ENH: Implement masked algorithm for value_counts

[phofl]

• closes #xxxx (Replace xxxx with the GitHub issue number)
• Tests added and passed if fixing a bug or adding a new feature
• All code checks passed.
• Added type annotations to new arguments/methods/functions.
• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

this adapts the same methods we are using for unique and friends. We get a 5-10% performance improvement as the Number of NAs grows

[WillAyd]

Does this close any open issues? Maybe #44946?

[phofl]

There is an issue that includes all of our algorithms (mode is missing), we can close after I finish with that

[WillAyd]

lgtm

[phofl]

merging this

[rhshadrach]

This patch may have induced a performance regression. If it was a necessary behavior change, this may have been expected and everything is okay.

Please check the links below. If any ASVs are parameterized, the combinations of parameters that a regression has been detected for appear as subbullets.

• algos.isin.IsInLongSeriesLookUpDominates.time_isin
  • dtype='float32'; MaxNumber=1000; series_type='random_hits'
  • dtype='int32'; MaxNumber=1000; series_type='random_misses'
  • dtype='float32'; MaxNumber=1000; series_type='monotone_hits'
• algos.isin.IsInFloat64.time_isin
  • dtype='Float64'; title='only_nans_values'
  • dtype=<class 'numpy.float64'>; title='only_nans_values'
  • dtype='Float64'; title='few_different_values'
  • dtype=<class 'numpy.float64'>; title='few_different_values'
• frame_methods.SeriesNuniqueWithNan.time_series_nunique_nan
• timeseries.DatetimeIndex.time_unique
  • index_type='repeated'

Subsequent benchmarks may have skipped some commits. The link below lists the commits that are between the two benchmark runs where the regression was identified.

c8e7a98...6f0cd8d

[phofl]

I agree that the regression look valid based on ASV, but 2 things point in another direction:

• I can not reproduce any of these regressions locally
• this pr didn't touch the code-path of isin at all

It's just very weird, not sure what to do

[rhshadrach]

Ah, I didn't think to check if a dependency changed. Will check that.

[phofl]

This is mostly going through our own cython code, but nothing that I’ve touched. It’s very weird