how to use glmnet without regularization

Question

I have read that glmnet can be used without regularization, i.e. it can be used as a regular glm. I am writing a thesis and trying to avoid using to many different packages, so it would be convenient to use glmnet to do a regular glm logistic regression fitting. Can anyone help me?

score 5 · Answer 1 · edited Aug 14 '15 at 14:21

5

Setting both tuning parameters alpha and lambda to zero should give you a "normal" logistic regression without regularization.

edited Aug 14 '15 at 14:21

durron597

31,968
17
99
158

answered Aug 14 '15 at 14:04

dende85

51
1
3

1

I would add, I think setting lambda to zero is sufficient. – Arnaud A Nov 10 '15 at 15:35

how to use glmnet without regularization

1 Answers1