ENH: Add numeric_only to resampler methods

[lorentzbao]

• closes ENH: resample.sum documents a numeric_only argument #46442
• Tests added and passed if fixing a bug or adding a new feature
• All code checks passed.
• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

After examining the reported issue #46442, I found the issue exists in all the following docs.
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.sum.html
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.prod.html
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.min.html
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.max.html
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.first.html
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.last.html

The problem seems to be caused by the fact that the docstrings of sum, prod, min, max, first, last methods
in the Resampler class borrow the _groupby_agg_method_template template from GroupBy class. I think one possible fix is
creating a _resample_agg_method_template and using this template to create docstrings for the methods mentioned above.

I am a first-time contributor. Appreciate any suggestions and ideas for fixing this issue.

[jreback]

can you show the rendered doc-string and point to the current one for comparison

[lorentzbao]

Thanks @jreback. The re-rendered doc-string of pandas.core.resample.Resampler.first looks as follows.

Resampler.first(_method='first', min_count=0, *args, **kwargs)[source]¶

Compute first of group values.

Parameters

min_count: int, default 0

The required number of valid values to perform the operation. If fewer than min_count non-NA values are present the result will be NA.

Returns

Series or DataFrame

Computed first of values within each group.

===========================================================================
For comparison, the current doc-string of pandas.core.resample.Resampler.first looks as follows.
https://pandas.pydata.org/docs/reference/api/pandas.core.resample.Resampler.first.html

Resampler.first(_method='first', min_count=0, *args, **kwargs)[source]¶

Compute first of group values.

Parameters

numeric_only: bool, default False

Include only float, int, boolean columns. If None, will attempt to use everything, then use only numeric data.

min_count: int, default -1

The required number of valid values to perform the operation. If fewer than min_count non-NA values are present the result will be NA.

Returns

Series or DataFrame

Computed first of values within each group.

[jreback]

cc @rhshadrach

[rhshadrach]

Thanks for the PR @lorentzbao; this is indeed the path forward mentioned in the linked issue. However, I did not realize that we'd want to deprecate the dropping of nuisance columns in resample similar to what we're doing with with Series/DataFrame/groupby. See #46560 for details (especially the two linked issues there). For this, it seems necessary to add numeric_only to these methods, so that when the deprecation is enforced users have an easy way to recover the 1.x behavior (namely, by specifying numeric_only=True).

Do you think you could add numeric_only to the methods here instead of removing the argument from the documentation?

[lorentzbao]

Thanks for the PR @lorentzbao; this is indeed the path forward mentioned in the linked issue. However, I did not realize that we'd want to deprecate the dropping of nuisance columns in resample similar to what we're doing with with Series/DataFrame/groupby. See #46560 for details (especially the two linked issues there). For this, it seems necessary to add numeric_only to these methods, so that when the deprecation is enforced users have an easy way to recover the 1.x behavior (namely, by specifying numeric_only=True).

Do you think you could add numeric_only to the methods here instead of removing the argument from the documentation?

Really appreciate your comments @rhshadrach. Let me go through the links you shared. I think I could have a try adding numeric_only to these methods.

[rhshadrach]

I should have also mentioned - see #46708 for similar examples where this argument was added to frame methods.

[lorentzbao]

I should have also mentioned - see #46708 for similar examples where this argument was added to frame methods.

Thanks @rhshadrach. I have added numeric_only to resample sum, prod, min, max, first, last methods, and created a test for this implementation. I'm not sure if I'm on the right track. Appreciate any comments or suggestions.

[rhshadrach]

Looking real good; I think the only test to add (in addition to request below) would be Series raising if numeric_only=True is passed.

[rhshadrach]

lgtm

[rhshadrach]

ping @jreback

[jreback]

lgtm failures look u related

[jreback]

thanks @lorentzbao