[MRG] final edits to the class

[solegalli]

Hi @gverbock

I made tiny edits to the docstrings and the class and added a few additional asserts and tests to the test file.

I have one final concern. I am not sure split_distrinct is working as I expect. Maybe my expectation is wrong.

For example, if df["A"] = [ A, A, A, A, A, A, B, C, D, E]

• if split_distinct=False and split_frac=0.5, I expect basis to be [ A, A, A, A, A] and test [ A, B, C, D, E]
• if split_distinct=True and split_frac=0.5, I expect basis to be [ A, A, A, A, A, A] and test [ B, C, D, E]

Also when split_distinct=True and the split_col is numerical, are we obtaining what we expect?

For example, if df["A"] = [ 0, 0, 0, 0, 0, 0, 1, 2, 3, 4]

• if split_distinct=False and split_frac=0.5, I expect basis to be [ 0, 0, 0, 0, 0] and test [0, 1, 2, 3, 4]
• if split_distinct=True and split_frac=0.5, I expect basis to be [ 0, 0, 0, 0, 0, 0] and test [1, 2, 3, 4]

Are we obtaining that? can we test? or could you tell me which one is the test for this? or maybe I am interpreting the functionality wrongly?

Sorry to ask, but I can't see it directly from the logic and I've been thinking about this and going back and forth for a while now, so my brain is burnt.

I also added comments in the test file in the original PR to elaborate more on my doubts. Could you have a look at those?

Thank you!