You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The flechette package provides faster and more flexible Arrow parsing than the JS reference implementation. It also provides a (more correct) implementation of the fast row object proxies used by our Arrow loader here. Should we switch out the apache-arrowtableFromIPC for the flechette version?
If so:
We can simply call flechette's version of tableFromIPC with the useProxy: true option to get fast row proxy objects.
We can add "duck typing" checks to pass existing tables (from either flechette or arrow-js) through to Vega, calling toArray() on them first to return Vega's desired array-of-objects format.
The flechette package provides faster and more flexible Arrow parsing than the JS reference implementation. It also provides a (more correct) implementation of the fast row object proxies used by our Arrow loader here. Should we switch out the
apache-arrow
tableFromIPC
for theflechette
version?If so:
tableFromIPC
with theuseProxy: true
option to get fast row proxy objects.toArray()
on them first to return Vega's desired array-of-objects format.useBigInt
,useMap
, etc).@domoritz any thoughts?
The text was updated successfully, but these errors were encountered: