combss: Continuous Optimisation Towards Best Subset Selection

Best subset selection in generalised linear models via continuous optimisation. Reformulates the NP-hard discrete subset selection problem as a continuous optimisation over the hypercube [0,1]^p, solved via a Frank-Wolfe homotopy algorithm with closed-form ridge inner solves. Supports linear (Gaussian), binary logistic, and multinomial regression. For methodological details see Moka, Liquet, Zhu and Muller (2024) <doi:10.1007/s11222-024-10387-8> and Mathur, Liquet, Muller and Moka (2026) <doi:10.48550/arXiv.2603.21952>.

Version: 0.1.0
Imports: glmnet (≥ 4.0), stats
Suggests: testthat (≥ 3.0.0), knitr, rmarkdown
Published: 2026-05-11
DOI: 10.32614/CRAN.package.combss (may not be active yet)
Author: Benoit Liquet ORCID iD [aut, cre], Anant Mathur [aut], Sarat Moka [aut]
Maintainer: Benoit Liquet <benoit.liquet at univ-pau.fr>
License: GPL-3
URL: https://github.com/benoit-liquet/combss
NeedsCompilation: no
CRAN checks: combss results

Documentation:

Reference manual: combss.html , combss.pdf
Vignettes: Best subset selection with combss (source, R code)

Downloads:

Package source: combss_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=combss to link to this page.