[R] Problems drawing a colored 'rug' in the Lattice 'densityplot'

[Phil Spector]

Mark -
    If I understand what you want, it can be done with a custom
panel function:

mypanel = function(x,subscripts,groups,...){
   panel.densityplot(x,plot.points=FALSE,groups=groups,subscripts=subscripts,...)
   panel.rug(x,col=trellis.par.get('superpose.line')$col[groups[subscripts]])
}

Then I think you'll get the result you want if you use

densityplot(~Value|Type, group=Category, data=d, panel=mypanel)

 					- Phil Spector
 					 Statistical Computing Facility
 					 Department of Statistics
 					 UC Berkeley
 					 spector at stat.berkeley.edu

On Thu, 16 Dec 2010, Mark Dalphin wrote:

> Hi All,
>
> I'm trying to add a 'rug' representation of my data to a plot created with 
> densityplot(). While I can do this in the simple case, I can't do it properly 
> when I include the "groups" argument. I have an example below. I am running a 
> reasonably new version of R.
>
> print(sessionInfo())
> R version 2.12.0 Patched (2010-11-07 r53537)
> Platform: i686-pc-linux-gnu (32-bit)
>
> locale:
> [1] LC_CTYPE=en_NZ.UTF-8       LC_NUMERIC=C               LC_TIME=en_NZ.UTF-8 
> LC_COLLATE=en_NZ.UTF-8    [5] LC_MONETARY=C 
> LC_MESSAGES=en_NZ.UTF-8    LC_PAPER=en_NZ.UTF-8       LC_NAME=C 
> [9] LC_ADDRESS=C               LC_TELEPHONE=C 
> LC_MEASUREMENT=en_NZ.UTF-8 LC_IDENTIFICATION=C 
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base 
> other attached packages:
> [1] lattice_0.19-13
>
> R code to show the problem:
>
> ## Setup - load package, set random seed & create fake dataset
> library(lattice)
> set.seed(1234)
> d <- data.frame(Type=rep(LETTERS[1:4], times=250),
>               Category=rep(LETTERS[22:26], times=200),
>               Value=c(rnorm(500), rnorm(300, 0.5), rnorm(200, 1))
>               )
>
> ## Basic "densityplot" using 'points' to show the data
> densityplot(~Value|Type, data=d)
>
> ## And I can plot a 'rug' for the simple density plot
> densityplot(~Value|Type, data=d, plot.points='rug')
>
> ## Now add a "groups" selector to show sub-grouping of data by 'Category'
> ## Note: the data points are in color
> densityplot(~Value|Type, group=Category, data=d)
>
> ## Finally, with the groups, and with a rug.
> ## Note: no color for the rug
> densityplot(~Value|Type, group=Category, data=d, plot.points='rug')
>
>
> So, I can draw a rug (which is an improvement over version 2.9.1 of R when I 
> got no rug), however, the color associated with the 'group' doesn't seem to 
> propagate through to the rug. Is there something I am doing wrong here or is 
> this a bug? Anyone have suggestions to work around this?
>
> Regards,
> Mark
> -- 
>