expand tests selection by target mean performance

[solegalli]

the tests available right now are a bit high level. It does not test internal functionality, for example, the calculation of the performances. It does not test threshold defaulting to None. And also, it does not test when we pass a subset of categories to the selector.

[AgustinFi]

Hi @solegalli,

I saw several RFE regarding measuring performance.
I was thinking about starting with a POC using pytest-benchmark. Is someone else working on something similar?

[solegalli]

Hi @AgustinFi

I think performance in this issue means the performance of the model based on the encoded feature. Not computational performance. Is this what you had in mind?

We have not considered pytest-benchmark yet. Not sure this is a priority?
@ChristopherGS

[Morgan-Sell]

Hi @solegalli,

Since I helped with the TargetMeanRegressor and TargetMeanClassifier, I'm happy to work on this issue. I recognize we first need to complete the classes. I believe there are a couple of items still outstanding.