cv.adaptive_lasso: Adaptive lasso with cross-validation, using lasso(BIC) as...

In pbreheny/hdrm: High-Dimensional Regression Modeling

Adaptive lasso with cross-validation, using lasso(BIC) as initial estimator

Description

Using cv.glmnet() to carry out the adaptive lasso will result in an underestimation of error unless you also apply cross-validation to the initial estimator; this function provides a wrapper to do that.

Usage

cv.adaptive_lasso(X, y, nfolds = 10, fold)

Arguments

X	Design matrix, as in glmnet()
y	Response vector, as in glmnet()
nfolds	Number of cv folds (default: 10)
fold	Which fold each observation belongs to. By default the observations are randomly assigned.

Examples

attach_data(pollution)
cvfit <- cv.adaptive_lasso(X, y)
plot(cvfit)
summary(cvfit)

pbreheny/hdrm documentation built on March 29, 2025, 5:18 a.m.