rclust.null {fossil}

Relational Clustering

Description

A Monte Carlo method for calculating a null/random clustering solution based on the type and arrangement of a known clustering solution.

Usage

rclust.null(groups, dist)

Arguments

groups

cluster identity vector

dist

original pairwise distance matrix

Details

This function calculates a random/null clustering solution based on a given solution. It resamples the data and reassigns samples to groups, keeping the same group sizes.

Value

A matrix equal in rows to the number of clusters originally given, with two columns for mean within group distance and standard deviation respectively.

Author(s)

Matthew Vavrek

See Also

To cluster the data, use the function rclust; see as well rclust.weights, rclust.dist

Examples

#a null solution for the fdata example data set
data(fdata.mat)
fd.dist <- dino.dist(fdata.mat)
fd.clust <- rclust(fd.dist, 2)
rclust.null(fd.clust, fd.dist)