rect.hclust.labels: Annotate the rectangles produced by rect.hclust using...

In drmjc/mjcgraphics: mjc's graphics library

Description

Why? plclust is great for plotting heirarchical clusters; rect.hclust is great for selecting clusters at a given distance threshold; what is lacking is knowing which clusters are which. This is espescially useful if you then do some downstream analysis on each cluster, if you end up referring to your clusters numerically.

Usage

  rect.hclust.labels(tree, k = NULL, which = NULL,
    x = NULL, h = NULL, border = 2, cluster = NULL)

Arguments

tree	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
k	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
which	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
x	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
h	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
border	see rect.hclust. NB this must be identical to those used in your call to rect.hclust
cluster	see rect.hclust. NB this must be identical to those used in your call to rect.hclust

Value

annotates a plot with indices

Author(s)

Mark Cowley, 2009-10-30

Examples

m <- matrix(rnorm(200), nrow=20)
hc <- hclust(dist(m))
plot(hc)
rect.hclust(hc, k=5)
rect.hclust.labels(hc, k=5)

drmjc/mjcgraphics documentation built on May 15, 2019, 2:40 p.m.