gespeR (version 1.4.2)

lasso.rand: Randomized Lasso

Description

Based on Meinshausen and Buehlmann (2009)

Usage

lasso.rand(x, y, weakness = 1, subsample = 1:nrow(x), dfmax = (ncol(x) +
  1), lambda = NULL, standardize = FALSE, intercept = FALSE, ...)

Arguments

x

The design matrix

y

The response vector

weakness

The weakness parameter

subsample

The data subsample (default: none)

dfmax

The maxiumum number of degrees of freedom

lambda

The regularisation parameter

standardize

Indicator, wheter to standardize the design matrix

intercept

Indicator, whether to fit an intercept

...

Additional arguments to glmnet

Value

• A glmnet object

Examples

y <- rnorm(50)
 x <- matrix(runif(50 * 20), ncol = 20)
 lasso.rand(x = x, y = y)