popsom: An Efficient Implementation of Kohonen's Self-Organizing Maps (SOMs) with Starburst Visualizations

Kohonen's self-organizing maps with a number of distinguishing features: (1) An efficient, single threaded, stochastic training algorithm inspired by ideas from tensor algebra. Provides significant speedups over traditional single-threaded training algorithms. No special accelerator hardware required (see <doi:10.1007/978-3-030-01057-7_60>). (2) Automatic centroid detection and visualization using starbursts. (3) Two models of the data: (a) a self organizing map model, (b) a centroid based clustering model. (4) A number of easily accessible quality metrics for the self organizing map and the centroid based cluster model (see <doi:10.1007/978-3-319-28518-4_4>).

Version: 6.0
Imports: fields, graphics, ggplot2, hash, stats, grDevices
Suggests: testthat (≥ 3.0.0)
Published: 2021-12-20
Author: Lutz Hamel [aut, cre], Benjamin Ott [aut], Gregory Breard [aut], Robert Tatoian [aut], Michael Eiger [aut], Vishakh Gopu [aut]
Maintainer: Lutz Hamel <lutzhamel at uri.edu>
BugReports: https://github.com/lutzhamel/popsom/issues
License: GPL-3
URL: https://github.com/lutzhamel/popsom
NeedsCompilation: yes
Materials: NEWS
CRAN checks: popsom results

Documentation:

Reference manual: popsom.pdf

Downloads:

Package source: popsom_6.0.tar.gz
Windows binaries: r-devel: popsom_6.0.zip, r-devel-UCRT: popsom_6.0.zip, r-release: popsom_5.2.zip, r-oldrel: popsom_5.2.zip
macOS binaries: r-release (arm64): popsom_6.0.tgz, r-release (x86_64): popsom_5.2.tgz, r-oldrel: popsom_5.2.tgz
Old sources: popsom archive

Linking:

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