Closed
Description
Code Sample
>>> import pandas as pd
>>> pd.__version__
'1.0.0rc0+183.g4da554f75'
>>> df = pd.DataFrame(
... {
... "A": [1, 1, 1, 1],
... "B": [0, 1, 2, 3]
... }
... )
>>> df
A B
0 1 0
1 1 1
2 1 2
3 1 3
>>> df.groupby("A").corrwith(df) # Clearly corrwith is a reduction rather than transformation
B A
A
1 1.0 NaN
On current master (1.0.0rc0+183.g4da554f75
) we have "corrwith" in the list of transformation_kernels
rather than reduction_kernels
within groupby.base
.
We probably have to move it and fix any impact it has.