Pandas QST: Why do frame GroupBy objects and have different capitalization from RollingGroupby, etc.?

Research

• [X] I have searched the [pandas] tag on StackOverflow for similar questions.

• [X] I have asked my usage related question on StackOverflow.

Link to question on StackOverflow

N/A, not a usage question

Question about pandas

I'm trying to write a pandas API for another execution engine, so I've been reading a lot more pandas (and pyspark.pandas) code. One question that came up as I was implementing group-by functionality is--why do GroupBy objects (including SeriesGroupBy and DataFrameGroupBy) have different capitalization for "group by" than BaseWindowGroupby, RollingGroupby, etc.? Is it just historical, or is there some justification that I'm missing?

Comment From: mroeschke

It's just historical. Could be changed to capitalize the b but since these are psudo-public not sure if its worth the churn to change in pandas

Comment From: rhshadrach

I agree these likely aren't used much by users (could be for type-hinting, but I'd guess that's uncommon), they may be used more by those who create a pandas-like API. If we are going to change, in my mind sooner would be better than later.

Comment From: jbrockmendel

I’d be OK with deprecating to get consistency.

Comment From: rhshadrach

@mroeschke - unless you have any objection, I think we should go ahead with the alias & deprecation in 2.x. I'm removing the Needs Discussion, please reinstate it if you wish.

Comment From: mroeschke

No objection on my end. Happy to deprecate it to get consistency too