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
This package is very appealing for the accelerated anynan,allnan etc functions. There is a common use case that this package is failing to consider though. And that is that following common numpy structure:
assertnp.all(np.isfinite(arr))
We may want to exclude inf and -inf as well as NaNs from the input. Would it be possible to add additional functions to this package that can do so? Or alternatively parameterize existing functions to enable this functionality with similar high-performance to anynan / allnan ?
The text was updated successfully, but these errors were encountered:
Skylion007
changed the title
[Enhancement] Add functions that extend anyNaN for non-finite values
[Enhancement] Add functions that extend anynan for non-finite values
Jan 30, 2021
I came here to request the same thing; found that in pandas we do np.isfinite(arr).all() and it'd be nice to short-circuit. (see also pandas-dev/pandas#32339)
This package is very appealing for the accelerated anynan,allnan etc functions. There is a common use case that this package is failing to consider though. And that is that following common numpy structure:
We may want to exclude inf and -inf as well as NaNs from the input. Would it be possible to add additional functions to this package that can do so? Or alternatively parameterize existing functions to enable this functionality with similar high-performance to anynan / allnan ?
The text was updated successfully, but these errors were encountered: