R help - Sep 2009 - Access to conditioning variables (lattice)

[using R version 2.8.1 (2008-12-22)]

Hello,

I'm trying to access the conditioning variables of an xyplot within a
'panel' function but I have not been able to figure out how to do so.
Here is a simple example that describes what I wish to do (the problem
lies with the commented line):

dataset <- data.frame(x = c(1,2), y = c(4,5), Type =
factor(c("a","b")))
xyplot( y ~ x | Type, dataset, 
        panel = function(...) {
            panel.xyplot(...)
#           do_something_with(conditioning_variables[which.packet()])
        })

The problem I am facing is that I do not know how to generically access
the conditioning variables within the panel function. In this simple
case, I can achieve what I want to do with the following call :

    do_something_with(Type[which.packet()])

but that requires the panel function call to have prior knowledge of the
object used as the conditioning variable, which is not flexible enough
for my needs.

Thank you,
Martin D. Lepage

> -----Original Message-----
> From: r-help-bounces at r-project.org 
> [mailto:r-help-bounces at r-project.org] On Behalf Of Martin D. Lepage
> Sent: Thursday, September 24, 2009 7:38 AM
> To: r-help at r-project.org
> Subject: [R] Access to conditioning variables (lattice)
> 
> [using R version 2.8.1 (2008-12-22)]
> 
> Hello,
> 
> I'm trying to access the conditioning variables of an xyplot within a
> 'panel' function but I have not been able to figure out how to do
so.
> Here is a simple example that describes what I wish to do (the problem
> lies with the commented line):
> 
> dataset <- data.frame(x = c(1,2), y = c(4,5), Type = 
> factor(c("a","b")))
> xyplot( y ~ x | Type, dataset, 
>         panel = function(...) {
>             panel.xyplot(...)
> #           do_something_with(conditioning_variables[which.packet()])
>         })
> 
> The problem I am facing is that I do not know how to 
> generically access
> the conditioning variables within the panel function. In this simple
> case, I can achieve what I want to do with the following call :
> 
>     do_something_with(Type[which.packet()])
If your panel function has an argument called 'subscripts' then
xyplot will pass it the row numbers of the data argument that
correspond this the current panel.   E.g.,

   > xyplot( y ~ x | Type, dataset, 
   +         panel = function(..., subscripts) {
   +             panel.xyplot(...)
   +             cat("subscripts=", deparse(subscripts),
":\n")
   +             print(dataset[subscripts,])
   +         })
   subscripts= 1L :
      x y Type
   1 1 4    a
   subscripts= 2L :
      x y Type
   2 2 5    b
You can use that information to add panel-specific information
to the plot.

Bill Dunlap
TIBCO Software Inc - Spotfire Division
wdunlap tibco.com  

> 
> but that requires the panel function call to have prior 
> knowledge of the
> object used as the conditioning variable, which is not flexible enough
> for my needs.
> 
> Thank you,
> Martin D. Lepage
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

> The problem I am facing is that I do not know how to generically access
> the conditioning variables within the panel function. In this simple
> case, I can achieve what I want to do with the following call :
>
> ? ?do_something_with(Type[which.packet()])
>
I just continued testing and I realized that the last line I wrote only
worked on my test session because I had a object named 'Type'
that existed in the workplace :

    Type <- dataset.Type

I cannot access the xyplot 'data$Type' object from within the panel
function (although dataset$Type obviously works).

I will continue my search through the documentation and report any
success.

Martin D. Lepage