bgns: Biweight Graph and Network Statistics

Provides memory-efficient biweight midcorrelation and exact bicor-based k-nearest-neighbor graph construction for dense and sparse numeric matrices. Dense, sparse, and mixed-input paths avoid materializing full dense similarity matrices for tidy and k-nearest-neighbor workflows where possible. The implementation supports pairwise finite-overlap handling and robust correlation-based graph construction for biological expression matrices and other high-dimensional numeric data.

Version:	0.4.3
Depends:	R (≥ 4.3.0)
Imports:	Matrix, methods, stats
Suggests:	knitr, rmarkdown, testthat (≥ 3.0.0)
Published:	2026-09-27
DOI:	10.32614/CRAN.package.bgns
Author:	Aditya Kshirsagar [aut, cre]
Maintainer:	Aditya Kshirsagar <adityaksh4 at gmail.com>
BugReports:	https://github.com/metaddict/bgns/issues
License:	GPL-3
URL:	https://github.com/metaddict/bgns
NeedsCompilation:	yes
SystemRequirements:	C++17, optional OpenMP
Citation:	bgns citation info
Materials:	README, NEWS
CRAN checks:	bgns results

Documentation:

Reference manual:	bgns.html , bgns.pdf
Vignettes:	bgns: Biweight Graph and Network Statistics (source, R code)

Downloads:

Package source:	bgns_0.4.3.tar.gz
Windows binaries:	r-devel: bgns_0.4.3.zip, r-release: bgns_0.4.3.zip, r-oldrel: bgns_0.4.3.zip
macOS binaries:	r-release (arm64): bgns_0.4.3.tgz, r-oldrel (arm64): bgns_0.4.3.tgz, r-release (x86_64): bgns_0.4.3.tgz, r-oldrel (x86_64): bgns_0.4.3.tgz

Linking:

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