sglssnal: Sparse-Group Lasso via Semismooth Newton Augmented Lagrangian

Implements the sparse-group lasso method of Zhang et al. (2020) <doi:10.1007/s10107-018-1329-6>. Unlike many widely available methods based on first-order descent, this method uses second-order information to solve the dual optimization problem via a semismooth Newton method.

Version: 0.1.0
Depends: R (≥ 4.4.0)
Imports: Matrix, Rcpp, RSpectra, methods, utils
LinkingTo: Rcpp, RcppArmadillo
Suggests: knitr, rmarkdown, spelling, testthat (≥ 3.0.0)
Published: 2026-09-11
DOI: 10.32614/CRAN.package.sglssnal (may not be active yet)
Author: Robin Liu [aut, cre], Yangjing Zhang [ctb] (Author of the original MATLAB SSNAL implementation this package ports)
Maintainer: Robin Liu <robin28liu at gmail.com>
BugReports: https://github.com/roobnloo/sglssnal/issues
License: MIT + file LICENSE
URL: https://github.com/roobnloo/sglssnal
NeedsCompilation: yes
Language: en-US
Citation: sglssnal citation info
Materials: NEWS
CRAN checks: sglssnal results

Documentation:

Reference manual: sglssnal.html , sglssnal.pdf
Vignettes: Introduction to sglssnal (source, R code)

Downloads:

Package source: sglssnal_0.1.0.tar.gz
Windows binaries: r-devel: not available, r-release: not available, r-oldrel: not available
macOS binaries: r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available

Linking:

Please use the canonical form https://CRAN.R-project.org/package=sglssnal to link to this page.