dswatson/cc_testr: Rigorously test cluster stability

CCtestr provides a hypothesis testing framework for cluster validation. Using a modified consensus clustering algorithm, PAC scores are calculated for data simulated from an input matrix in order to estimate the theoretical null distribution of each cluster number k's performance metrics. Results for real data are subsequently evaluated relative to these distributions. By systematically testing against well-formulated null hypotheses, test statistics and p-values are computed with high accuracy and known precision.

Getting started

Package details

LicenseGPL (>= 3)
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
dswatson/cc_testr documentation built on Aug. 18, 2017, 10:18 p.m.