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
In [3]: import pandas as pd
In [4]: s = pd.Series([True, False, True, False])
In [5]: s.where(s,1)
Out[5]:
0 1
1 1
2 1
3 1
dtype: int32
on windows the scalar is not coerced to int64 (as it is on other platforms). IIRC .where already has quite complicated logic to do this. How simple is this to fix?
oh it IS caused by numpy as that defaults to int32 when creating integers on windows. However pandas coerces everything internally to int64. So just need to investigate if this is by a bug in the logic somewhere. Or we can just change the test. I'll take a look.
xref #12841
this is a test in
test_coercion.py, L361
on windows the scalar is not coerced to int64 (as it is on other platforms). IIRC
.where
already has quite complicated logic to do this. How simple is this to fix?cc @sinhrks
The text was updated successfully, but these errors were encountered: