search

feature-engine / feature_engine

Feature engineering package with sklearn like functionality
https://feature-engine.trainindata.com/
BSD 3-Clause "New" or "Revised" License
1.79k stars 304 forks source link

create CONTRIBUTING.md #735

Open MetroCat69 opened 2 months ago

MetroCat69 commented 2 months ago

Maybe we can create a CONTRIBUTING.md file explaining stuff about the project structure and how for example you should create a new transformer. I am new to this project so I don't feel qualified

solegalli commented 2 months ago

We have contributing guidelines in the docs: https://feature-engine.trainindata.com/en/latest/contribute/index.html

I think it's fair to find it intimidating.

My suggestion would be: if you want to create a new selector class, which I think it's what you do from our discussion in a different issue, take any selector class that already exits and use it as template. I think that for MRMR, you can take the correlation selector, and then replace the fit method with the new logic. And we take it from there.

I can point you in the right direction all along.

Hope that helps Cheers