Terms of interest:
Also called Scikit-learn.
train_test_split
X and y are separate things (y is the target variable/column) and X is multiple is columns used to get y.
Given any pandas df use .to_numpy to convert first.
classifier score?
-0.018 bad 0.72 good
importdata_cleaning (puts all values between -1 and 1)
skileanr.pipeline allows you to combine steps.
to save model use pickle with
with open(out_file,"wb") as out:
pickle.dump(pipe,out)