[R] cex multiplier not exact

[Ali Tofigh]

On Mon, Oct 24, 2011 at 06:10, Prof Brian Ripley <ripley at stats.ox.ac.uk> wrote:
> On Mon, 24 Oct 2011, Uwe Ligges wrote:
>
>>
>>
>> On 23.10.2011 22:33, Ali Tofigh wrote:
>>>
>>> Hi,
>>>
>>> When I plot text and use cex to change the text size, I notice that the
>>> cex
>>> multiplier is not exact. It looks as if the real size of text can take
>>> only
>>> certain discrete values. Is there a workaround to get text to follow the
>>> cex
>>> value more closely, or at least to be able to figure out what the real
>>> cex value
>>> will be?
>>
>> Typically this is multiplied by the default pointsize of the current
>> device and rounded to an integer - the resulting pointsize that you see with
>> your code.
>>
>> If you want to know it absolute precisely: This may even depend on the
>> device and I think you will have to read the source code in order to find
>> out how it is actually implemented for your current device (which you have
>> not even stated).
>
> It may be in the help for that unstated device (the help for windows() says
> that it is rounded to bigpoints), and in some cases the rounding is in the
> OS services (it is for the X11() device, for example) and hence very hard to
> know precisely.

I was hoping to write code that would work on most devices. I'm trying
to write a function that plots and automatically adjusts the size of a
given text string to fit inside an predefined area (actually a
viewport from the grid package).

I try to do this by computing the width and height of the text using
the unit function: unit(1, "strwidth", x) and unit(1, "strheight, x").
>From these values I then compute the desired cex value. But from the
information you and Uwe provided, I'm beginning to think that such a
function might not be feasible at all?

Thank you for all the info,
Best wishes
/Ali

>
> Oh, and the cex for text and for symbols (pch = 0:25) may not work the same
> way.
>
>>
>> Uwe Ligges
>>
>>
>>>
>>> Here is an example that illustrates the problem:
>>>
>>> cex<- seq(0.5, 1, 0.01)
>>> x<- 0.05
>>> y<- (1:length(cex))/length(cex)
>>> labels<- paste("XXXXXX", cex)
>>> plot.new()
>>> text(x=x, y=y, labels=labels, pos=4, cex=cex)
>>>
>>> /Ali
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide
>>> http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>
> --
> 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
>