meta.SubClustering: immunoClust Model Refinement Step in iterative...

In immunoClust: immunoClust - Automated Pipeline for Population Detection in Flow Cytometry

Description

These function tests each meta-cluster of a model for refining it into more sub-clusters and returns the refined cluster memberships in an integer array.s

Usage

meta.SubClustering(P, N, W, M, S, label, tol=1e-5, bias=0.25, alpha=1.0, 
    EM.method=20)

meta.TestSubCluster(P, N, W, M, S, J=8, B=500, tol=1e-5, bias=0.5, alpha=1.0, 
    EM.method=2, HC.samples=2000) 

Arguments

P	The number of parameters.
N	The number of clusters.
W	The N-dimensional vector with cluster weights, i.e. numbers of events in a cluster.
M	The N x P-dimensional vector with cluster means.
S	The N x P x P-dimensional vector with the cluster covariance matrices.
label	The N-dimension integer vector with the cell-cluster to meta-cluster membership.
tol	The tolerance used to assess the convergence of the EM(t)-algorithms in Sub-Clustering.
bias	he ICL-bias used in the EMt-algorithm.
alpha	A value between 0 and 1 used to balance the bhattacharrya probabilities calculated with either the full covariance matrices or using only the diagonal elements of it.
J	The number of sub-models to be builded and tested for a particular cluster.
B	The maximum number of EM(t)-iterations in Sub-Clustering.
EM.method	0 = KL-minimization not weighted 1 = BC-maximization not weighted 10 = BC-maximization weighted 2 = EMt-classification not weighted 20 = EMt-classification weighted
HC.samples	The number of samples used for initial hierarchical clustering.

Details

These function are used internally by the meta-clustering procedures meta.process and meta.Clustering in immunoClust and are not intended to be used directly.

Value

An integer array of length N containing the cell-clusters meta-cluster memberships of the refined model.

Author(s)

Till Sörensen [e-mail]

References

Sörensen, T., Baumgart, S., Durek, P., Grützkau, A. and Häupl, T. immunoClust - an automated analysis pipeline for the identification of immunophenotypic signatures in high-dimensional cytometric datasets. Cytometry A (accepted).

See Also

meta.process, meta.Clustering, meta.hclust

Examples

data(dat.exp)
d <- meta.exprs(dat.exp)
label <- rep(1,sum(d$K))
label <- meta.SubClustering(d$P, sum(d$K), d$clsEvents, d$M, d$S, label=label)
## Not run: 
r1 <- meta.ME(d$P, d$N, d$K, d$clsEvents, d$M, d$S, label)

## End(Not run)

immunoClust documentation built on Nov. 8, 2020, 5:19 p.m.