supcluster-package: Supervised Cluster Anaysis

In supcluster: Supervised Cluster Analysis

Supervised Cluster Anaysis

Description

The function clusters features under the assumption that each cluster has a random effect and there is an outcome variable that is related to the random effects by a linear regression. In this way the cluster analysis is “supervised” by the outcome variable. An alternate specification is that features in each cluster have the same compound symetric normal distribution, and the conditional distribution of the outcome given the features has the same coefficient for each feature in a cluster.

Details

Package:	supcluster
Type:	Package
Version:	1.0
Date:	2015-03-24
License:	GPL-2

The package consists of a function supcluster which reads a data frame whose columns include features and an outcome. It then peforms a cluster analysis that is supervised by the outcome as described above. The cluster analysis is performed using a Markoff Chain Monte Carlo algorythm, the output is a matrix where each row is a parameter vector consisting of the parameters of the multivariate normal distribution described above as well as the cluster membership of each of the features.

In addition there is function concordmap which produces a array with the posterior probability that each pair of features are in the same cluster and a function compare.chains used to compare these arrays for two chains in order to determine whether different chains have converged to the same set of clusters.

Author(s)

David A. Schoenfeld, Jesse Hsu Maintainer: David A. Schoenfeld <[e-mail]> ~~ The author and/or maintainer of the package ~~

References

~~ Literature or other references for background information ~~

See Also

supcluster, concordmap, compare.chains,beta.by.gene

supcluster documentation built on May 20, 2022, 1:07 a.m.