DOC: Index.where(..., other) other could be a Series/Index but not documented as such

[loicdiridollou]

Pandas version checks

• I have checked that the issue still exists on the latest versions of the docs on main here

Location of the documentation

https://pandas.pydata.org/docs/dev/reference/api/pandas.Index.where.html#pandas.Index.where

Documentation problem

Right now the docs for where for an Index mention that the other argument can be an array-like. As opposed to Series.where and DataFrame.where where other can be a Series or DataFrame.

At runtime this works totally fine with a Series or an Index:

import pandas as pd
import numpy as np

datetime_index = pd.DatetimeIndex(pd.date_range(start="2025-01-01", freq="h", periods=48))
mask = np.ones(48, dtype=bool)

condition = datetime_index.where(mask,datetime_index - pd.Timedelta(days=1))

Suggested fix for documentation

Should we allow Index/Series type for other?

Happy to make the PR, we have added this in a PR in pandas-dev/pandas-stubs#1420 so waiting for confirmation of the pandas team to move forward with it or not.
Thanks!

[YilangHe]

Wonder if you still need help in this issue, interest to help

[loicdiridollou]

Hey @YilangHe, I need the confirmation from the pandas team before moving forward with anything.

[rhshadrach]

Both Series and Index are considered array-like. However I don't believe we define this anywhere in the docs, nor do I have a good idea on where it can go. pandas defines "array-like" as being "list-like with a dtype attribute".

[loicdiridollou]

Not sure if there is a good place to put it, we had a doubt on the definition of array-like and if Series and Index fell into that category, thanks for the precision @rhshadrach!
Let me think if there is a right place otherwise I will close the issue.