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
If dask.array.random is used with non-dask limits, say numpy floats, it intermediately creates the requested random arrays as non-dask. This can require much memory and is, I believe, not the expected behavior.
I would have created an issue, if it didn't seem explicitly designed this way. The origin is in _wrap_func in _broadcast_any where explixitly numpy's broadcast_to is used.
Is this actually the desired behavior, or could this be changed?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
If dask.array.random is used with non-dask limits, say numpy floats, it intermediately creates the requested random arrays as non-dask. This can require much memory and is, I believe, not the expected behavior.
I would have created an issue, if it didn't seem explicitly designed this way. The origin is in
_wrap_funcin_broadcast_anywhere explixitly numpy'sbroadcast_tois used.Is this actually the desired behavior, or could this be changed?
(There is a related issue for my use case -> Ouranosinc/xsdba#113)
Beta Was this translation helpful? Give feedback.
All reactions