Jaccard {DIscBIO}

Jaccard’s similarity

Description

Robustness of the clusters can be assessed by Jaccard’s similarity, which reflects the reproducibility of individual clusters across bootstrapping runs. Jaccard’s similarity is the intersect of two clusters divided by the union.

Usage

Jaccard(object, Clustering = "K-means", K, plot = TRUE, R = 100)

Arguments

object

DISCBIO class object.

Clustering

Clustering has to be one of the following: ["K-means","MB"]. Default is "K-means"

K

A numeric value of the number of clusters

plot

if 'TRUE', plots the mean Jaccard similarities

R

number of bootstrap replicates

Value

A plot of the mean Jaccard similarity coefficient per cluster.