profile_matrix_kmeans: Profile Matrix k-means Clustering

In lakhanp1/chipmine: Analyze ChIP-seq Data

Description

This function perform k-means clustering on the profile matrix using stats::kmeans() function. It tries to cluster the matrix by default "Hartigan-Wong" algorithm. If this algorithm fails at Quick-TRANSfer stage, clustering is done with "MacQueen"

Usage

profile_matrix_kmeans(
  mat,
  km,
  clustFile,
  name,
  kmIter = 100,
  kmStarts = 150,
  ...
)

Arguments

mat	profile matrix
km	number of cluster centers
clustFile	File to which cluster data will be stored
name	Sample name
kmIter	iter.max argument of stats::kmeans() function. Default: 100
kmStarts	nstart argument of stats::kmeans() function. Default: 150
...	other arguments for stats::kmeans() function

Value

A list with two elements:

• km: kmeans clustering object returned by stats::kmeans

• geneClusters: a dataframe with columns: geneId, cluster

Examples

NA

lakhanp1/chipmine documentation built on March 6, 2021, 9:06 a.m.