flow_OptimalClust: Optimal Number of Clusters for Flow Data

In ssmpsn2/flowAssist: User friendly manipulation and analysis of flow cytometry data

Description

Function implementing various statistical methods to determine the optimal number of clusters for flow cytometry data, using various clustering algorithms. ***WARNING*** extremely computationally intensive– very large data sets may not function and/or may take significant time to complete.

Usage

flow_OptimalClust(flowObj, algorithm, kMax, nBoot, plot = TRUE)

Arguments

flowObj	A data frame or flowFrame to be analyzed
algorithm	The clustering algorithm to utilize. Options (no quotations): kmeans, pam (Partitioning Around Mediods), clara (Clustering LARge Applications), hcut (Hierarchical clustering, cut tree into 'k' clusters)
kMax	Integer representing max number of possible clusters to consider
nBoot	Integer representing number of times to repeat the analysis from a random starting position, must be >1 for gap statistic analysis
plot	Logical indicating whether or not to generate graphical plots for wss, silhouette, and gap statistic methods

ssmpsn2/flowAssist documentation built on May 30, 2019, 7:12 p.m.