MetaClustering {FlowSOM} | R Documentation |
Cluster data with automatic number of cluster determination for several algorithms
MetaClustering(data, method, max = 20, nClus = NULL, ...)
data |
Matrix containing the data to cluster |
method |
Clustering method to use, given as a string. Options are metaClustering_consensus,metaClustering_hclust, metaClustering_kmeans,metaClustering_som |
max |
Maximum number of clusters to try out |
nClus |
Exact number of clusters to use. If not NULL, max will be ignored. |
... |
Extra parameters to pass along |
Numeric array indicating cluster for each datapoint
# Read from file, build self-organizing map and minimal spanning tree fileName <- system.file("extdata","lymphocytes.fcs",package="FlowSOM") flowSOM.res <- ReadInput(fileName, compensate=TRUE,transform=TRUE, scale=TRUE) flowSOM.res <- BuildSOM(flowSOM.res,colsToUse=c(9,12,14:18)) flowSOM.res <- BuildMST(flowSOM.res) # Apply metaclustering metacl <- MetaClustering(flowSOM.res$map$codes, "metaClustering_consensus", max=10) # Get metaclustering per cell flowSOM.clustering <- metacl[flowSOM.res$map$mapping[,1]]