CompositionalLASSO: LASSO Compositional Data Regression

LASSO (Least Absolute Shrinkage and Selection Operator) regression for compositional responses and Euclidean predictors, or Euclidean responses and compositional predictors with and without zero values present. The choice of the penalty parameter, lambda, takes place using K-fold cross-validation. If the compositional data appear on the predictors side the alpha-transformation is employed first. Two relevant papers are Li Z. et al. (2018) <doi:10.1007/s12561-018-9219-2> and Alenazi A.A. (2022) <doi:10.18187/pjsor.v18i4.3969>.

Version: 1.1
Depends: R (≥ 4.0)
Imports: Compositional, glmnet, Rfast, stats
Suggests: Rfast2
Published: 2026-11-20
DOI: 10.32614/CRAN.package.CompositionalLASSO
Author: Michail Tsagris [aut, cre]
Maintainer: Michail Tsagris <mtsagris at uoc.gr>
License: GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
NeedsCompilation: no
CRAN checks: CompositionalLASSO results

Documentation:

Reference manual: CompositionalLASSO.html , CompositionalLASSO.pdf

Downloads:

Package source: CompositionalLASSO_1.1.tar.gz
Windows binaries: r-devel: CompositionalLASSO_1.0.zip, r-release: CompositionalLASSO_1.0.zip, r-oldrel: CompositionalLASSO_1.0.zip
macOS binaries: r-release (arm64): CompositionalLASSO_1.1.tgz, r-oldrel (arm64): CompositionalLASSO_1.1.tgz, r-release (x86_64): CompositionalLASSO_1.1.tgz, r-oldrel (x86_64): CompositionalLASSO_1.1.tgz
Old sources: CompositionalLASSO archive

Linking:

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