optimTree.plot {compdiagTools}R Documentation

Produces a clusterplot with optimal rearrangement of samples

Description

Function is based on original clusterplot code by Mark Wilkinson. For a given expression matrix, hierarchical clustering with average linkage is performed on Euclidean distances. The samples (colums) are rearranged using optimTree.

Usage

optimTree.plot(x, optimize = TRUE, row.scale = TRUE, zlim = 2, max.samp = 15, barplot = FALSE, col = "color1")

Arguments

x Matrix of expression values with genes in rows and samples in columns OR an exprSet.
optimize Logical value. If TRUE, optimTree is used in succession with hclust.
row.scale Logical value. If TRUE, genes are scaled before plotting to derive nicer colors.
zlim Limit of plotted values. All values exceeding +-zlim are encoded by the brightest colors. Maybe increase the level if row.scale=FALSE.
max.samp Maximal sample size. If there are more samples, no optimization is done. Note that computation time increases with sample size.
barplot Logical value. If TRUE, the encoding color bar is plotted instead of a clusterplot.
col Color schemes color1 to color3.

Value

Graphical output.

Author(s)

Stefanie Scheid <stefanie.scheid@molgen.mpg.de>, clusterplot code by Mark Wilkinson

See Also

optimTree,optimTree.dendrogram

Examples

a <- matrix(rnorm(1000),ncol=10)
optimTree.plot(a)
optimTree.plot(a,barplot=TRUE)

[Package compdiagTools version 1.5.3 Index]