Spectrum: Versatile Ultra-Fast Spectral Clustering for Single and Multi-View Data

A versatile ultra-fast spectral clustering method for single or multi-view data. 'Spectrum' uses a new type of adaptive density aware kernel that strengthens local connections in dense regions in the graph. For integrating multi-view data and reducing noise we use a recently developed tensor product graph data integration and diffusion system. 'Spectrum' contains two techniques for finding the number of clusters (K); the classical eigengap method and a novel multimodality gap procedure. The multimodality gap analyses the distribution of the eigenvectors of the graph Laplacian to decide K and tune the kernel. 'Spectrum' is suited for clustering a wide range of complex data.

Version: 0.2
Depends: R (≥ 3.5.0)
Imports: ggplot2, Rtsne, ClusterR, umap, Rfast, RColorBrewer, diptest
Suggests: knitr
Published: 2019-02-11
Author: Christopher R John
Maintainer: Christopher R John <chris.r.john86 at gmail.com>
License: AGPL-3
NeedsCompilation: no
CRAN checks: Spectrum results

Downloads:

Reference manual: Spectrum.pdf
Vignettes: Spectrum
Package source: Spectrum_0.2.tar.gz
Windows binaries: r-devel: Spectrum_0.2.zip, r-release: Spectrum_0.2.zip, r-oldrel: not available
OS X binaries: r-release: Spectrum_0.2.tgz, r-oldrel: not available
Old sources: Spectrum archive

Linking:

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