Apply fit_transform() to each of the training folds, and transform to each of the validation folds in TabularPredictor #4085
Unanswered
alberto-jj
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi,
I am currently running Autogluon TabularPredictor on a dataset previously split into train and test sets.
I re-scaled my features using fit_transform() using an algorithm to correct batch effects (reComBat) before running the TabularPredictor.fit().
However, I would like to apply my re-scaling with fit_transform() to each of the training folds of TabularPredictor and transform each of the validation folds (AND afterward, run the rest of AutoGluon preprocessing as default).
I have seen the class autogluon.features.generators.AbstractFeatureGenerator, but I'm not sure if it will help me to get my desired output (or if I will end up just applying my "fit_transform()" re-scaling without the remaining Autogluon default preprocessing).
Here is the code I have with my re-scaling method.
Unfortunately, I cannot share the data due to ethical constraints, but its all continuous numerical features without missing values to handle, nor categories to encode.
Thanks a lot in advance for your help.
edit:format
Beta Was this translation helpful? Give feedback.
All reactions