提取由r中的glmnet或lars软件包选择的变量

Question

I'm dealing with a large p small n problem (p>1000, n=150) and have decided to use glmnet and lars to select possible explanatory variables (as I understand is their functionality) rather than do this manually.

I have gotten out both a lars, and a glmnet object from the functions, but now can't figure out how to extract from those objects, the names of the coefficients / the optimal model, that the functions have chosen.

Any help would be greatly appreciated!

Answer 1

That's what the coef function is for. See ?coef.glmnet for more details.

# fit a glmnet object
obj <- glmnet(x, y)

# matrix of coefficients
# each row is a variable, each column is one step in the glmnet path
coef(obj)

# coefficients for a specific model in the glmnet sequence
coef(obj, s=obj$lambda[1])
coef(obj, s=0.01)

Answer 2

I also think this should be easier, but here is one way assuming you know which spot one the regularization path you want to use, ie you know the Lambda of interest.

mod <- glmnet(x, y)
LambdaIndex <- 50 # just randomly picking one of the 100 Lambas tried
nonZerosIndices <- mod$beta[, LambdaIndex]!=0 
nonZeroVars <- names(mod$beta[nonZerosIndices, LambdaIndex])