bclust (version 1.5)

ditplot: dendrogram-image-teeth plot

Description

This is a handy function to plot a bclustvs object. The function attaches a coloured horizontal dendrogram to the left side of an image plot with the optimal grouping highlighted by a teethplot on the right.

Usage

ditplot(x, xlab = colnames(x$data), 
ylab = x$labels, xlab.cex = 1, 
ylab.cex = 1, dendrogram.lwd = 1, dendrogram.size = 2, 
xlab.mar = 3, ylab.mar = 3, image.col = rainbow(20), 
horizbar.plot = FALSE, 
horizbar.col = rev(c(heat.colors(5)[-4], "white")), 
horizbar.distance = 4,varimp = rep(0, ncol(x$data)), 
horizbar.size = 0.5, vertbar = NULL, 
vertbar.col = rainbow(max(vertbar)), 
teeth.size = 0.25, plot.width = 10)

Arguments

x

A bclustvs object.

xlab

A vector of strings elements. The labels for the clustering types automatically extracted from x.

ylab

A vector of strings. The variable labels automatically extracted from x.

xlab.cex

A positive value, the magnitude of the clustering type labels.

ylab.cex

A positive value, the magnitude of the variable labels.

dendrogram.lwd

A positive value, the thickness of lines used to plot the dendrogram.

dendrogram.size

A positive value, the size of the dendrogram plot.

xlab.mar

A positive value, the margin reserved to write variable labels.

ylab.mar

A positive value, the margin reserved to write type labels.

image.col

Colours used for the image plot.

horizbar.plot

A logical value, if TRUE a horizontal bar is plotted according to categorised varimp.

horizbar.col

Colours used for the horizontal bar.

horizbar.distance

A positive value, the distance between the horizbar and the image plot.

varimp

A numerical vector denoting the importance of variables. We propose to use the imp function to compute these values. If it is specified, the variables will be ordered respect to this vector.

horizbar.size

A positive value, the size of the horizontal bar.

vertbar

A positive integer vector that may be used to draw an additional vertical bar on the right of the teeth plot. This may be helpful to represent another optional grouping on the data.

vertbar.col

The colours used to plot the additional vertical bar.

teeth.size

A positive value, the size of the teeth plot.

plot.width

A positive value, the width of the whole plot. If the plot region is unbalanced in width and height, adjust this value.

Details

The varimp is assumed to be log Bayes factors and therefore categorised according to Kass and Raftery (1995) for a better visualisation.

References

Kass and Raftery (1995) Bayes Factors, Journal of the American Statistical Association, Vol. 90, pp. 773--795.

See Also

dptplot, teethplot,profileplot,viplot.

Examples

data(gaelle)

gaelle.bclust<-bclust(gaelle,
transformed.par=c(-1.84,-0.99,1.63,0.08,-0.16,-1.68))
ditplot(gaelle.bclust,varimp=imp(gaelle.bclust)$var,horizbar.plot=TRUE,
plot.width=5,horizbar.size=0.2,ylab.mar=4)
#unreplicated clustering

wildtype<-rep(1,55) #initiate a vector
wildtype[c(1:3,48:51,40:43)]<-2 #associate 2 to wildtypes
ditplot(gaelle.bclust,varimp=imp(gaelle.bclust)$var,horizbar.plot=TRUE,
plot.width=5,horizbar.size=0.2,vertbar=wildtype,
vertbar.col=c("white","violet"),ylab.mar=4)
#mark wildtype plants using violet 

gaelle.id<-rep(1:14,c(3,rep(4,13))) 
# first 3 rows replication of ColWT, 4 for the rest
gaelle.lab<-c("ColWT","d172","d263","isa2",
"sex4","dpe2","mex1","sex3","pgm","sex1","WsWT","tpt","RLDWT","ke103")
gaelle.bclust<-bclust(gaelle,rep.id=gaelle.id,
labels=gaelle.lab,transformed.par=c(-1.84,-0.99,1.63,0.08,-0.16,-1.68))
ditplot(gaelle.bclust,varimp=imp(gaelle.bclust)$var,horizbar.plot=TRUE)
#replicated clustering