PaulWestenthanner

Hi Will, I think I could do some work in the next couple of months. I guess starting with reviewing the quantile encoder #303 makes sense, does it?

That makes sense. I'll deal with it.

@iuiu34 I haven't had time to look at it yet (probably not before next week). If you want to create a PR for this please go for it. What exactly...

for 1. I think would make most sense to keep the `get_feature_names` function but flag it with a deprecation warning and let it just call the new `get_feature_names_out` rather than...

Thanks for that issue report. From the top of my head I'd guess this is because most encoders convert the input to dataframes and create a deep copy of it....

https://github.com/scikit-learn-contrib/category_encoders/blob/5e9e803c9131b377af305d5302723ba2415001da/category_encoders/one_hot.py#L340 that should be the relevant line

Hi @Timaos123 can you please provide the version of category encoders you are using as well as a reproducible code snippet leading to the issue you describe

Hi @nercisla Thanks for your work. I created a draft pull request just because it is easier and more transparent to make comments and have the tests run. I hope...

Hi @nercisla thanks for updating the PR. I think it looks really good now. The only open issue I see is the test that checks if a trivial hierarchy yields...

Great! Thanks for implementing the multilevel hierarchy. Maybe it would be nice to give an example of the hierarchy parameter in the docstrings since this is getting a little complicated...