Sparse Group Lasso Solver

This is the R package sglOptim version 1.3.7.

R-package Overview

This package doesn’t provide any direct functionality. It is a service package for other R packages, and it provides an implementation of an optimizer of sparse group lasso penalized objective functions. The package is used by the following other R packages:

• For fitting multivariate linear regression models using sparse group lasso see the lsgl R package.
• For fitting multinomial classification models using sparse group lasso see the msgl R package.

More details on the algorithm are given in the paper Sparse group lasso and high dimensional multinomial classification.

The loss/objective function must be defined in a C++ module.

Installation

Installation from souce requires installation of the necessary compiler tools for Mac or Windows.

Install the package from GitHub:

# install.packages("remotes") # uncomment if remotes is not already installed
remotes::install_github("nielsrhansen/sglOptim", build_vignettes = TRUE)

If you don’t want to build the vignettes when installing, just remove the build_vignettes = TRUE argument.

Author

Martin Vincent wrote the package. Niels Richard Hansen is the current maintainer.

License

GPL (>=2)