How to get shaded background in xyplot in R?

Question

using xyplot from the lattice package, I plot a time series over a number of years. I would to add a shaded area for some of these years to indicate that this time period was "special" (e.g. war).

Please apologize if this is trivial, but I could not figure out how to do that, so I would be happy if someone could help me out, or at least point me in the right direction. I think my main problem is that I don't really know how to approach this problem. I am still relatively new to R, and to lattice in particular.

Here a minimal example:

 xyplot( rnorm(100) ~ 1:100, type="l", col="black")

In the corresponding plot, I would like the color of the background (from say x-values of 45 until 65) from the bottom to the top of the plotting area be shaded in, say, light grey.

Note that solutions that I have found so far use base graphics and the polygon-function, but there the intention is to shade the area under or above a curve, which is different from what I would like to do. I don't "just" want to shade the area below my line, or above my line. Instead I would like to shade the entire background for a given time interval.

If anyone could help me out here, I would be very grateful!

Courlan

Answer 1

See ?panel.xblocks in the latticeExtra package:

library(latticeExtra)
x <- 1:100
xyplot( rnorm(100) ~ x, type="l", col="black") + 
         layer_(panel.xblocks(x, x > 20, col = "lightgrey"))

Dot

Appears rather powerful. y=x > 45 & y <0 gets you grey bands to the left of x=45, but only where y<0. I think you chose to use layer_ because it has a default of under=TRUE? – Lauraine

It just seemed rather nifty and ggplot2-ish but alternately one could use it in a conventional panel function: xyplot( rnorm(100) ~ x, type="l", col="black", panel = function(x, ...) { panel.xblocks(x, x > 20, col = "lightgrey"); panel.xyplot(x, ...) }) . – Dot

Answer 2

Try this:

 xyplot( 
    rnorm(100) ~ 1:100, type="l", col="black",
    panel=function (x,y,...){
    panel.rect(xleft=45, xright=65,ybottom=-3, ytop=3,col="grey")
    panel.xyplot(x,y,...)
    }
  )

The panel.rect() function controls the rectangle and is the lattice equivalent of the rect() function. It has a variety of settings that you may find useful. It is called first and then the xyplot() is put on top of it. You many need to play around a little to get your ybottom and ytop parameters to look as you like them.

Intracardiac

Wow, thanks guys, that was quick. Your solutions definitely help me. They also teach me some new things, which is why I love this website! I hope at some point in time I will be proficient enough so as to contribute here myself! Thanks and have a good weekend! – Courlan

Answer 3

trellis.focus("panel", 1, 1)
grid.rect(x =.55, , y=.5, w = .2, height=6,  
  gp = gpar(fill = "light grey"))
trellis.unfocus()

This differs from @JohnPaul's solution in a couple of ways (and I think his answer is better). This uses the center of the desired X-band for placement in "native coordinates" and calculates the width as 'range(xlim)/range(band)' and it modifies an existing plot. the grid.rect function is the grid packages lower level function that is used by panel.rect. I sometimes find this useful when integrating lattice panels inside the xyplot system defeats me.

Lauraine