rcorex-package: An Implementation of Total Correlation Explanation

rcorex-packageR Documentation

An Implementation of Total Correlation Explanation

Description

rcorex implements Total Correlation Explanation (CorEx) in R. CorEx is a computationally efficient algorithm which uses multivariate mutual information to discover latent factors which best explain correlations in a dataset. CorEx is computationally efficient and is easily extended to discover hierarchical structure.

The main function in the rcorex package is biocorex, model convergence can be visualised via plot.rcorex, and a tidygraph object of discovered structure can be created using make_corex_tidygraph.


jpkrooney/rcorex documentation built on July 25, 2022, 1:37 a.m.