sortClusters {segmenTier}

Sort clusters by similarity.

Description

Takes a "clustering" object as returned by clusterTimeseries and uses the cluster-cluster similarity matrix, item Ccc, to sort clusters by their similarity, starting with the cluster labeled ‘1’; the next cluster is the first cluster (lowest cluster label) with the highest similarity to cluster ‘1’, and proceeding from there. The final sorting is added as item "sorting" to the cset object and returned. This sorting is subsequently used to select cluster colors and in the plot method. This simply allows for more informative plots of the clustering underlying a segmentation but has no consequence on segmentation itself.

Usage

sortClusters(cset, sort = TRUE, verb = 0)

Arguments

cset

a clustering set as returned by clusterTimeseries

sort

if set to FALSE the clusters will be sorted merely numerically

verb

level of verbosity, 0: no output, 1: progress messages

Value

Returns the input "clustering" object with a list of vectors (named "sorting"), each providing a similarity-based sorting of cluster labels.