What is “random-state” in sklearn.model_selection.train_test_split example?

Question

I am really new to machine learning,i was going through some example on sklearn

Can someone explain me what really \"Random-state\" means in below e

Answer 1

If the random_state is always fixed (42), doesn't that go against the Machine Learning perspective, in that it supposed to use randomness to help it discover the best possible outcomes?

For debugging I understand a fixed randomizer.. But when doing the "real" training should we use a truly random seed?