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: | 0.3 |
Depends: | R (≥ 3.5.0) |
Imports: | ggplot2, Rtsne, ClusterR, umap, Rfast, RColorBrewer, diptest |
Suggests: | knitr |
Published: | 2019-02-18 |
Author: | Christopher R John, David Watson |
Maintainer: | Christopher R John <chris.r.john86 at gmail.com> |
License: | AGPL-3 |
NeedsCompilation: | no |
CRAN checks: | Spectrum results |
Reference manual: | Spectrum.pdf |
Vignettes: |
Spectrum |
Package source: | Spectrum_0.3.tar.gz |
Windows binaries: | r-devel: Spectrum_0.3.zip, r-release: Spectrum_0.3.zip, r-oldrel: not available |
OS X binaries: | r-release: Spectrum_0.3.tgz, r-oldrel: not available |
Old sources: | Spectrum archive |
Please use the canonical form https://CRAN.R-project.org/package=Spectrum to link to this page.