Mhorseshoe: Approximate Algorithm for Horseshoe Prior

Provides an approximate algorithm for the horseshoe estimator used in Bayesian linear models. By implementing a sampler with high computational cost in the 'Rcpp' package and using an approximate algorithm that reduces matrix calculation complexity, parameter estimation speed for high-dimensional sparse data is faster. The approximate algorithm is described in Johndrow et al. (2020) <https://www.jmlr.org/papers/v21/19-536.html>.

Version: 0.1.2
Imports: stats, Rcpp (≥ 1.0.11)
LinkingTo: Rcpp
Suggests: knitr, rmarkdown, ggplot2, horseshoe
Published: 2023-11-24
Author: Kang Mingi [aut, cre], Lee Kyoungjae [aut]
Maintainer: Kang Mingi <leehuimin115 at g.skku.edu>
License: MIT + file LICENSE
NeedsCompilation: yes
Materials: README NEWS
CRAN checks: Mhorseshoe results

Documentation:

Reference manual: Mhorseshoe.pdf
Vignettes: Mhorseshoe

Downloads:

Package source: Mhorseshoe_0.1.2.tar.gz
Windows binaries: r-devel: Mhorseshoe_0.1.2.zip, r-release: Mhorseshoe_0.1.2.zip, r-oldrel: Mhorseshoe_0.1.2.zip
macOS binaries: r-release (arm64): Mhorseshoe_0.1.2.tgz, r-oldrel (arm64): Mhorseshoe_0.1.2.tgz, r-release (x86_64): Mhorseshoe_0.1.2.tgz
Old sources: Mhorseshoe archive

Linking:

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