Allow standardization from glmnet

[frousseu]

Hello!

From what I have read, glmnet standardizes/scales variables internally and returns everything on the original scale. It would be useful if maxnet would allow setting standardize = TRUE to take advantage of the glmnet internal standardization. Currently, standardization is set to FALSE forcing the user to scale variables before the modeling (and descaling after). I'm not sure what would be the best way to add this functionality. Either through an explicit argument or through the ...?

Thanks!
François Rousseu