Scaling dataframe with categorical features, then fit_transform() and transform() for X_train / X_test

Question

I have a dataframe df:

cat1 | cat2 | cat3 | year | quartal | month | y
a    | aa   | aaa  | 2020 | 3       | 9     | 12
b    | bb   | bbb  | 2019 | 2       |