G_packet.panel.default: Associating Packets with Panels

Description

When a "trellis" object is plotted, panels are always drawn in an order such that columns vary the fastest, then rows and then pages. An optional function can be specified that determines, given the column, row and page and other relevant information, the packet (if any) which should be used in that panel. The function documented here implements the default behaviour, which is to match panel order with packet order, determined by varying the first conditioning variable the fastest, then the second, and so on. This matching is performed after any reordering and/or permutation of the conditioning variables.

Usage

packet.panel.default(layout, condlevels, page, row, column,
                     skip, all.pages.skip = TRUE)

Value

A suitable combination of levels of the conditioning variables in the form of a numeric vector as long as the number of conditioning variables, with each element an integer indexing the levels of the corresponding variable. Specifically, if the return value is

p, then the i-th conditioning variable will have level

condlevels[[i]][p[i]].

Arguments

layout: the layout argument in high level functions, suitably standardized.
condlevels: a list of levels of conditioning variables, after relevant permutations and/or reordering of levels
page, row, column: the location of the panel in the coordinate system of pages, rows and columns.
skip: the skip argument in high level functions
all.pages.skip: whether skip should be replicated over all pages. If FALSE, skip will be replicated to be only as long as the number of positions on a page, and that template will be used for all pages.

Author

Deepayan Sarkar Deepayan.Sarkar@R-project.org

Examples

Run this code


packet.panel.page <- function(n)
{
    ## returns a function that when used as the 'packet.panel'
    ## argument in print.trellis plots page number 'n' only
    function(layout, page, ...) {
        stopifnot(layout[3] == 1)
        packet.panel.default(layout = layout, page = n, ...)
    }
}

data(mtcars)
HP <- equal.count(mtcars$hp, 6)
p <- 
    xyplot(mpg ~ disp | HP * factor(cyl),
           mtcars, layout = c(0, 6, 1))

print(p, packet.panel = packet.panel.page(1))
print(p, packet.panel = packet.panel.page(2))