It is common to fit and transform using StandardScaling the training data

Question

It is common to fit and transform using StandardScaling the training data and only transform the test data. I am working on the project and how to apply it to scale if I want to use cross_validate? What is the best practice? Because I cannot fit transform the whole data before passing to cross_validate.

Rewriting the function manually comes to mind.

Answer 1

Why you don't want to fit transform the whole data before passing it to cross_validate?