[R] 3D-scatterplots - high quality rendering?
J.delasHeras at ed.ac.uk
J.delasHeras at ed.ac.uk
Thu Oct 21 19:23:11 CEST 2010
Quoting Duncan Murdoch <murdoch.duncan at gmail.com>:
> J.delasHeras at ed.ac.uk wrote:
>> I have just started using the package 'rgl' to explore my data as a
>> 3D scatterplot.
>>
>> It's a great tool. But I would like to be able to save some graph
>> views and it appears the only format available is 'png' and it
>> doesn't look that nice. It is excellent to work with on screen and
>> explore/play, but when it comes to producing a good quality
>> graphic I'm a bit disappointed.
>>
>> So the questions I have are:
>>
>> 1) is there a way to render the graph in the RGL device as a high
>> quality graphic?
>>
>
> You can use rgl.postscript for various other formats, but you may find
> the bitmap output is better. In particular, shading is not always done
> correctly.
>
> Duncan Murdoch
Hi Duncan,
thanks for that.
Stupidly I hadn't realised only part of the manual had printed so I
was missing a lot of interesting options, not only this one but others
to animate views.
The type of plots I'm producing are similar to these:
library(rgl)
m<-matrix(rnorm(900),ncol=3)
plot3d(m,col=rainbow(6),type="s",size=0.8)
I tried rgl.postscript to save to pdf, as it's become my default
format when I want smooth plots using standard R graphics X11()
It looks smoother, but I get some "artifacts" in that the edges and
colouring of a number of spheres are rough, or incompletely coloured
(imagine a small scratch on a photograph)
using 'rgl.snapshot' to output a bitmap (png) with small spheres it
doesn't look that great. The default squares look fine 'though.
I guess I'll just play with these limitations. I was just wondering if
there was a simple way to get significantly better results.
But I'm not complaining! :) the 'rgl' package is pretty impressive!
Thank you.
Jose de las Heras
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
More information about the R-help
mailing list