Plot of the central curves

Description

centralPlot plots distinctly the [np] most central observations, where [np] is the largest integer smaller than np, and the remaining most external ones, according to the modified band depth.

Usage

1
2
centralPlot(x, p=0.5,col.c='red',col.e='slategray',lty=c(1,3),gradient=FALSE,
            gradient.ramp=NULL,main=NULL,cex=1,...)

Arguments

x

a data matrix containing the observations (samples) by rows and the variables (genes) by columns.

p

proportion of most central samples to be displayed.

col.c

the color for the central samples, either as a character string or as a number. Ignored if gradient is TRUE.

col.e

the color for the external samples.

lty

a vector of two components with the line type of the central and external curves.

gradient

a logical value. If TRUE then the most central curves are plotted with colors according to the gradient.ramp parameter.

gradient.ramp

an optional vector of two components containing the first and last colors of the palette used to color the most central curves.

main

a character string for the plot title.

cex

the magnification to be used for the legend.

...

further graphical parameters to be passed to 'plot'.

Details

The centralPlot allows to visualise the most central curves within the dataset.

Author(s)

Sara Lopez-Pintado sl2929@columbia.edu and Aurora Torrente etorrent@est-econ.uc3m.es

References

Lopez-Pintado, S. et al. (2010). Robust depth-based tools for the analysis of gene expression data. Biostatistics, 11 (2), 254-264.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
  ## simulated data
  set.seed(0)  
  x <- matrix(rnorm(100),10,10)
  centralPlot(x,p=0.2)

  ## real data
  data(prostate)
  prost.x<-prostate[,1:100]
  prost.y<-prostate[,101]
  centralPlot(prost.x[prost.y==0,], p=0.5)  ## 50 % most central normal samples
  centralPlot(prost.x[prost.y==1,], p=0.5, gradient=TRUE, main='Tumor samples')  
                                            ## 50 % most central tumoral samples