rsvddpd: Robust Singular Value Decomposition using Density Power Divergence

Computing singular value decomposition with robustness is a challenging task. This package provides an implementation of computing robust SVD using density power divergence (<doi:10.48550/arXiv.2109.10680>). It combines the idea of robustness and efficiency in estimation based on a tuning parameter. It also provides utility functions to simulate various scenarios to compare performances of different algorithms.

Version:	1.0.1
Imports:	Rcpp (≥ 1.0.5), MASS, stats, utils, matrixStats
LinkingTo:	Rcpp, RcppArmadillo
Suggests:	knitr, rmarkdown, microbenchmark, pcaMethods, V8
Published:	2025-09-20
DOI:	10.32614/CRAN.package.rsvddpd
Author:	Subhrajyoty Roy [aut, cre]
Maintainer:	Subhrajyoty Roy <subhrajyotyroy at gmail.com>
BugReports:	https://github.com/subroy13/rsvddpd/issues
License:	MIT + file LICENSE
URL:	https://github.com/subroy13/rsvddpd
NeedsCompilation:	yes
Materials:	README, NEWS
CRAN checks:	rsvddpd results

Documentation:

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

Downloads:

Package source:	rsvddpd_1.0.1.tar.gz
Windows binaries:	r-devel: rsvddpd_1.0.1.zip, r-release: rsvddpd_1.0.1.zip, r-oldrel: rsvddpd_1.0.1.zip
macOS binaries:	r-release (arm64): rsvddpd_1.0.1.tgz, r-oldrel (arm64): rsvddpd_1.0.1.tgz, r-release (x86_64): rsvddpd_1.0.1.tgz, r-oldrel (x86_64): rsvddpd_1.0.1.tgz
Old sources:	rsvddpd archive

Linking:

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