log cp transformer should NOT add a constant when variables are strictly positive

feature-engine / feature_engine

Feature engineering package with sklearn like functionality

https://feature-engine.trainindata.com/

BSD 3-Clause "New" or "Revised" License

1.87k stars 310 forks source link

log cp transformer should NOT add a constant when variables are strictly positive #738

Closed solegalli closed 1 month ago

solegalli commented 5 months ago

that would be a better implementation than the current one.

Also, add a test to corroborate that the transformer can parse int variables into float after the transformation, which is deprecated in future pandas version. The functionality is already there, we just need a test.