On Tuesday 09 September 2003 10:35 am, Douglas Bates wrote:
> Ernesto Jardim <[EMAIL PROTECTED]> writes:
> > Hi,
> >
> > I want to plot 2 variables (confidence intervals) in a single xyplot. I
> > have a dataframe with variables Yup, Ylo, X and Z and I want to have a
> > xyplot with both variables ploted. I'm trying with
> >
> > xyplot(Yup~X|Z, panel=function(x){...})
> >
> > but this way I'm not able to pass variable Ylo into the function...
> >
> > How can I do this ?
>
> According to the "formula" description on the manual page for xyplot
> you can use
>
> xyplot(Ylo+Yup~X|Z, allow.multiple=TRUE)
>
> Did you try that?
Yes, this seems to work with the following panel function:
panel = function(x, y, groups, subscripts, ...) {
groups <- as.numeric(groups)
lsegments(x[groups[subscripts] == 1], y[ groups[subscripts] == 1],
x[groups[subscripts] == 2], y[ groups[subscripts] == 2], ...)
}
Deepayan
______________________________________________
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
