 29, 2002 and Feb 22, 2006.\n\n2. Sources\n(a) Creator: A. Gunes Koru\n(b) Date: February 23, 2007\n(c) Contact: gkoru AT umbc DOT edu Phone: +1 (410) 455 8843\n\n3. Donor: A. Gunes Koru\n\n4. Past Usage: This data set was used for:\n\nA. Gunes Koru, Dongsong Zhang, and Hongfang Liu, \"Modeling the\nEffect of Size on Defect Proneness for Open-Source Software\",\nPredictive Models in Software Engineering Workshop, PROMISE 2007,\nMay 20th 2007, Minneapolis, Minnesota, US.\n\nAbstract:\nQuality is becoming increasingly important with the continuous\nadoption of open-source software.  Previous research has found that\nthere is generally a positive relationship between module size and\ndefect proneness. Therefore, in open-source software development, it\nis important to monitor module size and understand its impact on\ndefect proneness. However, traditional approaches to quality\nmodeling, which measure specific system snapshots and obtain future\ndefect counts, are not well suited because open-source modules\nusually evolve and their size changes over time. In this study, we\nused Cox proportional hazards modeling with recurrent events to\nstudy the effect of class size on defect-proneness in the Mozilla\nproduct. We found that the effect of size was significant, and we\nquantified this effect on defect proneness.\n\nThe full paper can be downloaded from A. Gunes Koru\'s Website\nhttp://umbc.edu/~gkoru\nby following the Publications link or from the Web site of PROMISE 2007.\n\n5. Features:\n\nThis data set is used to create a conditional Cox Proportional\nHazards Model\n\nid: A numeric identification assigned to each separate C++ class\n(Note that the id\'s do not increment from the first to the last\ndata row)\n\nstart: A time infinitesimally greater than the time of the modification\nthat created this observation (practically, modification time). When a\nclass is introduced to a system, a new observation is entered with start=0\n\nend: Either the time of the next modification, or the end of the\nobservation period, or the time of deletion, whichever comes first.\n\nevent: event is set to 1 if a defect fix takes place\nat the time represented by \'end\', or 0 otherwise.  A class deletion\nis handled easily by entering a final observation whose event is set\nto 1 if the class is deleted for corrective maintenance, or 0 otherwise.\n\nsize: It is a time-dependent covariate and its column carries the\nnumber of source Lines of Code of the C++ classes\nat time \'start\'. Blank and comment lines are not counted.\n\nstate: Initially set to 0, and it becomes 1 after the class\nexperiences an event, and remains at 1 thereafter.', 2),
-(49, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE data set made publicly available in order to encourage\nrepeatable, verifiable, refutable, and/or improvable predictive models\nof software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC4/ defect prediction\n\n(c) 2007 : Tim Menzies  : tim@menzies.us\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\nFor more deatils on this data set, see\nhttp://promisedata.org/repository/data/kc2/kc2.arff', 2),
-(50, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE data set made publicly available in order to encourage\nrepeatable, verifiable, refutable, and/or improvable predictive models\nof software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC3/software defect prediction\n\n(c) 2007 : Tim Menzies  : tim@menzies.us\nThis data set is distributed under the\nCreative Commons Attribution-Share Alike 3.0 License\nhttp://creativecommons.org/licenses/by-sa/3.0/\n\nYou are free:\n\n* to Share -- copy, distribute and transmit the work\n* to Remix -- to adapt the work\n\nUnder the following conditions:\n\nAttribution. You must attribute the work in the manner specified by\nthe author or licensor (but not in any way that suggests that they endorse\nyou or your use of the work).\n\nShare Alike. If you alter, transform, or build upon this work, you\nmay distribute the resulting work only under the same, similar or a\ncompatible license.\n\n* For any reuse or distribution, you must make clear to others the\nlicense terms of this work.\n* Any of the above conditions can be waived if you get permission from\nthe copyright holder.\n* Apart from the remix rights granted under this license, nothing in\nthis license impairs or restricts the author\'s moral rights.\nFor more deatils on this data set, see\nhttp://promisedata.org/repository/data/kc2/kc2.arff', 2),
-(51, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nThis is a PROMISE data set made publicly available in order to encourage repeatable, verifiable, refutable, and/or improvable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please follow the acknowledgment guidelines posted on [the PROMISE repository web page](http://promise.site.uottawa.ca/SERepository).\n\n## Title/Topic\nJM1/software defect prediction\n\n\n## Sources\n* **Creators:**  NASA, then the NASA Metrics Data Program, http://mdp.ivv.nasa.gov. \n* **Contacts:** \n  * Mike Chapman, Galaxy Global Corporation (Robert.Chapman@ivv.nasa.gov) +1-304-367-8341\n  * Pat Callis, NASA, NASA project manager for MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n* **Donor:** Tim Menzies (tim@barmag.net)\n\n* **Date:**  December 2nd, 2004\n\n## Past usage\n**_How Good is Your Blind Spot Sampling Policy?_; 2003; Tim Menzies and Justin S. Di Stefano; 2004 IEEE Conference on High Assurance Software Engineering (http://menzies.us/pdf/03blind.pdf).**\n\n### Results:\n* Very simple learners (ROCKY) perform as well in this domain as more sophisticated methods (e.g. J48, model trees, model trees) for predicting detects\n* Many learners have very low false alarm rates.\n* Probability of detection (PD) rises with effort and rarely rises above it.\n* High PDs are associated with high PFs (probability of failure)\n* PD, PF, effort can change significantly while accuracy remains essentially stable\n* With two notable exceptions, detectors learned from one data set (e.g. KC2) have nearly they same properties when applied to another (e.g. PC2, KC2). Exceptions:\n* LinesOfCode measures generate wider inter-data-set variances;\n* Precision\'s inter-data-set variances vary wildly\n\n**_\"Assessing Predictors of Software Defects\"_, T. Menzies and J. DiStefano and A. Orrego and R. Chapman, 2004,**\n**Proceedings, workshop on Predictive Software Models, Chicago, Available from http://menzies.us/pdf/04psm.pdf.**\n\n### Results:\n* From JM1, Naive Bayes generated PDs of 25% with PF of 20%\n* Naive Bayes out-performs J48 for defect detection\n* When learning on more and more data, little improvement is seen after processing 300 examples.\n* PDs are much higher from data collected below the sub-sub-system level.\n* Accuracy is a surprisingly uninformative measure of success for a defect detector. Two detectors with the same accuracy can have widely varying PDs and PFs.\n\n## Relevant information\n* JM1 is written in \"C\" and is a real-time predictive ground system: Uses simulations to generate predictions\n* Data comes from McCabe and Halstead features extractors of source code.  These features were defined in the 70s in an attempt to objectively characterize code features that are associated with software quality. The nature of association is under dispute.\n\nNotes on McCabe and Halstead follow.\n\n* The McCabe and Halstead measures are \"module\"-based where a \"module\" is the smallest unit of functionality. In C or Smalltalk, \"modules\" would be called \"function\" or \"method\" respectively.\n\n* Defect detectors can be assessed according to the following measures:\n\n    module actually has defects\n    +-------------+------------+\n    |     no      |     yes    |\n    +-----+-------------+------------+\n    classifier predicts no defects |  no |      a      |      b     |\n    +-----+-------------+------------+\n    classifier predicts some defects | yes |      c      |      d     |\n    +-----+-------------+------------+\n\n\n    accuracy                   = acc          = (a+d)/(a+b+c+d\n    probability of detection   = pd  = recall = d/(b+d)\n    probability of false alarm = pf           = c/(a+c)\n    precision                  = prec         = d/(c+d)\n    effort                     = amount of code selected by detector = (c.LOC + d.LOC)/(Total LOC)\n\n\nIdeally, detectors have high PDs, low PFs, and low effort. This ideal state rarely happens:\n\n* PD and effort are linked. The more modules that trigger the detector, the higher the PD. However, effort also gets increases\n* High PD or low PF comes at the cost of high PF or low PD (respectively). This linkage can be seen in a standard receiver operator curve (ROC).  Suppose, for example, LOC> x is used as the detector (i.e. we assume large modules have more errors). LOC > x represents a family of detectors. At x=0, EVERY module is predicted to have errors. This detector has a high PD but also a high false alarm rate. At x=0, NO module is predicted to have errors. This detector has a low false alarm rate but won\'t detect anything at all. At 0<x<1, a set of detectors are generated as shown below:\n\n    pd\n    1 |           x  x  x                KEY:\n    |        x     .                   \".\"  denotes the line PD=PF\n    |     x      .                     \"x\"  denotes the roc curve\n    |   x      .                            for a set of detectors\n    |  x     .\n    | x    .\n    | x  .\n    |x .\n    |x\n    x------------------ pf\n    0                   1\n\nNote that:\n* The only way to make no mistakes (PF=0) is to do nothing (PD=0)\n* The only way to catch more detects is to make more mistakes (increasing PD means increasing PF).\n* Our detector bends towards the \"sweet spot\" of <PD=1,PF=0> but does not reach it.\n* The line pf=pd on the above graph represents the \"no information\" line. If pf=pd then the detector is pretty useless. The better the detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more error-prone.  His metrics therefore reflect the pathways within a code module.\n\n    @Article{mccabe76,\n    title  = \"A Complexity Measure\",\n    author  = \"T.J. McCabe\",\n    pages  = \"308--320\",\n    journal = \"IEEE Transactions on Software Engineering\",\n    year  = \"1976\",\n    volume  = \"2\",\n    month  = \"December\",\n    number  = \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be fault prone. Halstead estimates reading complexity by counting the number of concepts in a module; e.g. number of unique operators.\n\n    @Book{halstead77,\n    Author    = \"M.H. Halstead\",\n    Title    = \"Elements of Software Science\",\n    Publisher = \"Elsevier \",\n    Year    = 1977}\n\nWe study these static code measures since they are useful, easy to use, and widely used:\n\n* USEFUL: E.g. this data set can generate highly accurate predictors for defects\n* EASY TO USE: Static code measures (e.g. lines of code, the McCabe/Halstead measures) can be automatically and cheaply collected.\n* WIDELY USED: Many researchers use static measures to guide software quality predictions (see the reference list in the above \"blind spot\" paper. Verification and validation (V\\&V) textbooks advise using static code complexity measures to decide which modules are worthy of manual inspections.  Further, we know of several large government software contractors that won\'t review software modules _unless_ tools like McCabe predict that they are fault prone.  Hence, defect detectors have a major economic impact when they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely criticized.  Static code measures are hardly a complete characterization of the internals of a function. Fenton offers an insightful example where the same functionality is achieved using different programming language constructs resulting in different static measurements for that module. Fenton uses this example to argue the uselessness of static code measures.\n\n    @Book{fenton97,\n    author    = \"N.E. Fenton and S.L. Pfleeger\",\n    title     = {Software metrics: a Rigorous \\& Practical Approach},\n    publisher = {International Thompson Press},\n    year      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static measures can never be a definite and certain indicator of the presence of a fault.  Rather, defect detectors based on static measures are best viewed as probabilistic statements that the frequency of faults tends to increase in code modules that trigger the detector.  By definition, such probabilistic statements will are not categorical claims with some a non-zero false alarm rate. The research challenge for data miners is to ensure that these false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics: essential complexity, cyclomatic complexity, design complexity and LOC, Lines of Code.\n\n* Cyclomatic Complexity, or \"v(G)\", measures the number of \"linearly independent paths\". A set of paths is said to be linearly independent if no path in the set is a linear combination of any other paths in the set through a program\'s \"flowgraph\". A flowgraph is a directed graph where each node corresponds to a program statement, and each arc indicates the flow of control from one statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\" where \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in the flowgraph, and \"n\" is the number of nodes in the flowgraph. The standard McCabes rules (\"v(G)\">10), are used to identify fault-prone module.\n* Essential Complexity, or \"ev(G)$\" is the extent to which a flowgraph can be \"reduced\" by decomposing all the subflowgraphs of $G$ that are \"D-structured primes\". Such \"D-structured primes\" are also sometimes referred to as \"proper one-entry one-exit subflowgraphs\" (for a more thorough discussion of D-primes, see the Fenton text referenced above). \"ev(G)\" is calculated using \"ev(G)= v(G) - m\" where $m$ is the number of subflowgraphs of \"G\" that are D-structured primes.\n* Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a module\'s reduced flowgraph.  The flowgraph, \"G\", of a module is reduced to eliminate any complexity which does not influence the interrelationship between design modules.  According to McCabe, this complexity measurement reflects the modules calling patterns to its immediate subordinate modules.\n* Lines of code is measured according to McCabe\'s line counting conventions.\n\nThe Halstead falls into three groups: the base measures, the derived measures, and lines of code measures.\n\n* Base measures:\n  * mu1             = number of unique operators\n  * mu2             = number of unique operands\n  * N1              = total occurrences of operators\n  * N2              = total occurrences of operands\n  * length     = N  = N1 + N2\n  * vocabulary = mu = mu1 + mu2\n  * Constants set for each function:\n  * mu1\' =  2 = potential operator count (just the function name and the \"return\" operator)\n  * mu2\'      = potential operand count. (the number of arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\" operators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y), \"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique operands (w,x,y).\n\n* Derived measures:\n  * P = volume = V = N * log2(mu) (the number of mental comparisons needed to write\na program of length N)\n  * V* = volume on minimal implementation = (2 + mu2\')*log2(2 + mu2\')\n  * L  = program length = V*/N\n  * D  = difficulty = 1/L\n  * L\' = 1/D\n  * I  = intelligence = L\'*V\'\n  * E  = effort to write program = V/L\n  * T  = time to write program = E/18 seconds\n\n## Number of instances\n10885\n\n## Number of attributes\n22 (5 different lines of code measure, 3 McCabe metrics, 4 base Halstead measures, 8 derived Halstead measures, a branch-count, and 1 goal field)\n\n## Attribute Information\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. defects         : {false,true} % module has/has not one or more reported defects\n\n## Missing attributes\nNone\n\n## Class Distribution\nThe class value (defects) is discrete\nfalse: 2106 = 19.35%\ntrue:  8779 = 80.65%', 2),
-(52, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: KC2/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From KC2, Naive Bayes generated PDs of 50% with PF of 10%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n-- Data from C++ functions. Science data processing; another part\nof the same project as KC1; different personnel than KC1.  Shared\nsome third-party software libraries with KC1, but no other software\noverlap.\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 522\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. problems        : {no,yes}% module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (problems) is discrete\nyes: 105 = 20.5%\nno:  415 = 79.5%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 2),
-(53, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: KC1/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From KC2, Naive Bayes generated PDs of 45% with PF of 10%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n\n-- KC1 is a \"C++\" system implementing storage management for\nreceiving and processing ground data\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 2109\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. problems        : {false,true}% module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (problems) is discrete\nyes:  326 = 15.45%\nno:  1783 = 84.54%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 2),
-(54, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\n%-*- text -*-\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\nThis is a PROMISE Software Engineering Repository data set made publicly\navailable in order to encourage repeatable, verifiable, refutable, and/or\nimprovable predictive models of software engineering.\n\nIf you publish material based on PROMISE data sets then, please\nfollow the acknowledgment guidelines posted on the PROMISE repository\nweb page http://promise.site.uottawa.ca/SERepository .\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n1. Title/Topic: PC1/software defect prediction\n2. Sources:\n\n-- Creators:  NASA, then the NASA Metrics Data Program,\n-- http://mdp.ivv.nasa.gov. Contacts: Mike Chapman,\nGalaxy Global Corporation (Robert.Chapman@ivv.nasa.gov)\n+1-304-367-8341; Pat Callis, NASA, NASA project manager\nfor MDP (Patrick.E.Callis@ivv.nasa.gov) +1-304-367-8309\n\n-- Donor: Tim Menzies (tim@barmag.net)\n\n-- Date:  December 2 2004\n3. Past usage:\n\n1. How Good is Your Blind  Spot Sampling Policy?; 2003; Tim Menzies\nand Justin S. Di Stefano; 2004 IEEE Conference on High Assurance\nSoftware Engineering (http://menzies.us/pdf/03blind.pdf).\n\n-- Results:\n\n-- Very simple learners (ROCKY) perform as well in this domain\nas more sophisticated methods (e.g. J48, model trees, model\ntrees) for predicting detects\n\n-- Many learners have very low false alarm rates.\n\n-- Probability of detection (PD) rises with effort and rarely\nrises above it.\n\n-- High PDs are associated with high PFs (probability of\nfailure)\n\n-- PD, PF, effort can change significantly while accuracy\nremains essentially stable\n\n-- With two notable exceptions, detectors learned from one\ndata set (e.g. KC2) have nearly they same properties when\napplied to another (e.g. PC2, KC2). Exceptions:\n-- LinesOfCode measures generate wider inter-data-set variances;\n-- Precision\'s inter-data-set variances vary wildly\n\n2. \"Assessing Predictors of Software Defects\", T. Menzies and\nJ. DiStefano and A. Orrego and R. Chapman, 2004,\nProceedings, workshop on Predictive Software Models, Chicago,\nAvailable from http://menzies.us/pdf/04psm.pdf.\n-- Results:\n\n-- From JM1, Naive Bayes generated PDs of 20% with PF of 25%\n\n-- Naive Bayes out-performs J48 for defect detection\n\n-- When learning on more and more data, little improvement is\nseen after processing 300 examples.\n\n-- PDs are much higher from data collected below the sub-sub-\nsystem level.\n\n-- Accuracy is a surprisingly uninformative measure of success\nfor a defect detector. Two detectors with the same accuracy\ncan have widely varying PDs and PFs.\n4. Relevant information:\n\n-- Data from C  functions. flight software\nfor earth orbiting satellite.\n\n-- Data comes from McCabe and Halstead features extractors of\nsource code.  These features were defined in the 70s in an attempt\nto objectively characterize code features that are associated with\nsoftware quality.  The nature of association is under dispute.\nNotes on McCabe and Halstead follow.\n\n-- The McCabe and Halstead measures are \"module\"-based where a\n\"module\" is the smallest unit of functionality. In C or Smalltalk,\n\"modules\" would be called \"function\" or \"method\" respectively.\n\n-- Defect detectors can be assessed according to the following measures:\n\nmodule actually has defects\n+-------------+------------+\n|     no      |     yes    |\n+-----+-------------+------------+\nclassifier predicts no defects |  no |      a      |      b     |\n+-----+-------------+------------+\nclassifier predicts some defects | yes |      c      |      d     |\n+-----+-------------+------------+\n\naccuracy                   = acc          = (a+d)/(a+b+c+d\nprobability of detection   = pd  = recall = d/(b+d)\nprobability of false alarm = pf           = c/(a+c)\nprecision                  = prec         = d/(c+d)\neffort                     = amount of code selected by detector\n= (c.LOC + d.LOC)/(Total LOC)\n\nIdeally, detectors have high PDs, low PFs, and low\neffort. This ideal state rarely happens:\n\n-- PD and effort are linked. The more modules that trigger\nthe detector, the higher the PD. However, effort also gets\nincreases\n\n-- High PD or low PF comes at the cost of high PF or low PD\n(respectively). This linkage can be seen in a standard\nreceiver operator curve (ROC).  Suppose, for example, LOC> x\nis used as the detector (i.e. we assume large modules have\nmore errors). LOC > x represents a family of detectors. At\nx=0, EVERY module is predicted to have errors. This detector\nhas a high PD but also a high false alarm rate. At x=0, NO\nmodule is predicted to have errors. This detector has a low\nfalse alarm rate but won\'t detect anything at all. At 0<x<1,\na set of detectors are generated as shown below:\n\npd\n1 |           x  x  x                KEY:\n|        x     .                   \".\"  denotes the line PD=PF\n|     x      .                     \"x\"  denotes the roc curve\n|   x      .                            for a set of detectors\n|  x     .\n| x    .\n| x  .\n|x .\n|x\nx------------------ pf\n0                   1\n\nNote that:\n\n-- The only way to make no mistakes (PF=0) is to do nothing\n(PD=0)\n\n-- The only way to catch more detects is to make more\nmistakes (increasing PD means increasing PF).\n\n-- Our detector bends towards the \"sweet spot\" of\n<PD=1,PF=0> but does not reach it.\n\n-- The line pf=pd on the above graph represents the \"no information\"\nline. If pf=pd then the detector is pretty useless. The better\nthe detector, the more it rises above PF=PD towards the \"sweet spot\".\n\nNOTES ON MCCABE/HALSTEAD\n========================\nMcCabe argued that code with complicated pathways are more\nerror-prone.  His metrics therefore reflect the pathways within a\ncode module.\n@Article{mccabe76,\ntitle 	= \"A Complexity Measure\",\nauthor 	= \"T.J. McCabe\",\npages 	= \"308--320\",\njournal = \"IEEE Transactions on Software Engineering\",\nyear 	= \"1976\",\nvolume 	= \"2\",\nmonth 	= \"December\",\nnumber 	= \"4\"}\n\nHalstead argued that code that is hard to read is more likely to be\nfault prone. Halstead estimates reading complexity by counting the\nnumber of concepts in a module; e.g. number of unique operators.\n@Book{halstead77,\nAuthor 	  = \"M.H. Halstead\",\nTitle 	  = \"Elements of Software Science\",\nPublisher = \"Elsevier \",\nYear 	  = 1977}\n\nWe study these static code measures since they are useful, easy to\nuse, and widely used:\n\n-- USEFUL: E.g. this data set can generate highly accurate\npredictors for defects\n\n-- EASY TO USE: Static code measures (e.g. lines of code, the\nMcCabe/Halstead measures) can be automatically and cheaply\ncollected.\n\n-- WIDELY USED: Many researchers use static measures to guide\nsoftware quality predictions (see the reference list in the above\n\"blind spot\" paper. Verification and validation (V\\&V) textbooks\nadvise using static code complexity measures to decide which\nmodules are worthy of manual inspections.  Further, we know of\nseveral large government software contractors that won\'t review\nsoftware modules _unless_ tools like McCabe predict that they are\nfault prone.  Hence, defect detectors have a major economic impact\nwhen they may force programmers to rewrite code.\n\nNevertheless, the merits of these metrics has been widely\ncriticized.  Static code measures are hardly a complete\ncharacterization of the internals of a function. Fenton offers an\ninsightful example where the same functionality is achieved using\ndifferent programming language constructs resulting in different\nstatic measurements for that module. Fenton uses this example to\nargue the uselessness of static code measures.\n@book{fenton97,\nauthor    = \"N.E. Fenton and S.L. Pfleeger\",\ntitle     = {Software metrics: a Rigorous \\& Practical Approach},\npublisher = {International Thompson Press},\nyear      = {1997}}\n\nAn alternative interpretation of Fenton\'s example is that static\nmeasures can never be a definite and certain indicator of the\npresence of a fault.  Rather, defect detectors based on static\nmeasures are best viewed as probabilistic statements that the\nfrequency of faults tends to increase in code modules that trigger\nthe detector.  By definition, such probabilistic statements will\nare not categorical claims with some a non-zero false alarm\nrate. The research challenge for data miners is to ensure that\nthese false alarms do not cripple their learned theories.\n\nThe McCabe metrics are a collection of four software metrics:\nessential complexity, cyclomatic complexity, design complexity and\nLOC, Lines of Code.\n\n-- Cyclomatic Complexity, or \"v(G)\", measures the number of\n\"linearly independent paths\". A set of paths is said to be\nlinearly independent if no path in the set is a linear combination\nof any other paths in the set through a program\'s \"flowgraph\". A\nflowgraph is a directed graph where each node corresponds to a\nprogram statement, and each arc indicates the flow of control from\none statement to another. \"v(G)\" is calculated by \"v(G) = e - n + 2\"\nwhere \"G\" is a program\'s flowgraph, \"e\" is the number of arcs in\nthe flowgraph, and \"n\" is the number of nodes in the\nflowgraph. The standard McCabes rules (\"v(G)\">10), are used to\nidentify fault-prone module.\n\n-- Essential Complexity, or \"ev(G)$\" is the extent to which a\nflowgraph can be \"reduced\" by decomposing all the subflowgraphs\nof $G$ that are \"D-structured primes\". Such \"D-structured\nprimes\" are also sometimes referred to as \"proper one-entry\none-exit subflowgraphs\" (for a more thorough discussion of\nD-primes, see the Fenton text referenced above). \"ev(G)\" is\ncalculated using \"ev(G)= v(G) - m\" where $m$ is the number of\nsubflowgraphs of \"G\" that are D-structured primes.\n\n-- Design Complexity, or \"iv(G)\", is the cyclomatic complexity of a\nmodule\'s reduced flowgraph.  The flowgraph, \"G\", of a module is\nreduced to eliminate any complexity which does not influence the\ninterrelationship between design modules.  According to McCabe,\nthis complexity measurement reflects the modules calling patterns\nto its immediate subordinate modules.\n\n-- Lines of code is measured according to McCabe\'s line counting\nconventions.\n\nThe Halstead falls into three groups: the base measures, the\nderived measures, and lines of code measures.\n\n-- Base measures:\n-- mu1             = number of unique operators\n-- mu2             = number of unique operands\n-- N1              = total occurrences of operators\n-- N2              = total occurrences of operands\n-- length     = N  = N1 + N2\n-- vocabulary = mu = mu1 + mu2\n-- Constants set for each function:\n-- mu1\' =  2 = potential operator count (just the function\nname and the \"return\" operator)\n-- mu2\'      = potential operand count. (the number\nof arguments to the module)\n\nFor example, the expression \"return max(w+x,x+y)\" has \"N1=4\"\noperators \"return, max, +,+)\", \"N2=4\" operands (w,x,x,y),\n\"mu1=3\" unique operators (return, max,+), and \"mu2=3\" unique\noperands (w,x,y).\n\n-- Derived measures:\n-- P = volume = V = N * log2(mu) (the number of mental\ncomparisons needed to write\na program of length N)\n-- V* = volume on minimal implementation\n= (2 + mu2\')*log2(2 + mu2\')\n-- L  = program length = V*/N\n-- D  = difficulty = 1/L\n-- L\' = 1/D\n-- I  = intelligence = L\'*V\'\n-- E  = effort to write program = V/L\n-- T  = time to write program = E/18 seconds\n5. Number of instances: 1109\n6. Number of attributes: 22 (5 different lines of code measure,\n3 McCabe metrics, 4 base Halstead measures, 8 derived\nHalstead measures, a branch-count, and 1 goal field)\n7. Attribute Information:\n\n1. loc             : numeric % McCabe\'s line count of code\n2. v(g)            : numeric % McCabe \"cyclomatic complexity\"\n3. ev(g)           : numeric % McCabe \"essential complexity\"\n4. iv(g)           : numeric % McCabe \"design complexity\"\n5. n               : numeric % Halstead total operators + operands\n6. v               : numeric % Halstead \"volume\"\n7. l               : numeric % Halstead \"program length\"\n8. d               : numeric % Halstead \"difficulty\"\n9. i               : numeric % Halstead \"intelligence\"\n10. e               : numeric % Halstead \"effort\"\n11. b               : numeric % Halstead\n12. t               : numeric % Halstead\'s time estimator\n13. lOCode          : numeric % Halstead\'s line count\n14. lOComment       : numeric % Halstead\'s count of lines of comments\n15. lOBlank         : numeric % Halstead\'s count of blank lines\n16. lOCodeAndComment: numeric\n17. uniq_Op         : numeric % unique operators\n18. uniq_Opnd       : numeric % unique operands\n19. total_Op        : numeric % total operators\n20. total_Opnd      : numeric % total operands\n21: branchCount     : numeric % of the flow graph\n22. defects         : {false,true} % module has/has not one or more\n% reported defects\n8. Missing attributes: none\n9. Class Distribution: the class value (defects) is discrete\nfalse:   77 =  6.94%\ntrue:  1032 = 93.05%\n%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%', 2),
-(55, '1', '**Author**: Orange Telecom  \n**Source**: [ACM KDD Cup](http://www.sigkdd.org/kddcup/index.php) - 2009  \n**Please cite**: \n\nThe KDD Cup 2009 offers the opportunity to work on large marketing databases from the French Telecom company Orange to predict the propensity of customers to switch provider (churn). \n\nChurn (wikipedia definition): Churn rate is also sometimes called attrition rate. It is one of two primary factors that determine\nthe steady-state level of customers a business will support. In its broadest sense, churn rate is a measure of the number\nof individuals or items moving into or out of a collection over a specific period of time.\n\nThe term is used in many contexts, but is most widely applied in business with respect to a contractual customer base. For instance, it is an important factor for any business with a subscriber-based service model, including mobile telephone networks and pay TV operators. The term is also used to refer to participant turnover in peer-to-peer networks.\n\nThe training set contains 50,000 examples.\nThe first predictive 190 variables are numerical and the last 40 predictive variables are categorical.\nThe last target variable is binary {-1,1}.', 2),
-(56, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDatasets from ACM KDD Cup (http://www.sigkdd.org/kddcup/index.php)\n\nKDD Cup 2009\nhttp://www.kddcup-orange.com\n\nConverted to ARFF format by TunedIT\nCustomer Relationship Management (CRM) is a key element of modern marketing strategies. The KDD Cup 2009 offers the opportunity to work on large marketing databases from the French Telecom company Orange to predict the propensity of customers to switch provider (churn), buy new products or services (appetency), or buy upgrades or add-ons proposed to them to make the sale more profitable (up-selling).\nThe most practical way, in a CRM system, to build knowledge on customer is to produce scores. A score (the output of a model) is an evaluation for all instances of a target variable to explain (i.e. churn, appetency or up-selling). Tools which produce scores allow to project, on a given population, quantifiable information. The score is computed using input variables which describe instances. Scores are then used by the information system (IS), for example, to personalize the customer relationship. An industrial customer analysis platform able to build prediction models with a very large number of input variables has been developed by Orange Labs. This platform implements several processing methods for instances and variables selection, prediction and indexation based on an efficient model combined with variable selection regularization and model averaging method. The main characteristic of this platform is its ability to scale on very large datasets with hundreds of thousands of instances and thousands of variables. The rapid and robust detection of the variables that have most contributed to the output prediction can be a key factor in a marketing application.\nUp-selling (wikipedia definition): Up-selling is a sales technique whereby a salesman attempts to have the customer purchase more expensive\nitems, upgrades, or other add-ons in an attempt to make a more profitable sale.\nUp-selling usually involves marketing more profitable services or products, but up-selling can also be simply exposing the customer\nto other options he or she may not have considered previously.\nUp-selling can imply selling something additional, or selling something that is more profitable or otherwise preferable for the seller instead of the original sale.\nThe training set contains 50,000 examples.\nThe first predictive 190 variables are numerical and the last 40 predictive variables are categorical.\nThe last target variable is binary {-1,1}.', 2),
-(57, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDataset from the MLRR repository: http://axon.cs.byu.edu:5000/', 2),
-(58, '1', '**Author**:   \n**Source**: Unknown - Date unknown  \n**Please cite**:   \n\nDataset from the MLRR repository: http://axon.cs.byu.edu:5000/', 2),
-(59, '1', '**Author**: Nicholas Kushmerick  \n**Source**: [UCI](http://archive.ics.uci.edu/ml/datasets/Internet+Advertisements) - 1998  \n**Please cite**:   \n\nThis dataset represents a set of possible advertisements on Internet pages. The features encode the geometry of the image (if available) as well as phrases occurring in the URL, the image\'s URL and alt text, the anchor text, and words occurring near the anchor text. The task is to predict whether an image is an advertisement (\"ad\") or not (\"nonad\").\n\nRelevant Papers: N. Kushmerick (1999). \"Learning to remove Internet advertisements\", 3rd Int Conf Autonomous Agents.  \nAvailable at: http://rexa.info/paper/2fdc1cee89b7f4f2c9227d6f5d9b05d22c5ab3e9', 2),
-(60, '1', '**Author**: Tencent Inc.  \n**Source**: [KDD Cup](https://www.kddcup2012.org/) - 2012  \n**Please cite**:   \n\n**0.1% balanced subsample of the original KDD dataset**  \n\nThis data is derived from the 2012 KDD Cup. The data is subsampled to 0.1% of the original number of instances, downsampling the majority class (click=0) so that the target feature is reasonably balanced (5 to 1).\n\nThe data is about advertisements shown alongside search results in a search engine, and whether or not people clicked on these ads. \nThe task is to build the best possible model to predict whether a user will click on a given ad.\n\nA search session contains information on user id, the query issued by the user, ads displayed to the user, and target feature indicating whether a user clicked at least one of the ads in this session. The number of ads displayed to a user in a session is called ‘depth’. The order of an ad in the displayed list is called ‘position’.  An ad is displayed as a short text called ‘title’, followed by a slightly longer text called ’description’, and a URL  called ‘display URL’.   \nTo construct this dataset each session was split into multiple instances. Each instance describes an ad displayed under a certain setting  (‘depth’, ‘position’).  Instances with the same user id, ad id, query, and setting are merged. Each ad and each user have some additional properties located in separate data files that can be looked up using ids in the instances.\n\nThe dataset has the following features:  \n* Click – binary variable indicating whether a user clicked on at least one ad. \n* Impression - the number of search sessions in which AdID was impressed by UserID who issued Query.\n* Url_hash - URL is hashed for anonymity\n* AdID \n* AdvertiserID - some advertisers consistently optimize their ads, so the title and description of their ads are more attractive than those of others’ ads.\n* Depth - number of ads displayed to a user in a session\n* Position - order of an ad in the displayed list\n* QueryID - is the key of the data file \'queryid_tokensid.txt\'. (follow the link to the original KDD Cup page, track 2)\n* KeywordID - is the key of  \'purchasedkeyword_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* TitleID - is the key of \'titleid_tokensid.txt\'\n* DescriptionID - is the key of \'descriptionid_tokensid.txt\' (follow the link to the original KDD Cup page, track 2)\n* UserID – is also the key of \'userid_profile.txt\' (follow the link to the original KDD Cup page, track 2). 0 is a special value denoting that the user could be identified.', 2),
-(61, '1', '**Author**: H. Altay Guvenir, Burak Acar, Haldun Muderrisoglu    \n\n**Source**: UCI\n\n**Please cite**:   \n\nDataset artificially generated by using first order theory which describes structure of ten capital letters of English alphabet', 64),
-(62, '1', '**Author**: Paulo Cortez, Sérgio Moro   \n**Source**: UCI    \n**Please cite**: S. Moro, R. Laureano and P. Cortez. Using Data Mining for Bank Direct Marketing: An Application of the CRISP-DM Methodology. In P. Novais et al. (Eds.), Proceedings of the European Simulation and Modelling Conference - ESM\'2011, pp. 117-121, Guimarães, Portugal, October, 2011. EUROSIS.       \n\nAvailable at: [pdf] http://hdl.handle.net/1822/14838\n              [bib] http://www3.dsi.uminho.pt/pcortez/bib/2011-esm-1.txt\n \n1. Title: Bank Marketing\n \n2. Sources\n    Created by: Paulo Cortez (Univ. Minho) and Sérgio Moro (ISCTE-IUL) @ 2012\n    \n3. Past Usage:\n \n   The full dataset was described and analyzed in:\n \n S. Moro, R. Laureano and P. Cortez. Using Data Mining for Bank Direct Marketing: An Application of the CRISP-DM Methodology. In P. Novais et al. (Eds.), Proceedings of the European Simulation and Modelling Conference - ESM\'2011, pp. 117-121, Guimarães, Portugal, October, 2011. EUROSIS.\n\n4. Relevant Information:\n\nThe data is related with direct marketing campaigns of a Portuguese banking institution. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required, in order to access if the product (bank term deposit) would be (or not) subscribed. \n\nThe classification goal is to predict if the client will subscribe a term deposit (variable y).\n\n5. Attribute information:\n \nFor more information, read [Moro et al., 2011].\n\nInput variables:\n# bank client data:\n1 - age (numeric)\n2 - job : type of job (categorical: \"admin.\",\"unknown\",\"unemployed\",\"management\",\"housemaid\",\"entrepreneur\", \"student\",\"blue-collar\",\"self-employed\",\"retired\",\"technician\",\"services\") \n3 - marital : marital status (categorical: \"married\",\"divorced\",\"single\"; note: \"divorced\"  means divorced or widowed)\n4 - education (categorical: \"unknown\",\"secondary\",\"primary\",\"tertiary\")\n5 - default: has credit in default? (binary: \"yes\",\"no\")\n6 - balance: average yearly balance, in euros (numeric) \n7 - housing: has housing loan? (binary: \"yes\",\"no\")\n8 - loan: has personal loan? (binary: \"yes\",\"no\")\n# related with the last contact of the current campaign:\n9 - contact: contact communication type (categorical: \"unknown\",\"telephone\",\"cellular\") \n10 - day: last contact day of the month (numeric)\n11 - month: last contact month of year (categorical: \"jan\", \"feb\", \"mar\", ..., \"nov\", \"dec\")\n12 - duration: last contact duration, in seconds (numeric)\n# other attributes:\n13 - campaign: number of contacts performed during this campaign and for this client (numeric, includes last contact)\n14 - pdays: number of days that passed by after the client was last contacted from a previous campaign (numeric, -1 means client was not previously contacted)\n15 - previous: number of contacts performed before this campaign and for this client (numeric)\n16 - poutcome: outcome of the previous marketing campaign (categorical: \"unknown\",\"other\",\"failure\",\"success\")\n \nOutput variable (desired target):\n17 - y - has the client subscribed a term deposit? (binary: \"yes\",\"no\")\n', 64),
-(63, '1', '**Author**: Volker Lohweg  \n**Source**: UCI\n**Please cite**:   \n\nSource:\n\nOwner of database: Volker Lohweg (University of Applied Sciences, Ostwestfalen-Lippe, volker.lohweg \'@\' hs-owl.de) \n\nDonor of database: Helene Doerksen (University of Applied Sciences, Ostwestfalen-Lippe, helene.doerksen \'@\' hs-owl.de) . Date received: August, 2012 \n\nData Set Information:\n \nData were extracted from images that were taken from genuine and forged banknote-like specimens. For digitization, an industrial camera usually used for print inspection was used. The final images have 400x 400 pixels. Due to the object lens and distance to the investigated object gray-scale pictures with a resolution of about 660 dpi were gained. Wavelet Transform tool were used to extract features from images.\n\nAttribute Information:\n \n1. variance of Wavelet Transformed image (continuous) \n2. skewness of Wavelet Transformed image (continuous) \n3. curtosis of Wavelet Transformed image (continuous) \n4. entropy of image (continuous) \n5. class (integer)', 64),
-(64, '1', '**Author**: Prof. I-Cheng Yeh  \n**Source**: UCI\n**Please cite**: Yeh, I-Cheng, Yang, King-Jang, and Ting, Tao-Ming, \"Knowledge discovery on RFM model using Bernoulli sequence, \"Expert Systems with Applications, 2008 (doi:10.1016/j.eswa.2008.07.018).   \n\nTitle: Blood Transfusion Service Center Data Set\n\nAbstract: Data taken from the Blood Transfusion Service Center in Hsin-Chu City in Taiwan -- this is a classification problem.\n\n-----------------------------------------------------\nDate Donated: 2008-10-03\n-----------------------------------------------------\n  \nSource:\n  \nOriginal Owner and Donor\nProf. I-Cheng Yeh, Department of Information Management, Chung-Hua University,  Hsin Chu, Taiwan 30067, R.O.C.\ne-mail:icyeh \'at\' chu.edu.tw, TEL:886-3-5186511\nDate Donated: October 3, 2008 \n\n-----------------------------------------------------\n  \nData Set Information:\n  \nTo demonstrate the RFMTC marketing model (a modified version of RFM), this study adopted the donor database of Blood Transfusion Service Center in Hsin-Chu City in Taiwan. The center passes their blood transfusion service bus to one university in Hsin-Chu City to gather blood donated about every three months. To build a FRMTC model, we selected 748 donors at random from the donor database. These 748 donor data, each one included R (Recency - months since last donation), F (Frequency - total number of donation), M (Monetary - total blood donated in c.c.), T (Time - months since first donation), and a binary variable representing whether he/she donated blood in March 2007 (1 stand for donating blood; 0 stands for not donating blood).\n\n-----------------------------------------------------\n  \nAttribute Information:\n  \nGiven is the variable name, variable type, the measurement unit and a brief description. The \"Blood Transfusion Service Center\" is a classification problem. The order of this listing corresponds to the order of numerals along the rows of the database.\n\nR (Recency - months since last donation),\nF (Frequency - total number of donation),\nM (Monetary - total blood donated in c.c.),\nT (Time - months since first donation), and a binary variable representing whether he/she donated blood in March 2007 (1 stand for donating blood; 0 stands for not donating blood).\n \nCitation Request:\n  \nNOTE: Reuse of this database is unlimited with retention of copyright notice for Prof. I-Cheng Yeh and the following published paper:\n\nYeh, I-Cheng, Yang, King-Jang, and Ting, Tao-Ming, \"Knowledge discovery on RFM model using Bernoulli sequence, \"Expert Systems with Applications, 2008 (doi:10.1016/j.eswa.2008.07.018).', 64),
-(65, '1', '**Author**: J. P. Marques de Sá, J. Bernardes, D. Ayers de Campos.  \n**Source**: UCI  \n**Please cite**:     \n\n* Source:\n\nMarques de Sá, J.P., jpmdesa \'@\' gmail.com, Biomedical Engineering Institute, Porto, Portugal. \nBernardes, J., joaobern \'@\' med.up.pt, Faculty of Medicine, University of Porto, Portugal. \nAyres de Campos, D., sisporto \'@\' med.up.pt, Faculty of Medicine, University of Porto, Portugal.\n\n\n* Data Set Information:\n\n2126 fetal cardiotocograms (CTGs) were automatically processed and the respective diagnostic features measured. The CTGs were also classified by three expert obstetricians and a consensus classification label assigned to each of them. Classification was both with respect to a morphologic pattern (A, B, C. ...) and to a fetal state (N, S, P). Therefore the dataset can be used either for 10-class or 3-class experiments.\n\n\n* Attribute Information:\n\nLB - FHR baseline (beats per minute) \nAC - # of accelerations per second \nFM - # of fetal movements per second \nUC - # of uterine contractions per second \nDL - # of light decelerations per second \nDS - # of severe decelerations per second \nDP - # of prolongued decelerations per second \nASTV - percentage of time with abnormal short term variability \nMSTV - mean value of short term variability \nALTV - percentage of time with abnormal long term variability \nMLTV - mean value of long term variability \nWidth - width of FHR histogram \nMin - minimum of FHR histogram \nMax - Maximum of FHR histogram \nNmax - # of histogram peaks \nNzeros - # of histogram zeros \nMode - histogram mode \nMean - histogram mean \nMedian - histogram median \nVariance - histogram variance \nTendency - histogram tendency \nCLASS - FHR pattern class code (1 to 10) \nNSP - fetal state class code (N=normal; S=suspect; P=pathologic)\n\n\n* Relevant Papers:\n\nAyres de Campos et al. (2000) SisPorto 2.0 A Program for Automated Analysis of Cardiotocograms. J Matern Fetal Med 5:311-318 ', 64),
-(66, '1', '**Author**:   \n**Source**: UCI\n\n**Please cite**:   \n\nSource:\n\nD. Lucas (ddlucas .at. alum.mit.edu), Lawrence Livermore National Laboratory; R. Klein (rklein .at. astron.berkeley.edu), Lawrence Livermore National Laboratory & U.C. Berkeley; J. Tannahill (tannahill1 .at. llnl.gov), Lawrence Livermore National Laboratory; D. Ivanova (ivanova2 .at. llnl.gov), Lawrence Livermore National Laboratory; S. Brandon (brandon1 .at. llnl.gov), Lawrence Livermore National Laboratory; D. Domyancic (domyancic1 .at. llnl.gov), Lawrence Livermore National Laboratory; Y. Zhang (zhang24 .at. llnl.gov), Lawrence Livermore National Laboratory .\n\nThis data was constructed using LLNL\'s UQ Pipeline, was created under the auspices of the US Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344, was funded by LLNL\'s Uncertainty Quantification Strategic Initiative Laboratory Directed Research and Development Project under tracking code 10-SI-013, and is released under UCRL number LLNL-MISC-633994.\n\n\nData Set Information:\n\nThis dataset contains records of simulation crashes encountered during climate model uncertainty quantification (UQ) ensembles. Ensemble members were constructed using a Latin hypercube method in LLNL\'s UQ Pipeline software system to sample the uncertainties of 18 model parameters within the Parallel Ocean Program (POP2) component of the Community Climate System Model (CCSM4). Three separate Latin hypercube ensembles were conducted, each containing 180 ensemble members. 46 out of the 540 simulations failed for numerical reasons at combinations of parameter values. The goal is to use classification to predict simulation outcomes (fail or succeed) from input parameter values, and to use sensitivity analysis and feature selection to determine the causes of simulation crashes. Further details about the data and methods are given in the publication \'Failure Analysis of Parameter-Induced Simulation Crashes in Climate Models,\' Geoscientific Model Development ([Web Link]).\n\n\nAttribute Information:\n\nThe goal is to predict climate model simulation outcomes (column 21, fail or succeed) given scaled values of climate model input parameters (columns 3-20). \n\nColumn 1: Latin hypercube study ID (study 1 to study 3) \nColumn 2: simulation ID (run 1 to run 180) \nColumns 3-20: values of 18 climate model parameters scaled in the interval [0, 1] \nColumn 21: simulation outcome (0 = failure, 1 = success)\n\nRelevant Papers:\n\nLucas, D. D., Klein, R., Tannahill, J., Ivanova, D., Brandon, S., Domyancic, D., and Zhang, Y.: Failure analysis of parameter-induced simulation crashes in climate models, Geosci. Model Dev. Discuss., 6, 585-623, [Web Link], 2013.', 64),
-(67, '1', '**Author**: Patrick Marques Ciarelli, Elias Oliviera   \n**Source**: UCI\n\n**Please cite**:   \n\nSource:\n\nPatrick Marques Ciarelli, pciarelli \'@\' lcad.inf.ufes.br, Department of Electrical Engineering, Federal University of Espirito Santo \nElias Oliveira, elias \'@\' lcad.inf.ufes.br, Department of Information Science, Federal University of Espirito Santo\n\n\nData Set Information:\n\nThis is a data set containing 1080 documents of free text business descriptions of Brazilian companies categorized into a \nsubset of 9 categories cataloged in a table called National Classification of Economic Activities (Classificação Nacional de \nAtividade Econômicas - CNAE). The original texts were pre-processed to obtain the current data set: initially, it was kept only \nletters and then it was removed prepositions of the texts. Next, the words were transformed to their canonical form. Finally, \neach document was represented as a vector, where the weight of each word is its frequency in the document. This data set is \nhighly sparse (99.22% of the matrix is filled with zeros).\n\n\nAttribute Information:\n\nIn the data set there are 857 attributes, 1 attributes with the class of instance and 856 with word frequency: \n1. category: range 1 - 9 (integer) \n2 - 857. word frequency: (integer)\n\n\nRelevant Papers:\n\nPatrick Marques Ciarelli, Elias Oliveira, \'Agglomeration and Elimination of Terms for Dimensionality Reduction\', \nNinth International Conference on Intelligent Systems Design and Applications, pp.547-552, 2009 \n\nPatrick Marques Ciarelli, Elias Oliveira, Evandro O. T. Salles, \'An Evolving System Based on Probabilistic Neural Network\', \nBrazilian Symposium on Artificial Neural Network, 2010', 64),
-(68, '1', '**Author**: Oliver Roesler, it12148\'@\'lehre.dhbw-stuttgart.de  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/EEG+Eye+State), Baden-Wuerttemberg, Cooperative State University (DHBW), Stuttgart, Germany  \n**Please cite**:   \n\nAll data is from one continuous EEG measurement with the Emotiv EEG Neuroheadset. The duration of the measurement was 117 seconds. The eye state was detected via a camera during the EEG measurement and added later manually to the file after analysing the video frames. \'1\' indicates the eye-closed and \'0\' the eye-open state. All values are in chronological order with the first measured value at the top of the data.\n', 64),
-(69, '1', '**Author**: James P Bridge, Sean B Holden and Lawrence C Paulson \n  \n**Source**: UCI  \n\n**Please cite**: James P Bridge, Sean B Holden and Lawrence C Paulson . Machine learning for first-order theorem proving: learning to select a good heuristic. Journal of Automated Reasoning, Springer 2012/13. \n\nSource:\n\nJames P Bridge, Sean B Holden and Lawrence C Paulson \n\nUniversity of Cambridge \nComputer Laboratory \nWilliam Gates Building \n15 JJ Thomson Avenue \nCambridge CB3 0FD \nUK \n\n+44 (0)1223 763500 \nforename.surname \'@\' cl.cam.ac.uk\n\n\nData Set Information:\n\nSee the file dataset file.\n\n\nAttribute Information:\n\nThe attributes are a mixture of static and dynamic features derived from theorems to be proved. See the paper for full details.\n', 64),
-(70, '1', '**Author**: Alexander Vergara\n\n**Source**: UCI  \n\n**Please cite**: Alexander Vergara and Shankar Vembu and Tuba Ayhan and Margaret A. Ryan and Margie L. Homer and Ramón Huerta, Chemical gas sensor drift compensation using classifier ensembles, Sensors and Actuators B: Chemical (2012) doi: 10.1016/j.snb.2012.01.074. \n\nTitle: Gas Sensor Array Drift Dataset Data Set\n\nSource:\n\nCreators: Alexander Vergara (vergara \'@\' ucsd.edu) \nBioCircutis Institute \nUniversity of California San Diego \nSan Diego, California, USA \nDonors of the Dataset: Alexander Vergara (vergara \'@\' ucsd.edu) \nRamon Huerta (rhuerta \'@\' ucsd.edu) \n\n\nData Set Information:\n\nThis archive contains 13910 measurements from 16 chemical sensors utilized in simulations for drift compensation in a discrimination task of 6 gases at various levels of concentrations. The goal is to achieve good performance (or as low degradation as possible) over time, as reported in the paper mentioned below in Section 2: Data collection. The primary purpose of providing this dataset is to make it freely accessible on-line to the chemo-sensor research community and artificial intelligence to develop strategies to cope with sensor/concept drift. The dataset can be used exclusively for research purposes. Commercial purposes are fully excluded.\nThe dataset was gathered within January 2007 to February 2011 (36 months) in a gas delivery platform facility situated at the ChemoSignals Laboratory in the BioCircuits Institute, University of California San Diego. Being completely operated by a fully computerized environment —controlled by a LabVIEW–National Instruments software on a PC fitted with the appropriate serial data acquisition boards. The measurement system platform provides versatility for obtaining the desired concentrations of the chemical substances of interest with high accuracy and in a highly reproducible manner, minimizing thereby the common mistakes caused by human intervention and making it possible to exclusively concentrate on the chemical sensors for compensating real drift.\nThe resulting dataset comprises recordings from six distinct pure gaseous substances, namely Ammonia, Acetaldehyde, Acetone, Ethylene, Ethanol, and Toluene, each dosed at a wide variety of concentration values ranging from 5 to 1000 ppmv. See Tables 1 and 2 of the below cited manuscript for details on the gas identity name, concentration values, and time distribution sequence of the measurement recordings considered in this dataset.\n\nBatch10.dat was updated on 10/14/2013 to correct some corrupted values in the last 120 lines of the file.\n\nAn extension of this dataset with the concentration values is available at Gas Sensor Array Drift Dataset at Different Concentrations Data Set [Web Link]\n\n\nAttribute Information:\n\nThe response of the said sensors is read-out in the form of the resistance across the active layer of each sensor; hence each measurement produced a 16-channel time series, each of which represented by an aggregate of features reflecting all the dynamic processes occurring at the sensor surface in reaction to the chemical substance being evaluated. In particular, two distinct types of features were considered in the creation of this dataset: (i) The so-called steady-state feature (ΔR), defined as the difference of the maximal resistance change and the baseline and its normalized version expressed by the ratio of the maximal resistance and the baseline values when the chemical vapor is present in the test chamber. And (ii), an aggregate of features reflecting the sensor dynamics of the increasing/decaying transient portion of the sensor response during the entire measurement procedure under controlled conditions, namely the exponential moving average (emaα). These aggregate of features is a transform, borrowed from the field of econometrics originally introduced to the chemo-sensing community by Muezzinoglu et al. (2009), that converts the said transient portion into a real scalar, by estimating the maximum value —minimum for the decaying portion of the sensor response— of its exponential moving average (emaα), with an initial condition set to zero and a scalar smoothing parameter of the operator, α, that defines both the quality of the feature and the time of its occurrence along the time series the scalar, set to range between 0 and 1. In particular, three different values for α were set to obtain three different feature values from the pre-recorded rising portion of the sensor response and three additional features with the same α values but for the decaying portion of the sensor response, covering thus the entire sensor response dynamics. For a more detailed analysis and discussion on these features as well as a graphical illustration of them please refer to Section 2.3 and Figure 2, respectively of the annotated manuscript.\nOnce the abovementioned features are calculated, one is to form a feature vector containing the 8 features extracted from each particular sensor multiplied by the 16 sensors considered here. In the end, the resulting 128-dimensional feature vector containing all the features indicated above (8 features × 16 sensors) is organized as follows:\nΔR_1, |ΔR|_1, EMAi0.001_1, EMAi0.01_1, EMAi0.1_1, EMAd0.001_1, EMAd0.01_1, EMAd0.1_1, ΔR_2, |ΔR|_2, EMAi0.001_2, EMAi0.01_2, EMAi0.1_2, EMAd0.001_2, EMAd0.01_2, EMAd0.1_2,..., ΔR_16, |ΔR|_16, EMAi0.001_16, EMAi0.01_16, EMAi0.1_16, EMAd0.001_16, EMAd0.01_16, EMAd0.1_16,\n\nwhere: “ΔR_1” and “|ΔR|_1” is the ΔR and the normalized ΔR feature, respectively, “EMAi0.001_1”, “EMAi0.01_1”, and “EMAi0.1_1”, the emaα of the rising transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, and “EMAd0.001_1”, “EMAd0.01_1”, and “EMAd0.1_1”, the emaα of the decaying transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, all corresponding to sensor # 1; “ΔR_2” and “|ΔR|_2” is the ΔR and the normalized ΔR feature, respectively, “EMAi0.001_2”, “EMAi0.01_2”, and “EMAi0.1_2”, the emaα of the rising transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, and “EMAd0.001_2”, “EMAd0.01_2”, and “EMAd0.1_2”, the emaα of the decaying transient portion of the sensor response for α equals to 0.001, 0.01, and 0.1, respectively, all corresponding to sensor # 2; and so forth up until sensor # 16, forming thus the 128-dimensional feature vector that is to be fetched to the classifiers for training.\nFor processing purposes, the data is organized into ten batches, each containing the number of measurements per class and month indicated in the table below. This reorganization of data was done to ensure having a sufficient and as uniformly distributed as possible number of experiments in each class and month when training the classifier.\n\nDataset organization details. Each row corresponds to months that were combined to form a batch:\nBatch ID Month IDs\nBatch 1 Months 1 and 2\nBatch 2 Months 3, 4, 8, 9 and 10\nBatch 3 Months 11, 12, and 13\nBatch 4 Months 14 and 15\nBatch 5 Month 16\nBatch 6 Months 17, 18, 19, and 20\nBatch 7 Month 21\nBatch 8 Months 22 and 23\nBatch 9 Months 24 and 30\nBatch 10 Month 36\n\nThe data format follows the same coding style as in libsvm, in which one indicates the class each data point belongs to (1: Ethanol; 2: Ethylene; 3:Ammonia; 4: Acetaldehyde; 5: Acetone; 6: Toluene), and, then, the collection of features in a format x:v, where x stands for the feature number and v for the actual value of the feature. For example, in \n\n1 1:15596.162100 2:1.868245 3:2.371604 4:2.803678 5:7.512213 … 128:-2.654529 \n\nThe number “1” stands for the class number (in this case Ethanol), whereas the remaining 128 columns list the actual feature values for each measurement recording organized as described above. \nFinally, to make the results presented in the associated article reproducible for the reader, please use the following parameter values in the training task:\n• folds: 10\n• log2c = -5, 10, 1\n• log2g = -10, 5, 1\n• Scale the features in the training set appropriately to lie between -1 and +1.\n• And use the following cross validation parameters:\n\nBatch C Gamma (ɤ) Rate\n1 256.0 0.03125 98.8764\n2 64.0 0.00390625 99.7588\n3 128.0 0.03125 100.0\n4 1.0 0.25 100.0\n5 2.0 0.015625 99.4924\n6 256.0 0.0009765625 99.5217\n7 64.0 0.0625 99.9723\n8 1024.0 0.0078125 99.6599\n9 2.0 0.00390625 100.0', 64),
-(71, '1', '**Author**: Jorge L. Reyes-Ortiz, Davide Anguita, Alessandro Ghio, Luca Oneto and Xavier Parra\n\n**Source**: UCI\n\n**Please cite**: Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. A Public Domain Dataset for Human Activity Recognition Using Smartphones. 21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013. Bruges, Belgium 24-26 April 2013.   \n\nTitle: Human Activity Recognition Using Smartphones\n\nAbstract: Human Activity Recognition database built from the recordings of 30 subjects performing activities of daily living (ADL) while carrying a waist-mounted smartphone with embedded inertial sensors.\n\nSource:\n\nJorge L. Reyes-Ortiz(1,2), Davide Anguita(1), Alessandro Ghio(1), Luca Oneto(1) and Xavier Parra(2)\n1 - Smartlab - Non-Linear Complex Systems Laboratory DITEN - Università degli Studi di Genova, Genoa (I-16145), Italy. \n2 - CETpD - Technical Research Centre for Dependency Care and Autonomous Living Universitat Politècnica de Catalunya (BarcelonaTech). Vilanova i la Geltrú (08800), Spain, activityrecognition \'@\' smartlab.ws\n\n\nData Set Information:\n\nThe experiments have been carried out with a group of 30 volunteers within an age bracket of 19-48 years. Each person performed six activities (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING) wearing a smartphone (Samsung Galaxy S II) on the waist. Using its embedded accelerometer and gyroscope, we captured 3-axial linear acceleration and 3-axial angular velocity at a constant rate of 50Hz. The experiments have been video-recorded to label the data manually. The obtained dataset has been randomly partitioned into two sets, where 70% of the volunteers was selected for generating the training data and 30% the test data. \n\nThe sensor signals (accelerometer and gyroscope) were pre-processed by applying noise filters and then sampled in fixed-width sliding windows of 2.56 sec and 50% overlap (128 readings/window). The sensor acceleration signal, which has gravitational and body motion components, was separated using a Butterworth low-pass filter into body acceleration and gravity. The gravitational force is assumed to have only low frequency components, therefore a filter with 0.3 Hz cutoff frequency was used. From each window, a vector of features was obtained by calculating variables from the time and frequency domain.\n\nCheck the README.txt file for further details about this dataset. \n\nA video of the experiment including an example of the 6 recorded activities with one of the participants can be seen in the following link: [Web Link]\n\n\nAttribute Information:\n\nFor each record in the dataset it is provided: \n- Triaxial acceleration from the accelerometer (total acceleration) and the estimated body acceleration. \n- Triaxial Angular velocity from the gyroscope. \n- A 561-feature vector with time and frequency domain variables. \n- Its activity label. \n- An identifier of the subject who carried out the experiment.\n\n\nRelevant Papers:\n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. International Workshop of Ambient Assisted Living (IWAAL 2012). Vitoria-Gasteiz, Spain. Dec 2012 \n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra, Jorge L. Reyes-Ortiz. Energy Efficient Smartphone-Based Activity Recognition using Fixed-Point Arithmetic. Journal of Universal Computer Science. Special Issue in Ambient Assisted Living: Home Care. Volume 19, Issue 9. May 2013\n\nDavide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. 4th International Workshop of Ambient Assited Living, IWAAL 2012, Vitoria-Gasteiz, Spain, December 3-5, 2012. Proceedings. Lecture Notes in Computer Science 2012, pp 216-223. \n\nJorge Luis Reyes-Ortiz, Alessandro Ghio, Xavier Parra-Llanas, Davide Anguita, Joan Cabestany, Andreu Català. Human Activity and Motion Disorder Recognition: Towards Smarter Interactive Cognitive Environments. 21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013. Bruges, Belgium 24-26 April 2013.\n\n', 64),
-(72, '1', '**Author**: Lee Graham, Franz Oppacher\n**Source**: UCI\n\n**Please cite**:   \n\n1. Source:\n\nLee Graham (lee \'@\' stellaralchemy.com) \n\nFranz Oppacher (oppacher \'@\' scs.carleton.ca) \nCarleton University, Department of Computer Science \nIntelligent Systems Research Unit \n1125 Colonel By Drive, Ottawa, Ontario, Canada, K1S5B6 \n\n\n2. Attribute Information:\n\n1-100: Labeled “X##”. Floating point values (numeric) \n101: Labeled “class”. Binary {0, 1} representing {valley, hill} ', 64),
-(73, '1', '**Author**: Bendi Venkata Ramana, M. Surendra Prasad Babu, N. B. Venkateswarlu \n  \n**Source**: UCI  \n\n**Please cite**:   \n\nSource:\n\n1. Bendi Venkata Ramana,  \nramana.bendi \'@\' gmail.com \nAssociate Professor, \nDepartment of Information Technology, \nAditya Instutute of Technology and Management, \nTekkali - 532201, Andhra Pradesh, India. \n\n2. Prof. M. Surendra Prasad Babu \ndrmsprasadbabu \'@\' yahoo.co.in \nDeptartment of Computer Science & Systems Engineering, \nAndhra University College of Engineering, \nVisakhapatnam-530 003 Andhra Pradesh, India. \n\n3.Prof. N. B. Venkateswarlu \nvenkat_ritch \'@\' yahoo.com \nDepartment of Computer Science and Engineering, \nAditya Instutute of Technology and Management, \nTekkali - 532201, Andhra Pradesh, India.\n\n\nData Set Information:\n\nThis data set contains 416 liver patient records and 167 non liver patient records.The data set was collected from north east of Andhra Pradesh, India. Selector is a class label used to divide into groups(liver patient or not). This data set contains 441 male patient records and 142 female patient records. \n\nAny patient whose age exceeded 89 is listed as being of age \"90\".\n\nAttribute Information:\n\n1. Age Age of the patient \n2. Gender Gender of the patient \n3. TB Total Bilirubin \n4. DB Direct Bilirubin \n5. Alkphos Alkaline Phosphotase \n6. Sgpt Alamine Aminotransferase \n7. Sgot Aspartate Aminotransferase \n8. TP Total Protiens \n9. ALB Albumin \n10. A/G Ratio Albumin and Globulin Ratio \n11. Selector field used to split the data into two sets (labeled by the experts) \n\n\nRelevant Papers:\n\n1. Bendi Venkata Ramana, Prof. M. S. Prasad Babu and Prof. N. B. Venkateswarlu, “A Critical Comparative Study of Liver Patients from USA and INDIA: An Exploratory Analysis”, International Journal of Computer Science Issues, ISSN :1694-0784, May 2012. \n2. Bendi Venkata Ramana, Prof. M. S. Prasad Babu and Prof. N. B. Venkateswarlu, “A Critical Study of Selected Classification Algorithms for Liver Disease Diagnosis”, International Journal of Database Management Systems (IJDMS), Vol.3, No.2, ISSN : 0975-5705, PP 101-114, May 2011.', 64),
-(74, '1', '**Author**: Isabelle Guyon  \n**Source**: UCI  \n**Please cite**: Isabelle Guyon, Steve R. Gunn, Asa Ben-Hur, Gideon Dror, 2004. Result analysis of the NIPS 2003 feature selection challenge.\n\nAbstract: \n\nMADELON is an artificial dataset, which was part of the NIPS 2003 feature selection challenge. This is a two-class classification problem with continuous input variables. The difficulty is that the problem is multivariate and highly non-linear.\n\nSource:\n\nIsabelle Guyon \nClopinet \n955 Creston Road \nBerkeley, CA 90708 \nisabelle \'@\' clopinet.com \n\n\nData Set Information:\n\nMADELON is an artificial dataset containing data points grouped in 32 clusters placed on the vertices of a five dimensional hypercube and randomly labeled +1 or -1. The five dimensions constitute 5 informative features. 15 linear combinations of those features were added to form a set of 20 (redundant) informative features. Based on those 20 features one must separate the examples into the 2 classes (corresponding to the +-1 labels). We added a number of distractor feature called \'probes\' having no predictive power. The order of the features and patterns were randomized. \n\nThis dataset is one of five datasets used in the NIPS 2003 feature selection challenge. Our website is still open for post-challenge submissions.  All details about the preparation of the data are found in our technical report: Design of experiments for the NIPS 2003 variable selection benchmark, Isabelle Guyon, July 2003. Such information was made available only after the end of the challenge. \n\nThe data are split into training, validation, and test set. Target values are provided only for the 2 first sets. Test set performance results are obtained by submitting prediction results to: [Web Link]. \n\nThe data are in the following format: \ndataname.param: Parameters and statistics about the data \ndataname.feat: Identities of the features (in the order the features are found in the data). \ndataname_train.data: Training set (a space-delimited regular matrix, patterns in lines, features in columns). \ndataname_valid.data: Validation set. \ndataname_test.data: Test set. \ndataname_train.labels: Labels (truth values of the classes) for training examples. \ndataname_valid.labels: Validation set labels (withheld during the benchmark, but provided now). \ndataname_test.labels: Test set labels (withheld, so the data can still be use as a benchmark). \n\n\nAttribute Information:\nWe do not provide attribute information, to avoid biasing the feature selection process.\n\n\nRelevant Papers:\n\nThe best challenge entrants wrote papers collected in the book: \nIsabelle Guyon, Steve Gunn, Masoud Nikravesh, Lofti Zadeh (Eds.), Feature Extraction, Foundations and Applications. Studies in Fuzziness and Soft Computing. Physica-Verlag, Springer.\n\nSee also: \nIsabelle Guyon, et al, 2007. Competitive baseline methods set new standards for the NIPS 2003 feature selection benchmark. Pattern Recognition Letters 28 (2007) 1438–1444. \n\nand the associated technical report: \nIsabelle Guyon, et al. 2006. Feature selection with the CLOP package. Technical Report.\n', 64),
-(75, '1', '**Author**:   \n**Source**: UCI\n**Please cite**:   \n\nRelevant Papers: Laurent Candillier and Vincent Lemaire. Design and Analysis of the Nomao Challenge - Active Learning in the Real-World. In: Proceedings of the ALRA : Active Learning in Real-world Applications, Workshop ECML-PKDD 2012, Friday, September 28, 2012, Bristol, UK.\n\n1. Data set title:\nNomao Data Set \n\n\n2. Abstract: \nNomao collects data about places (name, phone, localization...) from many sources. Deduplication consists in detecting what data refer to the same place. Instances in the dataset compare 2 spots.\n\n3. Data Set Characteristics:  \n\n- Univariate\n- Area: Computer\n- Attribute Characteristics: Real\n- Associated Tasks: Classification\n- Missing Values?: Yes\n\n\n4. Source:\n\n(a) Original owner of database (name / phone / snail address / email address) \nNomao / 00 33 5 62 48 33 90 / 1 avenue Jean Rieux, 31500 Toulouse / challenge \'@\' nomao.com \n(b) Donor of database (name / phone / snail address / email address) \nLaurent Candillier / - / 1 avenue Jean Rieux, 31500 Toulouse / laurent \'@\' nomao.com\n\n\n5. Data Set Information:\n\nThe dataset has been enriched during the Nomao Challenge: organized along with the ALRA workshop (Active Learning in Real-world Applications): held at the ECML-PKDD 2012 conference.', 64),
-(76, '1', '**Author**: Kun Zhang, Wei Fan, XiaoJing Yuan\n\n**Source**: UCI\n\n**Please cite**:   \n\n\n1 . Abstract: \nTwo ground ozone level data sets are included in this collection. One is the eight hour peak set (eighthr.data), the other is the one hour peak set (onehr.data). Those data were collected from 1998 to 2004 at the Houston, Galveston and Brazoria area.\n\n2. Source:\n\nKun Zhang, zhang.kun05 \'@\' gmail.com, Department of Computer Science, Xavier University of Lousiana \nWei Fan, wei.fan \'@\' gmail.com, IBM T.J.Watson Research \nXiaoJing Yuan, xyuan \'@\' uh.edu, Engineering Technology Department, College of Technology, University of Houston \n\n\n3. Data Set Information:\n\nAll the attribute start with T means the temperature measured at different time throughout the day; and those starts with WS indicate the wind speed at various time. \n\nWSR_PK: continuous. peek wind speed -- resultant (meaning average of wind vector) \nWSR_AV: continuous. average wind speed \nT_PK: continuous. Peak T \nT_AV: continuous. Average T \nT85: continuous. T at 850 hpa level (or about 1500 m height) \nRH85: continuous. Relative Humidity at 850 hpa \nU85: continuous. (U wind - east-west direction wind at 850 hpa) \nV85: continuous. V wind - N-S direction wind at 850 \nHT85: continuous. Geopotential height at 850 hpa, it is about the same as height at low altitude \nT70: continuous. T at 700 hpa level (roughly 3100 m height) \n\nRH70: continuous. \nU70: continuous. \nV70: continuous. \nHT70: continuous. \n\nT50: continuous. T at 500 hpa level (roughly at 5500 m height) \n\nRH50: continuous. \nU50: continuous. \nV50: continuous. \nHT50: continuous. \n\nKI: continuous. K-Index [Web Link] \nTT: continuous. T-Totals [Web Link] \nSLP: continuous. Sea level pressure \nSLP_: continuous. SLP change from previous day \n\nPrecp: continuous. -- precipitation\n\n\n4. Attribute Information:\n\nThe following are specifications for several most important attributes that are highly valued by Texas Commission on Environmental Quality (TCEQ). More details can be found in the two relevant papers. \n\nO 3 - Local ozone peak prediction \nUpwind - Upwind ozone background level \nEmFactor - Precursor emissions related factor \nTmax - Maximum temperature in degrees F \nTb - Base temperature where net ozone production begins (50 F) \nSRd - Solar radiation total for the day \nWSa - Wind speed near sunrise (using 09-12 UTC forecast mode) \nWSp - Wind speed mid-day (using 15-21 UTC forecast mode) \n\n\n5. Relevant Papers:\n\nForecasting skewed biased stochastic ozone days: analyses, solutions and beyond, Knowledge and Information Systems, Vol. 14, No. 3, 2008. \n\nIt Discusses details about the dataset, its use as well as various experiments (both cross-validation and streaming) using many state-of-the-art methods. \nA shorter version of the paper (does not contain some detailed experiments as the journal paper above) is in: \nForecasting Skewed Biased Stochastic Ozone Days: Analyses and Solutions. ICDM 2006: 753-764 \n\n', 64),
-(77, '1', '**Author**:   \n**Source**: KEEL\n**Please cite**:   \n\n* Title:\n\nPhoneme dataset\n\n* Abstract:\n\nThe aim of this dataset is to distinguish between nasal (class 0) and oral sounds (class 1). The class distribution is 3,818 samples in class 0 and 1,586 samples in class 1. The phonemes are transcribed as follows: sh as in she, dcl as in dark, iy as the vowel in she, aa as the vowel in dark, and ao as the first vowel in water.\n\n* Attributes information:\n\n@relation phoneme\n@attribute Aa real [-1.7, 4.107]\n@attribute Ao real [-1.327, 4.378]\n@attribute Dcl real [-1.823, 3.199]\n@attribute Iy real [-1.581, 2.826]\n@attribute Sh real [-1.284, 2.719]\n@attribute Class {0, 1}\n@inputs Aa, Ao, Dcl, Iy, Sh\n@outputs Class', 64),
-(78, '1', '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \n\n**Source**: UCI \n\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \n\n1. One-hundred plant species leaves data set (class = margin).\n \n2. Sources:\n   (a) Original owners of colour Leaves Samples:\n\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\n The colour images are not included in this submission.\n The Leaves were collected in the Royal Botanic Gardens, Kew, UK.\n email: james.cope@kingston.ac.uk\n   \n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\n\n   (c) Date received 03/12/2012\n\n3. Past Usage:\n\n   (a) This is a new data set, provisional paper: \n\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\n\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \n\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\n Plant texture classification using gabor cooccurrences.\n Advances in Visual Computing,\n pages 669ñ677, 2010.\n\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\n Shape and texture based plant leaf classification. In\n Advanced Concepts for Intelligent Vision Systems,\n pages 345ñ353. Springer, 2010.\n\n4. Relevant Information Paragraph:\n The data directory contains the binary images (masks) of the leaf samples. The colour images are not included.\n The data set features are organised as the following:\n * \'data_Sha_64.txt\'\n  * \'data_Tex_64.txt\'\n * [THIS DATASET]\'data_Mar_64.txt\'\n\n One file for each 64-element feature vectors. Each row begins with the class label.\n The remaining 64 elements is the feature vector.\n\n5. Number of Instances\n\n 1600 samples each of three features (16 samples per leaf class).\n\n6. Number of Attributes\n\n    Three 64 element feature vectors per sample.\n\n7. Vectors\n There are three features: Shape, Margin and Texture. As discussed in the paper(s) above.\n For Each feature, a 64 element vector is given per sample of leaf.\n These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin).\n\n8. Missing Attribute Values: none\n\n9. Class Distribution: 16 instances per class', 64),
-(79, '1', '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \n\n**Source**: UCI \n\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \n\n1. One-hundred plant species leaves data set (class = shape).\n \n2. Sources:\n   (a) Original owners of colour Leaves Samples:\n\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\n The colour images are not included in this submission.\n The Leaves were collected in the Royal Botanic Gardens, Kew, UK.\n email: james.cope@kingston.ac.uk\n   \n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\n\n   (c) Date received 03/12/2012\n\n3. Past Usage:\n\n   (a) This is a new data set, provisional paper: \n\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\n\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \n\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\n Plant texture classification using gabor cooccurrences.\n Advances in Visual Computing,\n pages 669ñ677, 2010.\n\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\n Shape and texture based plant leaf classification. In\n Advanced Concepts for Intelligent Vision Systems,\n pages 345ñ353. Springer, 2010.\n\n4. Relevant Information Paragraph:\n The data directory contains the binary images (masks) of the leaf samples. The colour images are not included.\n The data set features are organised as the following:\n * [THIS DATASET]\'data_Sha_64.txt\'\n  * \'data_Tex_64.txt\'\n * \'data_Mar_64.txt\'\n\n One file for each 64-element feature vectors. Each row begins with the class label.\n The remaining 64 elements is the feature vector.\n\n5. Number of Instances\n\n 1600 samples each of three features (16 samples per leaf class).\n\n6. Number of Attributes\n\n    Three 64 element feature vectors per sample.\n\n7. Vectors\n There are three features: Shape, Margin and Texture. As discussed in the paper(s) above.\n For Each feature, a 64 element vector is given per sample of leaf.\n These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin).\n\n8. Missing Attribute Values: none\n\n9. Class Distribution: 16 instances per class', 64),
-(80, '1', '**Author**: James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/One-hundred+plant+species+leaves+data+set) - 2012  \r\n**Please cite**: Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press. 2013.   \r\n\r\n**One-hundred plant species leaves data set (class = texture)**  \r\nThe data directory contains the binary images (masks) of the leaf samples. The colour images are not included. There are three features: Shape, Margin and Texture. As discussed in the paper(s) above. For Each feature, a 64 element vector is given per sample of leaf. These vectors are taken as a contigous descriptors (for shape) or histograms (for texture and margin). The colour images are not included in this submission. The Leaves were collected in the Royal Botanic Gardens, Kew, UK. email: james.cope@kingston.ac.uk\r\n\r\n**Sources**  \r\n   (a) Original owners of colour Leaves Samples:\r\n James Cope, Thibaut Beghin, Paolo Remagnino, Sarah Barman.\r\n   (b) This dataset consists of work carried out by James Cope, Charles Mallah, and James Orwell.\r\n Donor of database Charles Mallah: charles.mallah@kingston.ac.uk; James Cope:  james.cope@kingston.ac.uk\r\n   (c) Date received 03/12/2012\r\n\r\n**Usage**\r\n   (a) This is a new data set, provisional paper: \r\n\r\n Charles Mallah, James Cope, James Orwell. Plant Leaf Classification Using Probabilistic Integration of Shape, Texture and Margin Features. Signal Processing, Pattern Recognition and Applications, in press.\r\n\r\n   (b) Previous parts of the data set relate to feature extraction of leaves from: \r\n\r\n J. Cope, P. Remagnino, S. Barman, and P. Wilkin.\r\n Plant texture classification using gabor cooccurrences.\r\n Advances in Visual Computing,\r\n pages 669ñ677, 2010.\r\n\r\n T. Beghin, J. Cope, P. Remagnino, and S. Barman.\r\n Shape and texture based plant leaf classification. In\r\n Advanced Concepts for Intelligent Vision Systems,\r\n pages 345ñ353. Springer, 2010.', 64),
-(81, '1', '**Author**:   \n**Source**: UCI  \n**Please cite**: Mansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878 \n\n\nQSAR biodegradation Data Set \n\n* Abstract: \n\nData set containing values for 41 attributes (molecular descriptors) used to classify 1055 chemicals into 2 classes (ready and not ready biodegradable).\n\n\n* Source:\n\nKamel Mansouri, Tine Ringsted, Davide Ballabio (davide.ballabio \'@\' unimib.it), Roberto Todeschini, Viviana Consonni, Milano Chemometrics and QSAR Research Group (http://michem.disat.unimib.it/chm/), Università degli Studi Milano – Bicocca, Milano (Italy)\n\n\n* Data Set Information:\n\nThe QSAR biodegradation dataset was built in the Milano Chemometrics and QSAR Research Group (Università degli Studi Milano – Bicocca, Milano, Italy). The research leading to these results has received funding from the European Community’s Seventh Framework Programme [FP7/2007-2013] under Grant Agreement n. 238701 of Marie Curie ITN Environmental Chemoinformatics (ECO) project. \nThe data have been used to develop QSAR (Quantitative Structure Activity Relationships) models for the study of the relationships between chemical structure and biodegradation of molecules. Biodegradation experimental values of 1055 chemicals were collected from the webpage of the National Institute of Technology and Evaluation of Japan (NITE). Classification models were developed in order to discriminate ready (356) and not ready (699) biodegradable molecules by means of three different modelling methods: k Nearest Neighbours, Partial Least Squares Discriminant Analysis and Support Vector Machines. Details on attributes (molecular descriptors) selected in each model can be found in the quoted reference: Mansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878.\n\n\n* Attribute Information:\n\n41 molecular descriptors and 1 experimental class: \n1) SpMax_L: Leading eigenvalue from Laplace matrix \n2) J_Dz(e): Balaban-like index from Barysz matrix weighted by Sanderson electronegativity \n3) nHM: Number of heavy atoms \n4) F01[N-N]: Frequency of N-N at topological distance 1 \n5) F04[C-N]: Frequency of C-N at topological distance 4 \n6) NssssC: Number of atoms of type ssssC \n7) nCb-: Number of substituted benzene C(sp2) \n8) C%: Percentage of C atoms \n9) nCp: Number of terminal primary C(sp3) \n10) nO: Number of oxygen atoms \n11) F03[C-N]: Frequency of C-N at topological distance 3 \n12) SdssC: Sum of dssC E-states \n13) HyWi_B(m): Hyper-Wiener-like index (log function) from Burden matrix weighted by mass \n14) LOC: Lopping centric index \n15) SM6_L: Spectral moment of order 6 from Laplace matrix \n16) F03[C-O]: Frequency of C - O at topological distance 3 \n17) Me: Mean atomic Sanderson electronegativity (scaled on Carbon atom) \n18) Mi: Mean first ionization potential (scaled on Carbon atom) \n19) nN-N: Number of N hydrazines \n20) nArNO2: Number of nitro groups (aromatic) \n21) nCRX3: Number of CRX3 \n22) SpPosA_B(p): Normalized spectral positive sum from Burden matrix weighted by polarizability \n23) nCIR: Number of circuits \n24) B01[C-Br]: Presence/absence of C - Br at topological distance 1 \n25) B03[C-Cl]: Presence/absence of C - Cl at topological distance 3 \n26) N-073: Ar2NH / Ar3N / Ar2N-Al / R..N..R \n27) SpMax_A: Leading eigenvalue from adjacency matrix (Lovasz-Pelikan index) \n28) Psi_i_1d: Intrinsic state pseudoconnectivity index - type 1d \n29) B04[C-Br]: Presence/absence of C - Br at topological distance 4 \n30) SdO: Sum of dO E-states \n31) TI2_L: Second Mohar index from Laplace matrix \n32) nCrt: Number of ring tertiary C(sp3) \n33) C-026: R--CX--R \n34) F02[C-N]: Frequency of C - N at topological distance 2 \n35) nHDon: Number of donor atoms for H-bonds (N and O) \n36) SpMax_B(m): Leading eigenvalue from Burden matrix weighted by mass \n37) Psi_i_A: Intrinsic state pseudoconnectivity index - type S average \n38) nN: Number of Nitrogen atoms \n39) SM6_B(m): Spectral moment of order 6 from Burden matrix weighted by mass \n40) nArCOOR: Number of esters (aromatic) \n41) nX: Number of halogen atoms \n42) experimental class: ready biodegradable (RB) and not ready biodegradable (NRB)\n\n\n* Relevant Papers:\n\nMansouri, K., Ringsted, T., Ballabio, D., Todeschini, R., Consonni, V. (2013). Quantitative Structure - Activity Relationship models for ready biodegradability of chemicals. Journal of Chemical Information and Modeling, 53, 867-878\n', 64),
-(82, '1', '**Author**: Ananda Freire, Marcus Veloso and Guilherme Barreto     \n**Source**: UCI   \n**Please cite**:   \n\n* Dataset Title: \n\nWall-Following Robot Navigation Data Data Set \n\n* Abstract:  \n\nThe data were collected as the SCITOS G5 robot navigates through the room following the wall in a clockwise direction, for 4 rounds, using 24 ultrasound sensors arranged circularly around its \'waist\'.\n\n* Source:\n\n(a) Creators: Ananda Freire, Marcus Veloso and Guilherme Barreto \nDepartment of Teleinformatics Engineering \nFederal University of Ceará \nFortaleza, Ceará, Brazil \n\n(b) Donors of database: Ananda Freire (anandalf \'@\' gmail.com) \nGuilherme Barreto (guilherme \'@\' deti.ufc.br)\n\n* Data Set Information:\n\nThe provided file contain the raw values of the measurements of all 24 ultrasound sensors and the corresponding class label. Sensor readings are sampled at a rate of 9 samples per second. \n\nIt is worth mentioning that the 24 ultrasound readings and the simplified distances were collected at the same time step, so each file has the same number of rows (one for each sampling time step). \n\nThe wall-following task and data gathering were designed to test the hypothesis that this apparently simple navigation task is indeed a non-linearly separable classification task. Thus, linear classifiers, such as the Perceptron network, are not able to learn the task and command the robot around the room without collisions. Nonlinear neural classifiers, such as the MLP network, are able to learn the task and command the robot successfully without collisions. \n\nIf some kind of short-term memory mechanism is provided to the neural classifiers, their performances are improved in general. For example, if past inputs are provided together with current sensor readings, even the Perceptron becomes able to learn the task and command the robot successfully. If a recurrent neural network, such as the Elman network, is used to learn the task, the resulting dynamical classifier is able to learn the task using less hidden neurons than the MLP network. \n\n* Attribute Information:\n\nNumber of Attributes: sensor_readings_24.data: 24 numeric attributes and the class. \n\nFor Each Attribute: \n-- File sensor_readings_24.data: \n1. US1: ultrasound sensor at the front of the robot (reference angle: 180°) - (numeric: real)  \n2. US2: ultrasound reading (reference angle: -165°) - (numeric: real)   \n3. US3: ultrasound reading (reference angle: -150°) - (numeric: real)   \n4. US4: ultrasound reading (reference angle: -135°) - (numeric: real)     \n5. US5: ultrasound reading (reference angle: -120°) - (numeric: real)   \n6. US6: ultrasound reading (reference angle: -105°) - (numeric: real)   \n7. US7: ultrasound reading (reference angle: -90°) - (numeric: real)   \n8. US8: ultrasound reading (reference angle: -75°) - (numeric: real)   \n9. US9: ultrasound reading (reference angle: -60°) - (numeric: real)   \n10. US10: ultrasound reading (reference angle: -45°) - (numeric: real)   \n11. US11: ultrasound reading (reference angle: -30°) - (numeric: real)   \n12. US12: ultrasound reading (reference angle: -15°) - (numeric: real)   \n13. US13: reading of ultrasound sensor situated at the back of the robot (reference angle: 0°) - (numeric: real)   \n14. US14: ultrasound reading (reference angle: 15°) - (numeric: real)   \n15. US15: ultrasound reading (reference angle: 30°) - (numeric: real)   \n16. US16: ultrasound reading (reference angle: 45°) - (numeric: real)   \n17. US17: ultrasound reading (reference angle: 60°) - (numeric: real)   \n18. US18: ultrasound reading (reference angle: 75°) - (numeric: real)   \n19. US19: ultrasound reading (reference angle: 90°) - (numeric: real)   \n20. US20: ultrasound reading (reference angle: 105°) - (numeric: real)   \n21. US21: ultrasound reading (reference angle: 120°) - (numeric: real)   \n22. US22: ultrasound reading (reference angle: 135°) - (numeric: real)   \n23. US23: ultrasound reading (reference angle: 150°) - (numeric: real)   \n24. US24: ultrasound reading (reference angle: 165°) - (numeric: real)   \n25. Class: {Move-Forward, Slight-Right-Turn, Sharp-Right-Turn, Slight-Left-Turn}   \n\n\n* Relevant Papers:\n\nAnanda L. Freire, Guilherme A. Barreto, Marcus Veloso and Antonio T. Varela (2009), \n\'Short-Term Memory Mechanisms in Neural Network Learning of Robot Navigation \nTasks: A Case Study\'. Proceedings of the 6th Latin American Robotics Symposium (LARS\'2009), \nValparaíso-Chile, pages 1-6, DOI: 10.1109/LARS.2009.5418323\n', 64),
-(83, '1', '**Author**: Semeion Research Center of Sciences of Communication     \n**Source**: UCI     \n**Please cite**: Semeion Research Center of Sciences of Communication, via Sersale 117, 00128 Rome, Italy \nTattile Via Gaetano Donizetti, 1-3-5,25030 Mairano (Brescia), Italy.    \n\n* Title:   \nSemeion Handwritten Digit Data Set \n\n* Abstract:   \n1593 handwritten digits from around 80 persons were scanned, stretched in a rectangular box 16x16 in a gray scale of 256 values.\n\n\n* Source:\n\nThe dataset was created by Tactile Srl, Brescia, Italy (http://www.tattile.it) and donated in 1994 to Semeion Research Center of Sciences of Communication, Rome, Italy (http://www.semeion.it), for machine learning research. \n\nFor any questions, e-mail Massimo Buscema (m.buscema \'@\' semeion.it) or Stefano Terzi (s.terzi \'@\' semeion.it)\n\n\n* Data Set Information:\n\n1593 handwritten digits from around 80 persons were scanned, stretched in a rectangular box 16x16 in a gray scale of 256 values.Then each pixel of each image was scaled into a bolean (1/0) value using a fixed threshold. \n\nEach person wrote on a paper all the digits from 0 to 9, twice. The commitment was to write the digit the first time in the normal way (trying to write each digit accurately) and the second time in a fast way (with no accuracy). \n\nThe best validation protocol for this dataset seems to be a 5x2CV, 50% Tune (Train +Test) and completly blind 50% Validation\n\n\n* Attribute Information:\n\nThis dataset consists of 1593 records (rows) and 256 attributes (columns). Each record represents a handwritten digit, orginally scanned with a resolution of 256 grays scale (28). Each pixel of the each original scanned image was first stretched, and after scaled between 0 and 1 (setting to 0 every pixel whose value was under tha value 127 of the grey scale (127 included) and setting to 1 each pixel whose orinal value in the grey scale was over 127). Finally, each binary image was scaled again into a 16x16 square box (the final 256 binary attributes).\n\n\n* Relevant Papers:\n\nM Buscema, MetaNet: The Theory of Independent Judges, in Substance Use & Misuse 33(2)1998, pp 439-461.', 64),
-(84, '1', '**Author**: Semeion, Research Center of Sciences of Communication, Rome, Italy.     \n**Source**: UCI     \n**Please cite**: Dataset provided by Semeion, Research Center of Sciences of Communication, Via Sersale 117, 00128, Rome, Italy. \n(www.semeion.it)   \n\n\n* Title: Steel Plates Faults Data Set \n\n* Abstract: \n\nA dataset of steel plates\' faults, classified into 7 different types. The goal was to train machine learning for automatic pattern recognition.\n\n* Source:\n\nSemeion, Research Center of Sciences of Communication, Via Sersale 117, 00128, Rome, Italy. \nwww.semeion.it\n\n\n* Data Set Information:\n\nType of dependent variables (7 Types of Steel Plates Faults): \n1.Pastry \n2.Z_Scratch \n3.K_Scatch \n4.Stains \n5.Dirtiness \n6.Bumps \n7.Other_Faults \n\n\nAttribute Information:\n\n27 independent variables: \nX_Minimum \nX_Maximum \nY_Minimum \nY_Maximum \nPixels_Areas \nX_Perimeter \nY_Perimeter \nSum_of_Luminosity \nMinimum_of_Luminosity \nMaximum_of_Luminosity \nLength_of_Conveyer \nTypeOfSteel_A300 \nTypeOfSteel_A400 \nSteel_Plate_Thickness \nEdges_Index \nEmpty_Index \nSquare_Index \nOutside_X_Index \nEdges_X_Index \nEdges_Y_Index \nOutside_Global_Index \nLogOfAreas \nLog_X_Index \nLog_Y_Index \nOrientation_Index \nLuminosity_Index \nSigmoidOfAreas \n\n\n* Relevant Papers:\n\n1.M Buscema, S Terzi, W Tastle, A New Meta-Classifier,in NAFIPS 2010, Toronto (CANADA),26-28 July 2010, 978-1-4244-7858-6/10 ©2010 IEEE \n2.M Buscema, MetaNet: The Theory of Independent Judges, in Substance Use & Misuse, 33(2), 439-461,1998\n\n', 64),
-(85, '1', '**Author**: K.Kalyani.    \n**Source**: UCI    \n**Please cite**:   \n\n* Title: Tamilnadu Electricity Board Hourly Readings Data Set \n\n* Abstract: \n\nThis data can be effectively produced the result to fewer parameter of the Load profile can be reduced in the Database\n\n* Source:\n\nK.Kalyani ,kkalyanims \'@\' gmail.com,T.U.K Arts College,Karanthai,Thanjavur.\n\n\n* Data Set Information:\n\nCollect the real time readings for residential,commercial,industrial,agriculure,to find the accuracy consumption in Tamil Nadu Around Thanajvur.\n\n\n* Attribute Information:\n\nforkva,forkw,type,sector,service\n\n\n* Relevant Papers:\n\nEfficient Electricity Utilization By IHBMO\n', 64),
-(86, '1', '**Author**: William H. Wolberg, W. Nick Street, Olvi L. Mangasarian    \n**Source**: UCI   \n**Please cite**:   \n\n* Title: \n\nBreast Cancer Wisconsin (Diagnostic) Data Set (WDBC)\n\n* Abstract: \n\nDiagnostic Wisconsin Breast Cancer Database\n\n* Source:\n\nCreators: \n\n1. Dr. William H. Wolberg, General Surgery Dept. \nUniversity of Wisconsin, Clinical Sciences Center \nMadison, WI 53792 \nwolberg \'@\' eagle.surgery.wisc.edu \n\n2. W. Nick Street, Computer Sciences Dept. \nUniversity of Wisconsin, 1210 West Dayton St., Madison, WI 53706 \nstreet \'@\' cs.wisc.edu 608-262-6619 \n\n3. Olvi L. Mangasarian, Computer Sciences Dept. \nUniversity of Wisconsin, 1210 West Dayton St., Madison, WI 53706 \nolvi \'@\' cs.wisc.edu \n\nDonor: Nick Street\n\n* Data Set Information:\n\nFeatures are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. They describe characteristics of the cell nuclei present in the image. A few of the images can be found at [Web Link] \n\nSeparating plane described above was obtained using Multisurface Method-Tree (MSM-T) [K. P. Bennett, \"Decision Tree Construction Via Linear Programming.\" Proceedings of the 4th Midwest Artificial Intelligence and Cognitive Science Society, pp. 97-101, 1992], a classification method which uses linear programming to construct a decision tree. Relevant features were selected using an exhaustive search in the space of 1-4 features and 1-3 separating planes. \n\nThe actual linear program used to obtain the separating plane in the 3-dimensional space is that described in: [K. P. Bennett and O. L. Mangasarian: \"Robust Linear Programming Discrimination of Two Linearly Inseparable Sets\", Optimization Methods and Software 1, 1992, 23-34]. \n\nThis database is also available through the UW CS ftp server: \nftp ftp.cs.wisc.edu \ncd math-prog/cpo-dataset/machine-learn/WDBC/\n\n\n* Attribute Information:\n\n1) ID number \n2) Diagnosis (M = malignant, B = benign) \n3-32) \n\nTen real-valued features are computed for each cell nucleus: \n\na) radius (mean of distances from center to points on the perimeter) \nb) texture (standard deviation of gray-scale values) \nc) perimeter \nd) area \ne) smoothness (local variation in radius lengths) \nf) compactness (perimeter^2 / area - 1.0) \ng) concavity (severity of concave portions of the contour) \nh) concave points (number of concave portions of the contour) \ni) symmetry \nj) fractal dimension (\"coastline approximation\" - 1)\n\n\n* Relevant Papers:\n\nW.N. Street, W.H. Wolberg and O.L. Mangasarian. Nuclear feature extraction for breast tumor diagnosis. IS&T/SPIE 1993 International Symposium on Electronic Imaging: Science and Technology, volume 1905, pages 861-870, San Jose, CA, 1993. \n[Web Link] \n\nO.L. Mangasarian, W.N. Street and W.H. Wolberg. Breast cancer diagnosis and prognosis via linear programming. Operations Research, 43(4), pages 570-577, July-August 1995. \n[Web Link] \n\nW.H. Wolberg, W.N. Street, and O.L. Mangasarian. Machine learning techniques to diagnose breast cancer from fine-needle aspirates. Cancer Letters 77 (1994) 163-171. \n\nW.H. Wolberg, W.N. Street, and O.L. Mangasarian. Image analysis and machine learning applied to breast cancer diagnosis and prognosis. Analytical and Quantitative Cytology and Histology, Vol. 17 No. 2, pages 77-87, April 1995. \n\nW.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. Computerized breast cancer diagnosis and prognosis from fine needle aspirates. Archives of Surgery 1995;130:511-516. \n\nW.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. Computer-derived nuclear features distinguish malignant from benign breast cytology. Human Pathology, 26:792--796, 1995. \n\n', 64),
-(87, '1', '**Author**: Pierre Mahé, Jean-Baptiste Veyrieras  \n**Source**: [original](http://www.openml.org/d/1514) - UCI   \n**Please cite**:   \n\n* Dataset Title: MicroMass - Pure (pure spectra version)   \n\n* Abstract:   \nA dataset to explore machine learning approaches for the identification of microorganisms from mass-spectrometry data.  \n\n* Source:\n\nPierre Mahé, pierre.mahe \'@\' biomerieux.com, bioMérieux\nJean-Baptiste Veyrieras, jean-baptiste.veyrieras \'@\' biomerieux.com, bioMérieux\n\n* Data Set Information:\n\nThis MALDI-TOF dataset consists in:\nA) A reference panel of 20 Gram positive and negative bacterial species covering 9 genera among which several species are known to be hard to discriminate by mass spectrometry (MALDI-TOF). Each species was represented by 11 to 60 mass spectra obtained from 7 to 20 bacterial strains, constituting altogether a dataset of 571 spectra obtained from 213 strains. The spectra were obtained according to the standard culture-based workflow used in clinical routine in which the microorganism was first grown on an agar plate for 24 to 48 hours, before a portion of colony was picked, spotted on a MALDI slide and a mass spectrum was acquired. \nB) Based on this reference panel, a dedicated in vitro mock-up mixture dataset was constituted. For that purpose we considered 10 pairs of species of various taxonomic proximity:\n* 4 mixtures, labelled A, B, C and D, involved species that belong to the same genus,  \n* 2 mixtures, labelled E and F, involved species that belong to distinct genera, but to the same Gram type,  \n* 4 mixtures, labelled G, H, I and J, involved species that belong to distinct Gram types.  \nEach mixture was represented by 2 pairs of strains, which were mixed according to the following 9 concentration ratios : 1:0, 10:1, 5:1, 2:1, 1:1, 1:2, 1:5, 1:10, 0:1. Two replicate spectra were acquired for each concentration ratio and each couple of strains, leading altogether to a dataset of 360 spectra, among which 80 are actually pure sample spectra.\n\n* Relevant Papers:\n\nMahé et al. (2014). Automatic identification of mixed bacterial species fingerprints in a MALDI-TOF mass-spectrum. Bioinformatics.\n\nVervier et al., A benchmark of support vector machines strategies for microbial identification by mass-spectrometry data, submitted\n\n\n\n\n\n', 64),
-(88, '1', '**Author**: Brian Johnson     \n**Source**: [UCI] (https://archive.ics.uci.edu/ml/datasets/Wilt)  \n**Please cite**: Johnson, B., Tateishi, R., Hoan, N., 2013. A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees. International Journal of Remote Sensing, 34 (20), 6969-6982.   \n\n\n* Dataset:  \nWilt Data Set \n\n* Abstract:    \nHigh-resolution Remote Sensing data set (Quickbird). Small number of training samples of diseased trees, large number for other land cover. Testing data set from stratified random sample of image.\n\n* Source:\n  \nBrian Johnson; \nInstitute for Global Environmental Strategies; \n2108-11 Kamiyamaguchi, Hayama, Kanagawa,240-0115 Japan; \nEmail: Johnson \'@\' iges.or.jp \n\n\n* Data Set Information:  \n\nThis data set contains some training and testing data from a remote sensing study by Johnson et al. (2013) that involved detecting diseased trees in Quickbird imagery. There are few training samples for the \'diseased trees\' class (74) and many for \'other land cover\' class (4265). \n\nThe data set consists of image segments, generated by segmenting the pansharpened image. The segments contain spectral information from the Quickbird multispectral image bands and texture information from the panchromatic (Pan) image band. The testing data set is for the row with “Segmentation scale 15” segments and “original multi-spectral image” Spectral information in Table 2 of the reference (i.e. row 5). Please see the reference below for more information on the data set, and please cite the reference if you use this data set. Enjoy! \n\n* Attribute Information:\n\nclass: \'w\' (diseased trees), \'n\' (all other land cover)   \nGLCM_Pan: GLCM mean texture (Pan band)   \nMean_G: Mean green value   \nMean_R: Mean red value   \nMean_NIR: Mean NIR value   \nSD_Pan: Standard deviation (Pan band)   \n\n\n* Relevant Papers:\n\nJohnson, B., Tateishi, R., Hoan, N., 2013. A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees. International Journal of Remote Sensing, 34 (20), 6969-6982.\n', 64),
-(89, '1', '**Author**: Ronny Kohavi and Barry Becker  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Adult) - 1996-05-01  \n**Please cite**: Ron Kohavi, \"Scaling Up the Accuracy of Naive-Bayes Classifiers: a Decision-Tree Hybrid\", Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, 1996  \n\n**Note: this is the original version from the UCI repository, with training and test sets merged.**\n\nPrediction task is to determine whether a person makes over 50K a year. Extraction was done by Barry Becker from the 1994 Census database. A set of reasonably clean records was extracted using the following conditions: ((AAGE>16) && (AGI>100) && (AFNLWGT>1)&& (HRSWK>0))\n\nRonny Kohavi and Barry Becker. Data Mining and Visualization, Silicon Graphics.  \ne-mail: ronnyk \'@\' live.com for questions. \n', 2),
-(90, '1', 'Data from the Kaggle Bioresponse challenge:\nhttps://www.kaggle.com/c/bioresponse\n\nThe objective of the competition is to help us build as good a model as possible so that we can, as optimally as this data allows, relate molecular information, to an actual biological response.\n\nWe have shared the data in the comma separated values (CSV) format. Each row in this data set represents a molecule. The first column contains experimental data describing an actual biological response; the molecule was seen to elicit this response (1), or not (0). The remaining columns represent molecular descriptors (d1 through d1776), these are calculated properties that can capture some of the characteristics of the molecule - for example size, shape, or elemental constitution. The descriptor matrix has been normalized.\n\nThe original training and test set were merged.', 2),
-(91, '1', 'Data from the Kaggle Amazon Employee Access Challenge:\nhttps://www.kaggle.com/c/amazon-employee-access-challenge\n\nWhen an employee at any company starts work, they first need to obtain the computer access necessary to fulfill their role. This access may allow an employee to read/manipulate resources through various applications or web portals. It is assumed that employees fulfilling the functions of a given role will access the same or similar resources. It is often the case that employees figure out the access they need as they encounter roadblocks during their daily work (e.g. not able to log into a reporting portal). A knowledgeable supervisor then takes time to manually grant the needed access in order to overcome access obstacles. As employees move throughout a company, this access discovery/recovery cycle wastes a nontrivial amount of time and money.\n\nThere is a considerable amount of data regarding an employee&rsquo;s role within an organization and the resources to which they have access. Given the data related to current employees and their provisioned access, models can be built that automatically determine access privileges as employees enter and leave roles within a company. These auto-access models seek to minimize the human involvement required to grant or revoke employee access.\n\nThe original training and test set were merged.', 2),
-(92, '1', '**Author**: Rami Mustafa A Mohammad ( University of Huddersfield\",\"rami.mohammad \'@\' hud.ac.uk\",\"rami.mustafa.a \'@\' gmail.com) Lee McCluskey (University of Huddersfield\",\"t.l.mccluskey \'@\' hud.ac.uk )  Fadi Thabtah (Canadian University of Dubai\",\"fadi \'@\' cud.ac.ae)  \n**Source**: UCI \n**Please cite**: Please refer to the Machine Learning Repository\'s citation policy  \n\nSource:\n\nRami Mustafa A Mohammad ( University of Huddersfield, rami.mohammad \'@\' hud.ac.uk, rami.mustafa.a \'@\' gmail.com)\nLee McCluskey (University of Huddersfield,t.l.mccluskey \'@\' hud.ac.uk )\n\nFadi Thabtah (Canadian University of Dubai,fadi \'@\' cud.ac.ae)\n\n\nData Set Information:\n\nOne of the challenges faced by our research was the unavailability of reliable training datasets. In fact this challenge faces any researcher in the field. However, although plenty of articles about predicting phishing websites have been disseminated these days, no reliable training dataset has been published publically, may be because there is no agreement in literature on the definitive features that characterize phishing webpages, hence it is difficult to shape a dataset that covers all possible features. \nIn this dataset, we shed light on the important features that have proved to be sound and effective in predicting phishing websites. In addition, we propose some new features.\n\n\nAttribute Information:\n\nFor Further information about the features see the features file in the data folder.\n\n\nRelevant Papers:\n\nMohammad, Rami, McCluskey, T.L. and Thabtah, Fadi (2012) An Assessment of Features Related to Phishing Websites using an Automated Technique. In: International Conferece For Internet Technology And Secured Transactions. ICITST 2012 . IEEE, London, UK, pp. 492-497. ISBN 978-1-4673-5325-0\n\nMohammad, Rami, Thabtah, Fadi Abdeljaber and McCluskey, T.L. (2014) Predicting phishing websites based on self-structuring neural network. Neural Computing and Applications, 25 (2). pp. 443-458. ISSN 0941-0643\n\nMohammad, Rami, McCluskey, T.L. and Thabtah, Fadi Abdeljaber (2014) Intelligent Rule based Phishing Websites Classification. IET Information Security, 8 (3). pp. 153-160. ISSN 1751-8709\n\n \n\nCitation Request:\n\nPlease refer to the Machine Learning Repository\'s citation policy', 874),
-(93, '1', '**Author**: enata Cristina Barros Madeo (Madeo\",\"R. C. B.)  Priscilla Koch Wagner (Wagner\",\"P. K.)  Sarajane Marques Peres (Peres\",\"S. M.)  {renata.si\",\"priscilla.wagner\",\"sarajane} at usp.br  http://each.uspnet.usp.br/sarajane/  \n**Source**: UCI \n**Please cite**: Please refer to the Machine Learning Repository\'s citation policy. \nAdditionally, the authors require a citation to one or more publications from those cited as relevant papers.  \n\nSource:\n\nCreators: \nRenata Cristina Barros Madeo (Madeo, R. C. B.) \nPriscilla Koch Wagner (Wagner, P. K.) \nSarajane Marques Peres (Peres, S. M.) \n{renata.si, priscilla.wagner, sarajane} at usp.br \nhttp://each.uspnet.usp.br/sarajane/ \n\nDonor: \nUniversity of Sao Paulo - Brazil\n\n\nData Set Information:\n\nThe dataset is composed by features extracted from 7 videos with people gesticulating, aiming at studying Gesture Phase Segmentation. \nEach video is represented by two files: a raw file, which contains the position of hands, wrists, head and spine of the user in each frame; and a processed file, which contains velocity and acceleration of hands and wrists. See the data set description for more information on the dataset.\n\n\nAttribute Information:\n\nRaw files: 18 numeric attributes (double), a timestamp and a class attribute (nominal). \nProcessed files: 32 numeric attributes (double) and a class attribute (nominal). \nA feature vector with up to 50 numeric attributes can be generated with the two files mentioned above.\n\n\nRelevant Papers:\n\n1. Madeo, R. C. B. ; Lima, C. A. M. ; PERES, S. M. . Gesture Unit Segmentation using Support Vector Machines: Segmenting \nGestures from Rest Positions. In: Symposium on Applied Computing (SAC), 2013, Coimbra. Proceedings of the 28th Annual \nACM Symposium on Applied Computing (SAC), 2013. p. 46-52. \n* In this paper, the videos A1 and A2 were studied. \n\n2. Wagner, P. K. ; PERES, S. M. ; Madeo, R. C. B. ; Lima, C. A. M. ; Freitas, F. A. . Gesture Unit Segmentation Using \nSpatial-Temporal Information and Machine Learning. In: 27th Florida Artificial Intelligence Research Society Conference \n(FLAIRS), 2014, Pensacola Beach. Proceedings of the 27th Florida Artificial Intelligence Research Society Conference \n(FLAIRS). Palo Alto : The AAAI Press, 2014. p. 101-106. \n* In this paper, the videos A1, A2, A3, B1, B3, C1 and C3 were studied. \n\n3. Madeo, R. C. B.. Support Vector Machines and Gesture Analysis: incorporating temporal aspects (in Portuguese). Master \nThesis - Universidade de Sao Paulo, Sao Paulo Researcher Foundation. 2013. \n* In this document, the videos named B1 and B3 in the document correspond to videos C1 and C3 in this dataset. Only \nfive videos were explored in this document: A1, A2, A3, C1 and C3. \n\n4. Wagner, P. K. ; Madeo, R. C. B. ; PERES, S. M. ; Lima, C. A. M. . Segmenta&Atilde;&sect;ao de Unidades Gestuais com Multilayer \nPerceptrons (in Portuguese). In: Encontro Nacional de Inteligencia Artificial e Computacional (ENIAC), 2013, Fortaleza. \nAnais do X Encontro Nacional de Inteligencia Artificial e Computacional (ENIAC), 2013. \n* In this paper, the videos A1, A2 and A3 were studied.\n\n\n\nCitation Request:\n\nPlease refer to the Machine Learning Repository\'s citation policy. \nAdditionally, the authors require a citation to one or more publications from those cited as relevant papers.', 874),
-(94, '1', '**Author**: lara Higuera Department of Software Engineering and Artificial Intelligence\",\"Faculty of Informatics and the Department of Biochemistry and Molecular Biology\",\"Faculty of Chemistry\",\"University Complutense\",\"Madrid\",\"Spain.  Email: clarahiguera \'@\' ucm.es   Katheleen J. Gardiner\",\"creator and owner of the protein expression data\",\"is currently with the Linda Crnic Institute for Down Syndrome\",\"Department of Pediatrics\",\"Department of Biochemistry and Molecular Genetics\",\"Human Medical Genetics and Genomics\",\"and Neuroscience Programs\",\"University of Colorado\",\"School of Medicine\",\"Aurora\",\"Colorado\",\"USA.  Email: katheleen.gardiner \'@\' ucdenver.edu   Krzysztof J. Cios is currently with the Department of Computer Science\",\"Virginia Commonwealth University\",\"Richmond\",\"Virginia\",\"USA\",\"and IITiS Polish Academy of Sciences\",\"Poland.  Email: kcios \'@\' vcu.edu  \n**Source**: UCI  \n**Please cite**: Higuera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126  \n\nAbstract: Expression levels of 77 proteins measured in the cerebral cortex of 8 classes of control and Down syndrome mice exposed to context fear conditioning, a task used to assess associative learning.\nSource:\n\nClara Higuera Department of Software Engineering and Artificial Intelligence, Faculty of Informatics and the Department of Biochemistry and Molecular Biology, Faculty of Chemistry, University Complutense, Madrid, Spain. \nEmail: clarahiguera \'@\' ucm.es \n\nKatheleen J. Gardiner, creator and owner of the protein expression data, is currently with the Linda Crnic Institute for Down Syndrome, Department of Pediatrics, Department of Biochemistry and Molecular Genetics, Human Medical Genetics and Genomics, and Neuroscience Programs, University of Colorado, School of Medicine, Aurora, Colorado, USA. \nEmail: katheleen.gardiner \'@\' ucdenver.edu \n\nKrzysztof J. Cios is currently with the Department of Computer Science, Virginia Commonwealth University, Richmond, Virginia, USA, and IITiS Polish Academy of Sciences, Poland. \nEmail: kcios \'@\' vcu.edu \n\n\nData Set Information:\n\nThe data set consists of the expression levels of 77 proteins/protein modifications that produced detectable signals in the nuclear fraction of cortex. There are 38 control mice and 34 trisomic mice (Down syndrome), for a total of 72 mice. In the experiments, 15 measurements were registered of each protein per sample/mouse. Therefore, for control mice, there are 38x15, or 570 measurements, and for trisomic mice, there are 34x15, or 510 measurements. The dataset contains a total of 1080 measurements per protein. Each measurement can be considered as an independent sample/mouse. \n\nThe eight classes of mice are described based on features such as genotype, behavior and treatment. According to genotype, mice can be control or trisomic. According to behavior, some mice have been stimulated to learn (context-shock) and others have not (shock-context) and in order to assess the effect of the drug memantine in recovering the ability to learn in trisomic mice, some mice have been injected with the drug and others have not. \n\nClasses: \nc-CS-s: control mice, stimulated to learn, injected with saline (9 mice) \nc-CS-m: control mice, stimulated to learn, injected with memantine (10 mice) \nc-SC-s: control mice, not stimulated to learn, injected with saline (9 mice) \nc-SC-m: control mice, not stimulated to learn, injected with memantine (10 mice) \n\nt-CS-s: trisomy mice, stimulated to learn, injected with saline (7 mice) \nt-CS-m: trisomy mice, stimulated to learn, injected with memantine (9 mice) \nt-SC-s: trisomy mice, not stimulated to learn, injected with saline (9 mice) \nt-SC-m: trisomy mice, not stimulated to learn, injected with memantine (9 mice) \n\nThe aim is to identify subsets of proteins that are discriminant between the classes. \n\n\nAttribute Information:\n\n1 Mouse ID \n2..78 Values of expression levels of 77 proteins; the names of proteins are followed by &acirc;&euro;&oelig;_n&acirc;&euro; indicating that they were measured in the nuclear fraction. For example: DYRK1A_n \n79 Genotype: control (c) or trisomy (t) \n80 Treatment type: memantine (m) or saline (s) \n81 Behavior: context-shock (CS) or shock-context (SC) \n82 Class: c-CS-s, c-CS-m, c-SC-s, c-SC-m, t-CS-s, t-CS-m, t-SC-s, t-SC-m \n\n\nRelevant Papers:\n\nThe posted data were analyzed by: \nHiguera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126 \n\nThe data are a subset of the data analyzed by: \nAhmed MM, Dhanasekaran AR, Block A, Tong S, Costa ACS, Stasko M, et al. (2015) Protein Dynamics Associated with Failed and Rescued Learning in the Ts65Dn Mouse Model of Down Syndrome. PLoS ONE 10(3): e0119491. [Web Link] \n\n\n\n\nCitation Request:\n\nHiguera C, Gardiner KJ, Cios KJ (2015) Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome. PLoS ONE 10(6): e0129126. [Web Link] journal.pone.0129126', 874),
-(95, '1', '**Author**: Bob Evans, RR Donnelley & Sons Co.  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Cylinder+Bands) - August, 1995  \n**Please cite**:   \n\n**Cylinder bands**  \nProcess delays known as cylinder banding in rotogravure printing were substantially mitigated using control rules discovered by decision tree induction.\n \nAttribute Information:\n>  \n   1. timestamp: numeric;19500101 - 21001231  \n   2. cylinder number: nominal  \n   3. customer: nominal;  \n   4. job number: nominal;   \n   5. grain screened: nominal; yes, no  \n   6. ink color: nominal;  key, type  \n   7. proof on ctd ink:  nominal;  yes, no   \n   8. blade mfg: nominal;  benton, daetwyler, uddeholm  \n   9. cylinder division: nominal;  gallatin, warsaw, mattoon  \n  10. paper type: nominal;  uncoated, coated, super  \n  11. ink type: nominal;  uncoated, coated, cover  \n  12. direct steam: nominal; use; yes, no *  \n  13. solvent type: nominal;  xylol, lactol, naptha, line, other  \n  14. type on cylinder:  nominal;  yes, no   \n  15. press type: nominal; use; 70 wood hoe, 70 motter, 70 albert, 94 motter  \n  16. press: nominal;  821, 802, 813, 824, 815, 816, 827, 828  \n  17. unit number: nominal;  1, 2, 3, 4, 5, 6, 7, 8, 9, 10  \n  18. cylinder size: nominal;  catalog, spiegel, tabloid  \n  19. paper mill location: nominal; north us, south us, canadian, \n      scandanavian, mid european  \n  20. plating tank: nominal; 1910, 1911, other  \n  21. proof cut: numeric;  0-100  \n  22. viscosity: numeric;  0-100  \n  23. caliper: numeric;  0-1.0  \n  24. ink temperature: numeric;  5-30  \n  25. humifity: numeric;  5-120  \n  26. roughness: numeric;  0-2  \n  27. blade pressure: numeric;  10-75  \n  28. varnish pct: numeric;  0-100  \n  29. press speed: numeric;  0-4000  \n  30. ink pct: numeric;  0-100  \n  31. solvent pct: numeric;  0-100  \n  32. ESA Voltage: numeric;  0-16  \n  33. ESA Amperage: numeric;  0-10  \n  34. wax: numeric ;  0-4.0  \n  35. hardener:  numeric; 0-3.0  \n  36. roller durometer:  numeric;  15-120  \n  37. current density:  numeric;  20-50  \n  38. anode space ratio:  numeric;  70-130  \n  39. chrome content: numeric; 80-120  \n  40. band type: nominal; class; band, no band  \n\nNotes:  \n* cylinder number is an identifier and should be ignored when modelling the data', 2),
-(96, '1', '**Author**: Dr. Fernando Camacho  \n**Source**: Unknown - 1995  \n**Please cite**: Camacho, F. and Arron, G. (1995)  Effects of the regulators paclobutrazol and flurprimidol on the growth of terminal sprouts formed on trimmed silver maple trees. Canadian Journal of Statistics 3(23).\n\nData on tree growth used in the Case Study published in the September, 1995 issue of the Canadian Journal of Statistics. This data set was been provided by Dr. Fernando Camacho, Ontario Hydro Technologies, 800 Kipling Ave, Toronto Canada M3Z 5S4. It forms the basis of the Case Study in Data Analysis published in the Canadian Journal of Statistics, September 1995. It can be freely used for noncommercial purposes, as long as proper acknowledgement to the source and to the Canadian Journal of Statistics is made.\n\n\nDescription\n\n\nThe effects of the Growth Regulators Paclobutrazol (PP 333)\nand Flurprimidol (EL-500) on the Number and Length of Internodes\nin Terminal Sprouts Formed on Trimmed Silver Maple Trees.\n \nIntroduction:\n \nThe trimming of trees under distribution lines on city streets and\nin rural areas is a major problem and expense for electrical\nutilities.  Such operations are routinely performed at intervals of\none to eight years depending upon the individual species growth rate\nand the amount of clearance required.  Ontario Hydro trims about\n500,000 trees per year at a cost of about $25 per tree.\n \nMuch effort has been spent in developing chemicals for the horticultural\nindustry to retard the growth of woody and herbaceous plants.  Recently,\na group of new growth regulators was introduced which was shown to be\neffective in controlling the growth of trees without producing\nnoticeable injury symptoms.  In this group are PP 333 ( common name\npaclobutrazol) (2RS, 3RS - 1 -(4-chlorophenyl) - 4,4 - dimethyl - 2 -\n(1,2,4-triazol-l-yl) pentan - 3- ol and EL-500 (common name flurprimidol\nand composition alpha - (1-methylethyl) - alpha - [4-(trifluromethoxyl)\nphenyl] - 5- pyrimidine - methanol).  Both EL-500 and PP-333 have been\nreported to control excessive sprout growth in a number of species\nwhen applied as a foliar spray, as a soil drench, or by trunk injection.\nSprout length is a function of both the number of internodes and\nthe length of the individual internodes in the sprout.  While there\nhave been many reports that both PP 333 and EL-500 cause a reduction\nin the length of internodes formed in sprouts on woody plants treated\nwith the growth regulators, there has been but one report that EL-500\napplication to apple trees resulted in a reduction of the number\nof internodes formed per sprout.\n \nThe purpose of the present study was to investigate the length of the\nterminal sprouts, the length of the individual internodes in those\nsprouts, and the number of internodes in trimmed silver maple trees\nfollowing trunk injection with the growth regulators PP 333 and EL-500.\n \nExperimental Details.\n \nMultistemmed 12-year-old silver maple trees growing at Wesleyville,\nOntario were trunk injected with methanolic solutions of EL-500\nand PP-333 in May of 1985 using a third generation Asplundh\ntree injector.\n \nTwo different application rates (20 g/L and 4 g/L) were used for each\nchemical.  The volume of solution (and hence the amount of active\ningredient) injected into each tree was determined from the diameter\nof the tree, using the formula: vol(mL) = (dbh)*(dbh)*.492 where dbh\nis the diameter at breast height.  Two sets of control trees were\nincluded in the experiment.  In one set, tree received no injection\n(control) and in a second set, the trees were injected with\nmethanol, the carrier in the growth regulator solutions.  Ten trees,\nchosen at random, were used in each of the control and experimental\nsets.  Prior to injection, all the trees were trimmed by a forestry\ncrew, with their heights being reduced by about one third.\n \nIn January 1987, twenty months after the trees were injected, between\nsix and eight limbs were removed at random from the bottom two-thirds\nof the canopy of each of the ten trees in each experimental and control\nset.  The limbs were returned to the laboratory and the length of all\nthe terminal sprouts, the lengths of the individual internodes, and\nthe number of internodes recorded.  Between one and 25 terminal\nsprouts were found on each limb collected.  Sprouts which had a\nlength of 1 cm or less were recorded as being 1 cm in length.\nIn such spouts, the internode lengths were not measured, but were\ncalculated from the total length of the sprout and the number\nof internodes counted.  Internode lengths were then expressed to one\ndecimal place.  In two instances, one of the ten trees in a set\ncould not be sampled because limb removal would have jeopardized the\nhealth of the tree over the long-term.\n \nData set:\n \nEach of the records represents a terminal sprout and contains the\nfollowing information:\n   N   the sprout number\n   TR  treatment 1  control\n                  2  methanol control\n                  3  PP 333 20g/L\n                  4  PP 333  4g/L\n                  5  EL 500 20g/L\n                  6  EL 500  4g/L\n   TREE  tree id\n   BR    branch id\n   TL    total sprout length (cm)\n   IN    number of internodes on the sprout\n   INTER a list of the lengths of the internodes in the sprout,\n          starting from the base of the sprout (129 entries)\n \nSprouts 1868 to 1879 do not have branch identification data.', 2),
-(97, '1', '**Author**: Muhammad Usman & Adeel Ahmed   \r\n**Source**: origin source at [UCI](https://archive.ics.uci.edu/ml/datasets/Dresses_Attribute_Sales)  \r\n**Please cite**:   \r\n\r\n####1. Summary\r\n\r\nThis dataset contain attributes of dresses and their recommendations according to their sales.Sales are monitor on the basis of alternate days. \r\n\r\nThe attributes present analyzed are: Style, Price, Rating, Size, Season, NeckLine, SleeveLength, waiseline, Material, FabricType, Decoration, Pattern, Type, Recommendation.\r\n\r\nContact:\r\n```\r\nMuhammad Usman & Adeel Ahmed, usman.madspot \'@\' gmail.com adeel.ahmed92 \'@\' gmail.com, Air University, Students at Air University.\r\n```\r\n\r\n####2: Attribute Information:\r\n\r\n```\r\nStyle: Bohemia,brief,casual,cute,fashion,flare,novelty,OL,party,sexy,vintage,work. \r\nPrice:Low,Average,Medium,High,Very-High \r\nRating:1-5 \r\nSize:S,M,L,XL,Free \r\nSeason:Autumn,winter,Spring,Summer \r\nNeckLine:O-neck,backless,board-neck,Bowneck,halter,mandarin-collor,open,peterpan-collor,ruffled,scoop,slash-neck,square-collar,sweetheart,turndowncollar,V-neck. \r\nSleeveLength:full,half,halfsleeves,butterfly,sleveless,short,threequarter,turndown,null \r\nwaiseline:dropped,empire,natural,princess,null. \r\nMaterial:wool,cotton,mix etc \r\nFabricType:shafoon,dobby,popline,satin,knitted,jersey,flannel,corduroy etc \r\nDecoration:applique,beading,bow,button,cascading,crystal,draped,embroridary,feathers,flowers etc \r\nPattern type: solid,animal,dot,leapard etc \r\nRecommendation:0,1 \r\n```\r\n', 64),
-(98, '1', '1. Title of Database: LED display domain\n\n2. Sources:\n   (a) Breiman,L., Friedman,J.H., Olshen,R.A., &amp; Stone,C.J. (1984). \n       Classification and Regression Trees.  Wadsworth International\n       Group: Belmont, California.  (see pages 43-49).\n   (b) Donor: David Aha \n   (c) Date: 11/10/1988\n\n3. Past Usage: (many)\n     1. CART book (above):\n        -- Optimal Bayes classification rate: 74%\n        -- CART decision tree algorithm: 71% (resubstitution estimate)\n        -- Nearest Neighbor Algorithm: 71%\n           -- 200 training and 5000 test instances\n     \n     2. Quinlan,J.R. (1987). Simplifying Decision Trees.  In International\n        Journal of Man-Machine Studies (to appear).\n        -- C4 decision tree algorithm: 72.6% (using pessimistic pruning)\n           -- 2000 training and 500 test instances\n     3. Tan,M. &amp; Eshelman,L. (1988). Using Weighted Networks to Represent\n        Classification Knowledge in Noisy Domains.  In Proceedings of the\n        5th International Conference on Machine Learning, 121-134, Ann\n        Arbor, Michigan: Morgan Kaufmann.  \n        -- IWN system: 73.3% (using the And-OR classification algorithm)\n           -- 400 training and 500 test cases\n\n4. Relevant Information Paragraph:\n     This simple domain contains 7 Boolean attributes and 10 concepts,\n     the set of decimal digits.  Recall that LED displays contain 7\n     light-emitting diodes -- hence the reason for 7 attributes.  The\n     problem would be easy if not for the introduction of noise.  In\n     this case, each attribute value has the 10% probability of having\n     its value inverted.  \n\n     It\'s valuable to know the optimal Bayes rate for these databases.\n     In this case, the misclassification rate is 26% (74% classification\n     accuracy).\n        \n5. Number of Instances: 500. But in the original URL you can find a C script and run it choosing the number of instances to be generated.\n\n6. Number of Attributes: 7 (all Boolean-valued)\n\n7. Attribute Information:\n   -- All attribute values are either 0 or 1, according to whether\n      the corresponding light is on or not for the decimal digit.\n   -- Each attribute (excluding the class attribute, which is an\n      integer ranging between 0 and 9 inclusive) has a 10% percent\n      chance of being inverted.\n\n8. Missing Attribute Values: None\n\n9. Class Distribution: 10% (Theoretical)\n   -- Each concept (digit) has the same theoretical probability\n      distribution.  The program randomly selects the attribute.', 64),
-(99, '1', '**Author**: Laboratory of Image Processing and Pattern Recognition (INPG-LTIRF), Grenoble - France.  \r\n**Source**: [original](https://www.elen.ucl.ac.be/neural-nets/Research/Projects/ELENA/databases/REAL/texture/) - ELENA project   \r\n**Please cite**:   \r\n\r\n####1. Summary\r\n\r\nThis database was generated by the Laboratory of Image Processing and Pattern Recognition (INPG-LTIRF) in the development of the Esprit project ELENA No. 6891 and the Esprit working group ATHOS No. 6620.\r\n```\r\n (a) Original source:\r\n\r\n   P. Brodatz \"Textures: A Photographic Album for Artists and Designers\",\r\n   Dover Publications,Inc.,New York, 1966.\r\n\r\n (b) Creation: Laboratory of Image Processing and Pattern Recognition\r\n\r\n   Institut National Polytechnique de Grenoble INPG\r\n   Laboratoire de Traitement d\'Image et de Reconnaissance de Formes LTIRF\r\n   Av. Felix Viallet, 46\r\n   F-38031 Grenoble Cedex\r\n   France\r\n\r\n (c) Contact: Dr. A. Guerin-Dugue, INPG-LTIRF, guerin@tirf.inpg.fr\r\n```\r\n\r\n####2. Past Usage:\r\n\r\nThis database has a private usage at the TIRF laboratory. It has been created in order to study the textures discrimination with high order statistics.\r\n\r\n```\r\nA.Guerin-Dugue, C. Aviles-Cruz, \"High Order Statistics from Natural Textured Images\",\r\nIn ATHOS workshop on System Identification and High Order Statistics, Sophia-Antipolis, France, September 1993.\r\n\r\nGuerin-Dugue, A. and others, Deliverable R3-B4-P - Task B4: Benchmarks, Technical report,\r\nElena-NervesII \"Enhanced Learning for Evolutive Neural Architecture\", ESPRIT-Basic Research Project  Number 6891,\r\nJune 1995.\r\n```\r\n\r\n####3. Relevant Information:\r\n\r\nThe aim is to distinguish between 11 different textures (Grass lawn, Pressed calf leather, Handmade paper, Raffia looped to a high pile, Cotton canvas, ...), each pattern (pixel) being characterised by 40 attributes built by the estimation of fourth order modified moments in four orientations: 0, 45, 90 and 135 degrees.\r\n\r\nA statistical method based on the extraction of fourth order moments for the characterization of natural micro-textures was developed called \"fourth order modified moments\" (mm4) [Guerin93], this method measures the deviation from first-order Gauss-Markov process, for each texture. The features were estimated in four directions to take into account the possible orientations of the textures (0, 45, 90 and 135 degrees). Only correlation between the current pixel, the first neighbourhood and the second neighbourhood are taken into account. This small neighbourhood is adapted to the fine grain property of the textures.\r\n\r\nThe data set contains 11 classes of 500 instances and each class refers to a type of texture in the Brodatz album.\r\n\r\nThe database dimension is 40 plus one for the class label. The 40 attributes were build respectively by the estimation of the following fourth order modified moments in four orientations: 0, 45, 90 and 135 degrees: mm4(000), mm4(001), mm4(002), mm4(011), mm4(012), mm4(022), mm4(111), mm4(112), mm4(122) and mm4(222).\r\n\r\n!! Patterns are always sorted by class and are presented in the increasing order of their class label in each dataset relative to the texture database (texture.dat, texture_CR.dat, texture_PCA.dat, texture_DFA.dat)\r\n\r\n####4. Class:\r\n\r\nThe class label is a code for the following classes:\r\n```\r\n                Class         Class label\r\n  2   Grass lawn                      (D09)  \r\n  3   Pressed calf leather            (D24) \r\n  4   Handmade paper                  (D57) \r\n  6   Raffia looped to a high pile:   (D84) \r\n  7   Cotton canvas                   (D77) \r\n  8   Pigskin                         (D92) \r\n  9   Beach sand:                     (D28) \r\n  10  Beach sand                      (D29) \r\n  12  Oriental straw cloth            (D53) \r\n  13  Oriental straw cloth            (D78) \r\n  14  Oriental grass fiber cloth      (D79) \r\n```\r\n\r\n####5. Summary Statistics:\r\n\r\nTable here below provides for each attribute of the database the dynamic (Min and Max values), the mean value and the standard deviation.\r\n\r\n```\r\nAttribute  Min      Max       Mean     Standard    \r\n                                       deviation   \r\n\r\n    1   -1.4495    0.7741   -1.0983    0.2034\r\n    2   -1.2004    0.3297   -0.5867    0.2055\r\n    3   -1.3099    0.3441   -0.5838    0.3135\r\n    4   -1.1104    0.5878   -0.4046    0.2302\r\n    5   -1.0534    0.4387   -0.3307    0.2360\r\n    6   -1.0029    0.4515   -0.2422    0.2225\r\n    7   -1.2076    0.5246   -0.6026    0.2003\r\n    8   -1.0799    0.3980   -0.4322    0.2210\r\n    9   -1.0570    0.4369   -0.3317    0.2361\r\n   10   -1.2580    0.3546   -0.5978    0.3268\r\n   11   -1.4495    0.7741   -1.0983    0.2034\r\n   12   -1.0831    0.3715   -0.5929    0.2056\r\n   13   -1.1194    0.6347   -0.4019    0.3368\r\n   14   -1.0182    0.1573   -0.6270    0.1390\r\n   15   -0.9435    0.1642   -0.4482    0.1952\r\n   16   -0.9944    0.0357   -0.5763    0.1587\r\n   17   -1.1722    0.0201   -0.7331    0.1955\r\n   18   -1.0174    0.1155   -0.4919    0.2335\r\n   19   -1.0044    0.0833   -0.4727    0.2257\r\n   20   -1.1800    0.4392   -0.4831    0.3484\r\n   21   -1.4495    0.7741   -1.0983    0.2034\r\n   22   -1.2275    0.5963   -0.7363    0.2220\r\n   23   -1.3412    0.4464   -0.7771    0.3290\r\n   24   -1.1774    0.6882   -0.5770    0.2646\r\n   25   -1.1369    0.4098   -0.5085    0.2538\r\n   26   -1.1099    0.3725   -0.4038    0.2515\r\n   27   -1.2393    0.6120   -0.7279    0.2278\r\n   28   -1.1540    0.4221   -0.5863    0.2446\r\n   29   -1.1323    0.3916   -0.5090    0.2526\r\n   30   -1.4224    0.4718   -0.7708    0.3264\r\n   31   -1.4495    0.7741   -1.0983    0.2034\r\n   32   -1.1789    0.5647   -0.6463    0.1890\r\n   33   -1.1473    0.6755   -0.4919    0.3304\r\n   34   -1.1228    0.3132   -0.6435    0.1441\r\n   35   -1.0145    0.3396   -0.4918    0.1922\r\n   36   -1.0298    0.1560   -0.5934    0.1704\r\n   37   -1.2534    0.0899   -0.7795    0.1641\r\n   38   -1.0966    0.1944   -0.5541    0.2111\r\n   39   -1.0765    0.2019   -0.5230    0.2015\r\n   40   -1.2155    0.4647   -0.5677    0.3091\r\n```\r\n\r\nThe dynamic of the attributes is in [-1.45 - 0.775]. The database resulting from the centering and reduction by attribute of the Texture database is on the ftp server in the `REAL/texture/texture_CR.dat.Z\' file.\r\n\r\n####6. Confusion matrix.\r\n\r\nThe following confusion matrix of the k_NN classifier was obtained with a Leave_One_Out error counting method on the texture_CR.dat database. k was set to 1 in order to reach the minimum mean error rate : 1.0 +/- 0.8%.\r\n\r\n```\r\n Class    2      3      4      6      7      8      9      10     12     13     14 \r\n 2      97.0    1.0    0.4    0.0    0.0    0.0    1.6    0.0    0.0    0.0    0.0   \r\n 3       0.2   99.0    0.0    0.0    0.0    0.0    0.4    0.0    0.0    0.0    0.4   \r\n 4       1.0    0.0   98.8    0.0    0.0    0.0    0.2    0.0    0.0    0.0    0.0   \r\n 6       0.0    0.0    0.0   99.4    0.0    0.0    0.0    0.6    0.0    0.0    0.0   \r\n 7       0.0    0.0    0.0    0.0  100.0    0.0    0.0    0.0    0.0    0.0    0.0   \r\n 8       0.0    0.0    0.0    0.0    0.0   98.6    0.0    1.4    0.0    0.0    0.0   \r\n 9       0.4    0.0    0.2    0.0    0.0    0.2   98.8    0.4    0.0    0.0    0.0   \r\n 10      0.0    0.0    0.0    0.0    0.0    1.4    0.0   98.6    0.0    0.0    0.0   \r\n 12      0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  100.0    0.0    0.0   \r\n 13      0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   99.8    0.2   \r\n 14      0.0    0.4    0.0    0.0    0.0    0.4    0.0    0.0    0.2    0.0   99.0   \r\n```\r\n\r\n7. Result of the Principal Component Analysis:\r\n\r\nThe Principal Components Analysis is a very classical method in pattern recognition [Duda73]. PCA reduces the sample dimension in a linear way for the best representation in lower dimensions keeping the maximum of inertia. The best axe for the representation is however not necessary the best axe for the discrimination. After PCA, features are selected according to the percentage of initial inertia which is covered by the different axes and the number of features is determined according to the percentage of initial inertia to keep for the classification process.\r\n\r\nThis selection method has been applied on the texture_CR database. When quasi-linear correlations exists between some initial features, these redundant dimensions are removed by PCA and this preprocessing is then recommended. In this case, before a PCA, the determinant of the data covariance matrix is near zero; this database is thus badly conditioned for all process which use this information (the quadratic classifier for example).\r\n\r\nThe following file is available for the texture database: \'\'texture_PCA.dat.Z\'\', it is the projection of the \'\'texture_CR\'\' database on its principal components (sorted in a decreasing order of the related inertia percentage; so, if you desire to work on the database projected on its x first principal components you only have to keep the x first attributes of the texture_PCA.dat database and the class labels (last attribute)).\r\n\r\nTable here below provides the inertia percentages associated to the eigenvalues corresponding to the principal component axis sorted in the decreasing order of the associated inertia percentage. 99.85 percent of the total database inertia will remain if the 20 first principal components are kept.\r\n\r\n```\r\n       Eigen Value   Inertia      Cumulated\r\n         value      percentage      inertia\r\n\r\n  1   30.267500000 75.6687000000  75.6687000000 \r\n  2   3.6512500000  9.1281300000  84.7969000000 \r\n  3   2.2937000000  5.7342400000  90.5311000000 \r\n  4   1.7039700000  4.2599300000  94.7910000000 \r\n  5   0.6716540000  1.6791300000  96.4702000000 \r\n  6   0.5015290000  1.2538200000  97.7240000000 \r\n  7   0.1922830000  0.4807070000  98.2047000000 \r\n  8   0.1561070000  0.3902670000  98.5950000000 \r\n  9   0.1099570000  0.2748920000  98.8699000000 \r\n  10  0.0890891000  0.2227230000  99.0926000000 \r\n  11  0.0656016000  0.1640040000  99.2566000000 \r\n  12  0.0489988000  0.1224970000  99.3791000000 \r\n  13  0.0433819000  0.1084550000  99.4875000000 \r\n  14  0.0345022000  0.0862554000  99.5738000000 \r\n  15  0.0299203000  0.0748007000  99.6486000000 \r\n  16  0.0248857000  0.0622141000  99.7108000000 \r\n  17  0.0167901000  0.0419752000  99.7528000000 \r\n  18  0.0161633000  0.0404083000  99.7932000000 \r\n  19  0.0128898000  0.0322246000  99.8254000000 \r\n  20  0.0113884000  0.0284710000  99.8539000000 \r\n  21  0.0078481400  0.0196204000  99.8735000000 \r\n  22  0.0071527800  0.0178820000  99.8914000000 \r\n  23  0.0067661400  0.0169153000  99.9083000000 \r\n  24  0.0053149500  0.0132874000  99.9216000000 \r\n  25  0.0051102600  0.0127757000  99.9344000000 \r\n  26  0.0047116600  0.0117792000  99.9461000000 \r\n  27  0.0036193700  0.0090484300  99.9552000000 \r\n  28  0.0033222000  0.0083054900  99.9635000000 \r\n  29  0.0030722400  0.0076806100  99.9712000000 \r\n  30  0.0026373300  0.0065933300  99.9778000000 \r\n  31  0.0020996800  0.0052492000  99.9830000000 \r\n  32  0.0019376500  0.0048441200  99.9879000000 \r\n  33  0.0015642300  0.0039105700  99.9918000000 \r\n  34  0.0009679080  0.0024197700  99.9942000000 \r\n  35  0.0009578000  0.0023945000  99.9966000000 \r\n  36  0.0007379780  0.0018449400  99.9984000000 \r\n  37  0.0006280250  0.0015700600  100.000000000\r\n  38  0.0000000040  0.0000000099  100.000000000 \r\n  39  0.0000000001  0.0000000003  100.000000000 \r\n  40  0.0000000008  0.0000000019  100.000000000 \r\n\r\n```\r\n\r\nThis matrix can be found in the texture_EV.dat file.\r\n\r\nThe Discriminant Factorial Analysis (DFA) can be applied to a learning database where each learning sample belongs to a particular class [Duda73]. The number of discriminant features selected by DFA is fixed in function of the number of classes (c) and of the number of input dimensions (d); this number is equal to the minimum between d and c-1. In the usual case where d is greater than c, the output dimension is fixed equal to the number of classes minus one and the discriminant axes are selected in order to maximize the between-variance and to minimize the within-variance of the classes.\r\n\r\nThe discrimination power (ratio of the projected between-variance over the projected within-variance) is not the same for each discriminant axis: this ratio decreases for each axis. So for a problem with many classes, this preprocessing will not be always efficient as the last output features will not be so discriminant. This analysis uses the information of the inverse of the global covariance matrix, so the covariance matrix must be well conditioned (for example, a preliminary PCA must be applied to remove the linearly correlated dimensions).\r\n\r\nThe Discriminant Factorial Analysis (DFA) has been applied on the 18 first principal components of the texture_PCA database (thus by keeping only the 18 first attributes of these databases before to apply the DFA preprocessing) in order to build the texture_DFA.dat.Z database file, having 10 dimensions (the texture database having 11 classes). In the case of the texture database, experiments shown that a DFA preprocessing is very useful and most of the time improved the classifiers performances.\r\n\r\n[Duda73] Duda, R.O. and Hart, P.E.,Pattern Classification and Scene Analysis, John Wiley & Sons, 1973.\r\n', 64),
-(100, '1', '**Author**: Confidential. Donated by Ross Quinlan  \r\n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Statlog+(Australian+Credit+Approval))/LibSVM - 2014-11-14  \r\n**Please cite**:   \r\n\r\nThis is the famous Australian dataset, retrieved 2014-11-14 from the libSVM site.\r\nIt was normalized.\r\n\r\nThe original version is from [UCI](https://archive.ics.uci.edu/ml/datasets/Statlog+(Australian+Credit+Approval)).\r\n\r\nThis file concerns credit card applications. All attribute names and values have been changed to meaningless symbols to protect confidentiality of the data. \r\n\r\nThis dataset is interesting because there is a good mix of attributes -- continuous, nominal with small numbers of values, and nominal with larger numbers of values. There are also a few missing values. \r\n\r\n\r\nSource: Statlog / Australian\r\n# of classes: 2\r\n# of data: 690\r\n# of features: 14', 2),
-(101, '1', '**Author**: Terran Lane (terran@ecn.purdue.edu)   \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/UNIX+User+Data) - Date unknown  \n**Please cite**:   \n\nThis file contains 9 sets of sanitized user data drawn from the command histories of 8 UNIX computer users at Purdue over the course of up to 2 years (USER0 and USER1 were generated by the same person, working on different platforms and different projects).  The data is drawn from tcsh(1) history files and has been parsed and sanitized to remove filenames, user names, directory structures, web addresses, host names, and other possibly identifying items.  Command names, flags, and shell metacharacters have been preserved.  Additionally, **SOF** and **EOF** tokens have been inserted at the start and end of\nshell sessions, respectively.  Sessions are concatenated by date order and tokens appear in the order issued within the shell session, but no timestamps are included in this data.  For example, the two sessions:\n\ncd ~/private/docs  \nls -laF | more  \ncat foo.txt bar.txt zorch.txt > somewhere  \nexit  \n\ncd ~/games/  \nxquake &  \nfg  \nvi scores.txt  \nmailx john_doe@somewhere.com  \nexit  \n\nwould be represented by the token stream  \n\n**SOF**  \ncd  \n\\<1\\>                  (one \"file name\" argument)  \nls  \n-laF  \n|  \nmore  \ncat  \n\\<3\\>                     (three \"file\" arguments)\n\\>  \n\\<1\\>  \nexit  \n**EOF**  \n**SOF**  \ncd  \n\\<1\\>  \nxquake  \n&  \nfg  \nvi  \n\\<1\\>  \nmailx  \n\\<1\\>  \nexit  \n**EOF**\n\nThis data is made available under conditions of anonymity for the contributing users and may be used for research purposes only. Summaries and research results employing this data may be published, but literal tokens or token sequences from the data may not be published except with express consent of the originators of the data. No portion of this data may be released with or included in a commercial product, nor may any portion of this data be sold or redistributed for profit or as part of of a profit-making endeavor.', 2),
-(102, '1', '**Author**: Tin Kam Ho and Eugene M. Kleinberg.  \nlibSVM\",\"AAD group  \n**Source**: [original](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary.html) - Date unknown  \n**Please cite**:   \n\n#Dataset from the LIBSVM data repository.\n\nPreprocessing: transform to two-class', 402),
-(103, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis is data set is concerned with the forward kinematics of an 8 link\n robot arm. Among the existing variants of this data set we have used\n the variant 8nm, which is known to be highly non-linear and medium\n noisy.\n\n Original source: DELVE repository of data. \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 8192 cases, 9 attributes (0 nominal, 9 continuous).', 1),
-(104, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 1),
-(105, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Wisconsin Prognostic Breast Cancer (WPBC)\n \n 2. Source Information\n \n a) Creators: \n \n 	Dr. William H. Wolberg, General Surgery Dept., University of\n 	Wisconsin,  Clinical Sciences Center, Madison, WI 53792\n 	wolberg@eagle.surgery.wisc.edu\n \n 	W. Nick Street, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	street@cs.wisc.edu  608-262-6619\n \n 	Olvi L. Mangasarian, Computer Sciences Dept., University of\n 	Wisconsin, 1210 West Dayton St., Madison, WI 53706\n 	olvi@cs.wisc.edu \n \n b) Donor: Nick Street\n \n c) Date: December 1995\n \n 3. Past Usage:\n \n 	Various versions of this data have been used in the following\n 	publications: \n \n 	(i) W. N. Street, O. L. Mangasarian, and W.H. Wolberg. \n 	An inductive learning approach to prognostic prediction. \n 	In A. Prieditis and S. Russell, editors, Proceedings of the\n 	Twelfth International Conference on Machine Learning, pages\n 	522--530, San Francisco, 1995. Morgan Kaufmann.\n \n 	(ii) O.L. Mangasarian, W.N. Street and W.H. Wolberg. \n 	Breast cancer diagnosis and prognosis via linear programming. \n 	Operations Research, 43(4), pages 570-577, July-August 1995. \n \n 	(iii) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computerized breast cancer diagnosis and prognosis from fine\n 	needle aspirates.  Archives of Surgery 1995;130:511-516. \n \n 	(iv) W.H. Wolberg, W.N. Street, and O.L. Mangasarian. \n 	Image analysis and machine learning applied to breast cancer\n 	diagnosis and prognosis. Analytical and Quantitative Cytology\n 	and Histology, Vol. 17 No. 2, pages 77-87, April 1995.\n \n 	(v) W.H. Wolberg, W.N. Street, D.M. Heisey, and O.L. Mangasarian. \n 	Computer-derived nuclear ``grade\'\' and breast cancer prognosis. \n 	Analytical and Quantitative Cytology and Histology, Vol. 17,\n 	pages 257-264, 1995. \n \n See also:\n 	http://www.cs.wisc.edu/~olvi/uwmp/mpml.html\n 	http://www.cs.wisc.edu/~olvi/uwmp/cancer.html\n \n Results:\n \n 	Two possible learning problems:\n \n 	1) Predicting field 2, outcome: R = recurrent, N = nonrecurrent\n 	- Dataset should first be filtered to reflect a particular\n 	endpoint; e.g., recurrences before 24 months = positive,\n 	nonrecurrence beyond 24 months = negative.\n 	- 86.3% accuracy estimated accuracy on 2-year recurrence using\n 	previous version of this data.  Learning method: MSM-T (see\n 	below) in the 4-dimensional space of Mean Texture, Worst Area,\n 	Worst Concavity, Worst Fractal Dimension.\n \n 	2) Predicting Time To Recur (field 3 in recurrent records)\n 	- Estimated mean error 13.9 months using Recurrence Surface\n 	Approximation. (See references (i) and (ii) above)\n \n 4. Relevant information\n \n 	Each record represents follow-up data for one breast cancer\n 	case.  These are consecutive patients seen by Dr. Wolberg\n 	since 1984, and include only those cases exhibiting invasive\n 	breast cancer and no evidence of distant metastases at the\n 	time of diagnosis. \n \n 	The first 30 features are computed from a digitized image of a\n 	fine needle aspirate (FNA) of a breast mass.  They describe\n 	characteristics of the cell nuclei present in the image.\n 	A few of the images can be found at\n 	http://www.cs.wisc.edu/~street/images/\n \n 	The separation described above was obtained using\n 	Multisurface Method-Tree (MSM-T) [K. P. Bennett, \"Decision Tree\n 	Construction Via Linear Programming.\" Proceedings of the 4th\n 	Midwest Artificial Intelligence and Cognitive Science Society,\n 	pp. 97-101, 1992], a classification method which uses linear\n 	programming to construct a decision tree.  Relevant features\n 	were selected using an exhaustive search in the space of 1-4\n 	features and 1-3 separating planes.\n \n 	The actual linear program used to obtain the separating plane\n 	in the 3-dimensional space is that described in:\n 	[K. P. Bennett and O. L. Mangasarian: \"Robust Linear\n 	Programming Discrimination of Two Linearly Inseparable Sets\",\n 	Optimization Methods and Software 1, 1992, 23-34].\n \n 	The Recurrence Surface Approximation (RSA) method is a linear\n 	programming model which predicts Time To Recur using both\n 	recurrent and nonrecurrent cases.  See references (i) and (ii)\n 	above for details of the RSA method. \n \n 	This database is also available through the UW CS ftp server:\n \n 	ftp ftp.cs.wisc.edu\n 	cd math-prog/cpo-dataset/machine-learn/WPBC/\n \n 5. Number of instances: 198\n \n 6. Number of attributes: 34 (ID, outcome, 32 real-valued input features)\n \n 7. Attribute information\n \n 1) ID number\n 2) Outcome (R = recur, N = nonrecur)\n 3) Time (recurrence time if field 2 = R, disease-free time if \n 	field 2	= N)\n 4-33) Ten real-valued features are computed for each cell nucleus:\n \n 	a) radius (mean of distances from center to points on the perimeter)\n 	b) texture (standard deviation of gray-scale values)\n 	c) perimeter\n 	d) area\n 	e) smoothness (local variation in radius lengths)\n 	f) compactness (perimeter^2 / area - 1.0)\n 	g) concavity (severity of concave portions of the contour)\n 	h) concave points (number of concave portions of the contour)\n 	i) symmetry \n 	j) fractal dimension (\"coastline approximation\" - 1)\n \n Several of the papers listed above contain detailed descriptions of\n how these features are computed. \n \n The mean, standard error, and \"worst\" or largest (mean of the three\n largest values) of these features were computed for each image,\n resulting in 30 features.  For instance, field 4 is Mean Radius, field\n 14 is Radius SE, field 24 is Worst Radius.\n \n Values for features 4-33 are recoded with four significant digits.\n \n 34) Tumor size - diameter of the excised tumor in centimeters\n 35) Lymph node status - number of positive axillary lymph nodes\n observed at time of surgery\n \n 8. Missing attribute values: \n 	Lymph node status is missing in 4 cases.\n \n 9. Class distribution: 151 nonrecur, 47 recur\n\n-----------------------------------------------------------------------------------------------------------\n Luis Torgo\'s version: (reconstructed)\n - removed the four instances with unknown values of the last attribute\n - exchanged the attribute position of attributes n.3 (Time) and n.35\n   (Lymph node).\n - removed the attribute outcome as it is the class attribute if the\n   problem is treated as a classification one\n-----------------------------------------------------------------------------------------------------------', 1),
-(106, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 1),
-(107, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n \n SUMMARY:\n \n Data from an experiment on the affects of machine adjustments on\n the time to count bolts.  Data appear as the STATS (Issue 10) Challenge.\n \n DATA:\n \n Submitted by W. Robert Stephenson, Iowa State University\n                             email: wrstephe@iastate.edu\n \n A manufacturer of automotive accessories provides hardware, e.g. nuts,\n bolts, washers and screws, to fasten the accessory to the car or truck.\n Hardware is counted and packaged automatically.  Specifically, bolts\n are dumped into a large metal dish.  A plate that forms the bottom of\n the dish rotates counterclockwise.  This rotation forces bolts to the\n outside of the dish and up along a narrow ledge.  Due to the vibration\n of the dish caused by the spinning bottom plate, some bolts fall off \n the ledge and back into the dish.  The ledge spirals up to a point \n where the bolts are allowed to drop into a pan on a conveyor belt.  \n As a bolt drops, it passes by an electronic eye that counts it.  When \n the electronic counter reaches the preset number of bolts, the\n rotation is stopped and the conveyor belt is moved forward.  \n \n There are several adjustments on the machine that affect its operation.  \n These include; a speed setting that controls the speed of rotation\n (SPEED1) of the plate at the bottom of the dish, a total number of \n bolts (TOTAL) to be counted, a second speed setting (SPEED2) that is \n used to change the speed of rotation (usually slowing it down) for the\n last few bolts, the number of bolts to be counted at this second speed\n (NUMBER2), and the sensitivity of the electronic eye (SENS).  The \n sensitivity setting is to insure that the correct number of bolts are \n counted.  Too few bolts packaged causes customer complaints.  Too many\n bolts packaged increases costs.  For each run conducted in this \n experiment the correct number of bolts was counted.  From an\n engineering standpoint if the correct number of bolts is counted, the \n sensitivity should not affect the time to count bolts.  The measured \n response is the time (TIME), in seconds, it takes to count the desired\n number of bolts.  In order to put times on a equal footing the\n response to be analyzed is the time to count 20 bolts (T20BOLT).\n Below are the data for 40 combinations of settings.  RUN is the order \n in which the data were collected.\n \n Analyze the data.  What adjustments have the greatest effect on the \n time to count 20 bolts?  How would you adjust the machine to get\n the shortest time to count 20 bolts?  Are there any unusual features\n to the data?\n \n The data description and data may be freely used for non-commercial\n purposes and can be freely distributed.  Copyright remains with the\n author and STATS Magazine.', 1),
-(108, '1', '**Author**: Andras Janosi, M.D.  \nDonor: David W. Aha (aha@ics.uci.edu)  \n**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/Heart+Disease) - July 1988  \n**Please cite**: The author request that any publications resulting from the use of the data include the name of the author.\n\n**Heart Disease Databases: Cleveland**  \nThis database contains 76 attributes, but all published experiments refer to using a subset of 14 of them.  In particular, the Cleveland database is the only one that has been used by ML researchers to this date.  The \"goal\" field refers to the presence of heart disease in the patient.  It is integer valued from 0 (no presence) to 4. Experiments with the Cleveland database have concentrated on simply attempting to distinguish presence (values 1,2,3,4) from absence (value 0).  \n    \nThe names and social security numbers of the patients were recently removed from the database, replaced with dummy values.\nOne file has been \"processed\", that one containing the Cleveland database. \n     \nAttribute documentation:  \n>\n       1 id: patient identification number\n       2 ccf: social security number (I replaced this with a dummy value of 0)\n       3 age: age in years\n       4 sex: sex (1 = male; 0 = female)\n       5 painloc: chest pain location (1 = substernal; 0 = otherwise)\n       6 painexer (1 = provoked by exertion; 0 = otherwise)\n       7 relrest (1 = relieved after rest; 0 = otherwise)\n       8 pncaden (sum of 5, 6, and 7)\n       9 cp: chest pain type\n         -- Value 1: typical angina\n         -- Value 2: atypical angina\n         -- Value 3: non-anginal pain\n         -- Value 4: asymptomatic\n      10 trestbps: resting blood pressure (in mm Hg on admission to the \n         hospital)\n      11 htn\n      12 chol: serum cholestoral in mg/dl\n      13 smoke: I believe this is 1 = yes; 0 = no (is or is not a smoker)\n      14 cigs (cigarettes per day)\n      15 years (number of years as a smoker)\n      16 fbs: (fasting blood sugar > 120 mg/dl)  (1 = true; 0 = false)\n      17 dm (1 = history of diabetes; 0 = no such history)\n      18 famhist: family history of coronary artery disease (1 = yes; 0 = no)\n      19 restecg: resting electrocardiographic results\n         -- Value 0: normal\n         -- Value 1: having ST-T wave abnormality (T wave inversions and/or ST \n                     elevation or depression of > 0.05 mV)\n         -- Value 2: showing probable or definite left ventricular hypertrophy\n                     by Estes\' criteria\n      20 ekgmo (month of exercise ECG reading)\n      21 ekgday(day of exercise ECG reading)\n      22 ekgyr (year of exercise ECG reading)\n      23 dig (digitalis used furing exercise ECG: 1 = yes; 0 = no)\n      24 prop (Beta blocker used during exercise ECG: 1 = yes; 0 = no)\n      25 nitr (nitrates used during exercise ECG: 1 = yes; 0 = no)\n      26 pro (calcium channel blocker used during exercise ECG: 1 = yes; 0 = no)\n      27 diuretic (diuretic used used during exercise ECG: 1 = yes; 0 = no)\n      28 proto: exercise protocol\n           1 = Bruce     \n           2 = Kottus\n           3 = McHenry\n           4 = fast Balke\n           5 = Balke\n           6 = Noughton \n           7 = bike 150 kpa min/min  (Not sure if \"kpa min/min\" is what was \n               written!)\n           8 = bike 125 kpa min/min  \n           9 = bike 100 kpa min/min\n          10 = bike 75 kpa min/min\n          11 = bike 50 kpa min/min\n          12 = arm ergometer\n      29 thaldur: duration of exercise test in minutes\n      30 thaltime: time when ST measure depression was noted\n      31 met: mets achieved\n      32 thalach: maximum heart rate achieved\n      33 thalrest: resting heart rate\n      34 tpeakbps: peak exercise blood pressure (first of 2 parts)\n      35 tpeakbpd: peak exercise blood pressure (second of 2 parts)\n      36 dummy\n      37 trestbpd: resting blood pressure\n      38 exang: exercise induced angina (1 = yes; 0 = no)\n      39 xhypo: (1 = yes; 0 = no)\n      40 oldpeak = ST depression induced by exercise relative to rest\n      41 slope: the slope of the peak exercise ST segment\n         -- Value 1: upsloping\n         -- Value 2: flat\n         -- Value 3: downsloping\n      42 rldv5: height at rest\n      43 rldv5e: height at peak exercise\n      44 ca: number of major vessels (0-3) colored by flourosopy\n      45 restckm: irrelevant\n      46 exerckm: irrelevant\n      47 restef: rest raidonuclid (sp?) ejection fraction\n      48 restwm: rest wall (sp?) motion abnormality\n         0 = none\n         1 = mild or moderate\n         2 = moderate or severe\n         3 = akinesis or dyskmem (sp?)\n      49 exeref: exercise radinalid (sp?) ejection fraction\n      50 exerwm: exercise wall (sp?) motion \n      51 thal: 3 = normal; 6 = fixed defect; 7 = reversable defect\n      52 thalsev: not used\n      53 thalpul: not used\n      54 earlobe: not used\n      55 cmo: month of cardiac cath (sp?)  (perhaps \"call\")\n      56 cday: day of cardiac cath (sp?)\n      57 cyr: year of cardiac cath (sp?)\n      58 num: diagnosis of heart disease (angiographic disease status)\n         -- Value 0: < 50% diameter narrowing\n         -- Value 1: > 50% diameter narrowing\n         (in any major vessel: attributes 59 through 68 are vessels)\n      59 lmt\n      60 ladprox\n      61 laddist\n      62 diag\n      63 cxmain\n      64 ramus\n      65 om1\n      66 om2\n      67 rcaprox\n      68 rcadist\n      69 lvx1: not used\n      70 lvx2: not used\n      71 lvx3: not used\n      72 lvx4: not used\n      73 lvf: not used\n      74 cathef: not used\n      75 junk: not used\n      76 name: last name of patient \n         (I replaced this with the dummy string \"name\")', 1),
-(109, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis data set consists of three types of entities:\n (a) the specification of an auto in terms of various characteristics;\n (b) its assigned insurance risk rating,;\n (c) its normalized losses in use as compared to other cars. \n The second rating corresponds to the degree to which the auto is more risky than its price indicates. Cars are initially\n assigned a risk factor symbol associated with its price. Then, if it is more risky (or less), this symbol is adjusted by\n moving it up (or down) the scale. Actuarians call this process \"symboling\". A value of +3 indicates that the auto is\n risky, -3 that it is probably pretty safe.The third factor is the relative average loss payment per insured vehicle year.\n This value is normalized for all autos within a particular size classification (two-door small, station wagons,\n sports/speciality, etc...), and represents the average loss per car per year.\n - Note: Several of the attributes in the database could be used as a \"class\" attribute.\n The original data (from the UCI repository) (http://www.ics.uci.edu/~mlearn/MLSummary.html) has 205 instances\n described by 26 attributes :\n - 15 continuous\n - 1 integer\n - 10 nominal\n The following provides more information on these attributes:\n \n   1. symboling:                 -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback,convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n The original data also has some missing attribute values denoted by \"?\" : \n \n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4\n \n I\'ve changed the original data in the following way :\n - All instances with unknowns were removed giving 159 instances.\n - The goal variable is \"price\"\n - All nominal attributes (10) were removed.\n \n Original source: UCI machine learning repository. (http://www.ics.uci.edu/~mlearn/MLSummary.html). \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 159 cases; 14 continuous variables; 1 nominal vars..', 1),
-(110, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identifier attribute deleted.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n\n 1. Title: Auto-Mpg Data\n \n 2. Sources:\n    (a) Origin:  This dataset was taken from the StatLib library which is\n                 maintained at Carnegie Mellon University. The dataset was \n                 used in the 1983 American Statistical Association Exposition.\n    (c) Date: July 7, 1993\n \n 3. Past Usage:\n     -  See 2b (above)\n     -  Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning.\n        In Proceedings on the Tenth International Conference of Machine \n        Learning, 236-243, University of Massachusetts, Amherst. Morgan\n        Kaufmann.\n \n 4. Relevant Information:\n \n    This dataset is a slightly modified version of the dataset provided in\n    the StatLib library.  In line with the use by Ross Quinlan (1993) in\n    predicting the attribute \"mpg\", 8 of the original instances were removed \n    because they had unknown values for the \"mpg\" attribute.  The original \n    dataset is available in the file \"auto-mpg.data-original\".\n \n    \"The data concerns city-cycle fuel consumption in miles per gallon,\n     to be predicted in terms of 3 multivalued discrete and 5 continuous\n     attributes.\" (Quinlan, 1993)\n \n 5. Number of Instances: 398\n \n 6. Number of Attributes: 9 including the class attribute\n \n 7. Attribute Information:\n \n     1. mpg:           continuous\n     2. cylinders:     multi-valued discrete\n     3. displacement:  continuous\n     4. horsepower:    continuous\n     5. weight:        continuous\n     6. acceleration:  continuous\n     7. model year:    multi-valued discrete\n     8. origin:        multi-valued discrete\n     9. car name:      string (unique for each instance)\n \n 8. Missing Attribute Values:  horsepower has 6 missing values', 1),
-(111, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe Computer Activity databases are a collection of computer systems\n activity measures. The data was collected from a Sun Sparcstation\n 20/712 with 128 Mbytes of memory running in a multi-user university\n department. Users would typically be doing a large variety of tasks\n ranging from accessing the internet, editing files or running very\n cpu-bound programs.  The data was collected continuously on two\n separate occasions. On both occassions, system activity was gathered\n every 5 seconds. The final dataset is taken from both occasions with\n equal numbers of observations coming from each collection epoch.\n \n System measures used:\n 1. lread - Reads (transfers per second ) between system memory and user memory.\n 2. lwrite - writes (transfers per second) between system memory and user memory.\n 3. scall - Number of system calls of all types per second.\n 4. sread - Number of system read calls per second.\n 5. swrite - Number of system write calls per second . \n 6. fork - Number of system fork calls per second. \n 7. exec - Number of system exec calls per second. \n 8. rchar - Number of characters transferred per second by system read calls.\n 9. wchar - Number of characters transfreed per second by system write calls. \n 10. pgout - Number of page out requests per second.\n 11. ppgout - Number of pages, paged out per second. \n 12. pgfree - Number of pages per second placed on the free list. \n 13. pgscan - Number of pages checked if they can be freed per second.\n 14. atch - Number of page attaches (satisfying a page fault by reclaiming a page in memory) per second.\n 15. pgin - Number of page-in requests per second.\n 16. ppgin - Number of pages paged in per second.\n 17. pflt - Number of page faults caused by protection errors (copy-on-writes). \n 18. vflt - Number of page faults caused by address translation. \n 19. runqsz - Process run queue size.\n 20. freemem - Number of memory pages available to user processes.\n 21. freeswap - Number of disk blocks available for page swapping. \n 22. usr - Portion of time (%) that cpus run in user mode.\n 23. sys - Portion of time (%) that cpus run in system mode.\n 24. wio - Portion of time (%) that cpus are idle waiting for block IO.\n 25. idle - Portion of time (%) that cpus are otherwise idle.\n \n The two different regression tasks obtained from these databases are:\n \n CompAct \n Predict usr, the portion of time that cpus run in user mode from all attributes 1-21.\n \n CompAct(s) \n Predict usr using a restricted number (excluding the paging information (10-18)\n \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original source: DELVE repository of data. \n Characteristics: 8192 cases, 22 continuous attributes', 1),
-(112, '1', '**Author**: Rui Camacho (rcamacho@garfield.fe.up.pt)  \n**Source**: [Regression datasets collection Luis Torgo](http://www.dcc.fc.up.pt/~ltorgo/Regression/DataSets.html)  \n**Please cite**:   \n\nThis data set is also obtained from the task of controlling the ailerons of a F16 aircraft, although the target variable and attributes are different from the ailerons domain. The target variable here is a variation instead of an absolute value, and there was some pre-selection of the attributes.', 1),
-(113, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identifier attribute deleted.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  Sexual activity and the lifespan of male fruitflies\n TYPE:  Designed (almost factorial) experiment\n SIZE:  125 observations, 5 variables\n \n DESCRIPTIVE ABSTRACT:\n A cost of increased reproduction in terms of reduced longevity has been\n shown for female fruitflies, but not for males.  The flies used were an\n outbred stock.  Sexual activity was manipulated by supplying individual\n males with one or eight receptive virgin females per day.  The\n longevity of these males was compared with that of two control types.\n The first control consisted of two sets of individual males kept with\n one or eight newly inseminated females.  Newly inseminated females will\n not usually remate for at least two days, and thus served as a control\n for any effect of competition with the male for food or space.  The\n second control was a set of individual males kept with no females.\n There were 25 males in each of the five groups, which were treated\n identically in number of anaesthetizations (using CO2) and provision of\n fresh food medium.\n \n SOURCE:\n Figure 2 in the article \"Sexual Activity and the Lifespan of Male\n Fruitflies\" by Linda Partridge and Marion Farquhar.  _Nature_, 294,\n 580-581, 1981.\n \n VARIABLE DESCRIPTIONS:\n Columns  Variable    Description\n -------  --------    -----------\n  1- 2    ID          Serial No. (1-25) within each group of 25\n                      (the order in which data points were abstracted)\n \n  4       PARTNERS    Number of companions (0, 1 or 8)\n \n  6       TYPE        Type of companion\n                        0: newly pregnant female\n                        1: virgin female\n                        9: not applicable (when PARTNERS=0)\n \n  8- 9    LONGEVITY   Lifespan, in days\n \n 11-14    THORAX      Length of thorax, in mm (x.xx)\n \n 16-17    SLEEP       Percentage of each day spent sleeping\n \n \n SPECIAL NOTES:\n `Compliance\' of the males in the two experimental groups was documented\n as follows:  On two days per week throughout the life of each\n experimental male, the females that had been supplied as virgins to\n that male were kept and examined for fertile eggs.  The insemination\n rate declined from approximately 7 females/day at age one week to just\n under 2/day at age eight weeks in the males supplied with eight virgin\n females per day, and from just under 1/day at age one week to\n approximately 0.6/day at age eight weeks in the males supplied with one\n virgin female per day.  These `compliance\' data were not supplied for\n individual males, but the authors say that \"There were no significant\n differences between the individual males within each experimental\n group.\"\n \n STORY BEHIND THE DATA:\n James Hanley found this dataset in _Nature_ and was attracted by the\n way the raw data were presented in classical analysis of covariance\n style in Figure 2.  He read the data points from the graphs and brought\n them to the attention of a colleague with whom he was teaching the\n applied statistics course.  Dr. Liddell thought that with only three\n explanatory variables (THORAX, plus PARTNERS and TYPE to describe the\n five groups), it would not be challenging enough as a data-analysis\n project.  He suggested adding another variable.  James Hanley added\n SLEEP, a variable not mentioned in the published article.  Teachers can\n contact us about the construction of this variable.  (We prefer to\n divulge the details at the end of the data-analysis project.)\n \n Further discussion of the background and pedagogical use of this\n dataset can be found in Hanley (1983) and in Hanley and Shapiro\n (1994).  To obtain the Hanley and Shapiro article, send the one-line\n e-mail message:\n send jse/v2n1/datasets.hanley\n to the address archive@jse.stat.ncsu.edu\n \n PEDAGOGICAL NOTES:\n This has been the most successful and the most memorable dataset we\n have used in an \"applications of statistics\" course, which we have\n taught for ten years.  The most common analysis techniques have been\n analysis of variance, classical analysis of covariance, and multiple\n regression.  Because the variable THORAX is so strong (it explains\n about 1/3 of the variance in LONGEVITY), it is important to consider it\n to increase the precision of between-group contrasts.  When students\n first check and find that the distributions of thorax length, and in\n particular, the mean thorax length, are very similar in the different\n groups, many of them are willing to say (in epidemiological\n terminology) that THORAX is not a confounding variable, and that it can\n be omitted from the analysis.\n \n There is usually lively discussion about the primary contrast.  The\n five groups and their special structure allow opportunities for\n students to understand and verbalize what we mean by the term\n \"statistical interaction.\"\n \n There is also much debate as to whether one should take the SLEEP\n variable into account.  Some students say that it is an `intermediate\'\n variable.  Some students formally test the mean level of SLEEP across\n groups, find one pair where there is a statistically significant\n difference, and want to treat it as a confounding variable.  A few\n students muse about how it was measured.\n \n There is heteroscedasticity in the LONGEVITY variable.\n \n One very observant student (now a professor) argued that THORAX cannot\n be used as a predictor or explanatory variable for the LONGEVITY\n outcome since fruitflies who die young may not be fully grown, i.e., it\n is also an intermediate variable.  One Ph.D. student who had studied\n entomology assured us that fruitflies do not grow longer after birth;\n therefore, the THORAX length is not time-dependent!\n \n Curiously, the dataset has seldom been analyzed using techniques from\n survival analysis.  The fact that there are no censored observations is\n not really an excuse, and one could easily devise a way to introduce\n censoring of LONGEVITY.\n \n REFERENCES:\n Hanley, J. A. (1983), \"Appropriate Uses of Multivariate Analysis,\"\n _Annual Review of Public Health_, 4, 155-180.\n \n Hanley, J. A., and Shapiro, S. H. (1994), \"Sexual Activity and the\n Lifespan of Male Fruitflies:  A Dataset That Gets Attention,\" _Journal\n of Statistics Education_, Volume 2, Number 1.\n \n SUBMITTED BY:\n James A. Hanley and Stanley H. Shapiro\n Department of Epidemiology and Biostatistics\n McGill University\n 1020 Pine Avenue West\n Montreal, Quebec, H3A 1A2\n Canada\n tel: +1 (514) 398-6270 (JH) \n      +1 (514) 398-6272 (SS)\n fax: +1 (514) 398-4503\n INJH@musicb.mcgill.ca, StanS@epid.lan.mcgill.ca', 1),
-(114, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Case number deleted. X treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  PBC Data\n SIZE:  418 observations, 20 variables\n \n \n \n DESCRIPTIVE ABSTRACT:\n \n Below is a description of the variables recorded from the Mayo Clinic trial \n in primary biliary cirrhosis (PBC) of the liver conducted between 1974 and \n 1984.  A total of 424 PBC patients, referred to Mayo Clinic during\n that ten-year interval, met eligibility criteria for the randomized placebo \n controlled trial of the drug D-penicillamine. The first 312 cases in the data \n set participated in the randomized trial, and contain largely complete data. \n The additional 112 cases did not participate in the clinical trial, but \n consented to have basic measurements recorded and to be followed for survival.\n Six of those cases were lost to follow-up shortly after diagnosis, so there \n are data here on an additional 106 cases as well as the 312 randomized \n participants. Missing data items are denoted by \".\".  At least one space \n separates each variable in the .DAT file.  Censoring was due to liver \n transplantation for twenty-five subjects with the following case numbers: \n 5, 105, 111, 120, 125, 158, 183, 241, 246, 247, 254, 263, 264, 265, 274, \n 288, 291, 295, 297, 345, 361, 362, 375, 380, 383.\n \n \n \n SOURCE:  Counting Processes and Survival Analysis by T. Fleming & \n          D. Harrington, (1991),  published by John Wiley & Sons.\n \n \n \n VARIABLE DESCRIPTIONS:\n \n The data are in free format.  That is, at least one blank space separates\n each variable.  The variables contained in the .DAT are:\n \n \n N:   Case number.\n X:   The number of days between registration and the earlier of\n      death, liver transplantation, or study analysis time in July, 1986.\n D:   1 if X is time to death, 0 if time to censoring\n Z1:  Treatment Code, 1 = D-penicillamine, 2 = placebo.\n Z2:  Age in years. For the first 312 cases, age was calculated by\n      dividing the number of days between birth and study registration by 365.\n Z3:  Sex, 0 = male, 1 = female.\n Z4:  Presence of ascites, 0 = no, 1 = yes.\n Z5:  Presence of hepatomegaly, 0 = no, 1 = yes.\n Z6:  Presence of spiders 0 = no, 1 = Yes.\n Z7:  Presence of edema, 0 = no edema and no diuretic therapy for\n      edema; 0.5 = edema present for which no diuretic therapy was given, or \n      edema resolved with diuretic therapy; 1 = edema despite diuretic therapy\n Z8:  Serum bilirubin, in mg/dl.\n Z9:  Serum cholesterol, in mg/dl.\n Z10: Albumin, in gm/dl.\n Z11: Urine copper, in mg/day.\n Z12: Alkaline phosphatase, in U/liter.\n Z13: SGOT, in U/ml.\n Z14: Triglycerides, in mg/dl.\n Z15: Platelet count; coded value is number of platelets\n      per-cubic-milliliter of blood divided by 1000.\n Z16: Prothrombin time, in seconds.\n Z17: Histologic stage of disease, graded 1, 2, 3, or 4.\n \n \n \n \n STORY BEHIND THE DATA:\n \n Between January, 1974 and May, 1984, the Mayo Clinic conducted a\n double-blinded randomized trial in primary biliary cirrhosis of the liver\n (PBC), comparing the drug D-penicillamine (DPCA) with a placebo. There\n were 424 patients who met the eligibility criteria seen at the Clinic while\n the trial was open for patient registration. Both the treating physician and\n the patient agreed to participate in the randomized trial in 312 of the 424\n cases. The date of randomization and a large number of clinical, biochemical,\n serologic, and histologic parameters were recorded for each of the 312\n clinical trial patients. The data from the trial were analyzed in 1986 for\n presentation in the clinical literature. For that analysis, disease and \n survival status as of July, 1986, were recorded for as many patients as \n possible.  By that date, 125 of the 312 patients had died, with only 11 \n not attributable to PBC.  Eight patients had been lost to follow up, and 19 \n had undergone liver transplantation. \n \n PBC is a rare but fatal chronic liver disease of unknown cause,\n with a prevalence of about 50-cases-per-million population. The primary\n pathologic event appears to be the destruction of interlobular bile ducts,\n which may be mediated by immunologic mechanisms. The data discussed here are\n important in two respects. First, controlled clinical trials are difficult to\n complete in rare diseases, and this case series of patients uniformly\n diagnosed, treated, and followed is the largest existing for PBC. The\n treatment comparison in this trial is more precise than in similar trials\n having fewer participants and avoids the bias that may arise in comparing\n a case series to historical controls. Second, the data present an\n opportunity to study the natural history of the disease. We will see that, \n despite the immunosuppressive properties of DPCA, there are no detectable\n differences between the distributions of survival times for the DPCA and\n placebo treatment groups. This suggests that these groups can be combined\n in studying the association between survival time from randomization and\n clinical and other measurements. In the early to mid 1980s, the rate of \n successful liver transplant increased substantially, and transplant has \n become an effective therapy for PBC. The Mayo Clinic data set is therefore \n one of the last allowing a study of the natural history of PBC in patients \n who were treated with only supportive care or its equivalent. The PBC data \n can be used to: estimate a survival distribution; test for differences \n between two groups; and estimate covariate effects via a regression\n model.', 1),
-(115, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis is a commercial application described in Weiss & Indurkhya (1995). \n The data describes a telecommunication problem. No further information\n is available.\n \n Characteristics: (10000+5000) cases, 49 continuous attributes \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original Source: The data in the original format can be obtained \n from http://www.cs.su.oz.au/~nitin', 1),
-(116, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Horsepower treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n 1. Title: 1985 Auto Imports Database\n \n 2. Source Information:\n    -- Creator/Donor: Jeffrey C. Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n    -- Date: 19 May 1987\n    -- Sources:\n      1) 1985 Model Import Car and Truck Specifications, 1985 Ward\'s\n         Automotive Yearbook.\n      2) Personal Auto Manuals, Insurance Services Office, 160 Water\n         Street, New York, NY 10038 \n      3) Insurance Collision Report, Insurance Institute for Highway\n         Safety, Watergate 600, Washington, DC 20037\n\n 3. Past Usage:\n    -- Kibler,~D., Aha,~D.~W., & Albert,~M. (1989).  Instance-based prediction\n       of real-valued attributes.  {it Computational Intelligence}, {it 5},\n       51--57.\n          -- Predicted price of car using all numeric and Boolean attributes\n          -- Method: an instance-based learning (IBL) algorithm derived from a\n             localized k-nearest neighbor algorithm.  Compared with a\n             linear regression prediction...so all instances\n             with missing attribute values were discarded.  This resulted with\n             a training set of 159 instances, which was also used as a test\n             set (minus the actual instance during testing).\n          -- Results: Percent Average Deviation Error of Prediction from Actual\n             -- 11.84% for the IBL algorithm\n             -- 14.12% for the resulting linear regression equation\n \n 4. Relevant Information:\n    -- Description\n       This data set consists of three types of entities: (a) the\n       specification of an auto in terms of various characteristics, (b)\n       its assigned insurance risk rating, (c) its normalized losses in use\n       as compared to other cars.  The second rating corresponds to the\n       degree to which the auto is more risky than its price indicates.\n       Cars are initially assigned a risk factor symbol associated with its\n       price.   Then, if it is more risky (or less), this symbol is\n       adjusted by moving it up (or down) the scale.  Actuarians call this\n       process \"symboling\".  A value of +3 indicates that the auto is\n       risky, -3 that it is probably pretty safe.\n \n       The third factor is the relative average loss payment per insured\n       vehicle year.  This value is normalized for all autos within a\n       particular size classification (two-door small, station wagons,\n       sports/speciality, etc...), and represents the average loss per car\n       per year.\n \n    -- Note: Several of the attributes in the database could be used as a\n             \"class\" attribute.\n \n 5. Number of Instances: 205\n \n 6. Number of Attributes: 26 total\n    -- 15 continuous\n    -- 1 integer\n    -- 10 nominal\n \n 7. Attribute Information:     \n      Attribute:                Attribute Range:\n      ------------------        -----------------------------------------------\n   1. symboling:                -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback, convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n 8. Missing Attribute Values: (denoted by \"?\")\n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4%', 1),
-(117, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Identification code deleted. \n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n NAME:  LOW BIRTH WEIGHT DATA\n KEYWORDS:  Logistic Regression\n SIZE:  189 observations, 11 variables\n \n NOTE:\n         These data come from Appendix 1 of Hosmer and Lemeshow (1989).\n These data are copyrighted and must be acknowledged and used accordingly.\n \n DESCRIPTIVE ABSTRACT:\n         The goal of this study was to identify risk factors associated with\n giving birth to a low birth weight baby (weighing less than 2500 grams).\n Data were collected on 189 women, 59 of which had low birth weight babies\n and 130 of which had normal birth weight babies.  Four variables which were\n thought to be of importance were age, weight of the subject at her last\n menstrual period, race, and the number of physician visits during the first\n trimester of pregnancy.\n \n \n SOURCE:\n          Data were collected at Baystate Medical Center, Springfield,\n Massachusetts, during 1986.\n \n \n NOTE:\n           This data set consists of the complete data.  A paired data set\n created from this low birth weight data may be found in plowbwt.dat and\n a 3 to 1 matched data set created from the low birth weight data may be\n found in mlowbwt.dat.\n \n \n \n Table:  Code Sheet for the Variables in the Low Birth Weight Data Set.\n \n Columns   Variable                                              Abbreviation\n -----------------------------------------------------------------------------\n 2-4     Identification Code                                     ID\n    \n 10      Low Birth Weight (0 = Birth Weight ge 2500g,            LOW\n                           l = Birth Weight < 2500g)\n   \n 17-18   Age of the Mother in Years                              AGE\n      \n 23-25   Weight in Pounds at the Last Menstrual Period           LWT\n      \n 32      Race (1 = White, 2 = Black, 3 = Other)                  RACE\n      \n 40      Smoking Status During Pregnancy (1 = Yes, 0 = No)       SMOKE\n      \n 48      History of Premature Labor (0 = None, 1 = One, etc.)    PTL\n      \n 55      History of Hypertension (1 = Yes, 0 = No)               HT\n      \n 61      Presence of Uterine Irritability (1 = Yes, 0 = No)      UI\n      \n 67      Number of Physician Visits During the First Trimester   FTV\n                 (0 = None, 1 = One, 2 = Two, etc.)\n      \n 73-76   Birth Weight in Grams                                   BWT\n -----------------------------------------------------------------------------\n \n PEDAGOGICAL NOTES:\n         These data have been used as an example of fitting a multiple\n logistic regression model.\n \n STORY BEHIND THE DATA:\n         Low birth weight is an outcome that has been of concern to physicians\n for years. This is due to the fact that infant mortality rates and birth\n defect rates are very high for low birth weight babies. A woman\'s behavior\n during pregnancy (including diet, smoking habits, and receiving prenatal care)\n can greatly alter the chances of carrying the baby to term and, consequently,\n of delivering a baby of normal birth weight.\n         The variables identified in the code sheet given in the table have been\n shown to be associated with low birth weight in the obstetrical literature. The\n goal of the current study was to ascertain if these variables were important\n in the population being served by the medical center where the data were\n collected.\n \n \n References:\n \n 1. Hosmer and Lemeshow, Applied Logistic Regression, Wiley, (1989).', 1),
-(118, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Cholesterol treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Publication Request: \n    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n    This file describes the contents of the heart-disease directory.\n \n    This directory contains 4 databases concerning heart disease diagnosis.\n    All attributes are numeric-valued.  The data was collected from the\n    four following locations:\n \n      1. Cleveland Clinic Foundation (cleveland.data)\n      2. Hungarian Institute of Cardiology, Budapest (hungarian.data)\n      3. V.A. Medical Center, Long Beach, CA (long-beach-va.data)\n      4. University Hospital, Zurich, Switzerland (switzerland.data)\n \n    Each database has the same instance format.  While the databases have 76\n    raw attributes, only 14 of them are actually used.  Thus I\'ve taken the\n    liberty of making 2 copies of each database: one with all the attributes\n    and 1 with the 14 attributes actually used in past experiments.\n \n    The authors of the databases have requested:\n \n       ...that any publications resulting from the use of the data include the \n       names of the principal investigator responsible for the data collection\n       at each institution.  They would be:\n \n        1. Hungarian Institute of Cardiology. Budapest: Andras Janosi, M.D.\n        2. University Hospital, Zurich, Switzerland: William Steinbrunn, M.D.\n        3. University Hospital, Basel, Switzerland: Matthias Pfisterer, M.D.\n        4. V.A. Medical Center, Long Beach and Cleveland Clinic Foundation:\n           Robert Detrano, M.D., Ph.D.\n \n    Thanks in advance for abiding by this request.\n \n    David Aha\n    July 22, 1988\n    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n \n 1. Title: Heart Disease Databases\n \n 2. Source Information:\n    (a) Creators: \n        -- 1. Hungarian Institute of Cardiology. Budapest: Andras Janosi, M.D.\n        -- 2. University Hospital, Zurich, Switzerland: William Steinbrunn, M.D.\n        -- 3. University Hospital, Basel, Switzerland: Matthias Pfisterer, M.D.\n        -- 4. V.A. Medical Center, Long Beach and Cleveland Clinic Foundation:\n              Robert Detrano, M.D., Ph.D.\n    (b) Donor: David W. Aha (aha@ics.uci.edu) (714) 856-8779   \n    (c) Date: July, 1988\n \n 3. Past Usage:\n     1. Detrano,~R., Janosi,~A., Steinbrunn,~W., Pfisterer,~M., Schmid,~J.,\n        Sandhu,~S., Guppy,~K., Lee,~S., & Froelicher,~V. (1989).  {it \n        International application of a new probability algorithm for the \n        diagnosis of coronary artery disease.}  {it American Journal of \n        Cardiology}, {it 64},304--310.\n        -- International Probability Analysis \n        -- Address: Robert Detrano, M.D.\n                    Cardiology 111-C\n                    V.A. Medical Center\n                    5901 E. 7th Street\n                    Long Beach, CA 90028\n        -- Results in percent accuracy: (for 0.5 probability threshold)\n              Data Name:  CDF    CADENZA\n           -- Hungarian   77     74\n              Long beach  79     77\n              Swiss       81     81\n           -- Approximately a 77% correct classification accuracy with a\n              logistic-regression-derived discriminant function\n     2. David W. Aha & Dennis Kibler\n        -- \n           \n           \n           -- Instance-based prediction of heart-disease presence with the \n              Cleveland database\n              -- NTgrowth: 77.0% accuracy\n              --       C4: 74.8% accuracy\n     3. John Gennari\n        -- Gennari, J.~H., Langley, P, & Fisher, D. (1989). Models of\n           incremental concept formation. {it Artificial Intelligence, 40},\n           11--61.\n        -- Results: \n           -- The CLASSIT conceptual clustering system achieved a 78.9% accuracy\n              on the Cleveland database.\n \n 4. Relevant Information:\n      This database contains 76 attributes, but all published experiments\n      refer to using a subset of 14 of them.  In particular, the Cleveland\n      database is the only one that has been used by ML researchers to \n      this date.  The \"goal\" field refers to the presence of heart disease\n      in the patient.  It is integer valued from 0 (no presence) to 4.\n      Experiments with the Cleveland database have concentrated on simply\n      attempting to distinguish presence (values 1,2,3,4) from absence (value\n      0).  \n    \n      The names and social security numbers of the patients were recently \n      removed from the database, replaced with dummy values.\n \n      One file has been \"processed\", that one containing the Cleveland \n      database.  All four unprocessed files also exist in this directory.\n     \n 5. Number of Instances: \n         Database:    # of instances:\n           Cleveland: 303\n           Hungarian: 294\n         Switzerland: 123\n       Long Beach VA: 200\n \n 6. Number of Attributes: 76 (including the predicted attribute)\n \n 7. Attribute Information:\n    -- Only 14 used\n       -- 1. #3  (age)       \n       -- 2. #4  (sex)       \n       -- 3. #9  (cp)        \n       -- 4. #10 (trestbps)  \n       -- 5. #12 (chol)      \n       -- 6. #16 (fbs)       \n       -- 7. #19 (restecg)   \n       -- 8. #32 (thalach)   \n       -- 9. #38 (exang)     \n       -- 10. #40 (oldpeak)   \n       -- 11. #41 (slope)     \n       -- 12. #44 (ca)        \n       -- 13. #51 (thal)      \n       -- 14. #58 (num)       (the predicted attribute)\n \n    -- Complete attribute documentation:\n       1 id: patient identification number\n       2 ccf: social security number (I replaced this with a dummy value of 0)\n       3 age: age in years\n       4 sex: sex (1 = male; 0 = female)\n       5 painloc: chest pain location (1 = substernal; 0 = otherwise)\n       6 painexer (1 = provoked by exertion; 0 = otherwise)\n       7 relrest (1 = relieved after rest; 0 = otherwise)\n       8 pncaden (sum of 5, 6, and 7)\n       9 cp: chest pain type\n         -- Value 1: typical angina\n         -- Value 2: atypical angina\n         -- Value 3: non-anginal pain\n         -- Value 4: asymptomatic\n      10 trestbps: resting blood pressure (in mm Hg on admission to the \n         hospital)\n      11 htn\n      12 chol: serum cholestoral in mg/dl\n      13 smoke: I believe this is 1 = yes; 0 = no (is or is not a smoker)\n      14 cigs (cigarettes per day)\n      15 years (number of years as a smoker)\n      16 fbs: (fasting blood sugar > 120 mg/dl)  (1 = true; 0 = false)\n      17 dm (1 = history of diabetes; 0 = no such history)\n      18 famhist: family history of coronary artery disease (1 = yes; 0 = no)\n      19 restecg: resting electrocardiographic results\n         -- Value 0: normal\n         -- Value 1: having ST-T wave abnormality (T wave inversions and/or ST \n                     elevation or depression of > 0.05 mV)\n         -- Value 2: showing probable or definite left ventricular hypertrophy\n                     by Estes\' criteria\n      20 ekgmo (month of exercise ECG reading)\n      21 ekgday(day of exercise ECG reading)\n      22 ekgyr (year of exercise ECG reading)\n      23 dig (digitalis used furing exercise ECG: 1 = yes; 0 = no)\n      24 prop (Beta blocker used during exercise ECG: 1 = yes; 0 = no)\n      25 nitr (nitrates used during exercise ECG: 1 = yes; 0 = no)\n      26 pro (calcium channel blocker used during exercise ECG: 1 = yes; 0 = no)\n      27 diuretic (diuretic used used during exercise ECG: 1 = yes; 0 = no)\n      28 proto: exercise protocol\n           1 = Bruce     \n           2 = Kottus\n           3 = McHenry\n           4 = fast Balke\n           5 = Balke\n           6 = Noughton \n           7 = bike 150 kpa min/min  (Not sure if \"kpa min/min\" is what was \n               written!)\n           8 = bike 125 kpa min/min  \n           9 = bike 100 kpa min/min\n          10 = bike 75 kpa min/min\n          11 = bike 50 kpa min/min\n          12 = arm ergometer\n      29 thaldur: duration of exercise test in minutes\n      30 thaltime: time when ST measure depression was noted\n      31 met: mets achieved\n      32 thalach: maximum heart rate achieved\n      33 thalrest: resting heart rate\n      34 tpeakbps: peak exercise blood pressure (first of 2 parts)\n      35 tpeakbpd: peak exercise blood pressure (second of 2 parts)\n      36 dummy\n      37 trestbpd: resting blood pressure\n      38 exang: exercise induced angina (1 = yes; 0 = no)\n      39 xhypo: (1 = yes; 0 = no)\n      40 oldpeak = ST depression induced by exercise relative to rest\n      41 slope: the slope of the peak exercise ST segment\n         -- Value 1: upsloping\n         -- Value 2: flat\n         -- Value 3: downsloping\n      42 rldv5: height at rest\n      43 rldv5e: height at peak exercise\n      44 ca: number of major vessels (0-3) colored by flourosopy\n      45 restckm: irrelevant\n      46 exerckm: irrelevant\n      47 restef: rest raidonuclid (sp?) ejection fraction\n      48 restwm: rest wall (sp?) motion abnormality\n         0 = none\n         1 = mild or moderate\n         2 = moderate or severe\n         3 = akinesis or dyskmem (sp?)\n      49 exeref: exercise radinalid (sp?) ejection fraction\n      50 exerwm: exercise wall (sp?) motion \n      51 thal: 3 = normal; 6 = fixed defect; 7 = reversable defect\n      52 thalsev: not used\n      53 thalpul: not used\n      54 earlobe: not used\n      55 cmo: month of cardiac cath (sp?)  (perhaps \"call\")\n      56 cday: day of cardiac cath (sp?)\n      57 cyr: year of cardiac cath (sp?)\n      58 num: diagnosis of heart disease (angiographic disease status)\n         -- Value 0: < 50% diameter narrowing\n         -- Value 1: > 50% diameter narrowing\n         (in any major vessel: attributes 59 through 68 are vessels)\n      59 lmt\n      60 ladprox\n      61 laddist\n      62 diag\n      63 cxmain\n      64 ramus\n      65 om1\n      66 om2\n      67 rcaprox\n      68 rcadist\n      69 lvx1: not used\n      70 lvx2: not used\n      71 lvx3: not used\n      72 lvx4: not used\n      73 lvf: not used\n      74 cathef: not used\n      75 junk: not used\n      76 name: last name of patient \n         (I replaced this with the dummy string \"name\")\n \n 9. Missing Attribute Values: Several.  Distinguished with value -9.0.\n \n 10. Class Distribution:\n         Database:      0   1   2   3   4 Total\n           Cleveland: 164  55  36  35  13   303\n           Hungarian: 188  37  26  28  15   294\n         Switzerland:   8  48  32  30   5   123\n       Long Beach VA:  51  56  41  42  10   200', 1),
-(119, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n Data from which conclusions  were  drawn  in  the  article  \"Sleep  in \n Mammals: Ecological and Constitutional Correlates\" by Allison, T.  and \n Cicchetti, D. (1976), _Science_, November 12, vol. 194,  pp.  732-734. \n Includes brain and body  weight,  life  span,  gestation  time,  time \n sleeping, and predation and danger indices for 62 mammals.\n \n \n \n Variables below (from left to right) for Mammals Data Set:\n \n species of animal\n \n body weight in kg\n \n brain weight in g\n \n slow wave (\"nondreaming\") sleep (hrs/day)\n \n paradoxical (\"dreaming\") sleep (hrs/day)\n \n total sleep (hrs/day)  (sum of slow wave and paradoxical sleep)\n \n maximum life span (years)\n \n gestation time (days)\n \n predation index (1-5)\n                 1 = minimum (least likely to be preyed upon)\n                 5 = maximum (most likely to be preyed upon)\n \n sleep exposure index (1-5)\n                 1 = least exposed (e.g. animal sleeps in a \n                     well-protected den)\n                 5 = most exposed\n \n overall danger index (1-5)\n                 (based on the above two indices and other information)\n                 1 = least danger (from other animals)\n                 5 = most danger (from other animals)\n \n Note: Missing values denoted by -999.0\n \n \n For more details, see\n \n Allison, Truett and Cicchetti, Domenic V. (1976), \"Sleep  in  Mammals: \n Ecological and Constitutional  Correlates\",  _Science_,  November  12, \n vol. 194, pp. 732-734.\n \n The above data set can be freely used for non-commercial purposes  and \n can be freely distributed (permission in  writing  obtained  from  Dr. \n Truett Allison).\n \n Submitted by Roger Johnson\n rwjohnso@silver.sdsmt.edu\n\n Total sleep treated as the class attribute. Attributes for slow\n wave and paradoxical sleep have been deleted. (The animal\'s\n name has also been deleted.)', 1),
-(120, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThe problem is to learn a regression equation/rule/tree to predict the\n activity from the descriptive structural attributes.  The data and\n methodology is described in detail in: - King, Ross .D., Hurst,\n Jonathan. D., and Sternberg, Michael.J.E. A comparison of artificial\n intelligence methods for modelling QSARs Applied Artificial\n Intelligence, 1994 (in press).  - Hurst, Jonathan. D., King, Ross\n .D. and Sternberg, Michael.J.E. Quantitative Structure-Activity\n Relationships by neural networks and inductive logic programming:\n 2. The inhibition of dihydrofolate reductase by triazines. Journal of\n Computer Aided Molecular Design, 1994 (in press).\n \n Original source: ?. \n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Characteristics: 186 cases; 61 continuous variables', 1),
-(121, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n All nominal attributes and instances with missing values are deleted.\n Price treated as the class attribute.\n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n 1. Title: 1985 Auto Imports Database\n \n 2. Source Information:\n    -- Creator/Donor: Jeffrey C. Schlimmer (Jeffrey.Schlimmer@a.gp.cs.cmu.edu)\n    -- Date: 19 May 1987\n    -- Sources:\n      1) 1985 Model Import Car and Truck Specifications, 1985 Ward\'s\n         Automotive Yearbook.\n      2) Personal Auto Manuals, Insurance Services Office, 160 Water\n         Street, New York, NY 10038 \n      3) Insurance Collision Report, Insurance Institute for Highway\n         Safety, Watergate 600, Washington, DC 20037\n\n 3. Past Usage:\n    -- Kibler,~D., Aha,~D.~W., & Albert,~M. (1989).  Instance-based prediction\n       of real-valued attributes.  {it Computational Intelligence}, {it 5},\n       51--57.\n          -- Predicted price of car using all numeric and Boolean attributes\n          -- Method: an instance-based learning (IBL) algorithm derived from a\n             localized k-nearest neighbor algorithm.  Compared with a\n             linear regression prediction...so all instances\n             with missing attribute values were discarded.  This resulted with\n             a training set of 159 instances, which was also used as a test\n             set (minus the actual instance during testing).\n          -- Results: Percent Average Deviation Error of Prediction from Actual\n             -- 11.84% for the IBL algorithm\n             -- 14.12% for the resulting linear regression equation\n \n 4. Relevant Information:\n    -- Description\n       This data set consists of three types of entities: (a) the\n       specification of an auto in terms of various characteristics, (b)\n       its assigned insurance risk rating, (c) its normalized losses in use\n       as compared to other cars.  The second rating corresponds to the\n       degree to which the auto is more risky than its price indicates.\n       Cars are initially assigned a risk factor symbol associated with its\n       price.   Then, if it is more risky (or less), this symbol is\n       adjusted by moving it up (or down) the scale.  Actuarians call this\n       process \"symboling\".  A value of +3 indicates that the auto is\n       risky, -3 that it is probably pretty safe.\n \n       The third factor is the relative average loss payment per insured\n       vehicle year.  This value is normalized for all autos within a\n       particular size classification (two-door small, station wagons,\n       sports/speciality, etc...), and represents the average loss per car\n       per year.\n \n    -- Note: Several of the attributes in the database could be used as a\n             \"class\" attribute.\n \n 5. Number of Instances: 205\n \n 6. Number of Attributes: 26 total\n    -- 15 continuous\n    -- 1 integer\n    -- 10 nominal\n \n 7. Attribute Information:     \n      Attribute:                Attribute Range:\n      ------------------        -----------------------------------------------\n   1. symboling:                -3, -2, -1, 0, 1, 2, 3.\n   2. normalized-losses:        continuous from 65 to 256.\n   3. make:                     alfa-romero, audi, bmw, chevrolet, dodge, honda,\n                                isuzu, jaguar, mazda, mercedes-benz, mercury,\n                                mitsubishi, nissan, peugot, plymouth, porsche,\n                                renault, saab, subaru, toyota, volkswagen, volvo\n   4. fuel-type:                diesel, gas.\n   5. aspiration:               std, turbo.\n   6. num-of-doors:             four, two.\n   7. body-style:               hardtop, wagon, sedan, hatchback, convertible.\n   8. drive-wheels:             4wd, fwd, rwd.\n   9. engine-location:          front, rear.\n  10. wheel-base:               continuous from 86.6 120.9.\n  11. length:                   continuous from 141.1 to 208.1.\n  12. width:                    continuous from 60.3 to 72.3.\n  13. height:                   continuous from 47.8 to 59.8.\n  14. curb-weight:              continuous from 1488 to 4066.\n  15. engine-type:              dohc, dohcv, l, ohc, ohcf, ohcv, rotor.\n  16. num-of-cylinders:         eight, five, four, six, three, twelve, two.\n  17. engine-size:              continuous from 61 to 326.\n  18. fuel-system:              1bbl, 2bbl, 4bbl, idi, mfi, mpfi, spdi, spfi.\n  19. bore:                     continuous from 2.54 to 3.94.\n  20. stroke:                   continuous from 2.07 to 4.17.\n  21. compression-ratio:        continuous from 7 to 23.\n  22. horsepower:               continuous from 48 to 288.\n  23. peak-rpm:                 continuous from 4150 to 6600.\n  24. city-mpg:                 continuous from 13 to 49.\n  25. highway-mpg:              continuous from 16 to 54.\n  26. price:                    continuous from 5118 to 45400.\n \n 8. Missing Attribute Values: (denoted by \"?\")\n    Attribute #:   Number of instances missing a value:\n    2.             41\n    6.             2\n    19.            4\n    20.            4\n    22.            2\n    23.            2\n    26.            4%', 1),
-(122, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n This is the data set called `DETROIT\' in the book `Subset selection in\n regression\' by Alan J. Miller published in the Chapman & Hall series of\n monographs on Statistics & Applied Probability, no. 40.   The data are\n unusual in that a subset of three predictors can be found which gives a\n very much better fit to the data than the subsets found from the Efroymson\n stepwise algorithm, or from forward selection or backward elimination.\n \n The original data were given in appendix A of `Regression analysis and its\n application: A data-oriented approach\' by Gunst & Mason, Statistics\n textbooks and monographs no. 24, Marcel Dekker.   It has caused problems\n because some copies of the Gunst & Mason book do not contain all of the data,\n and because Miller does not say which variables he used as predictors and\n which is the dependent variable.   (HOM was the dependent variable, and the\n predictors were FTP ... WE)\n \n The data were collected by J.C. Fisher and used in his paper: \"Homicide in\n Detroit: The Role of Firearms\", Criminology, vol.14, 387-400 (1976)\n \n \n The data are on the homicide rate in Detroit for the years 1961-1973.\n FTP    - Full-time police per 100,000 population\n UEMP   - %  unemployed in the population\n MAN    - number of manufacturing workers in thousands\n LIC    - Number of handgun licences per 100,000 population\n GR     - Number of handgun registrations per 100,000 population\n CLEAR  - %  homicides cleared by arrests\n WM     - Number of white males in the population\n NMAN   - Number of non-manufacturing workers in thousands\n GOV    - Number of government workers in thousands\n HE     - Average hourly earnings\n WE     - Average weekly earnings\n \n HOM    - Number of homicides per 100,000 of population\n ACC    - Death rate in accidents per 100,000 population\n ASR    - Number of assaults per 100,000 population\n \n N.B. Each case takes two lines.', 1),
-(123, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nDataset from Smoothing Methods in Statistics \n (ftp stat.cmu.edu/datasets)\n\n Simonoff, J.S. (1996). Smoothing Methods in Statistics. New York: Springer-Verlag.', 1),
-(124, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n These data are those collected in a cloud-seeding experiment in Tasmania\n between mid-1964 and January 1971.   Their analysis, using regression\n techniques and permutation tests, is discussed in:\n \n       Miller, A.J., Shaw, D.E., Veitch, L.G. & Smith, E.J. (1979).\n       `Analyzing the results of a cloud-seeding experiment in Tasmania\',\n       Communications in Statistics - Theory & Methods, vol.A8(10),\n       1017-1047.\n \n The rainfalls are period rainfalls in inches.   TE and TW are the east and\n west target areas respectively, while NC, SC and NWC are the corresponding\n rainfalls in the north, south and north-west control areas respectively.\n S = seeded, U = unseeded.\n\n Rain in eastern target region is being treated\n as the class attribute. (Attribute for rain\n in the western target region has been deleted.)', 1),
-(125, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nData from StatLib (ftp stat.cmu.edu/datasets)\n\n The infamous Longley data, \"An appraisal of least-squares programs from\n  the point of view of the user\", JASA, 62(1967) p819-841.\n\n Variables are: Number of people employed   (usually the y variable)\n                GNP implicit price deflator\n                GNP\n                Unemployed\n                Armed forces\n                Non-institutionalized population >=14 years of age\n                Year\n\n Employment is being treated as the class\n attribute.', 1),
-(126, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\nThis data set concerns the study of the factors affecting patterns of\n insulin-dependent diabetes mellitus in children.  The objective is to\n investigate the dependence of the level of serum C-peptide on the\n various other factors in order to understand the patterns of residual\n insulin secretion. The response measurement is the logarithm of\n C-peptide concentration (pmol/ml) at the diagnosis, and the predictor\n measurements age and base deficit, a measure of acidity.\n\n Source: collection of regression datasets by Luis Torgo (ltorgo@ncc.up.pt) at\n http://www.ncc.up.pt/~ltorgo/Regression/DataSets.html\n Original source: Book Generalized Additive Models (p.304) by Hastie &\n Tibshirani, Chapman & Hall.  \n Characteristics: 43 cases; 3 continuous variables', 1),
-(127, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Case number deleted. \n\n As used by Kilpatrick, D. & Cameron-Jones, M. (1998). Numeric prediction\n using instance-based learning with encoding length selection. In Progress\n in Connectionist-Based Information Systems. Singapore: Springer-Verlag.\n\n !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n\n Name:  Pharynx (A clinical Trial in the Trt. of Carcinoma of the Oropharynx).\n SIZE:  195 observations, 13 variables.\n \n \n \n DESCRIPTIVE ABSTRACT:\n \n The .dat file gives the data for a part of a large clinical trial\n carried out by the Radiation Therapy Oncology Group in the United States. \n The full study included patients with squamous carcinoma of 15 sites in \n the mouth and throat, with 16 participating institutions, though only data \n on three sites in the oropharynx reported by the six largest institutions \n are considered here. Patients entering the study were randomly assigned to \n one of two treatment groups, radiation therapy alone or radiation therapy \n together with a chemotherapeutic agent.  One objective of the study was to \n compare the two treatment policies with respect to patient survival.\n \n \n \n SOURCE:  The Statistical Analysis of Failure Time Data, by JD Kalbfleisch\n          & RL Prentice, (1980),  Published by John Wiley & Sons \n \n \n \n VARIABLE DESCRIPTIONS:\n \n The data are in free format.  That is, at least one blank space separates \n each variable in the .dat file.  The variables are as follows:\n \n \n Case:         Case Number\n Inst:         Participating Institution\n sex:          1=male, 2=female\n Treatment:    1=standard, 2=test\n Grade:        1=well differentiated, 2=moderately differentiated, \n               3=poorly differentiated,  9=missing\n Age:          In years at time of diagnosis\n Condition:    1=no disability, 2=restricted work, 3=requires assistance with\n               self care, 4=bed confined,  9=missing\n Site:         1=faucial arch, 2=tonsillar fossa, 3=posterior pillar,\n               4=pharyngeal tongue, 5=posterior wall\n T staging:    1=primary tumor measuring 2 cm or less in largest diameter,\n               2=primary tumor measuring 2 cm to 4 cm in largest diameter with\n               minimal infiltration in depth, 3=primary tumor measuring more \n               than 4 cm, 4=massive invasive tumor\n N staging:    0=no clinical evidence of node metastases, 1=single positive\n               node 3 cm or less in diameter, not fixed, 2=single positive\n               node more than 3 cm in diameter, not fixed, 3=multiple\n               positive nodes or fixed positive nodes \n Entry Date:   Date of study entry: Day of year and year\n Status:       0=censored,  1=dead\n Time:         Survival time in days from day of diagnosis \n \n \n \n \n \n \n STORY BEHIND THE DATA:\n \n Approximately 30% of the survival times are censored owing primarily to \n patients surviving to the time of analysis. Some patients were lost\n to follow-up because the patient moved or transferred to an institution not\n participating in the study, though these cases were relatively rare. From \n a statistical point of view, an important feature of these data is the \n considerable lack of homogeneity between individuals being studied. \n Of course, as part of the study design, certain criteria for patient \n eligibility had to be met which eliminated extremes in the extent of disease, \n but still many factors are not controlled.\n \n This study included measurements of many covariates which would be expected \n to relate to survival experience. Six such variables are given in the data \n (sex, T staging, N staging, age, general condition, and grade).   The site \n of the primary tumor and possible differences between participating \n institutions require consideration as well.\n      \n The T,N staging classification gives a measure of the extent of the tumor at \n the primary site and at regional lymph nodes. T=1, refers to a small primary \n tumor, 2 centimeters or less in largest diameter, whereas T=4 is a massive \n tumor with extension to adjoining tissue. T=2 and T=3 refer to intermediate\n cases. N=0 refers to there being no clinical evidence of a lymph node \n metastasis and N=1, N=2, N=3 indicate, in increasing magnitude, the extent of \n existing lymph node involvement. Patients with classifications T=1,N=0; \n T=1,N=1;  T=2,N=0; or T=2,N=1, or with distant metastases were excluded \n from study.\n \n The variable general condition gives a measure of the functional capacity of \n the patient at the time of diagnosis (1 refers to no disability whereas\n 4 denotes bed confinement; 2 and 3 measure intermediate levels). The variable\n grade is a measure of the degree of differentiation of the tumor (the degree\n to which the tumor cell resembles the host cell) from 1 (well differentiated) \n to 3 (poorly differentiated)\n \n In addition to the primary question whether the combined treatment mode is\n preferable to the conventional radiation therapy, it is of considerable \n interest to determine the extent to which the several covariates relate to\n subsequent survival.  It is also imperative in answering the primary question \n to adjust the survivals for possible imbalance that may be present in the \n study with regard to the other covariates. Such problems are similar to those \n encountered in the classical theory of linear regression and the analysis of \n covariance.  Again, the need to accommodate censoring is an important \n distinguishing point. In many situations it is also important to develop \n nonparametric and robust procedures since there is frequently little empirical\n or theoretical work to support a particular family of failure time \n distributions.', 1),
-(128, '1', 'This data sets consists of 3 different types of irises\' (Setosa, Versicolour, and Virginica) petal and sepal length, stored in a 150x4 numpy.ndarray', 1),
-(129, '1', '', 1),
-(130, '1', '', 16);
+(1, '1', '**Author**:   \n**Source**: Unknown -   \n**Please cite**:   \n\n1. Title: Chess End-Game -- King+Rook versus King+Pawn on a7\n    (usually abbreviated KRKPA7).  The pawn on a7 means it is one square\n    away from queening.  It is the King+Rook\'s side (white) to move.\n \n 2. Sources:\n     (a) Database originally generated and described by Alen Shapiro.\n     (b) Donor/Coder: Rob Holte (holte@uottawa.bitnet).  The database\n         was supplied to Holte by Peter Clark of the Turing Institute\n         in Glasgow (pete@turing.ac.uk).\n     (c) Date: 1 August 1989\n \n 3. Past Usage:\n      - Alen D. Shapiro (1983,1987), \"Structured Induction in Expert Systems\",\n        Addison-Wesley.  This book is based on Shapiro\'s Ph.D. thesis (1983)\n        at the University of Edinburgh entitled \"The Role of Structured\n        Induction in Expert Systems\".\n      - Stephen Muggleton (1987), \"Structuring Knowledge by Asking Questions\",\n        pp.218-229 in \"Progress in Machine Learning\", edited by I. Bratko\n        and Nada Lavrac, Sigma Press, Wilmslow, England  SK9 5BB.\n      - Robert C. Holte, Liane Acker, and Bruce W. Porter (1989),\n        \"Concept Learning and the Problem of Small Disjuncts\",\n        Proceedings of IJCAI.  Also available as technical report AI89-106,\n        Computer Sciences Department, University of Texas at Austin,\n        Austin, Texas 78712.\n \n 4. Relevant Information:\n       The dataset format is described below.  Note: the format of this\n     database was modified on 2/26/90 to conform with the format of all\n     the other databases in the UCI repository of machine learning databases.\n \n 5. Number of Instances: 3196 total\n \n 6. Number of Attributes: 36\n \n 7. Attribute Summaries:\n     Classes (2):  -- White-can-win (\"won\") and White-cannot-win (\"nowin\").\n           I believe that White is deemed to be unable to win if the Black pawn\n           can safely advance.\n     Attributes: see Shapiro\'s book.\n \n 8. Missing Attributes: --  none\n \n 9. Class Distribution:\n     In 1669 of the positions (52%), White can win.\n     In 1527 of the positions (48%), White cannot win.\n \n The format for instances in this database is a sequence of 37 attribute values.\n Each instance is a board-descriptions for this chess endgame.  The first\n 36 attributes describe the board.  The last (37th) attribute is the\n classification: \"win\" or \"nowin\".  There are 0 missing values.\n A typical board-description is\n \n f,f,f,f,f,f,f,f,f,f,f,f,l,f,n,f,f,t,f,f,f,f,f,f,f,t,f,f,f,f,f,f,f,t,t,n,won\n \n The names of the features do not appear in the board-descriptions.\n Instead, each feature correponds to a particular position in the\n feature-value list.  For example, the head of this list is the value\n for the feature \"bkblk\".  The following is the list of features, in\n the order in which their values appear in the feature-value list:\n \n [bkblk,bknwy,bkon8,bkona,bkspr,bkxbq,bkxcr,bkxwp,blxwp,bxqsq,cntxt,dsopp,dwipd,\n  hdchk,katri,mulch,qxmsq,r2ar8,reskd,reskr,rimmx,rkxwp,rxmsq,simpl,skach,skewr,\n  skrxp,spcop,stlmt,thrsk,wkcti,wkna8,wknck,wkovl,wkpos,wtoeg]\n \n In the file, there is one instance (board position) per line.\n \n \n Num Instances:     3196\n Num Attributes:    37\n Num Continuous:    0 (Int 0 / Real 0)\n Num Discrete:      37\n Missing values:    0 /  0.0%\n\n     name                      type enum ints real     missing    distinct  (1)\n   1 \'bkblk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   2 \'bknwy\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   3 \'bkon8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   4 \'bkona\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   5 \'bkspr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   6 \'bkxbq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   7 \'bkxcr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   8 \'bkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n   9 \'blxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  10 \'bxqsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  11 \'cntxt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  12 \'dsopp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  13 \'dwipd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  14 \'hdchk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  15 \'katri\'                   Enum 100%   0%   0%     0 /  0%     3 /  0%   0% \n  16 \'mulch\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  17 \'qxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  18 \'r2ar8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  19 \'reskd\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  20 \'reskr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  21 \'rimmx\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  22 \'rkxwp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  23 \'rxmsq\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  24 \'simpl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  25 \'skach\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  26 \'skewr\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  27 \'skrxp\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  28 \'spcop\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  29 \'stlmt\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  30 \'thrsk\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  31 \'wkcti\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  32 \'wkna8\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  33 \'wknck\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  34 \'wkovl\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  35 \'wkpos\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  36 \'wtoeg\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0% \n  37 \'class\'                   Enum 100%   0%   0%     0 /  0%     2 /  0%   0%', 1);
 
diff --git a/openml_OS/config/BASE_CONFIG-BLANK_ENV.php b/openml_OS/config/BASE_CONFIG-BLANK_ENV.php
new file mode 100644
index 000000000..4714059e8
--- /dev/null
+++ b/openml_OS/config/BASE_CONFIG-BLANK_ENV.php
@@ -0,0 +1,205 @@
+<?php
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+*  After editing this file, rename to BASE_CONFIG.php
+* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Sets the local time to a specific zone. Can be removed, if it is
+ * set in php.ini.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+setlocale(LC_TIME, 'nl_NL');
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * BASE_URL: The domain of the instance, along with the folder where
+ * it can be found. Also with a tailing slash. For the main instance
+ * of OpenML, this would be http://www.openml.org/. For an instance on
+ * a localhost, this would be http://localhost/.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'BASE_URL', $_ENV["BASE_URL"] );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DATA_URL: The subdirectory where the data can be accessed. Will be
+ * deprecated in a few updates.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'DATA_URL', BASE_URL . 'data/' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * MINIO_URL: MINIO Server URL. Currently this is 'http://openml1.win.tue.nl/'.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'MINIO_URL', 'fill_in' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * PATH: The directory on the hard disk where the instance of OpenML
+ * can be found, with tailing slash. Typically, this would be
+ * something like /var/www/ (on Ubuntu installations)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'PATH', $_ENV["DIRECTORY_PATH"] );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DATA_PATH: The directory on whether the uploaded data is stored.
+ * It might be a good idea to locate this directory outside of the
+ * webserver directory. Should be writeable for webserver.
+ * IMPORTANT NOTE: If you choose to change the value indicated below
+ * (PATH . 'data/') into something different, make sure to also move
+ * the content of the data directory that was provided at installation.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+// define( 'DATA_PATH', $_ENV["DATA_PATH"]PATH . 'data/' );
+define( 'DATA_PATH', $_ENV["DATA_PATH"] );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * TMP_PATH: Directory in which temporary files can be places. LOCK
+ * files are an example of these kind of files. /tmp/ is generally a
+ * good place on Unix based systems.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'TMP_PATH', '/tmp/' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Paths for server-side scripts
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'LIB_PATH', '/var/lib/' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Configuration details for the experiment database
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_NAME: The name of the database
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_HOST: The host of the database (typically: localhost)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_USER: The username to connect to the database. IMPORTANT: This
+ * account is used by the free query interface. This username should
+ * only contain READ (SELECT, DESCRIBE) access to the database
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_PASS: The password that belongs to the username
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_USER_WRITING: A user account that can also WRITE (SELECT,
+ * UPDATE, INSERT, DELETE) to the database. Administrative privileges
+ * are discouraged.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_PASS_WRITING: The password that belongs to the DB_USER_WRITING
+ * account.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'DB_NAME_EXPDB', $_ENV["DB_NAME_EXPDB"] );
+// define( 'DB_HOST_EXPDB', 'mysql' );
+define( 'DB_HOST_EXPDB', $_ENV["DB_HOST_EXPDB"] );
+define( 'DB_USER_EXPDB_READ', $_ENV["DB_USER_EXPDB_READ"] );
+define( 'DB_PASS_EXPDB_READ', $_ENV["DB_PASS_EXPDB_READ"] );
+define( 'DB_USER_EXPDB_WRITE', $_ENV["DB_USER_EXPDB_WRITE"] );
+define( 'DB_PASS_EXPDB_WRITE', $_ENV["DB_PASS_EXPDB_WRITE"] );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Configuration details for the OpenML database (User accounts, etc.)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_NAME_COMMUNITY: The name of the database
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_HOST_COMMUNITY: The host of the database (typically: localhost)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_USER_COMMUNITY: The username to connect to the database. Should
+ * have SELECT, INSERT, UPDATE and DELETE privileges.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * DB_PASS: The password that belongs to the username
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'DB_NAME_OPENML', $_ENV["DB_NAME_OPENML"] );
+// define( 'DB_HOST_OPENML', 'mysql' );
+define( 'DB_HOST_OPENML', $_ENV["DB_HOST_OPENML"] );
+define( 'DB_USER_OPENML', $_ENV["DB_USER_OPENML"] );
+define( 'DB_PASS_OPENML', $_ENV["DB_PASS_OPENML"] );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Configuration details for the OpenML API (Username, password)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * API_USERNAME: The username under which the system may perform API
+ * calls.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * API_KEY: The API_KEY that belongs to the username
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'API_USERNAME', 'FILL_IN_USERNAME' );
+define( 'API_KEY', 'FILL_IN_KEY' );
+
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Configuration details for the ElasticSearch server (Username, password)
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * ES_USERNAME and ES_PASSWORD can be set by ElasticSearch or a protected proxy
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+// define( 'ES_URL', 'elasticsearch:9200' );
+define( 'ES_URL', $_ENV["ES_URL"] );
+// define( 'ES_PUBLIC_URL', 'localhost:9200' );
+define( 'ES_PUBLIC_URL', $_ENV["ES_PUBLIC_URL"] );
+define( 'ES_USERNAME', '' );
+define( 'ES_PASSWORD', '' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Configuration details for the Wiki server (Username, password)
+ * We use the Gollem Wiki Engine
+ * WIKI_USERNAME and WIKI_PASSWORD can be set by a protected proxy
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'WIKI_URL', 'localhost:443/wiki' );
+define( 'WIKI_USERNAME', 'username' );
+define( 'WIKI_PASSWORD', 'password' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * * * * * *
+ * DEBUG: Will produce errors and warnings on the screen. Set this
+ * to true when developing. Set to false on the production server
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'DEBUG', TRUE );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * * * * * *
+ * DEBUG_XSD_EMAIL: Will send debug information for failed XSD
+ * validations to the user who initiated it
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'DEBUG_XSD_EMAIL', FALSE );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * CMD_PREFIX: Adds a prefix to all shell commands; needed for bugfix
+ * on some MAMP (OSX) systems. Can be left empty on all other systems.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'CMD_PREFIX', '');
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EMAIL_FROM: The email address from which the emails will be send.
+ * Use the domain name specified in the BASE_URL
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'EMAIL_FROM', 'noreply@openml.org' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EMAIL_API_LOG: The email address to which critical API errors
+ * get reported
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'EMAIL_API_LOG', 'api@openml.org' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL API SETTINGS
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL_API_PROXY: Set this to true if you want to use a proxy
+ * for communication with external proxies. In general, you don't
+ * need this (and can skip this block by setting this to false).
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL_API_PROXY_PORT: The port of the proxy
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL_API_PROXY_URL: The URL of the proxy
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL_API_PASSWORD: The password of the proxy
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'EXTERNAL_API_PROXY', FALSE );
+define( 'EXTERNAL_API_PROXY_PORT', '' );
+define( 'EXTERNAL_API_PROXY_URL', '' );
+define( 'EXTERNAL_API_PASSWORD', '' );
+
+/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * EXTERNAL API KEYS - The public and private keys of the external api's
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
+ * Typically, an external api provides us with a public keyb
+ * (FACEBOOK_APP_ID, TWITTER_CONSUMER_KEY, GOOGLE_CLIENT_ID) and a
+ * private key (FACEBOOK_APP_SECRET, TWITTER_CONSUMER_SECRET,
+ * GOOGLE_CLIENT_SECRET). These are to be filled in here.
+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
+define( 'FACEBOOK_APP_ID', 'xxx' );
+define( 'FACEBOOK_APP_SECRET', 'xxx' );
+define( 'TWITTER_CONSUMER_KEY', 'xxx' );
+define( 'TWITTER_CONSUMER_SECRET', 'xxx' );
+define( 'GOOGLE_CLIENT_ID', 'xxx.apps.googleusercontent.com' );
+define( 'GOOGLE_CLIENT_SECRET', 'xxx' );
+?>
diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index eaea9c58e..c8e10dc46 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -197,7 +197,7 @@ function move_run_files($start_index, $end_index) {
   }
 
   function install_database() {
-    echo "\r\nInstall database...";
+    echo "\r\nInstall database...\n";
     // note that this one does not come from DATA folder, as they are stored in github
     $models = directory_map('data/sql/', 1);
     $manipulated_order = array('file.sql', 'groups.sql', 'users.sql', 'implementation.sql', 'algorithm_setup.sql', 
@@ -210,7 +210,7 @@ function install_database() {
       }
     }
     $models = array_unique($models);
-
+    
     foreach ($models as $m) {
       $modelname = ucfirst(substr($m, 0, strpos($m, '.')));
       if ($this->load->is_model_loaded($modelname) == false) { $this->load->model($modelname); }

From afccf80ab2c42de985cab622c1d7e577a83a8bf8 Mon Sep 17 00:00:00 2001
From: Joost van der Gaag <joost@smartfarmsensing.com>
Date: Tue, 25 Oct 2022 10:15:51 +0200
Subject: [PATCH 14/15] Temporarily remove MinIO reference to datasets as this
 is not yet implemented in the docker image

---
 openml_OS/config/BASE_CONFIG-BLANK_ENV.php |  4 +--
 openml_OS/models/api/v1/Api_data.php       | 30 +++++++++++++++++-----
 2 files changed, 25 insertions(+), 9 deletions(-)

diff --git a/openml_OS/config/BASE_CONFIG-BLANK_ENV.php b/openml_OS/config/BASE_CONFIG-BLANK_ENV.php
index 4714059e8..7abb95448 100644
--- a/openml_OS/config/BASE_CONFIG-BLANK_ENV.php
+++ b/openml_OS/config/BASE_CONFIG-BLANK_ENV.php
@@ -113,8 +113,8 @@
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
  * API_KEY: The API_KEY that belongs to the username
  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
-define( 'API_USERNAME', 'FILL_IN_USERNAME' );
-define( 'API_KEY', 'FILL_IN_KEY' );
+define( 'API_USERNAME', $_ENV["API_USERNAME"] );
+define( 'API_KEY', $_ENV["API_KEY"] );
 
 
 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
diff --git a/openml_OS/models/api/v1/Api_data.php b/openml_OS/models/api/v1/Api_data.php
index 6e686b2ff..041ee2f02 100644
--- a/openml_OS/models/api/v1/Api_data.php
+++ b/openml_OS/models/api/v1/Api_data.php
@@ -690,49 +690,62 @@ private function data_description_list($data_id) {
    *)
    */
   private function data($data_id) {
+    // echo "check if data ID is provided \n";
     if( $data_id == false ) {
       $this->returnError( 110, $this->version );
       return;
     }
-
+    // echo "[Success] check if data ID is provided \n";
+    // echo "Get dataset by ID \n";
     $dataset = $this->Dataset->getById( $data_id );
     if( $dataset === false ) {
       $this->returnError( 111, $this->version );
       return;
     }
-
+    // echo "[Success] Get dataset by ID \n";
+    // echo "Check dataset visibility or if user is the uploader \n";
     if($dataset->visibility != 'public' &&
        $dataset->uploader != $this->user_id &&
        !$this->user_has_admin_rights) {
       $this->returnError( 112, $this->version );
       return;
     }
-
+    // echo "[Success] Check dataset visibility or if user is the uploader \n";
+    // echo "Override URL field \n";
     // overwrite url field
     if ($dataset->file_id != NULL) {
       $dataset->url = BASE_URL . 'data/v1/download/' . $dataset->file_id . '/' . htmlspecialchars($dataset->name) . '.' . strtolower($dataset->format);
     }
+    // echo "[Success] Override URL field \n";
 
-
+    // echo "Get file \n";
     $file = $this->File->getById($dataset->file_id);
     if (!$file) {
       $this->returnError(113, $this->version);
       return;
     }
+    // echo "[Success] get file \n";
 
     $dataset->md5_checksum = $file->md5_hash;
 
+    // echo "Get tags \n";
     $tags = $this->Dataset_tag->getColumnWhere('tag', 'id = ' . $dataset->did);
     $dataset->tag = $tags != false ? '"' . implode( '","', $tags ) . '"' : array();
+    // echo "[Success] Get tags \n";
 
+    // echo "Get description \n";
     $description_record = $this->Dataset_description->getWhereSingle('did =' . $data_id, 'version DESC');
     $dataset->description_version = $description_record->version;
     $dataset->description = $description_record->description;
+    // echo "[Success] Get description \n";
 
+    // echo "Get csv \n";
     foreach( $this->xml_fields_dataset['csv'] as $field ) {
       $dataset->{$field} = getcsv( $dataset->{$field} );
     }
+    // echo "[Success] Get csv \n";
 
+    // echo "Get data processed \n";
     $data_processed = $this->Data_processed->getById(array($data_id, $this->config->item('default_evaluation_engine_id')));
     $relevant_fields = array('processing_date', 'error', 'warning');
     foreach ($relevant_fields as $field) {
@@ -742,15 +755,18 @@ private function data($data_id) {
         $dataset->{$field} = null;
       }
     }
+    // echo "[Success] Get data processed \n";
 
+    // echo "Get status \n";
     $dataset->status = $this->config->item('default_dataset_status');
     $data_status = $this->Dataset_status->getWhereSingle('did =' . $data_id, 'status_date DESC');
     if ($data_status != false) {
       $dataset->status = $data_status->status;
     }
-    if ($dataset->format != 'Sparse_ARFF') {
-      $dataset->minio_url = 'http://openml1.win.tue.nl/dataset' . $data_id . '/dataset_' . $data_id . '.pq';
-    }
+    // echo "[Success] Get status \n";
+    // if ($dataset->format != 'Sparse_ARFF') {
+    //   $dataset->minio_url = 'http://openml1.win.tue.nl/dataset' . $data_id . '/dataset_' . $data_id . '.pq';
+    // }
       $this->xmlContents( 'data-get', $this->version, $dataset );
   }
 

From 749198369263bd5922a3fd8cdb551b17ddf34f0a Mon Sep 17 00:00:00 2001
From: Joost van der Gaag <joost@smartfarmsensing.com>
Date: Fri, 13 Jan 2023 08:50:24 +0100
Subject: [PATCH 15/15] Small update

---
 openml_OS/controllers/Cron.php        |  173 ++--
 openml_OS/libraries/ElasticSearch.php | 1319 +++++++++++++------------
 2 files changed, 805 insertions(+), 687 deletions(-)

diff --git a/openml_OS/controllers/Cron.php b/openml_OS/controllers/Cron.php
index c8e10dc46..5dc7daf0e 100644
--- a/openml_OS/controllers/Cron.php
+++ b/openml_OS/controllers/Cron.php
@@ -1,11 +1,13 @@
 <?php
-class Cron extends CI_Controller {
+class Cron extends CI_Controller
+{
 
-  function __construct() {
+  function __construct()
+  {
     parent::__construct();
 
-    $this->controller = strtolower(get_class ($this));
-    if(!$this->input->is_cli_request()) {
+    $this->controller = strtolower(get_class($this));
+    if (!$this->input->is_cli_request()) {
       die('Cron Controller can only be accessed by CLI. ');
     }
 
@@ -21,7 +23,7 @@ function __construct() {
     $this->load->helper('arff');
 
     $this->load->library('email');
-    $this->email->from( EMAIL_FROM, 'The OpenML Team');
+    $this->email->from(EMAIL_FROM, 'The OpenML Team');
 
     $this->load->model('Algorithm_setup');
     $this->load->model('File');
@@ -45,55 +47,59 @@ function __construct() {
   }
 
   // indexes a single es item, unless $id = false
-  public function index($type, $id = false){
-      $time_start = microtime(true);
+  public function index($type, $id = false)
+  {
+    $time_start = microtime(true);
 
-      if(!$id){
-        echo "\r\n Starting ".$type." indexer... ";
-        echo $this->elasticsearch->index($type);
-      } else {
-        echo "\r\n Starting ".$type." indexer for id ".$id." ";
-        echo $this->elasticsearch->index($type, $id);
-      }
+    if (!$id) {
+      echo "\r\n Starting " . $type . " indexer... ";
+      echo $this->elasticsearch->index($type);
+    } else {
+      echo "\r\n Starting " . $type . " indexer for id " . $id . " ";
+      echo $this->elasticsearch->index($type, $id);
+    }
 
-      $time_end = microtime(true);
-      $time = $time_end - $time_start;
-      echo "\nIndexing done in $time seconds\n";
+    $time_end = microtime(true);
+    $time = $time_end - $time_start;
+    echo "\nIndexing done in $time seconds\n";
   }
 
   // indices a range of es items, starting by $id (or 0 if id == false)
-  public function indexfrom($type, $id = false, $verbosity = 0){
-      if(!$id){
-        // TODO: I guess this function enables the exact same hevaiour as index($type, false) (JvR)
-        echo "\r\n Starting ".$type." indexer... ";
-        echo $this->elasticsearch->index($type);
-      } else {
-        echo "\r\n Starting ".$type." indexer from id ".$id."... ";
-        echo $this->elasticsearch->index_from($type, $id, $verbosity);
-      }
+  public function indexfrom($type, $id = false, $verbosity = 0)
+  {
+    if (!$id) {
+      // TODO: I guess this function enables the exact same hevaiour as index($type, false) (JvR)
+      echo "\r\n Starting " . $type . " indexer... ";
+      echo $this->elasticsearch->index($type);
+    } else {
+      echo "\r\n Starting " . $type . " indexer from id " . $id . "... ";
+      echo $this->elasticsearch->index_from($type, $id, $verbosity);
+    }
   }
 
-  public function update_tag($type, $id){
-      echo "tagging ".$type." ".$id;
-      $this->elasticsearch->update_tags($type, $id);
+  public function update_tag($type, $id)
+  {
+    echo "tagging " . $type . " " . $id;
+    $this->elasticsearch->update_tags($type, $id);
   }
 
 
   // Create admin user
-  public function create_local_admin() {
+  public function create_local_admin()
+  {
     echo "\r\Fetching local admin user from user table...";
- 
-    $record = $this->Users->getWhereSingle('id = ' . 1);  
-    $username= $record->username;
+
+    $record = $this->Users->getWhereSingle('id = ' . 1);
+    $username = $record->username;
     $session_hash = $record->session_hash;
     // $session_hash=md5(rand());
-    $password = "admin";    
+    $password = "admin";
     // $this->Users->update(1, array('session_hash' => $session_hash));
-    echo "\r\nUser Name:\t" .$username; 
+    echo "\r\nUser Name:\t" . $username;
     echo "\r\nPassword:\t" . $password;
-    echo "\r\nAPI key:\t" .$session_hash;
+    echo "\r\nAPI key:\t" . $session_hash;
     echo "\r\nSaving key in shared config...config/api_key.txt";
-    file_put_contents("/openmlconfig/api_key.txt",$session_hash);     
+    file_put_contents("/openmlconfig/api_key.txt", $session_hash);
     echo "\r\nDone! ";
     // $user_data = array(
     //   'first_name' => "Admin",
@@ -107,50 +113,53 @@ public function create_local_admin() {
     //   'created_on' => time(),
     //   'active' => 1
     // );
-   
-     // # Add to admin group
+
+    // # Add to admin group
     // $group = array('1');
     // $adminId=$this->ion_auth->register($username, $password, $email,$additional_data, $group);
   }
 
-  
+
   // builds all es indexes
-  public function build_es_indices() {
+  public function build_es_indices()
+  {
     echo "\r\nBuild_es_indices...";
-    foreach($this->es_indices as $index) {
+    foreach ($this->es_indices as $index) {
       $this->elasticsearch->initialize_index($index);
     }
-    foreach($this->es_indices as $index) {
+    foreach ($this->es_indices as $index) {
       $this->indexfrom($index, 1);
     }
     echo "\r\nDone!";
   }
-  
-  public function arff_parses($file_id) {
+
+  public function arff_parses($file_id)
+  {
     $file = $this->File->getById($file_id);
     if ($file === false) {
       die('File does not exists.');
     }
-    
+
     $ext = strtolower($file->extension);
     if ($ext != 'arff' && $ext != 'sparse_arff') {
       die('File does not have an arff extension.');
     }
-    
+
     if ($file->type == 'url') {
       $result = ARFFcheck($file->filepath, 100);
     } else {
       $result = ARFFcheck(DATA_PATH . $file->filepath, 100);
     }
-    
+
     if ($result === TRUE) {
       die('Valid arff. ');
     } else {
       die($result);
     }
   }
-  
-  function missing_file_report() {
+
+  function missing_file_report()
+  {
     $start_time = now();
     $batch_size = 10000;
     $missing_files = array();
@@ -166,22 +175,23 @@ function missing_file_report() {
         }
       }
     }
-    
+
     $to = EMAIL_API_LOG;
     $subject = 'OpenML Cronjob report - Missing files';
     $content = 'Start time: ' . $start_time . "\nFinish time: " . now() . "\nServer: " . BASE_URL . "\nMissing files from records with the following ID's: " . implode(', ', $missing_files);
     sendEmail($to, $subject, $content, 'text');
   }
-  
+
   // temp 
-  function move_run_files($start_index, $end_index) {
+  function move_run_files($start_index, $end_index)
+  {
     $this->load->model('Runfile');
     $results = $this->Runfile->getWhere("source >= " . $start_index . ' AND source < ' . $end_index);
     $run_path = 'run_structured/';
-    
+
     foreach ($results as $result) {
       $file = $this->File->getById($result->file_id);
-      
+
       if (substr($file->filepath, 0, strlen($run_path)) === $run_path) {
         continue;
       }
@@ -196,12 +206,15 @@ function move_run_files($start_index, $end_index) {
     }
   }
 
-  function install_database() {
+  function install_database()
+  {
     echo "\r\nInstall database...\n";
     // note that this one does not come from DATA folder, as they are stored in github
     $models = directory_map('data/sql/', 1);
-    $manipulated_order = array('file.sql', 'groups.sql', 'users.sql', 'implementation.sql', 'algorithm_setup.sql', 
-      'dataset.sql', 'task_type.sql', 'task.sql', 'study.sql');
+    $manipulated_order = array(
+      'file.sql', 'groups.sql', 'users.sql', 'implementation.sql', 'algorithm_setup.sql',
+      'dataset.sql', 'task_type.sql', 'task.sql', 'study.sql'
+    );
 
     // moves elements of $manipulated_order to the start of the models array
     foreach (array_reverse($manipulated_order) as $name) {
@@ -210,17 +223,19 @@ function install_database() {
       }
     }
     $models = array_unique($models);
-    
+
     foreach ($models as $m) {
       $modelname = ucfirst(substr($m, 0, strpos($m, '.')));
-      if ($this->load->is_model_loaded($modelname) == false) { $this->load->model($modelname); }
+      if ($this->load->is_model_loaded($modelname) == false) {
+        $this->load->model($modelname);
+      }
       if ($this->$modelname->get() === false) {
         $sql = file_get_contents('data/sql/' . $m);
-        
+
         echo 'inserting ' . $modelname . ', with ' . strlen($sql) . ' characters... ' . "\n";
         // might need to adapt this, because not all models are supposed to write
         $result = $this->$modelname->query($sql);
-        
+
         if ($result === false) {
           echo 'failure: ' . $this->$modelname->mysqlErrorMessage() . "\n";
         }
@@ -232,7 +247,8 @@ function install_database() {
   }
 
   // Runs all local env init steps
-  public function init_local_env() {
+  public function init_local_env()
+  {
     $this->install_database();
     // $this->create_es_openml_index();
     // $this->initialize_es_indices();
@@ -240,7 +256,8 @@ public function init_local_env() {
     $this->build_es_indices();
   }
 
-  function create_meta_dataset($id = false) {
+  function create_meta_dataset($id = false)
+  {
     if ($id == false) {
       $meta_dataset = $this->Meta_dataset->getWhere('processed IS NULL');
       echo 'No id specified, requesting first dataset in queue.' . "\n";
@@ -252,7 +269,7 @@ function create_meta_dataset($id = false) {
     if ($meta_dataset) {
       $meta_dataset = $meta_dataset[0];
       echo 'Processing meta-dataset with id ' . $meta_dataset->id . ".\n";
-      $this->Meta_dataset->update( $meta_dataset->id, array('processed' => now()));
+      $this->Meta_dataset->update($meta_dataset->id, array('processed' => now()));
       $dataset_constr = ($meta_dataset->datasets) ? 'AND d.did IN (' . $meta_dataset->datasets . ') ' : '';
       $task_constr = ($meta_dataset->tasks) ? 'AND t.task_id IN (' . $meta_dataset->tasks . ') ' : '';
       $flow_constr = ($meta_dataset->flows) ? 'AND i.id IN (' . $meta_dataset->flows . ') ' : '';
@@ -268,12 +285,12 @@ function create_meta_dataset($id = false) {
         $evaluation_keys = array('e.repeat', 'e.fold', 'e.sample', 'e.sample_size', 'm.name');
         $evaluation_column = 'evaluation_sample';
       }
-      
+
       if (file_exists(DATA_PATH . $this->dir_suffix) == false) {
         mkdir(DATA_PATH . $this->dir_suffix, $this->config->item('content_directories_mode'), true);
       }
 
-      $tmp_path = '/tmp/' . rand_string( 20 ) . '.csv';
+      $tmp_path = '/tmp/' . rand_string(20) . '.csv';
 
       if ($meta_dataset->type == 'qualities') {
         $quality_keys_string = '';
@@ -285,14 +302,14 @@ function create_meta_dataset($id = false) {
         $header = '"data_id","task_id","name","quality",' . $quality_keys_key_string . '"value" ' . "\n";
         $sql =
           'SELECT d.did, t.task_id, d.name, q.quality, ' . $quality_keys_string . 'q.value ' .
-          'FROM dataset d, '.$quality_colum.' q, task t, task_inputs i ' .
+          'FROM dataset d, ' . $quality_colum . ' q, task t, task_inputs i ' .
           'WHERE t.task_id = i.task_id ' .
           'AND i.input = "source_data" ' .
           'AND i.value = q.data AND q.evaluation_engine_id = 1 ' .
           'AND d.did = q.data ' .
           'AND t.ttid = "' . $meta_dataset->task_type . '" ' .
           $dataset_constr . $task_constr .
-          'INTO OUTFILE "'. $tmp_path .'" ' .
+          'INTO OUTFILE "' . $tmp_path . '" ' .
           'FIELDS TERMINATED BY "," ' .
           'ENCLOSED BY "\"" ' .
           'LINES TERMINATED BY "\n" ' .
@@ -304,8 +321,8 @@ function create_meta_dataset($id = false) {
           'SELECT MAX(r.rid) AS run_id, s.sid AS setup_id, t.task_id AS task_id, ' .
           implode(', ', $evaluation_keys) . ', MAX(e.value) ' .
           ', CONCAT("Task_", t.task_id, "_", MAX(d.name)) AS task_name, MAX(i.fullName) as flow_name ' .
-//          ',s.setup_string ' .
-//          ',CONCAT(i.fullName, " on ", d.name) as textual ' .
+          //          ',s.setup_string ' .
+          //          ',CONCAT(i.fullName, " on ", d.name) as textual ' .
           'FROM run r, task t, task_inputs v, dataset d, algorithm_setup s, implementation i, ' . $evaluation_column . ' e, math_function m ' .
           'WHERE r.task_id = t.task_id AND v.task_id = t.task_id  ' .
           'AND v.input = "source_data" AND v.value = d.did ' .
@@ -314,9 +331,9 @@ function create_meta_dataset($id = false) {
           'AND e.function_id = m.id AND e.evaluation_engine_id = 1 ' .
           'AND t.ttid = "' . $meta_dataset->task_type . '"' .
           $dataset_constr . $task_constr . $flow_constr . $setup_constr . $function_constr .
-           /* the GROUP BY line makes stuff slower, we might want to comment it out. */
+          /* the GROUP BY line makes stuff slower, we might want to comment it out. */
           'GROUP BY r.setup, r.task_id, ' . implode(',', $evaluation_keys) . ' ' .
-          'INTO OUTFILE "' . $tmp_path .'" ' .
+          'INTO OUTFILE "' . $tmp_path . '" ' .
           'FIELDS TERMINATED BY "," ' .
           'ENCLOSED BY "\"" ' .
           'LINES TERMINATED BY "\n" ' .
@@ -331,7 +348,7 @@ function create_meta_dataset($id = false) {
           'AND is.setup = s.sid ' .
           'AND s.implementation_id = i.id ' .
           $flow_constr . $setup_constr .
-          'INTO OUTFILE "'. $tmp_path .'" ' .
+          'INTO OUTFILE "' . $tmp_path . '" ' .
           'FIELDS TERMINATED BY "," ' .
           'ENCLOSED BY "\"" ' .
           'LINES TERMINATED BY "\n" ' .
@@ -370,7 +387,7 @@ function create_meta_dataset($id = false) {
 
       $file_id = $this->File->register_created_file($this->dir_suffix, $filename, $meta_dataset->user_id, 'dataset', 'text/csv', 'private');
 
-      $this->Meta_dataset->update($meta_dataset->id, array( 'file_id' => $file_id));
+      $this->Meta_dataset->update($meta_dataset->id, array('file_id' => $file_id));
 
       $user = $this->ion_auth->user($meta_dataset->user_id)->row();
       $this->email->to($user->email);
@@ -378,11 +395,12 @@ function create_meta_dataset($id = false) {
       $this->email->message("This is an automatically generated email. The your requested meta-dataset was created successfully and can be downloaded from " . BASE_URL);
       $this->email->send();
     } else {
-      echo 'No meta-dataset to process. '."\n";
+      echo 'No meta-dataset to process. ' . "\n";
     }
   }
 
-  private function _error_meta_dataset($id, $msg, $user_id) {
+  private function _error_meta_dataset($id, $msg, $user_id)
+  {
     echo $msg . "\n";
     $this->Meta_dataset->update($id, array('error_message' => $msg));
 
@@ -394,4 +412,3 @@ private function _error_meta_dataset($id, $msg, $user_id) {
     $this->email->send();
   }
 }
-?>
diff --git a/openml_OS/libraries/ElasticSearch.php b/openml_OS/libraries/ElasticSearch.php
index f3d73f81b..fa797f18e 100644
--- a/openml_OS/libraries/ElasticSearch.php
+++ b/openml_OS/libraries/ElasticSearch.php
@@ -3,9 +3,11 @@
 if (!defined('BASEPATH'))
     exit('No direct script access allowed');
 
-class ElasticSearch {
+class ElasticSearch
+{
 
-    public function __construct() {
+    public function __construct()
+    {
         $this->CI = &get_instance();
         $this->CI->load->model('Dataset');
         $this->CI->load->model('Author');
@@ -33,7 +35,8 @@ public function __construct() {
         $this->init_indexer = False;
     }
 
-    public function initialize(){
+    public function initialize()
+    {
         $this->data_names = $this->CI->Dataset->getAssociativeArray('did', 'name', 'name IS NOT NULL');
         $this->flow_names = $this->CI->Implementation->getAssociativeArray('id', 'fullName', 'name IS NOT NULL');
         $this->procedure_names = $this->CI->Estimation_procedure->getAssociativeArray('id', 'name', 'name IS NOT NULL');
@@ -48,7 +51,8 @@ public function initialize(){
             'properties' => array(
                 'time' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss')
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                )
             )
         );
 
@@ -56,7 +60,8 @@ public function initialize(){
             'properties' => array(
                 'time' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss')
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                )
             )
         );
 
@@ -64,7 +69,8 @@ public function initialize(){
             'properties' => array(
                 'time' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss')
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                )
             )
         );
 
@@ -72,7 +78,8 @@ public function initialize(){
             'properties' => array(
                 'time' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss')
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                )
             )
         );
 
@@ -81,10 +88,12 @@ public function initialize(){
                 'date' => array(
                     'type' => 'date',
                     'format' => 'yyyy-MM-dd HH:mm:ss',
-		    'fields' => array(
-            		'keyword' => array(
-			   'type' => 'keyword'))
-		),
+                    'fields' => array(
+                        'keyword' => array(
+                            'type' => 'keyword'
+                        )
+                    )
+                ),
                 'uploader' => array(
                     'type' => 'text',
                     'analyzer' => 'keyword'
@@ -96,28 +105,36 @@ public function initialize(){
                     'type' => 'nested',
                     'properties' => array(
                         'tag' => array('type' => 'text'),
-                        'uploader' => array('type' => 'text'))),
+                        'uploader' => array('type' => 'text')
+                    )
+                ),
                 'last_update' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss'),
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                ),
                 'description' => array(
                     'type' => 'text',
-                    'analyzer' => 'snowball'),
+                    'analyzer' => 'snowball'
+                ),
                 'data_id' => array('type' => 'long'),
                 'version' => array('type' => 'float'),
                 'name' => array(
                     'type' => 'text',
-                    'analyzer' => 'snowball'),
+                    'analyzer' => 'snowball'
+                ),
                 'visibility' => array(
-                    'type' => 'keyword'),
+                    'type' => 'keyword'
+                ),
                 'format' => array(
-                    'type' => 'keyword'),
+                    'type' => 'keyword'
+                ),
                 'suggest' => array(
                     'type' => 'completion',
-                    'analyzer' => 'standard')
+                    'analyzer' => 'standard'
+                )
             )
         );
-	$this->mappings['flow'] = array(
+        $this->mappings['flow'] = array(
             'properties' => array(
                 'date' => array(
                     'type' => 'date',
@@ -130,7 +147,9 @@ public function initialize(){
                     'type' => 'nested',
                     'properties' => array(
                         'tag' => array('type' => 'text'),
-                        'uploader' => array('type' => 'text'))),
+                        'uploader' => array('type' => 'text')
+                    )
+                ),
                 'flow_id' => array('type' => 'long'),
                 'version' => array('type' => 'float'),
                 'last_update' => array(
@@ -159,7 +178,7 @@ public function initialize(){
                 )
             )
         );
-	$this->mappings['user'] = array(
+        $this->mappings['user'] = array(
             'properties' => array(
                 'date' => array(
                     'type' => 'date',
@@ -175,11 +194,12 @@ public function initialize(){
                 )
             )
         );
-	$this->mappings['task'] = array(
+        $this->mappings['task'] = array(
             'properties' => array(
                 'date' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss'),
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                ),
                 'suggest' => array(
                     'type' => 'completion',
                     'analyzer' => 'standard'
@@ -188,18 +208,22 @@ public function initialize(){
                     'type' => 'nested',
                     'properties' => array(
                         'tag' => array('type' => 'text'),
-                        'uploader' => array('type' => 'text'))),
+                        'uploader' => array('type' => 'text')
+                    )
+                ),
                 'collections' => array(
                     'type' => 'nested',
                     'properties' => array(
                         'id' => array('type' => 'long'),
-                        'type' => array('type' => 'text'))),
+                        'type' => array('type' => 'text')
+                    )
+                ),
                 'task_id' => array('type' => 'long'),
                 'tasktype.tt_id' => array('type' => 'long'),
-		        'runs' => array('type' => 'long')
+                'runs' => array('type' => 'long')
             )
         );
-	$this->mappings['task_type'] = array(
+        $this->mappings['task_type'] = array(
             'properties' => array(
                 'description' => array(
                     'type' => 'text',
@@ -211,14 +235,15 @@ public function initialize(){
                 ),
                 'date' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss'),
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                ),
                 'suggest' => array(
                     'type' => 'completion',
                     'analyzer' => 'standard'
                 )
             )
         );
-	$this->mappings['run'] = array(
+        $this->mappings['run'] = array(
             'properties' => array(
                 'date' => array(
                     'type' => 'date',
@@ -228,12 +253,16 @@ public function initialize(){
                     'type' => 'nested',
                     'properties' => array(
                         'tag' => array('type' => 'text'),
-                        'uploader' => array('type' => 'text'))),
+                        'uploader' => array('type' => 'text')
+                    )
+                ),
                 'collections' => array(
                     'type' => 'nested',
                     'properties' => array(
                         'id' => array('type' => 'long'),
-                        'type' => array('type' => 'text'))),
+                        'type' => array('type' => 'text')
+                    )
+                ),
                 'run_id' => array('type' => 'long'),
                 'run_flow.flow_id' => array('type' => 'long'),
                 'run_flow.name'  => array('type' => 'text', 'fielddata' => true),
@@ -254,13 +283,15 @@ public function initialize(){
                 )
             )
         );
-	$this->mappings['study'] = array(
+        $this->mappings['study'] = array(
             'properties' => array(
                 'name' => array(
                     'type' => 'text',
-                    'analyzer' => 'snowball'),
+                    'analyzer' => 'snowball'
+                ),
                 'visibility' => array(
-                    'type' => 'keyword'),
+                    'type' => 'keyword'
+                ),
                 'date' => array(
                     'type' => 'date',
                     'format' => 'yyyy-MM-dd HH:mm:ss'
@@ -281,7 +312,7 @@ public function initialize(){
             )
         );
 
-	$this->mappings['measure'] = array(
+        $this->mappings['measure'] = array(
             'properties' => array(
                 'description' => array(
                     'type' => 'text',
@@ -297,40 +328,44 @@ public function initialize(){
                 ),
                 'date' => array(
                     'type' => 'date',
-                    'format' => 'yyyy-MM-dd HH:mm:ss'),
+                    'format' => 'yyyy-MM-dd HH:mm:ss'
+                ),
                 'suggest' => array(
                     'type' => 'completion',
                     'analyzer' => 'standard'
                 )
             )
-  );
-  $this->init_indexer = True;
+        );
+        $this->init_indexer = True;
     }
 
-    public function test() {
+    public function test()
+    {
         return $this->client->ping();
     }
 
-    public function get_types() {
+    public function get_types()
+    {
         $params['index'] = '_all';
-	$array_data = $this->client->indices()->getMapping($params);
+        $array_data = $this->client->indices()->getMapping($params);
         return array_keys($array_data);
     }
 
-    public function index($type, $id = false, $altmetrics=True, $verbosity=0) {
+    public function index($type, $id = false, $altmetrics = True, $verbosity = 0)
+    {
         //bootstrap
         $indexParams['index'] = $type;
-        if(! $this->client->indices()->getMapping($indexParams))
-          echo $this->initialize_index($type);
-        elseif (! $this->init_indexer)
-          $this->initialize();
+        if (!$this->client->indices()->getMapping($indexParams))
+            echo $this->initialize_index($type);
+        elseif (!$this->init_indexer)
+            $this->initialize();
 
         $method_name = 'index_' . $type;
         if (method_exists($this, $method_name)) {
             try {
                 return $this->$method_name($id, $altmetrics, $verbosity);
-	    } catch (Exception $e) {
-		echo $e->getMessage();
+            } catch (Exception $e) {
+                echo $e->getMessage();
                 // TODO: log?
             }
         } else {
@@ -338,20 +373,21 @@ public function index($type, $id = false, $altmetrics=True, $verbosity=0) {
         }
     }
 
-    public function index_from($type, $id = false, $verbosity=1, $altmetrics=True) {
+    public function index_from($type, $id = false, $verbosity = 1, $altmetrics = True)
+    {
         //bootstrap
         $indexParams['index'] = $type;
-        if(! $this->client->indices()->getMapping($indexParams))
-          echo $this->initialize_index($type);
-        elseif (! $this->init_indexer)
-          $this->initialize();
+        if (!$this->client->indices()->getMapping($indexParams))
+            echo $this->initialize_index($type);
+        elseif (!$this->init_indexer)
+            $this->initialize();
 
         $method_name = 'index_' . $type;
         if (method_exists($this, $method_name)) {
             try {
                 return $this->$method_name(false, $id, $altmetrics, $verbosity);
-	    } catch (Exception $e) {
-		echo $e->getMessage();
+            } catch (Exception $e) {
+                echo $e->getMessage();
                 // TODO: log?
             }
         } else {
@@ -359,16 +395,18 @@ public function index_from($type, $id = false, $verbosity=1, $altmetrics=True) {
         }
     }
 
-    public function delete($type, $id = false) {
+    public function delete($type, $id = false)
+    {
         $deleteParams = array();
         $deleteParams['index'] = $type;
-	$deleteParams['type'] = $type;
+        $deleteParams['type'] = $type;
         $deleteParams['id'] = $id;
         $response = $this->client->delete($deleteParams);
         return $response;
     }
 
-    public function initialize_settings() {
+    public function initialize_settings()
+    {
 
         $params['index'] = '_all';
         $params['body']['index']['analysis']['analyzer']['keyword-ci'] = array('tokenizer' => 'keyword', 'filter' => 'lowercase');
@@ -377,37 +415,39 @@ public function initialize_settings() {
         return 'Successfully updated settings';
     }
 
-    public function initialize_index($t) {
-        if(!$this->init_indexer)
-             $this->initialize();
-	$createparams = array(
-	    'index' => $t,
-    	    'body' => array(
-        	'settings' => array(
-		    'index' => array(
-            		'number_of_shards' => 1,
-            		'number_of_replicas' => 0
-		    )
-		),
-		'mappings' => array(
+    public function initialize_index($t)
+    {
+        if (!$this->init_indexer)
+            $this->initialize();
+        $createparams = array(
+            'index' => $t,
+            'body' => array(
+                'settings' => array(
+                    'index' => array(
+                        'number_of_shards' => 1,
+                        'number_of_replicas' => 0
+                    )
+                ),
+                'mappings' => array(
                     $t => $this->mappings[$t]
-		)
-	    )
+                )
+            )
         );
-	$this->client->indices()->create($createparams);
-        return '[Initialized mapping for ' . $t. '] ';
+        $this->client->indices()->create($createparams);
+        return '[Initialized mapping for ' . $t . '] ';
     }
 
-    public function index_downvote($id, $start_id = 0, $altmetrics=True, $verbosity=0){
+    public function index_downvote($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'downvote';
-	$params['type'] = 'downvote';
+        $params['type'] = 'downvote';
 
         $downvotes = $this->CI->Downvote->getDownvote($id);
 
-        if ($id and ! $downvotes)
+        if ($id and !$downvotes)
             return 'Error: downvote ' . $id . ' is unknown';
-        elseif (! $downvotes)
+        elseif (!$downvotes)
             return 'Nothing to index';
         foreach ($downvotes as $d) {
             $params['body'][] = array(
@@ -423,16 +463,17 @@ public function index_downvote($id, $start_id = 0, $altmetrics=True, $verbosity=
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($downvotes) . ' downvotes.';
     }
 
-    public function index_like($id, $start_id = 0, $altmetrics=True, $verbosity=0){
+    public function index_like($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'like';
         $params['type'] = 'like';
         $likes = $this->db->query('select * from likes' . ($id ? ' where lid=' . $id : ''));
 
-        if ($id and ! $likes)
+        if ($id and !$likes)
             return 'Error: like ' . $id . ' is unknown';
 
-        elseif (! $likes)
+        elseif (!$likes)
             return 'Nothing to index';
 
         foreach ($likes as $l) {
@@ -449,7 +490,8 @@ public function index_like($id, $start_id = 0, $altmetrics=True, $verbosity=0){
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($likes) . ' likes.';
     }
 
-    private function build_downvote($d){
+    private function build_downvote($d)
+    {
         $downvote = array(
             'downvote_id' => $d->did,
             'user_id' => $d->user_id,
@@ -464,7 +506,8 @@ private function build_downvote($d){
         return $downvote;
     }
 
-    private function build_like($l){
+    private function build_like($l)
+    {
         $like = array(
             'like_id' => $l->lid,
             'user_id' => $l->user_id,
@@ -475,15 +518,16 @@ private function build_like($l){
         return $like;
     }
 
-    public function index_download($id, $start_id = 0, $altmetrics=True, $verbosity=0){
+    public function index_download($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
         $params['index'] = 'download';
         $params['type'] = 'download';
         $downloads = $this->db->query('select * from downloads' . ($id ? ' where did=' . $id : ''));
 
-        if ($id and ! $downloads)
+        if ($id and !$downloads)
             return 'Error: download ' . $id . ' is unknown';
 
-        elseif (! $downloads)
+        elseif (!$downloads)
             return 'Nothing to index';
 
         foreach ($downloads as $d) {
@@ -500,7 +544,8 @@ public function index_download($id, $start_id = 0, $altmetrics=True, $verbosity=
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($downloads) . ' downloads.';
     }
 
-    private function build_download($d){
+    private function build_download($d)
+    {
         $download = array(
             'download_id' => $d->did,
             'user_id' => $d->user_id,
@@ -510,16 +555,16 @@ private function build_download($d){
             'time' => $d->time
         );
         return $download;
-
     }
 
-    public function index_user($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_user($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'user';
         $params['type'] = 'user';
         $users = $this->userdb->query('select id, first_name, last_name, email, company, country, bio, image, created_on, gamification_visibility from users where active="1"' . ($id ? ' and id=' . $id : ''));
 
-        if ($id and ! $users)
+        if ($id and !$users)
             return 'Error: user ' . $id . ' is unknown';
 
         foreach ($users as $d) {
@@ -536,7 +581,8 @@ public function index_user($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($users) . ' users.';
     }
 
-    private function build_user($d) {
+    private function build_user($d)
+    {
 
         $user = array(
             'user_id' => $d->id,
@@ -562,21 +608,21 @@ private function build_user($d) {
         $task_up = 0;
         $run_up = 0;
         $nr_of_uploads = 0;
-        if($uploads){
-            foreach($uploads as $up){
-                if($up->kt == 'd'){
-                    $data_up+=$up->count;
-                }else if($up->kt =='f'){
-                    $flow_up+=$up->count;
-                }else if($up->kt == 't'){
-                    $task_up+=$up->count;
-                }else if($up->kt == 'r'){
-                    $run_up+=$up->count;
+        if ($uploads) {
+            foreach ($uploads as $up) {
+                if ($up->kt == 'd') {
+                    $data_up += $up->count;
+                } else if ($up->kt == 'f') {
+                    $flow_up += $up->count;
+                } else if ($up->kt == 't') {
+                    $task_up += $up->count;
+                } else if ($up->kt == 'r') {
+                    $run_up += $up->count;
                 }
-                $nr_of_uploads+=$up->count;
+                $nr_of_uploads += $up->count;
             }
-            $user['nr_of_uploads'] =$nr_of_uploads;
-        }else{
+            $user['nr_of_uploads'] = $nr_of_uploads;
+        } else {
             $user['nr_of_uploads'] = 0;
         }
         $user['datasets_uploaded'] = $data_up;
@@ -586,16 +632,16 @@ private function build_user($d) {
 
 
         $runs_data = $this->db->query('select count(rid) as count FROM run r, task_inputs t, dataset d WHERE r.task_id=t.task_id and t.input="source_data" and t.value=d.did and r.uploader<>d.uploader and d.uploader=' . $d->id);
-        if ($runs_data){
+        if ($runs_data) {
             $user['runs_on_datasets'] = $runs_data[0]->count;
-        }else{
+        } else {
             $user['runs_on_datasets'] = 0;
         }
 
         $runs_flows = $this->db->query('select count(rid) as count FROM run r, algorithm_setup s, implementation i WHERE r.setup=s.sid and s.implementation_id=i.id and r.uploader<>i.uploader and i.uploader=' . $d->id);
-        if ($runs_flows){
+        if ($runs_flows) {
             $user['runs_on_flows'] = $runs_flows[0]->count;
-        }else{
+        } else {
             $user['runs_on_flows'] = 0;
         }
         // Activity: how many likes and downloads
@@ -611,32 +657,32 @@ private function build_user($d) {
         $nr_of_downloads_flow = 0;
         $nr_of_downloads_task = 0;
         $nr_of_downloads_run = 0;
-        if($ld_of_user){
-            foreach($ld_of_user as $ld){
-                if($ld->ldt=='l'){
-                    if($ld->knowledge_type=='d'){
-                        $nr_of_likes_data+=$ld->sum;
-                    }else if($ld->knowledge_type=='f'){
-                        $nr_of_likes_flow+=$ld->sum;
-                    }else if($ld->knowledge_type=='t'){
-                        $nr_of_likes_task+=$ld->sum;
-                    }else if($ld->knowledge_type=='r'){
-                        $nr_of_likes_run+=$ld->sum;
+        if ($ld_of_user) {
+            foreach ($ld_of_user as $ld) {
+                if ($ld->ldt == 'l') {
+                    if ($ld->knowledge_type == 'd') {
+                        $nr_of_likes_data += $ld->sum;
+                    } else if ($ld->knowledge_type == 'f') {
+                        $nr_of_likes_flow += $ld->sum;
+                    } else if ($ld->knowledge_type == 't') {
+                        $nr_of_likes_task += $ld->sum;
+                    } else if ($ld->knowledge_type == 'r') {
+                        $nr_of_likes_run += $ld->sum;
                     }
-                    $likes_of_user+=$ld->sum;
-                }else if($ld->ldt=='d'){
-                    if($ld->knowledge_type=='d'){
-                        $nr_of_downloads_data+=$ld->sum;
-                    }else if($ld->knowledge_type=='f'){
-                        $nr_of_downloads_flow+=$ld->sum;
-                    }else if($ld->knowledge_type=='t'){
-                        $nr_of_downloads_task+=$ld->sum;
-                    }else if($ld->knowledge_type=='r'){
-                        $nr_of_downloads_run+=$ld->sum;
+                    $likes_of_user += $ld->sum;
+                } else if ($ld->ldt == 'd') {
+                    if ($ld->knowledge_type == 'd') {
+                        $nr_of_downloads_data += $ld->sum;
+                    } else if ($ld->knowledge_type == 'f') {
+                        $nr_of_downloads_flow += $ld->sum;
+                    } else if ($ld->knowledge_type == 't') {
+                        $nr_of_downloads_task += $ld->sum;
+                    } else if ($ld->knowledge_type == 'r') {
+                        $nr_of_downloads_run += $ld->sum;
                     }
-                    $downloads_of_user+=$ld->sum;
+                    $downloads_of_user += $ld->sum;
                 }
-                $total_downloads+=$ld->sum;
+                $total_downloads += $ld->sum;
             }
         }
         $user['nr_of_likes'] = $likes_of_user;
@@ -650,8 +696,8 @@ private function build_user($d) {
         $user['nr_of_downloads_flow'] = $nr_of_downloads_flow;
         $user['nr_of_downloads_task'] = $nr_of_downloads_task;
         $user['nr_of_downloads_run'] = $nr_of_downloads_run;
-        if($d->gamification_visibility=='show'){
-            $user['activity'] = $this->CI->Gamification->getActivityFromParts($user['nr_of_uploads'],$user['nr_of_likes'],$user['nr_of_downloads']);
+        if ($d->gamification_visibility == 'show') {
+            $user['activity'] = $this->CI->Gamification->getActivityFromParts($user['nr_of_uploads'], $user['nr_of_likes'], $user['nr_of_downloads']);
         }
 
         // Reach: how many likes and downloads on uploads
@@ -666,32 +712,32 @@ private function build_user($d) {
         $downloads_received_flow = 0;
         $downloads_received_task = 0;
         $downloads_received_run = 0;
-        if($ld_received){
-            foreach($ld_received as $ld){
-                if($ld->ldt=='l'){
-                    if($ld->kt=='d'){
-                        $likes_received_data+=$ld->sum;
-                    }else if($ld->kt=='f'){
-                        $likes_received_flow+=$ld->sum;
-                    }else if($ld->kt=='t'){
-                        $likes_received_task+=$ld->sum;
-                    }else if($ld->kt=='r'){
-                        $likes_received_run+=$ld->sum;
+        if ($ld_received) {
+            foreach ($ld_received as $ld) {
+                if ($ld->ldt == 'l') {
+                    if ($ld->kt == 'd') {
+                        $likes_received_data += $ld->sum;
+                    } else if ($ld->kt == 'f') {
+                        $likes_received_flow += $ld->sum;
+                    } else if ($ld->kt == 't') {
+                        $likes_received_task += $ld->sum;
+                    } else if ($ld->kt == 'r') {
+                        $likes_received_run += $ld->sum;
                     }
-                    $likes_received+=$ld->sum;
-                }else if($ld->ldt=='d'){
-                  if (property_exists($ld, 'kt')) { // TODO: THIS IS A BUG! How can it not have a knowlegde piece ? (issue #70 on github)
-                    if($ld->kt=='d'){
-                        $downloads_received_data+=$ld->sum;
-                    }else if($ld->kt=='f'){
-                        $downloads_received_flow+=$ld->sum;
-                    }else if($ld->kt=='t'){
-                        $downloads_received_task+=$ld->sum;
-                    }else if($ld->kt=='r'){
-                        $downloads_received_run+=$ld->sum;
+                    $likes_received += $ld->sum;
+                } else if ($ld->ldt == 'd') {
+                    if (property_exists($ld, 'kt')) { // TODO: THIS IS A BUG! How can it not have a knowlegde piece ? (issue #70 on github)
+                        if ($ld->kt == 'd') {
+                            $downloads_received_data += $ld->sum;
+                        } else if ($ld->kt == 'f') {
+                            $downloads_received_flow += $ld->sum;
+                        } else if ($ld->kt == 't') {
+                            $downloads_received_task += $ld->sum;
+                        } else if ($ld->kt == 'r') {
+                            $downloads_received_run += $ld->sum;
+                        }
                     }
-                  }
-                  $downloads_received+=$ld->sum;
+                    $downloads_received += $ld->sum;
                 }
             }
         }
@@ -707,10 +753,10 @@ private function build_user($d) {
         $user['downloads_received_run'] = $downloads_received_run;
 
 
-        if($d->gamification_visibility=='show'){
-            $user['reach'] = $this->CI->Gamification->getReachFromParts($user['likes_received'],$user['downloads_received']);
+        if ($d->gamification_visibility == 'show') {
+            $user['reach'] = $this->CI->Gamification->getReachFromParts($user['likes_received'], $user['downloads_received']);
 
-            $impact_struct = $this->CI->Gamification->getTotalImpact('u',$d->id,"2013-1-1",date("Y-m-d"));
+            $impact_struct = $this->CI->Gamification->getTotalImpact('u', $d->id, "2013-1-1", date("Y-m-d"));
 
             $user['reuse'] = $impact_struct['reuse'];
 
@@ -722,9 +768,9 @@ private function build_user($d) {
 
             $user['badges'] = array();
             $badges = $this->CI->Badge->getBadgesOfUser($d->id);
-            if($badges){
-                foreach($badges as $b){
-                    $badge = array('badge_id'=>$b->badge_id,'rank'=>$b->rank);
+            if ($badges) {
+                foreach ($badges as $b) {
+                    $badge = array('badge_id' => $b->badge_id, 'rank' => $b->rank);
                     $user['badges'][] = $badge;
                 }
             }
@@ -733,15 +779,16 @@ private function build_user($d) {
         return $user;
     }
 
-    public function index_study($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_study($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'study';
         $params['type'] = 'study';
         $studies = $this->db->query('select * from study' . ($id ? ' where id=' . $id : ''));
 
-        if ($id and ! $studies)
+        if ($id and !$studies)
             return 'Error: study ' . $id . ' is unknown';
-        elseif (! $studies)
+        elseif (!$studies)
             return 'Nothing to index';
         foreach ($studies as $s) {
             $params['body'][] = array(
@@ -755,15 +802,16 @@ public function index_study($id, $start_id = 0, $altmetrics=True, $verbosity=0)
 
         $responses = $this->client->bulk($params);
 
-        if($responses['errors'] == True and array_key_exists('error', $responses['items'][0]['index'])){
-          $err = $responses['items'][0]['index']['error'];
-          return 'ERROR: Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
+        if ($responses['errors'] == True and array_key_exists('error', $responses['items'][0]['index'])) {
+            $err = $responses['items'][0]['index']['error'];
+            return 'ERROR: Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
         }
 
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($studies) . ' studies.';
     }
 
-    private function build_study($d) {
+    private function build_study($d)
+    {
         $study = array(
             'study_id' => $d->id,
             'name' => $d->name,
@@ -789,48 +837,49 @@ private function build_study($d) {
         $flows_tagged = NULL;
         $runs_tagged = NULL;
 
-        if($d->legacy == 'y'){
-          $data_tagged = $this->db->query("select id from dataset_tag where tag='study_" . $d->id . "'");
-          $task_tagged = $this->db->query("select id from task_tag where tag='study_" . $d->id . "'");
-          $flows_tagged = $this->db->query("select id from implementation_tag where tag='study_" . $d->id . "'");
-          $runs_tagged = $this->db->query("select id from run_tag where tag='study_" . $d->id . "'");
+        if ($d->legacy == 'y') {
+            $data_tagged = $this->db->query("select id from dataset_tag where tag='study_" . $d->id . "'");
+            $task_tagged = $this->db->query("select id from task_tag where tag='study_" . $d->id . "'");
+            $flows_tagged = $this->db->query("select id from implementation_tag where tag='study_" . $d->id . "'");
+            $runs_tagged = $this->db->query("select id from run_tag where tag='study_" . $d->id . "'");
         } elseif ($d->main_entity_type == 'task') {
-          $task_tagged = $this->db->query("select task_id as id from task_study where study_id=" . $d->id);
-          $data_tagged = $this->db->query("select value as id from task_inputs where input='source_data' and task_id in (select task_id from task_study where study_id=" . $d->id . ")");
+            $task_tagged = $this->db->query("select task_id as id from task_study where study_id=" . $d->id);
+            $data_tagged = $this->db->query("select value as id from task_inputs where input='source_data' and task_id in (select task_id from task_study where study_id=" . $d->id . ")");
         } elseif ($d->main_entity_type == 'run') {
-          $runs_tagged = $this->db->query("select run_id as id from run_study where study_id=" . $d->id);
-          $flows_tagged = $this->db->query("select distinct implementation_id as id from algorithm_setup where sid in (select setup from run where rid in (select run_id from run_study where study_id=" . $d->id . "))");
-          $task_tagged = $this->db->query("select distinct task_id as id from run where rid in (select run_id from run_study where study_id=" . $d->id . ")");
-          $data_tagged = $this->db->query("select distinct value as id from task_inputs where input='source_data' and task_id in (select task_id from run where rid in (select run_id from run_study where study_id=" . $d->id . "))");
+            $runs_tagged = $this->db->query("select run_id as id from run_study where study_id=" . $d->id);
+            $flows_tagged = $this->db->query("select distinct implementation_id as id from algorithm_setup where sid in (select setup from run where rid in (select run_id from run_study where study_id=" . $d->id . "))");
+            $task_tagged = $this->db->query("select distinct task_id as id from run where rid in (select run_id from run_study where study_id=" . $d->id . ")");
+            $data_tagged = $this->db->query("select distinct value as id from task_inputs where input='source_data' and task_id in (select task_id from run where rid in (select run_id from run_study where study_id=" . $d->id . "))");
         }
-        if ($data_tagged){
-          $study['datasets_included'] = count($data_tagged);
-          foreach ($data_tagged as $t){
-              $this->index_data($t->id);
+        if ($data_tagged) {
+            $study['datasets_included'] = count($data_tagged);
+            foreach ($data_tagged as $t) {
+                $this->index_data($t->id);
             }
         }
-        if ($task_tagged){
-          $study['tasks_included'] = count($task_tagged);
-          foreach ($task_tagged as $t){
-              $this->index_task($t->id);
+        if ($task_tagged) {
+            $study['tasks_included'] = count($task_tagged);
+            foreach ($task_tagged as $t) {
+                $this->index_task($t->id);
             }
         }
-        if ($flows_tagged){
-          $study['flows_included'] = count($flows_tagged);
-          foreach ($flows_tagged as $t){
-              $this->index_flow($t->id);
+        if ($flows_tagged) {
+            $study['flows_included'] = count($flows_tagged);
+            foreach ($flows_tagged as $t) {
+                $this->index_flow($t->id);
             }
         }
-        if ($runs_tagged){
-          $study['runs_included'] = count($runs_tagged);
-          foreach ($runs_tagged as $t){
-              $this->index_run($t->id);
+        if ($runs_tagged) {
+            $study['runs_included'] = count($runs_tagged);
+            foreach ($runs_tagged as $t) {
+                $this->index_run($t->id);
             }
         }
         return $study;
     }
 
-    public function index_task($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_task($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
         $params['index'] = 'task';
         $params['type'] = 'task';
         $taskmaxquery = $this->db->query('SELECT min(task_id) as mintask, max(task_id) as maxtask from task' . ($id ? ' where task_id=' . $id : ''));
@@ -843,11 +892,11 @@ public function index_task($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         $submitted = 0;
         $incr = min(50, $taskcount);
         if ($verbosity)
-          echo "Processing task          ";
+            echo "Processing task          ";
         while ($task_id <= $taskmax) {
             if ($verbosity) {
-              echo "\033[9D";
-              echo str_pad($task_id, 9, ' ', STR_PAD_RIGHT);
+                echo "\033[9D";
+                echo str_pad($task_id, 9, ' ', STR_PAD_RIGHT);
             }
             $tasks = null;
             $params['body'] = array();
@@ -871,16 +920,18 @@ public function index_task($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         return 'Successfully indexed ' . $submitted . ' out of ' . $taskcount . ' tasks.';
     }
 
-    private function build_single_task($id) {
+    private function build_single_task($id)
+    {
         $time = microtime(true);
         $task = $this->db->query('select a.*, b.runs from (SELECT t.task_id, tt.ttid, tt.name, t.creation_date, t.embargo_end_date, t.creator FROM task t, task_type tt where t.ttid=tt.ttid and task_id=' . $id . ') as a left outer join (select task_id, count(rid) as runs from run r group by task_id) as b on a.task_id=b.task_id');
         $t = $this->build_task($task[0]);
-        echo "Build task: ".(microtime(true) - $time)."\r\n";
+        echo "Build task: " . (microtime(true) - $time) . "\r\n";
         $time = microtime(true);
         return $t;
     }
 
-    private function build_task($d) {
+    private function build_task($d)
+    {
 
         $newdata = array(
             'task_id' => $d->task_id,
@@ -946,13 +997,15 @@ private function build_task($d) {
             foreach ($tags as $t => $u) {
                 $newdata['tags'][] = array(
                     'tag' => $t,
-                    'uploader' => $u);
-                if(substr( $t, 0, 6 ) === "study_"){
+                    'uploader' => $u
+                );
+                if (substr($t, 0, 6) === "study_") {
                     $study_id = substr($t, strpos($t, "_") + 1);
                     $studies[] = $study_id;
                     $newdata['collections'][] = array(
                         'type' => 'task',
-                        'id' => $study_id);
+                        'id' => $study_id
+                    );
                 }
             }
         }
@@ -961,26 +1014,32 @@ private function build_task($d) {
         $new_studies = array();
         $task_studies = $this->db->query("select study_id from task_study where task_id=" . $d->task_id);
         if ($task_studies != false) {
-            foreach ($task_studies as $t) { 
-                if (!in_array($t->study_id, $studies)){ 
+            foreach ($task_studies as $t) {
+                if (!in_array($t->study_id, $studies)) {
                     $new_studies[] = $t->study_id;
                     $newdata['collections'][] = array(
                         'type' => 'task',
-                        'id' => $t->study_id);
-                }}
+                        'id' => $t->study_id
+                    );
+                }
+            }
         }
         $run_studies = $this->db->query("select distinct study_id from run_study where run_id in (select rid from run where task_id=" . $d->task_id . ")");
         if ($run_studies != false) {
-            foreach ($run_studies as $t) { 
-                if (!in_array($t->study_id, $studies)){ 
-                    $new_studies[] = $t->study_id; 
+            foreach ($run_studies as $t) {
+                if (!in_array($t->study_id, $studies)) {
+                    $new_studies[] = $t->study_id;
                     $newdata['collections'][] = array(
                         'type' => 'run',
-                        'id' => $t->study_id);
-                }}
+                        'id' => $t->study_id
+                    );
+                }
+            }
         }
         if ($new_studies) {
-            foreach ($new_studies as $t) { $new_data['tags'][] = array('tag' => 'study_' . $t, 'uploader' => '0'); }
+            foreach ($new_studies as $t) {
+                $new_data['tags'][] = array('tag' => 'study_' . $t, 'uploader' => '0');
+            }
         }
 
         $newdata['suggest'] = array(
@@ -989,33 +1048,33 @@ private function build_task($d) {
         );
 
         $nr_of_downvotes = 0;
-        $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('t',$d->task_id,1);
-        if($nr_of_issues){
+        $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('t', $d->task_id, 1);
+        if ($nr_of_issues) {
             $newdata['nr_of_issues'] = count($nr_of_issues);
-            $nr_of_downvotes+=count($nr_of_issues);
-            $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('t',$d->task_id,0);
-            if($downvote_agrees){
-                $nr_of_downvotes+=count($downvote_agrees);
+            $nr_of_downvotes += count($nr_of_issues);
+            $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('t', $d->task_id, 0);
+            if ($downvote_agrees) {
+                $nr_of_downvotes += count($downvote_agrees);
             }
-        } else{
+        } else {
             $newdata['nr_of_issues'] = 0;
         }
         $newdata['nr_of_downvotes'] = $nr_of_downvotes;
 
-        $ld_task = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('t',$d->task_id);
+        $ld_task = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('t', $d->task_id);
         $reach = 0;
         $nr_of_likes = 0;
         $nr_of_downloads = 0;
         $total_downloads = 0;
-        if($ld_task){
-            foreach($ld_task as $ld){
-                if($ld->ldt=='l'){
-                    $reach += $this->CI->Gamification->getReachFromParts($ld->count,0);
-                    $nr_of_likes+=$ld->count;
-                }else if($ld->ldt=='d'){
-                    $reach += $this->CI->Gamification->getReachFromParts(0,$ld->count);
-                    $nr_of_downloads+=$ld->count;
-                    $total_downloads+=$ld->sum;
+        if ($ld_task) {
+            foreach ($ld_task as $ld) {
+                if ($ld->ldt == 'l') {
+                    $reach += $this->CI->Gamification->getReachFromParts($ld->count, 0);
+                    $nr_of_likes += $ld->count;
+                } else if ($ld->ldt == 'd') {
+                    $reach += $this->CI->Gamification->getReachFromParts(0, $ld->count);
+                    $nr_of_downloads += $ld->count;
+                    $total_downloads += $ld->sum;
                 }
             }
         }
@@ -1024,7 +1083,7 @@ private function build_task($d) {
         $newdata['total_downloads'] = $total_downloads;
         $newdata['reach'] = $reach;
 
-        $impact_struct = $this->CI->Gamification->getImpact('t',$d->task_id,"2013-1-1",date("Y-m-d"));
+        $impact_struct = $this->CI->Gamification->getImpact('t', $d->task_id, "2013-1-1", date("Y-m-d"));
 
         $newdata['reuse'] = $impact_struct['reuse'];
 
@@ -1037,7 +1096,8 @@ private function build_task($d) {
         return $newdata;
     }
 
-    private function fetch_setups($id = false) {
+    private function fetch_setups($id = false)
+    {
         $index = array();
         $setups = $this->db->query('SELECT s.setup, f.fullName, i.name, s.value FROM input_setting s, input i, implementation f where i.id=s.input_id AND i.implementation_id = f.id' . ($id ? ' and s.setup=' . $id : ''));
         if ($setups)
@@ -1046,122 +1106,125 @@ private function fetch_setups($id = false) {
                 $parameter_fullname = $v->fullName . '_' . $v->name;
                 $index[$v->setup][] = array('parameter' => $parameter_fullname, 'value' => $v->value);
             }
-        elseif($id != false)
-          $index[$id] = array();
+        elseif ($id != false)
+            $index[$id] = array();
         return $index;
     }
 
-    private function fetch_tasks($id = false) {
+    private function fetch_tasks($id = false)
+    {
         $index = array();
-	$tasks = $this->db->query("SELECT t.task_id, t.embargo_end_date, tt.name as ttname, i.value AS did, d.name AS dname, ep.name AS epname FROM task_inputs i, task_type tt, dataset d, task t LEFT JOIN task_inputs i2 on (t.task_id = i2.task_id AND i2.input = 'estimation_procedure') LEFT JOIN estimation_procedure ep on ep.id = i2.value WHERE t.task_id = i.task_id AND i.input = 'source_data' AND t.ttid = tt.ttid AND d.did = i.value" . ($id ? ' and t.task_id=' . $id : ''));
-        if ($tasks){
-            if($id)
-              $targets = $this->fetch_classes($tasks[0]->did);
+        $tasks = $this->db->query("SELECT t.task_id, t.embargo_end_date, tt.name as ttname, i.value AS did, d.name AS dname, ep.name AS epname FROM task_inputs i, task_type tt, dataset d, task t LEFT JOIN task_inputs i2 on (t.task_id = i2.task_id AND i2.input = 'estimation_procedure') LEFT JOIN estimation_procedure ep on ep.id = i2.value WHERE t.task_id = i.task_id AND i.input = 'source_data' AND t.ttid = tt.ttid AND d.did = i.value" . ($id ? ' and t.task_id=' . $id : ''));
+        if ($tasks) {
+            if ($id)
+                $targets = $this->fetch_classes($tasks[0]->did);
             else
-              $targets = $this->fetch_classes();
+                $targets = $this->fetch_classes();
             foreach ($tasks as $v) {
-              $index[$v->task_id]['task_id'] = $v->task_id;
-              $index[$v->task_id]['visibility'] = ((strtotime($v->embargo_end_date) < time()) ? 'public' : 'private');
-              $index[$v->task_id]['tasktype']['name'] = $v->ttname;
-              $index[$v->task_id]['source_data']['data_id'] = $v->did;
-              $index[$v->task_id]['source_data']['name'] = $v->dname;
-              $index[$v->task_id]['estimation_procedure']['name'] = $v->epname;
-
-              if($v->ttname == "Supervised Classification"){
-                if(array_key_exists($v->did,$targets)){ //check whether the task is valid (uses an existing dataset)
-                  $index[$v->task_id]['target_values'] = $targets[$v->did]['target_values'];
+                $index[$v->task_id]['task_id'] = $v->task_id;
+                $index[$v->task_id]['visibility'] = ((strtotime($v->embargo_end_date) < time()) ? 'public' : 'private');
+                $index[$v->task_id]['tasktype']['name'] = $v->ttname;
+                $index[$v->task_id]['source_data']['data_id'] = $v->did;
+                $index[$v->task_id]['source_data']['name'] = $v->dname;
+                $index[$v->task_id]['estimation_procedure']['name'] = $v->epname;
+
+                if ($v->ttname == "Supervised Classification") {
+                    if (array_key_exists($v->did, $targets)) { //check whether the task is valid (uses an existing dataset)
+                        $index[$v->task_id]['target_values'] = $targets[$v->did]['target_values'];
+                    }
                 }
-              }
             }
-          }
+        }
         return $index;
     }
 
-    private function fetch_classes($id = false) {
+    private function fetch_classes($id = false)
+    {
         $index = array();
         $data = $this->db->query("SELECT did, data_type, ClassDistribution FROM data_feature WHERE is_target='true'" . ($id ? ' and did=' . $id : ''));
         if ($data)
             foreach ($data as $v) {
-              if ($v->data_type == "nominal") {
-                  $distr = json_decode($v->ClassDistribution);
-                  if ($distr) {
-                      $index[$v->did]['target_values'] = $distr[0];
-                  }
-              }
+                if ($v->data_type == "nominal") {
+                    $distr = json_decode($v->ClassDistribution);
+                    if ($distr) {
+                        $index[$v->did]['target_values'] = $distr[0];
+                    }
+                }
             }
         return $index;
     }
 
-    private function fetch_runfiles($min, $max) {
+    private function fetch_runfiles($min, $max)
+    {
         $index = array();
         $runfiles = $this->db->query('SELECT source, field, name, format, file_id from runfile where source >= ' . $min . ' and source < ' . $max);
         if ($runfiles)
-          foreach ($runfiles as $r) {
-              $index[$r->source][$r->field]['url'] = BASE_URL . 'data/download/' . $r->file_id . '/' . $r->name;
-              $index[$r->source][$r->field]['format'] = $r->format;
-          }
+            foreach ($runfiles as $r) {
+                $index[$r->source][$r->field]['url'] = BASE_URL . 'data/download/' . $r->file_id . '/' . $r->name;
+                $index[$r->source][$r->field]['format'] = $r->format;
+            }
         return $index;
     }
 
-    function roundnum($i) {
+    function roundnum($i)
+    {
         if (is_numeric($i))
             return round($i, 4);
         else
             return $i;
     }
 
-    private function fetch_evaluations($min, $max, $include_folds=True) {
+    private function fetch_evaluations($min, $max, $include_folds = True)
+    {
         $index = array();
-  if($include_folds){
-        $folddata = $this->db->query('SELECT e.source, m.name AS function, e.fold, e.`repeat`, e.value FROM evaluation_fold e, math_function m WHERE e.function_id = m.id AND source >= ' . $min . ' and source < ' . $max);
-  $allfolds = array();
-
-        if ($folddata) {
-            $curr_src = array();
-            $folds = array();
-            $curr_fold = array();
-            $rp = 0;
-            $src = 0;
-            $fct = "";
-            foreach ($folddata as $f) {
-                if ($f->source != $src) {
-                    if (!empty($curr_fold))
-                        $folds[] = $curr_fold;
-                    if (!empty($folds))
-                        $curr_src[$fct] = $folds;
-                    if (!empty($curr_src))
-                        $allfolds[$src] = $curr_src;
-                    $src = $f->source;
-                    $curr_src = array();
-                    $folds = array();
-                    $curr_fold = array();
-                    $fct = $f->function;
-                    $rp = $f->repeat;
-                }
-                elseif ($f->function != $fct) {
-                    if (!empty($curr_fold))
-                        $folds[] = $curr_fold;
-                    if (!empty($folds))
-                        $curr_src[$fct] = $folds;
-                    $folds = array();
-                    $curr_fold = array();
-                    $fct = $f->function;
-                    $rp = $f->repeat;
-                }
-                elseif ($f->repeat != $rp) {
-                    if (!empty($curr_fold))
-                        $folds[] = $curr_fold;
-                    $rp = $f->repeat;
-                    $curr_fold = array();
+        if ($include_folds) {
+            $folddata = $this->db->query('SELECT e.source, m.name AS function, e.fold, e.`repeat`, e.value FROM evaluation_fold e, math_function m WHERE e.function_id = m.id AND source >= ' . $min . ' and source < ' . $max);
+            $allfolds = array();
+
+            if ($folddata) {
+                $curr_src = array();
+                $folds = array();
+                $curr_fold = array();
+                $rp = 0;
+                $src = 0;
+                $fct = "";
+                foreach ($folddata as $f) {
+                    if ($f->source != $src) {
+                        if (!empty($curr_fold))
+                            $folds[] = $curr_fold;
+                        if (!empty($folds))
+                            $curr_src[$fct] = $folds;
+                        if (!empty($curr_src))
+                            $allfolds[$src] = $curr_src;
+                        $src = $f->source;
+                        $curr_src = array();
+                        $folds = array();
+                        $curr_fold = array();
+                        $fct = $f->function;
+                        $rp = $f->repeat;
+                    } elseif ($f->function != $fct) {
+                        if (!empty($curr_fold))
+                            $folds[] = $curr_fold;
+                        if (!empty($folds))
+                            $curr_src[$fct] = $folds;
+                        $folds = array();
+                        $curr_fold = array();
+                        $fct = $f->function;
+                        $rp = $f->repeat;
+                    } elseif ($f->repeat != $rp) {
+                        if (!empty($curr_fold))
+                            $folds[] = $curr_fold;
+                        $rp = $f->repeat;
+                        $curr_fold = array();
+                    }
+                    if ($f->value)
+                        $curr_fold[] = round($f->value, 4);
                 }
-                if ($f->value)
-                    $curr_fold[] = round($f->value, 4);
+                $folds[] = $curr_fold;
+                $curr_src[$fct] = $folds;
+                $allfolds[$src] = $curr_src;
             }
-            $folds[] = $curr_fold;
-            $curr_src[$fct] = $folds;
-            $allfolds[$src] = $curr_src;
         }
-  }
         $evals = $this->db->query('SELECT e.source, m.name AS `function`, e.value, e.stdev, e.array_data FROM evaluation e, math_function m WHERE e.function_id = m.id AND source >= ' . $min . ' and source < ' . $max);
         if ($evals) {
             foreach ($evals as $r) {
@@ -1192,9 +1255,10 @@ private function fetch_evaluations($min, $max, $include_folds=True) {
     }
 
     //update task, dataset, flow to make sure that their indexed run counts are up to date? Only needed for sorting on number of runs.
-    private function update_runcounts($run) {
+    private function update_runcounts($run)
+    {
         $runparams['index'] = 'run';
-	$runparams['type'] = 'run';
+        $runparams['type'] = 'run';
         $runparams['body']['query']['match']['run_task.task_id'] = $run->task_id;
         $result = $this->client->search($runparams);
         $runcount = $this->checkNumeric($result['hits']['total']);
@@ -1220,14 +1284,15 @@ private function update_runcounts($run) {
     }
 
     //update tags for given type and id
-    public function update_tags($type, $id) {
+    public function update_tags($type, $id)
+    {
         $tagtable = $this->CI->Run_tag;
-        if($type == 'flow')
-          $tagtable = $this->CI->Implementation_tag;
-        elseif($type == 'data')
-          $tagtable = $this->CI->Dataset_tag;
-        elseif($type == 'task')
-          $tagtable = $this->CI->Task_tag;
+        if ($type == 'flow')
+            $tagtable = $this->CI->Implementation_tag;
+        elseif ($type == 'data')
+            $tagtable = $this->CI->Dataset_tag;
+        elseif ($type == 'task')
+            $tagtable = $this->CI->Task_tag;
 
         $ts = array();
         $tags = $tagtable->getAssociativeArray('tag', 'uploader', 'id = ' . $id);
@@ -1235,7 +1300,8 @@ public function update_tags($type, $id) {
             foreach ($tags as $t => $u) {
                 $ts[] = array(
                     'tag' => $t,
-                    'uploader' => $u);
+                    'uploader' => $u
+                );
             }
         }
 
@@ -1247,7 +1313,8 @@ public function update_tags($type, $id) {
     }
 
     //update topics for given type and id
-    public function update_topics($id) {
+    public function update_topics($id)
+    {
         $topictable = $this->CI->Dataset_topic;
         $ts = array();
         $topics = $topictable->getAssociativeArray('topic', 'uploader', 'id = ' . $id);
@@ -1255,7 +1322,8 @@ public function update_topics($id) {
             foreach ($topics as $t => $u) {
                 $ts[] = array(
                     'topic' => $t,
-                    'uploader' => $u);
+                    'uploader' => $u
+                );
             }
         }
 
@@ -1266,7 +1334,8 @@ public function update_topics($id) {
         $this->client->update($params);
     }
 
-    private function index_single_run($id) {
+    private function index_single_run($id)
+    {
 
         $params['index'] = 'run';
         $params['type'] = 'run';
@@ -1285,19 +1354,20 @@ private function index_single_run($id) {
         $responses = $this->client->index($params);
         $this->update_runcounts($run[0]);
 
-        return 'Successfully indexed run '. $id;
+        return 'Successfully indexed run ' . $id;
     }
 
-    public function index_run($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_run($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
         if ($id)
             return $this->index_single_run($id);
 
         $params['index'] = 'run';
         $params['type'] = 'run';
-  $setups = array();
+        $setups = array();
         $tasks = array();
 
-  $runmaxquery = $this->db->query('SELECT max(rid) as maxrun from run');
+        $runmaxquery = $this->db->query('SELECT max(rid) as maxrun from run');
         $runcountquery = $this->db->query('SELECT count(rid) as runcount from run');
         $runmax = intval($runmaxquery[0]->maxrun);
         $runcount = intval($runcountquery[0]->runcount);
@@ -1306,58 +1376,62 @@ public function index_run($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         $submitted = 0;
         $incr = 100;
         if ($verbosity) {
-          echo "Processing run ";
+            echo "Processing run ";
         }
-        while ($rid < $runmax) {
-          if ($verbosity) {
-            echo str_pad($rid, 9, ' ', STR_PAD_RIGHT);
-          }
-          set_time_limit(6000);
-          $runs = null;
-          $runfiles = null;
-          $evals = null;
-          $params['body'] = array();
-          $runs = $this->db->query('SELECT rid, uploader, setup, implementation_id, task_id, start_time, re.error, error_message, run_details FROM run r,run_evaluated re, algorithm_setup s where r.rid=re.run_id and s.sid=r.setup and rid>=' . $rid . ' and rid<' . ($rid + $incr));
-          if($runs){
-            $runfiles = $this->fetch_runfiles($rid, $rid + $incr);
-            $evals = $this->fetch_evaluations($rid, $rid + $incr);
-
-            foreach ($runs as $r) {
-              try {
-                $params['body'][] = array(
-                  'index' => array(
-                     '_id' => $r->rid
-                  )
-                );
-              $params['body'][] = $this->build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics);
-              } catch (Exception $e) {
-                if ($verbosity) {
-                  echo $e->getMessage();
-                  return $e->getMessage();
+
+        while ($rid <= $runmax) {
+            if ($verbosity) {
+                echo str_pad($rid, 9, ' ', STR_PAD_RIGHT);
+            }
+            set_time_limit(6000);
+            $runs = null;
+            $runfiles = null;
+            $evals = null;
+            $responses = null;
+            $params['body'] = array();
+            $runs = $this->db->query('SELECT rid, uploader, setup, implementation_id, task_id, start_time, re.error, error_message, run_details FROM run r,run_evaluated re, algorithm_setup s where r.rid=re.run_id and s.sid=r.setup and rid>=' . $rid . ' and rid<' . ($rid + $incr));
+
+            if ($runs) {
+                $runfiles = $this->fetch_runfiles($rid, $rid + $incr);
+                $evals = $this->fetch_evaluations($rid, $rid + $incr);
+
+                foreach ($runs as $r) {
+                    try {
+                        $params['body'][] = array(
+                            'index' => array(
+                                '_id' => $r->rid
+                            )
+                        );
+                        $params['body'][] = $this->build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics);
+                    } catch (Exception $e) {
+                        if ($verbosity) {
+                            echo $e->getMessage();
+                            return $e->getMessage();
+                        }
+                    }
+                    $responses = $this->client->bulk($params);
+                    $submitted += sizeof($responses['items']);
+                    if ($verbosity) {
+                        #echo "-  completed ".str_pad($submitted, 9, ' ', STR_PAD_RIGHT);
+                        echo "\033[31D";
+                    }
                 }
-              }
-              $responses = $this->client->bulk($params);
-              $submitted += sizeof($responses['items']);
-              if ($verbosity) {
-                #echo "-  completed ".str_pad($submitted, 9, ' ', STR_PAD_RIGHT);
-                echo "\033[31D";
-              }
-           }
-         } elseif($verbosity) {
-           echo "\033[9D";
-         }
-         $rid += $incr;
-       }
-       return 'Successfully indexed ' . $submitted . ' out of ' . $runcount . ' runs.';
+            } elseif ($verbosity) {
+                echo "\033[9D";
+            }
+            $rid += $incr;
+        }
+        return 'Successfully indexed ' . $submitted . ' out of ' . $runcount . ' runs.';
     }
 
-    private function build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics=True) {
+    private function build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics = True)
+    {
         // build dictionary of setups and tasks to eliminate duplicate calls
-        if(!array_key_exists($r->setup, $setups))
-          $setups[$r->setup] = $this->fetch_setups($r->setup)[$r->setup];
-        if(!array_key_exists($r->task_id, $tasks))
-          $tasks[$r->task_id] = $this->fetch_tasks($r->task_id)[$r->task_id];
-        if(!array_key_exists($r->task_id,$tasks) or !array_key_exists($r->implementation_id,$this->flow_names)){ // catch faulty runs
+        if (!array_key_exists($r->setup, $setups))
+            $setups[$r->setup] = $this->fetch_setups($r->setup)[$r->setup];
+        if (!array_key_exists($r->task_id, $tasks))
+            $tasks[$r->task_id] = $this->fetch_tasks($r->task_id)[$r->task_id];
+        if (!array_key_exists($r->task_id, $tasks) or !array_key_exists($r->implementation_id, $this->flow_names)) { // catch faulty runs
             return array();
         }
         $new_data = array(
@@ -1387,13 +1461,15 @@ private function build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics=T
             foreach ($tags as $t => $u) {
                 $new_data['tags'][] = array(
                     'tag' => $t,
-                    'uploader' => $u);
-                if(substr( $t, 0, 6 ) === "study_"){
+                    'uploader' => $u
+                );
+                if (substr($t, 0, 6) === "study_") {
                     $study_id = substr($t, strpos($t, "_") + 1);
                     $studies[] = $study_id;
                     $new_data['collections'][] = array(
                         'type' => 'run',
-                        'id' => $study_id);
+                        'id' => $study_id
+                    );
                 }
             }
         }
@@ -1401,12 +1477,13 @@ private function build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics=T
         $run_studies = $this->db->query("select distinct study_id from run_study where run_id=" . $r->rid);
         if ($run_studies != false) {
             foreach ($run_studies as $t) {
-              if (!in_array($t->study_id, $studies)){
-                 $new_data['tags'][] = array('tag' => 'study_' . $t->study_id, 'uploader' => '0');                    
-                 $new_data['collections'][] = array(
-                    'type' => 'run',
-                    'id' => $t->study_id);
-              }
+                if (!in_array($t->study_id, $studies)) {
+                    $new_data['tags'][] = array('tag' => 'study_' . $t->study_id, 'uploader' => '0');
+                    $new_data['collections'][] = array(
+                        'type' => 'run',
+                        'id' => $t->study_id
+                    );
+                }
             }
         }
 
@@ -1420,55 +1497,57 @@ private function build_run($r, $setups, $tasks, $runfiles, $evals, $altmetrics=T
         $new_data['impact_of_reuse'] = 0;
         $new_data['reach_of_reuse'] = 0;
         $new_data['impact'] = 0;
-        if($altmetrics){
-              $nr_of_downvotes = 0;
-              $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('r',$r->rid,1);
-              if($nr_of_issues){
-                  $new_data['nr_of_issues'] = count($nr_of_issues);
-                  $nr_of_downvotes+=count($nr_of_issues);
-                  $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('r',$r->rid,0);
-                  if($downvote_agrees){
-                      $nr_of_downvotes+=count($downvote_agrees);
-                  }
-              }else{
-                  $new_data['nr_of_issues'] = 0;
-              }
-              $new_data['nr_of_downvotes'] = $nr_of_downvotes;
-
-              $ld_run = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('r',$r->rid);
-              $reach = 0;
-              $nr_of_likes = 0;
-              $nr_of_downloads = 0;
-              $total_downloads = 0;
-              if($ld_run){
-                  foreach($ld_run as $ld){
-                      if($ld->ldt=='l'){
-                          $reach += $this->CI->Gamification->getReachFromParts($ld->count,0);
-                          $nr_of_likes+=$ld->count;
-                      }else if($ld->ldt=='d'){
-                          $reach += $this->CI->Gamification->getReachFromParts(0,$ld->count);
-                          $nr_of_downloads+=$ld->count;
-                          $total_downloads+=$ld->sum;
-                      }
-                  }
-              }
-              $new_data['nr_of_likes'] = $nr_of_likes;
-              $new_data['nr_of_downloads'] = $nr_of_downloads;
-              $new_data['total_downloads'] = $total_downloads;
-              $new_data['reach'] = $reach;
+        if ($altmetrics) {
+            $nr_of_downvotes = 0;
+            $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('r', $r->rid, 1);
+            if ($nr_of_issues) {
+                $new_data['nr_of_issues'] = count($nr_of_issues);
+                $nr_of_downvotes += count($nr_of_issues);
+                $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('r', $r->rid, 0);
+                if ($downvote_agrees) {
+                    $nr_of_downvotes += count($downvote_agrees);
+                }
+            } else {
+                $new_data['nr_of_issues'] = 0;
+            }
+            $new_data['nr_of_downvotes'] = $nr_of_downvotes;
+
+            $ld_run = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('r', $r->rid);
+            $reach = 0;
+            $nr_of_likes = 0;
+            $nr_of_downloads = 0;
+            $total_downloads = 0;
+            if ($ld_run) {
+                foreach ($ld_run as $ld) {
+                    if ($ld->ldt == 'l') {
+                        $reach += $this->CI->Gamification->getReachFromParts($ld->count, 0);
+                        $nr_of_likes += $ld->count;
+                    } else if ($ld->ldt == 'd') {
+                        $reach += $this->CI->Gamification->getReachFromParts(0, $ld->count);
+                        $nr_of_downloads += $ld->count;
+                        $total_downloads += $ld->sum;
+                    }
+                }
+            }
+            $new_data['nr_of_likes'] = $nr_of_likes;
+            $new_data['nr_of_downloads'] = $nr_of_downloads;
+            $new_data['total_downloads'] = $total_downloads;
+            $new_data['reach'] = $reach;
         }
         return $new_data;
     }
 
-    public function index_task_type($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_task_type($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'task_type';
         $params['type'] = 'task_type';
-        $types = $this->db->query('SELECT tt.ttid, tt.name, tt.description, count(task_id) as tasks, tt.creationDate as date FROM task_type tt, task t where tt.ttid=t.ttid' . ($id ? ' and tt.ttid=' . $id : '') . ' group by tt.ttid');
+        $types = $this->db->query('SELECT tt.ttid, tt.name, tt.description, count(task_id) as tasks, tt.creationDate as date FROM task_type tt LEFT JOIN task t ON t.ttid = tt.ttid' . ($id ? ' WHERE tt.ttid=' . $id : '') . ' GROUP BY tt.ttid');
+        // $types = $this->db->query('SELECT tt.ttid, tt.name, tt.description, count(task_id) as tasks, tt.creationDate as date FROM task_type tt, task t where tt.ttid=t.ttid' . ($id ? ' and tt.ttid=' . $id : '') . ' group by tt.ttid');
 
-        if ($id and ! $types)
+        if ($id and !$types)
             return 'Error: task type ' . $id . ' is unknown';
-        elseif (! $types)
+        elseif (!$types)
             return 'Nothing to index';
         foreach ($types as $d) {
             $params['body'][] = array(
@@ -1485,7 +1564,8 @@ public function index_task_type($id, $start_id = 0, $altmetrics=True, $verbosity
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($types) . ' task types.';
     }
 
-    private function build_task_type($d) {
+    private function build_task_type($d)
+    {
         $new_data = array(
             'tt_id' => $d->ttid,
             'name' => $d->name,
@@ -1501,28 +1581,29 @@ private function build_task_type($d) {
 
         $inputs = $this->db->query('SELECT name, type, description, io, requirement FROM task_type_inout where ttid=' . $d->ttid);
         if ($inputs)
-          foreach ($inputs as $i) {
-              $new_data['input'][] = array(
-                  'name' => $i->name,
-                  'type' => $i->type,
-                  'description' => $i->description,
-                  'io' => $i->io,
-                  'requirement' => $i->requirement
-              );
-          }
+            foreach ($inputs as $i) {
+                $new_data['input'][] = array(
+                    'name' => $i->name,
+                    'type' => $i->type,
+                    'description' => $i->description,
+                    'io' => $i->io,
+                    'requirement' => $i->requirement
+                );
+            }
         return $new_data;
     }
 
-    public function index_flow($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_flow($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'flow';
         $params['type'] = 'flow';
         $flows = $this->db->query('select i.*, count(rid) as runs from implementation i left join algorithm_setup s on (s.implementation_id=i.id) left join run r on (r.setup=s.sid)' . ($id ? ' where i.id=' . $id : '') . ' group by i.id');
 
 
-        if ($id and ! $flows)
+        if ($id and !$flows)
             return 'Error: flow ' . $id . ' is unknown';
-        elseif (! $flows)
+        elseif (!$flows)
             return 'Nothing to index';
         foreach ($flows as $d) {
             $params['body'][] = array(
@@ -1536,19 +1617,20 @@ public function index_flow($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
 
         $responses = $this->client->bulk($params);
 
-        if($responses['errors'] == True){
-    foreach ($responses['items'] as $res){
-    if(array_key_exists('error',$res['index'])){
-      $err = $res['index']['error'];
-      return 'ERROR for ID ' . $res['index']['_id'] . ' : Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
-      }
-    }
-  }
+        if ($responses['errors'] == True) {
+            foreach ($responses['items'] as $res) {
+                if (array_key_exists('error', $res['index'])) {
+                    $err = $res['index']['error'];
+                    return 'ERROR for ID ' . $res['index']['_id'] . ' : Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
+                }
+            }
+        }
 
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . sizeof($flows) . ' flows.';
     }
 
-    private function build_flow($d) {
+    private function build_flow($d)
+    {
         $new_data = array(
             'flow_id' => $d->id,
             'name' => $d->name,
@@ -1585,18 +1667,19 @@ private function build_flow($d) {
             foreach ($tags as $t => $u) {
                 $new_data['tags'][] = array(
                     'tag' => $t,
-                    'uploader' => $u);
-                if(substr( $t, 0, 6 ) === "study_")
-                  $studies[] = substr($t, strpos($t, "_") + 1);
+                    'uploader' => $u
+                );
+                if (substr($t, 0, 6) === "study_")
+                    $studies[] = substr($t, strpos($t, "_") + 1);
             }
         }
 
         $run_studies = $this->db->query("select distinct study_id from run_study where run_id in (select rid from run where setup in (select sid from algorithm_setup where implementation_id=" . $d->id . "))");
         if ($run_studies != false) {
             foreach ($run_studies as $t) {
-              if (!in_array($t->study_id, $studies)){
-                 $new_data['tags'][] = array('tag' => 'study_' . $t->study_id, 'uploader' => '0');
-              }
+                if (!in_array($t->study_id, $studies)) {
+                    $new_data['tags'][] = array('tag' => 'study_' . $t->study_id, 'uploader' => '0');
+                }
             }
         }
 
@@ -1633,33 +1716,33 @@ private function build_flow($d) {
         }
 
         $nr_of_downvotes = 0;
-        $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('f',$d->id,1);
-        if($nr_of_issues){
+        $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('f', $d->id, 1);
+        if ($nr_of_issues) {
             $new_data['nr_of_issues'] = count($nr_of_issues);
-            $nr_of_downvotes+=count($nr_of_issues);
-            $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('f',$d->id,0);
-            if($downvote_agrees){
-                $nr_of_downvotes+=count($downvote_agrees);
+            $nr_of_downvotes += count($nr_of_issues);
+            $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('f', $d->id, 0);
+            if ($downvote_agrees) {
+                $nr_of_downvotes += count($downvote_agrees);
             }
-        }else{
+        } else {
             $new_data['nr_of_issues'] = 0;
         }
         $new_data['nr_of_downvotes'] = $nr_of_downvotes;
 
-        $ld_flow = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('f',$d->id);
+        $ld_flow = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('f', $d->id);
         $reach = 0;
         $nr_of_likes = 0;
         $nr_of_downloads = 0;
         $total_downloads = 0;
-        if($ld_flow){
-            foreach($ld_flow as $ld){
-                if($ld->ldt=='l'){
-                    $reach += $this->CI->Gamification->getReachFromParts($ld->count,0);
-                    $nr_of_likes+=$ld->count;
-                }else if($ld->ldt=='d'){
-                    $reach += $this->CI->Gamification->getReachFromParts(0,$ld->count);
-                    $nr_of_downloads+=$ld->count;
-                    $total_downloads+=$ld->sum;
+        if ($ld_flow) {
+            foreach ($ld_flow as $ld) {
+                if ($ld->ldt == 'l') {
+                    $reach += $this->CI->Gamification->getReachFromParts($ld->count, 0);
+                    $nr_of_likes += $ld->count;
+                } else if ($ld->ldt == 'd') {
+                    $reach += $this->CI->Gamification->getReachFromParts(0, $ld->count);
+                    $nr_of_downloads += $ld->count;
+                    $total_downloads += $ld->sum;
                 }
             }
         }
@@ -1668,7 +1751,7 @@ private function build_flow($d) {
         $new_data['total_downloads'] = $total_downloads;
         $new_data['reach'] = $reach;
 
-        $impact_struct = $this->CI->Gamification->getImpact('f',$d->id,"2013-1-1",date("Y-m-d"));
+        $impact_struct = $this->CI->Gamification->getImpact('f', $d->id, "2013-1-1", date("Y-m-d"));
 
         $new_data['reuse'] = $impact_struct['reuse'];
 
@@ -1680,7 +1763,8 @@ private function build_flow($d) {
         return $new_data;
     }
 
-    public function index_measure($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_measure($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
 
         $params['index'] = 'measure';
         $params['type'] = 'measure';
@@ -1735,14 +1819,15 @@ public function index_measure($id, $start_id = 0, $altmetrics=True, $verbosity=0
                 $params['body'][] = $this->build_flowq($d);
             }
 
-        if ($id and ! array_key_exists('body', $params))
+        if ($id and !array_key_exists('body', $params))
             return "No measure found with id " . $id;
 
         $responses = $this->client->bulk($params);
         return 'Successfully indexed ' . sizeof($responses['items']) . ' out of ' . (($procs ? sizeof($procs) : 0) + ($funcs ? sizeof($funcs) : 0) + ($dataqs ? sizeof($dataqs) : 0) + ($flowqs ? sizeof($flowqs) : 0)) . ' measures (' . ($procs ? sizeof($procs) : 0) . ' procedures, ' . ($funcs ? sizeof($funcs) : 0) . ' functions, ' . ($dataqs ? sizeof($dataqs) : 0) . ' data qualities, ' . ($flowqs ? sizeof($flowqs) : 0) . ' flow qualities).';
     }
 
-    private function build_procedure($d) {
+    private function build_procedure($d)
+    {
         return array(
             'proc_id' => $d->id,
             'measure_type' => 'estimation_procedure',
@@ -1762,11 +1847,12 @@ private function build_procedure($d) {
         );
     }
 
-    private function build_function($d) {
+    private function build_function($d)
+    {
         $id = str_replace("_", "-", $d->name);
         $desc = $d->description;
-        $desc = str_replace("<math>","$$",$desc);
-        $desc = str_replace("</math>","$$",$desc);
+        $desc = str_replace("<math>", "$$", $desc);
+        $desc = str_replace("</math>", "$$", $desc);
 
         return array(
             'eval_id' => $id,
@@ -1787,7 +1873,8 @@ private function build_function($d) {
         );
     }
 
-    private function build_dataq($d) {
+    private function build_dataq($d)
+    {
         $id = str_replace("_", "-", $d->name);
         return array(
             'quality_id' => $id,
@@ -1803,7 +1890,8 @@ private function build_dataq($d) {
         );
     }
 
-    private function build_flowq($d) {
+    private function build_flowq($d)
+    {
         $id = str_replace("_", "-", $d->name);
         return array(
             'quality_id' => $id,
@@ -1819,38 +1907,40 @@ private function build_flowq($d) {
         );
     }
 
-    public function index_single_dataset($id) {
+    public function index_single_dataset($id)
+    {
 
-      	$params['index'] = 'data';
-      	$params['type'] = 'data';
+        $params['index'] = 'data';
+        $params['type'] = 'data';
         $status_sql_variable = 'IFNULL(`s`.`status`, \'' . $this->CI->config->item('default_dataset_status') . '\')';
         $datasets = $this->db->query('select d.*, ' . $status_sql_variable . ' AS `status`, count(rid) as runs, GROUP_CONCAT(dp.error) as error_message from dataset d left join (SELECT `did`, MAX(`status`) AS `status` FROM `dataset_status` GROUP BY `did`) s ON s.did = d.did left join task_inputs t on (t.value=d.did and t.input="source_data") left join run r on (r.task_id=t.task_id) left join data_processed dp on (d.did=dp.did)' . ($id ? ' where d.did=' . $id : '') . ' group by d.did');
 
-        if ($id and ! $datasets)
+        if ($id and !$datasets)
             return 'Error: data set ' . $id . ' is unknown';
 
         if ($datasets)
-          foreach ($datasets as $d) {
-              $params['body'][] = array(
-                  'index' => array(
-                      '_id' => $d->did
-                  )
-              );
+            foreach ($datasets as $d) {
+                $params['body'][] = array(
+                    'index' => array(
+                        '_id' => $d->did
+                    )
+                );
 
-              $params['body'][] = $this->build_data($d);
-          }
+                $params['body'][] = $this->build_data($d);
+            }
 
         $responses = $this->client->bulk($params);
 
-        if($responses['errors'] == True){
-          $err = $responses['items'][0]['index']['error'];
-          return 'ERROR: Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
+        if ($responses['errors'] == True) {
+            $err = $responses['items'][0]['index']['error'];
+            return 'ERROR: Type:' . $err['type'] . ' Reason: ' . $err['reason'] . (array_key_exists('caused_by', $err) ? ' Caused by: ' . $err['caused_by']['reason'] : '');
         }
 
-        return 'Successfully indexed dataset '.$id;
+        return 'Successfully indexed dataset ' . $id;
     }
 
-    public function index_data($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
+    public function index_data($id, $start_id = 0, $altmetrics = True, $verbosity = 0)
+    {
         if ($id)
             return $this->index_single_dataset($id);
 
@@ -1865,11 +1955,11 @@ public function index_data($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         $submitted = 0;
         $incr = 10;
         if ($verbosity) {
-          echo "Processing dataset ";
+            echo "Processing dataset ";
         }
         while ($did < $datamax) {
             if ($verbosity) {
-              echo $did." ";
+                echo $did . " ";
             }
             set_time_limit(600);
             $datasets = null;
@@ -1877,38 +1967,36 @@ public function index_data($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
             $valid_ids = array();
             $status_sql_variable = 'IFNULL(`s`.`status`, \'' . $this->CI->config->item('default_dataset_status') . '\')';
             $datasets = $this->db->query('select d.*, ' . $status_sql_variable . 'AS `status`, count(rid) as runs, GROUP_CONCAT(dp.error) as error_message from dataset d left join (SELECT `did`, MAX(`status`) AS `status` FROM `dataset_status` GROUP BY `did`) s ON d.did = s.did left join task_inputs t on (t.value=d.did and t.input="source_data") left join run r on (r.task_id=t.task_id) left join data_processed dp on (d.did=dp.did) where d.did>=' . $did . ' and d.did<' . ($did + $incr) . ' group by d.did');
-            if($datasets){
-              foreach ($datasets as $d) {
-                try {
-                  $params['body'][] = array(
-                      'index' => array(
-                          '_id' => $d->did
-                      )
-                  );
-                  $valid_ids[] = $d->did;
-                  $params['body'][] = $this->build_data($d, $altmetrics);
-
-                } catch (Exception $e) {
-                    return $e->getMessage();
+            if ($datasets) {
+                foreach ($datasets as $d) {
+                    try {
+                        $params['body'][] = array(
+                            'index' => array(
+                                '_id' => $d->did
+                            )
+                        );
+                        $valid_ids[] = $d->did;
+                        $params['body'][] = $this->build_data($d, $altmetrics);
+                    } catch (Exception $e) {
+                        return $e->getMessage();
+                    }
+                }
+
+                $responses = $this->client->bulk($params);
+                $submitted += sizeof($responses['items']);
+
+                //clean up, just to be sure
+                $params_del['index'] = 'data';
+                foreach (array_diff(range($did, $did + $incr - 1), $valid_ids) as $delid) {
+                    $params_del['id'] = $delid;
+                    try {
+                        $this->client->delete($params_del);
+                    } catch (Exception $e) {
+                        $result = json_decode($e);
+                        if ($result['found'] == 'true' && $verbosity)
+                            echo "deleted_" . $delid . " ";
+                    }
                 }
-              }
-
-              $responses = $this->client->bulk($params);
-              $submitted += sizeof($responses['items']);
-
-        //clean up, just to be sure
-              $params_del['index'] = 'data';
-        foreach(array_diff(range($did,$did+$incr-1),$valid_ids) as $delid){
-            $params_del['id'] = $delid;
-      try{
-        $this->client->delete($params_del);
-      }
-      catch (Exception $e) {
-        $result = json_decode($e);
-                    if($result['found']=='true' && $verbosity)
-                        echo "deleted_".$delid." ";
-                  }
-         }
             }
 
             $did += $incr;
@@ -1917,7 +2005,8 @@ public function index_data($id, $start_id = 0, $altmetrics=True, $verbosity=0) {
         return 'Successfully indexed ' . $submitted . ' out of ' . $datacount . ' datasets.';
     }
 
-    private function build_data($d, $altmetrics=True) {
+    private function build_data($d, $altmetrics = True)
+    {
         $description_record = $this->CI->Dataset_description->getWhereSingle('did =' . $d->did, 'version DESC');
         $headless_description = trim(preg_replace('/\s+/', ' ', preg_replace('/^\*{2,}.*/m', '', $description_record->description)));
         $new_data = array(
@@ -1928,7 +2017,7 @@ private function build_data($d, $altmetrics=True) {
             'version_label' => $d->version_label,
             'description' => $description_record->description,
             'format' => $d->format,
-            'uploader' => array_key_exists($d->uploader, $this->user_names) ? $this->user_names[$d->uploader]: 'unknown',
+            'uploader' => array_key_exists($d->uploader, $this->user_names) ? $this->user_names[$d->uploader] : 'unknown',
             'uploader_id' => intval($d->uploader),
             'visibility' => $d->visibility,
             'creator' => $d->creator,
@@ -1954,9 +2043,9 @@ private function build_data($d, $altmetrics=True) {
         $new_data['qualities'] = array();
         $qualities = $this->CI->Data_quality->getQualitiesOrderedByPriority($d->did);
         if ($qualities != false)
-          foreach($qualities as $q){
-            $new_data['qualities'][$q->name] = $this->checkNumeric($q->value);
-          }
+            foreach ($qualities as $q) {
+                $new_data['qualities'][$q->name] = $this->checkNumeric($q->value);
+            }
 
         $new_data['tags'] = array();
         $studies = array();
@@ -1965,9 +2054,10 @@ private function build_data($d, $altmetrics=True) {
             foreach ($tags as $t => $u) {
                 $new_data['tags'][] = array(
                     'tag' => $t,
-                    'uploader' => $u);
-                if(substr( $t, 0, 6 ) === "study_")
-                  $studies[] = substr($t, strpos($t, "_") + 1);
+                    'uploader' => $u
+                );
+                if (substr($t, 0, 6) === "study_")
+                    $studies[] = substr($t, strpos($t, "_") + 1);
             }
         }
 
@@ -1976,21 +2066,32 @@ private function build_data($d, $altmetrics=True) {
             foreach ($topics as $t => $u) {
                 $new_data['topics'][] = array(
                     'topic' => $t,
-                    'uploader' => $u);                
+                    'uploader' => $u
+                );
             }
         }
         // replace with study list in new indexer
         $new_studies = array();
         $task_studies = $this->db->query("select study_id from task_study where task_id in (select task_id from task_inputs where input='source_data' and value=" . $d->did . ")");
         if ($task_studies != false) {
-            foreach ($task_studies as $t) { if (!in_array($t->study_id, $studies)){ $new_studies[] = $t->study_id; }}
+            foreach ($task_studies as $t) {
+                if (!in_array($t->study_id, $studies)) {
+                    $new_studies[] = $t->study_id;
+                }
+            }
         }
         $run_studies = $this->db->query("select distinct study_id from run_study where run_id in (select rid from run where task_id in (select task_id from task_inputs where input='source_data' and value=" . $d->did . "))");
         if ($run_studies != false) {
-            foreach ($run_studies as $t) { if (!in_array($t->study_id, $studies)){ $new_studies[] = $t->study_id; }}
+            foreach ($run_studies as $t) {
+                if (!in_array($t->study_id, $studies)) {
+                    $new_studies[] = $t->study_id;
+                }
+            }
         }
         if ($new_studies) {
-            foreach ($new_studies as $t) { $new_data['tags'][] = array('tag' => 'study_' . $t, 'uploader' => '0'); }
+            foreach ($new_studies as $t) {
+                $new_data['tags'][] = array('tag' => 'study_' . $t, 'uploader' => '0');
+            }
         }
 
         $new_data['features'] = array();
@@ -2017,79 +2118,80 @@ private function build_data($d, $altmetrics=True) {
                     $feat['stdev'] = $f->StandardDeviation;
                 } elseif ($f->data_type == "nominal") {
                     $distr = json_decode($f->ClassDistribution);
-                    if(is_array($distr))
-                       $feat['distr'] = $this->array_map_recursive('strval',$distr);
+                    if (is_array($distr))
+                        $feat['distr'] = $this->array_map_recursive('strval', $distr);
                     else
-                       $feat['distr'] = [];
+                        $feat['distr'] = [];
                 }
                 $new_data['features'][] = $feat;
             }
         }
-      $new_data['nr_of_issues'] = 0;
-      $new_data['nr_of_downvotes'] = 0;
-      $new_data['nr_of_likes'] = 0;
-      $new_data['nr_of_downloads'] = 0;
-      $new_data['total_downloads'] = 0;
-      $new_data['reach'] = 0;
-      $new_data['reuse'] = 0;
-      $new_data['impact_of_reuse'] = 0;
-      $new_data['reach_of_reuse'] = 0;
-      $new_data['impact'] = 0;
-
-      if($altmetrics){
-
-        $nr_of_downvotes = 0;
-        $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('d',$d->did,1);
-        if($nr_of_issues){
-            $new_data['nr_of_issues'] = count($nr_of_issues);
-            $nr_of_downvotes+=count($nr_of_issues);
-            $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('d',$d->did,0);
-            if($downvote_agrees){
-                $nr_of_downvotes+=count($downvote_agrees);
-            }
-        }else{
-            $new_data['nr_of_issues'] = 0;
-        }
-        $new_data['nr_of_downvotes'] = $nr_of_downvotes;
+        $new_data['nr_of_issues'] = 0;
+        $new_data['nr_of_downvotes'] = 0;
+        $new_data['nr_of_likes'] = 0;
+        $new_data['nr_of_downloads'] = 0;
+        $new_data['total_downloads'] = 0;
+        $new_data['reach'] = 0;
+        $new_data['reuse'] = 0;
+        $new_data['impact_of_reuse'] = 0;
+        $new_data['reach_of_reuse'] = 0;
+        $new_data['impact'] = 0;
 
+        if ($altmetrics) {
 
-        $ld_data = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('d',$d->did);
-        $reach = 0;
-        $nr_of_likes = 0;
-        $nr_of_downloads = 0;
-        $total_downloads = 0;
-        if($ld_data){
-            foreach($ld_data as $ld){
-                if($ld->ldt=='l'){
-                    $reach += $this->CI->Gamification->getReachFromParts($ld->count,0);
-                    $nr_of_likes+=$ld->count;
-                }else if($ld->ldt=='d'){
-                    $reach += $this->CI->Gamification->getReachFromParts(0,$ld->count);
-                    $nr_of_downloads+=$ld->count;
-                    $total_downloads+=$ld->sum;
+            $nr_of_downvotes = 0;
+            $nr_of_issues = $this->CI->Downvote->getDownvotesByKnowledgePiece('d', $d->did, 1);
+            if ($nr_of_issues) {
+                $new_data['nr_of_issues'] = count($nr_of_issues);
+                $nr_of_downvotes += count($nr_of_issues);
+                $downvote_agrees = $this->CI->Downvote->getDownvotesByKnowledgePiece('d', $d->did, 0);
+                if ($downvote_agrees) {
+                    $nr_of_downvotes += count($downvote_agrees);
                 }
+            } else {
+                $new_data['nr_of_issues'] = 0;
             }
-        }
-        $new_data['nr_of_likes'] = $nr_of_likes;
-        $new_data['nr_of_downloads'] = $nr_of_downloads;
-        $new_data['total_downloads'] = $total_downloads;
-        $new_data['reach'] = $reach;
+            $new_data['nr_of_downvotes'] = $nr_of_downvotes;
+
+
+            $ld_data = $this->CI->KnowledgePiece->getNumberOfLikesAndDownloadsOnUpload('d', $d->did);
+            $reach = 0;
+            $nr_of_likes = 0;
+            $nr_of_downloads = 0;
+            $total_downloads = 0;
+            if ($ld_data) {
+                foreach ($ld_data as $ld) {
+                    if ($ld->ldt == 'l') {
+                        $reach += $this->CI->Gamification->getReachFromParts($ld->count, 0);
+                        $nr_of_likes += $ld->count;
+                    } else if ($ld->ldt == 'd') {
+                        $reach += $this->CI->Gamification->getReachFromParts(0, $ld->count);
+                        $nr_of_downloads += $ld->count;
+                        $total_downloads += $ld->sum;
+                    }
+                }
+            }
+            $new_data['nr_of_likes'] = $nr_of_likes;
+            $new_data['nr_of_downloads'] = $nr_of_downloads;
+            $new_data['total_downloads'] = $total_downloads;
+            $new_data['reach'] = $reach;
 
-        $impact_struct = $this->CI->Gamification->getImpact('d',$d->did,"2013-1-1",date("Y-m-d"));
+            $impact_struct = $this->CI->Gamification->getImpact('d', $d->did, "2013-1-1", date("Y-m-d"));
 
-        $new_data['reuse'] = $impact_struct['reuse'];
+            $new_data['reuse'] = $impact_struct['reuse'];
 
-        $new_data['impact_of_reuse'] = floor($impact_struct['recursive_impact']);
+            $new_data['impact_of_reuse'] = floor($impact_struct['recursive_impact']);
 
-        $new_data['reach_of_reuse'] = floor($impact_struct['reuse_reach']);
+            $new_data['reach_of_reuse'] = floor($impact_struct['reuse_reach']);
 
-        $new_data['impact'] = floor($impact_struct['impact']);
-       }
+            $new_data['impact'] = floor($impact_struct['impact']);
+        }
 
         return $new_data;
     }
 
-    public function checkNumeric($v) {
+    public function checkNumeric($v)
+    {
         if (is_integer($v))
             return intval($v);
         else if (is_numeric($v))
@@ -2098,13 +2200,12 @@ public function checkNumeric($v) {
             return $v;
     }
 
-    public function array_map_recursive($callback, $array) {
-       $func = function ($item) use (&$func, &$callback) {
-         return is_array($item) ? array_map($func, $item) : call_user_func($callback, $item);
-       };
+    public function array_map_recursive($callback, $array)
+    {
+        $func = function ($item) use (&$func, &$callback) {
+            return is_array($item) ? array_map($func, $item) : call_user_func($callback, $item);
+        };
 
-       return array_map($func, $array);
+        return array_map($func, $array);
     }
-
 }
-?>