Lasso model of the prostate cancer data set

The lars package was used to obtain a full path of lasso solutions.

Usage

data("prostateLasso")

Format

A list of 4 data.frames: path for the piecewise linear coefficient path, residuals for the prediction error of every model at every data point, models with one row per regularization parameter, error for total train/validation prediction error for each regularization parameter.

Source

https://github.com/tdhock/animint-book/blob/master/Ch11-lasso.Rmd