[FEA] Support the min_count argument in groupby aggregations

[shwina]

Updated 5/13/2024: numeric_only is now supported (as of #10629). min_count is not yet supported.

In Pandas, groupby aggregations (e.g., max) accept the following arguments:

• min_count: the minimum number of non-null values required per group in order for the result to be non-null
• numeric_only: only aggregate numeric columns

It would be nice for cuDF to support these as well:

In [6]: df = cudf.DataFrame({'a': [1, 1, 1, 2, 2], 'b': ['a', 'b', 'c', 'd', 'e'], 'c': [1, 2, 3, 4, 5]})

In [7]: df
Out[7]:
   a  b  c
0  1  a  1
1  1  b  2
2  1  c  3
3  2  d  4
4  2  e  5

In [8]: df.groupby('a').max(numeric_only=True)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-8-612714d07c42> in <module>
----> 1 df.groupby('a').max(numeric_only=True)

TypeError: max() got an unexpected keyword argument 'numeric_only'

In [9]: df.to_pandas().groupby('a').max(numeric_only=True)
Out[9]:
   c
a
1  3
2  5

[shwina]

@vyasr assigned you based on our chat offline :)

[github-actions]

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

[github-actions]

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

[vyasr]

As of #10629 the numeric_only functionality is addressed. min_count remains unsupported for now.