Set intercept = FALSE change dramatically the result

marjoleinF / pre

an R package for deriving Prediction Rule Ensembles

58 stars 17 forks source link

Set intercept = FALSE change dramatically the result #20

Closed elier closed 4 years ago

elier commented 5 years ago

When I set "intercept = FALSE", the result change dramatically. I was expecting only that the (Intercept) line disappears and result stay very similar.

marjoleinF commented 5 years ago

In any regression model, the estimated model will change after removing the intercept. This seems not an issue with pre(), but the result I would expect by passing intercept = FALSE to cv.glmnet(). Why do you want to exclude the intercept?