BisectKmeansClustering: Bisect k-means clustering of the data

In mridulaprasad/CorrDrugTumorMSI: A pipeline to combine drug distribution with tumor tissue-types in mass spectrometry imaging (MSI) data

Description

Perform bisect k-means clustering on a data-matrix.

Usage

BisectKmeansClustering(Inputdata, cluster.size = 0.5)

Arguments

cluster.size	The fraction of pixels in individual cluster. Particular cluster will not split further if contains this fracion of data-points with respect total data-points.
x	A numeric matrix of data.

Details

The data given by x is clustered by the k-means algorimthm based correlation distance. The number of clusters automatically decides according calinhara internal cluster validity index. The split of cluster into another cluster stop if the size of become less than cluster.size vlaue which the fraction of data points into compare compare to total size of input data.

Value

A vector of integers indicating the cluster to which each point is allocated.

See Also

optimalClust

Examples

x <- rbind(matrix(rnorm(500,sd=0.3),ncol=2),
          matrix(rnorm(500,mean=1,sd=0.3),ncol=2))
cl <- BisectKmeansClustering(x)
plot(x,col=cl)               

mridulaprasad/CorrDrugTumorMSI documentation built on Oct. 26, 2020, 5:45 a.m.