[R] using glmnet for the dataset with numerical and categorical

[yan]

Dear R users,

if all my numerical variables in my datasets having the same units, may I
leave them unnormalized, just do cv.glmnet
directly(cv.glmnet(data,standardize=FALSE))?
i know normally if there is a mixture of numerical and categorical , one has
to standardize the numerical part before applying  cv.glmnet with
standardize=fase, but that's due to the different units in the numerical
part, right? so if all the units are the same, could I skip the pre
standardize part? 

I tried both ways(standardize the numerical part and not standardize
numerical part), the results are very similar, what I don't understand is
for the coefficients of categorical variables, why they are so different in
two cases?

Many thanks

Yan

--
View this message in context: http://r.789695.n4.nabble.com/using-glmnet-for-the-dataset-with-numerical-and-categorical-tp4636279.html
Sent from the R help mailing list archive at Nabble.com.