autoplot.cv.glmnet: Plot cross-validated elastic net paths

In alexpghayes/hayeslib: Hayeslib

Description

Plots paths for features in the final model in color and provides a legend for these features only. Features not in the final are plotted in grey and not included in the legend.

Usage

## S3 method for class 'cv.glmnet'
autoplot(object, rule = "1-se", ...)

Arguments

object	Object of class cv.glmnet
rule	Either "min" or "1-se". Specifies whether lambda should be selected to minimize CV error, or to find the sparsest model within one standard deviation of the minimum error. Default cv.glmnet behavior is to use the one standard error rule, following Breiman.
...	Other arguments to autoplot. Ignored.

Value

ggplot2 object with elastic net paths

Examples

library(glmnet)
data(Boston, package = "MASS")

X <- model.matrix(medv ~ ., Boston)
y <- Boston$medv
cv_model <- cv.glmnet(X, y)

autoplot(cv_model, rule = "min")

alexpghayes/hayeslib documentation built on May 5, 2019, 4:47 p.m.