R help - Feb 2003 - Fw: Plotting in subareas using par(fig=) parameter

Any idea why I can no longer plot two graphs on the same graphics device
using the par(fig=) parameter?
A simpler par(mfrow=c(1,2)) does work, showing the two plots side-by-side,
but I would like the first
to be larger.   This simple example fails:


x<-c(1,1,NA,2,2,NA,3,3)
y<-c(2,4,NA,3,5,NA,1,4)

par(fig=c(0,2/3,0,1))
plot(x,y)

par(fig=c(2/3,1,0,1))
qqnorm(x)

When plotted, the last figure qqnorm(x) is alone on the graphics display.
The first figure plot(x,y) disapears.
I'm even more puzzled in that I'm sure that I had this type of command
working before.

REX

Try:

R> par(fig=c(0,2/3,0,1))
R> plot(x,y)
R> par(fig=c(2/3,1,0,1), new = TRUE)
R> qqnorm(x)
R>

 From ?par:

      `new' logical, defaulting to `FALSE'.  If set to `TRUE', the
next
           high-level plotting command (actually `plot.new') should not
           clean the frame before drawing ``as if it was on a new
           device''.

Regards,
Sundar

Rex_Bryan at urscorp.com wrote:
> Any idea why I can no longer plot two graphs on the same graphics device
> using the par(fig=) parameter?
> A simpler par(mfrow=c(1,2)) does work, showing the two plots side-by-side,
> but I would like the first
> to be larger.   This simple example fails:
> 
> 
> x<-c(1,1,NA,2,2,NA,3,3)
> y<-c(2,4,NA,3,5,NA,1,4)
> 
> par(fig=c(0,2/3,0,1))
> plot(x,y)
> 
> par(fig=c(2/3,1,0,1))
> qqnorm(x)
> 
> When plotted, the last figure qqnorm(x) is alone on the graphics display.
> The first figure plot(x,y) disapears.
> I'm even more puzzled in that I'm sure that I had this type of
command
> working before.
> 
> REX
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> http://www.stat.math.ethz.ch/mailman/listinfo/r-help
>

On Thu, 6 Feb 2003, Rex_Bryan at urscorp.com wrote:
> Any idea why I can no longer plot two graphs on the same graphics device
> using the par(fig=) parameter?
> A simpler par(mfrow=c(1,2)) does work, showing the two plots side-by-side,
> but I would like the first
> to be larger.   This simple example fails:
> 
> 
> x<-c(1,1,NA,2,2,NA,3,3)
> y<-c(2,4,NA,3,5,NA,1,4)
> 
> par(fig=c(0,2/3,0,1))
> plot(x,y)
> 
> par(fig=c(2/3,1,0,1))
add par(new=TRUE) here, as you want to plot again in the same device 
region.
> qqnorm(x)
> 
> When plotted, the last figure qqnorm(x) is alone on the graphics display.
> The first figure plot(x,y) disapears.
> I'm even more puzzled in that I'm sure that I had this type of
command
> working before.

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595