[R] dev.size() does not return correct size on 4K screen

[Martijn Tennekes]

Thank you, Peter!


 From xdpyinfo I got

screen #0:
   dimensions:    5120x2880 pixels (903x508 millimeters)
   resolution:    144x144 dots per inch

According to these specs my screen should have a sqrt((5120/144)^2 + 
(2880/144)^2) = 40.8" diagonal.

However, in practice it is 27", so that perfectly explains the 1.5 
factor difference.


In RStudio, dev.size() is not equal to par("fin"), which caused the 
problem for me. I the R terminal, these two are equal. I'll post this in 
the RStudio issue list.


Best,

Martijn



On 3/2/22 23:11, peter dalgaard wrote:
> I don't know about RStudio, but in X11 this sort of thing can happen due to misconfiguration of the display itself, i.e. not in R as such. Basically, it relies on getting the correct dots-per-inch value from the display, and may otherwise use some standard value.
>
> What happens if you run xdpyinfo in a terminal window?
>
> With XQuartz on an MB Air, I see
>
> ...
> screen #0:
>    dimensions:    1440x876 pixels (381x232 millimeters)
>    resolution:    96x96 dots per inch
> ...
>
> which is a multi-way lie! It is a 2560x1600 retina display, the resolution is way higher than 96x96, and it is 13.3", whereas the stated dimensions corresponds to a 17" diagonal.
>
> However, systems may be "lying in the users best interest", since e.g. the default for an X11() plot window is 7in square, which might not fit on a small laptop. You are somewhat more likely to want the same relative size across different laptops.
>
> - pd
>
>> On 2 Mar 2022, at 10:52 , Martijn Tennekes <mtennekes using gmail.com> wrote:
>>
>> dev.size() does not return the correct screen size when using a 4K
>> screen. With the R console and a x11() device, it seems to overestimate
>> with factor 1.5, and in RStudio underestimate with factor 2.
>>
>> See also https://github.com/rstudio/rstudio/issues/10723
>>
>> Is this a bug, or is there something I can do?
>>
>> Best,
>>
>> Martijn Tennekes
>>
>> 	[[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> 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.