Standardisation issue - Githubissues

pbreheny / grpreg

Regularization paths for regression models with grouped covariates

http://pbreheny.github.io/grpreg/

34 stars 14 forks source link

Standardisation issue #60

Open dannychu1108 opened 4 months ago

dannychu1108 commented 4 months ago

Hi,

A lot of variables in X were in binary and it may not be the best idea to standardize them. May I know if there is any modification that you can make to make sure my data was not standardized?

Also, I was trying to type in one lambda value to fit my Cox regression, but what I had was

Error in matrix(0, nrow = length(XG$scale), ncol = ncol(b)) : 
  non-numeric matrix extent

But the matrix worked fine when I didn't state the lambda value.

pbreheny commented 4 months ago

Standardization: I would argue that one should always standardize, no exceptions. Whether the variables are binary or not makes no difference. If you want to change the amount of penalization, you can always modify the penalty.factor option.
Error: Supply a reproducible example.