Ckmeans.1d.dp: Optimal and Fast Univariate k-Means Clustering

A dynamic programming algorithm for optimal univariate k-means clustering. Minimizing the sum of squares of within-cluster distances, the algorithm guarantees optimality and reproducibility. Its advantage over heuristic clustering algorithms in efficiency and accuracy is increasingly pronounced as the number of clusters k increases. With optional weights, the algorithm can also analyze 1-D signals for segmentation and peak calling. An auxiliary function can generate adaptive histograms to make patterns in data stand out. For univariate data analysis, the package provides a powerful alternative to heuristic clustering algorithms.

Version: 3.4.6-5
Depends: R (≥ 2.10.0)
Suggests: testthat, knitr, rmarkdown
Published: 2016-12-05
Author: Joe Song [aut, cre], Haizhou Wang [aut]
Maintainer: Joe Song <joemsong at>
License: LGPL (≥ 3)
NeedsCompilation: yes
Citation: Ckmeans.1d.dp citation info
Materials: NEWS
CRAN checks: Ckmeans.1d.dp results


Reference manual: Ckmeans.1d.dp.pdf
Vignettes: Tutorial: Optimal univariate k-means clustering
Tutorial: adaptive versus regular histograms
Package source: Ckmeans.1d.dp_3.4.6-5.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X Mavericks binaries: r-release: Ckmeans.1d.dp_3.4.6-5.tgz, r-oldrel: Ckmeans.1d.dp_3.4.6-5.tgz
Old sources: Ckmeans.1d.dp archive

Reverse dependencies:

Reverse imports: gsrc
Reverse suggests: FunChisq, xgboost


Please use the canonical form to link to this page.