Allow for some column specification in feature transformations
ngriffiths13 opened this issue · comments
Is your feature request related to a problem? Please describe.
I want to be able to apply transformations to only specific columns when I build feature transforms into a model. For example, I may have 10 features, and I know 9 of them have no lagged information I want. But one of them does. I just want to add some lag transformations to that one variable.
Describe the solution you'd like
I want transformers to all take in an optional parameter of column names. When that column name is set to None, we do all the columns.
Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.
Additional context
This is something scikit-learn makes pretty easy in their pipelines that I think we need to mimic.