R help - Sep 2008 - restricting lattice pages to one

Dear list members,

I'd like to be able to restrict the number of pages in a lattice display to
one without having to specify explicity the number of rows and columns in
the display -- that is, having forced one page, I'd like the number of rows
and columns to be determined automatically.

For example, the following code produces two pages of output, with the
second replacing the first on the trellis device:

  A <- factor(rep(c("a1", "a2"), c(20, 20)))
  B <- factor(rep(rep(c("b1", "b2"), c(10, 10)), 2))
  C <- factor(rep(rep(c("c1", "c2"), c(5, 5)), 4))
  X <- rnorm(40)
  Y <- X + 0.2*rnorm(40)
  trellis.device()
  xyplot(Y ~ X | A + B + C)

I can get what I want via

  xyplot(Y ~ X | A + B + C, layout=c(4, 2, 1))

but in the application I have in mind, I'd like to avoid having to compute
the rows and columns.

Any help would be appreciated.

John

------------------------------
John Fox, Professor
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
web: socserv.mcmaster.ca/jfox

On Thu, Sep 25, 2008 at 7:33 AM, John Fox <jfox at mcmaster.ca>
wrote:
> Dear list members,
>
> I'd like to be able to restrict the number of pages in a lattice
display to
> one without having to specify explicity the number of rows and columns in
> the display -- that is, having forced one page, I'd like the number of
rows
> and columns to be determined automatically.
>
> For example, the following code produces two pages of output, with the
> second replacing the first on the trellis device:
>
>  A <- factor(rep(c("a1", "a2"), c(20, 20)))
>  B <- factor(rep(rep(c("b1", "b2"), c(10, 10)), 2))
>  C <- factor(rep(rep(c("c1", "c2"), c(5, 5)), 4))
>  X <- rnorm(40)
>  Y <- X + 0.2*rnorm(40)
>  trellis.device()
>  xyplot(Y ~ X | A + B + C)
>
> I can get what I want via
>
>  xyplot(Y ~ X | A + B + C, layout=c(4, 2, 1))
>
> but in the application I have in mind, I'd like to avoid having to
compute
> the rows and columns.
A general (but two-step) solution is

p = xyplot(Y ~ X | A + B + C)
update(p, layout = c(0, prod(dim(p))))

Another solution that is effectively the same (with different strip
annotation) is to use an interaction as the single conditioning
variable.

xyplot(Y ~ X | A:B:C)

-Deepayan