ENH: implement `nlargest` and `nsmallest` for DataFrameGroupBy like SeriesGroupBy

[yoch]

The DataFrameGroupBy should get a new method nlargest to allow selecting the N largest rows for each group.

Currently this is doable by using df.sort_values(...).groupby(...).head(n) but I want to use the keep='all' parameter of nlargest which cannot be obtained with head

[yoch]

In fact, this can be done exactly in the same way than SeriesGroupBy.nlargest and SeriesGroupBy.nsmallest

pandas/pandas/core/groupby/generic.py

Lines 744 to 760 in 4bfe3d0

	@doc(Series.nlargest)
	def nlargest(self, n: int = 5, keep: str = "first"):
	f = partial(Series.nlargest, n=n, keep=keep)
	data = self._obj_with_exclusions
	# Don't change behavior if result index happens to be the same, i.e.
	# already ordered and n >= all group sizes.
	result = self._python_apply_general(f, data, not_indexed_same=True)
	return result
	@doc(Series.nsmallest)
	def nsmallest(self, n: int = 5, keep: str = "first"):
	f = partial(Series.nsmallest, n=n, keep=keep)
	data = self._obj_with_exclusions
	# Don't change behavior if result index happens to be the same, i.e.
	# already ordered and n >= all group sizes.
	result = self._python_apply_general(f, data, not_indexed_same=True)
	return result

[lorentzbao]

If still open, I would like to work on this issue.

[lorentzbao]

take