Skip to content

Plot's aspect ratio and pty

5 messages · Uwe Ligges, Marcin Kozak, Brian Ripley

#
Dear All,

I've been playing with pty, and it seems it does not produce square
plots as it is expected to (or at least as I expect it to). Consider
this simple example:

par(pty="s"); plot(1:10, 1:10)

This should produce a square plot, right? Well, if you have a look at
the graph, it is not square! So, maybe the limits?

par(pty="s"); plot(1:10, 1:10, xlim = c(0,11), ylim=c(0,11))

No, again not. So let's try and help to equalize everything, just to be sure:

windows(6, 6); par(mar=c(3, 3, 3, 3), pty="s"); plot(1:10, 1:10, xlim
= c(0, 11), ylim = c(0, 11))

Again not!

pty = "s" is to generate a square plotting region, and it does not
seem to do that. Where is my mistake?

Thanks in advance,
Marcin
#
It does for me under R-2.12.0 32-bit on Windows with the windows() 
device, so:

Which version of R, which OS, which device do you use?

Uwe Ligges
On 13.12.2010 07:30, Marcin Kozak wrote:
#
R-11.1 on both 32-bit and 64-bit on Windows with the windows() device.

Best
Marcin



2010/12/13 Uwe Ligges <ligges at statistik.tu-dortmund.de>:
#
On 13.12.2010 11:29, Marcin Kozak wrote:
This one does not exist.
Please try R-2.12.0 (but it also worked with R-2.11.1 if you meant that).

My guess is that you are confusing plotting region with device region.
In order to get a squared device region, you have to ask the device 
function for a squared region.

Best,
Uwe Ligges





on both 32-bit and 64-bit on Windows with the windows() device.
#
On Mon, 13 Dec 2010, Uwe Ligges wrote:

            
Since (s)he used windows() we know the OS.  But I think one possible 
explanation is on the help page, arguments 'xpinch' and 'ypinch': most 
likely this is one of those devices with a mendacious Windows driver.

Of course. we simply do not kwow what (s)her saw, and it is possible 
that this is simply a misunderstanding of 'plot region'.