Re: [S] Alternative to xyplot for conditioning scatter plots

Prof Brian Ripley (
Fri, 3 Apr 1998 07:39:15 +0100 (BST)

John Leathwick wrote:
> Dear S-Plus users,
> Can anyone remind me of the alternative to xyplot for
> doing scatter plots of two continuous variables conditioned
> by intervals along one or more additional variables - I have
> graphs produced in earlier days using an alternative function
> which provided much more control than equal.count over the
> intervals used for the conditioning variables, i.e., by providing
> a 2 by n matrix of values, but foolishly didn't keep the details
> of how I produced them! I am using 4.0 under Windows 95,
> mainly with the command line, finding the gui pathway to
> produce sometimes inconsistent results.

I think you are referring to coplot(). But actually I think Trellis
(xyplot) has just as much control: you don't have to use equal.count
as you can create a shingle directly -- look up help for the
`shingle' function. You need something like

xyplot( z ~ y | shingle(z, cbind(zlower, zupper)), ...)

The GUI graphics are a completely different system rather than a pathway:
that is they have been designed to give similar results to Trellis but do
not use Trellis.

> John Leathwick,
> Landcare Research NZ Ltd,
> Hamilton,
> New Zealand.

Brian D. Ripley,        
Professor of Applied Statistics,
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595
This message was distributed by  To unsubscribe
send e-mail to with the BODY of the
message:  unsubscribe s-news