[R] Exporting an rgl graph

[David Winsemius]

On Apr 15, 2010, at 12:34 PM, luke at stat.uiowa.edu wrote:

> The current issue of JCGS (Vol 18 No 1,
> http://pubs.amstat.org/toc/jcgs/19/1 ) has an editorial on including
> animations, 3D visualizations, and movies in on-line PDF files
> supporting JCGS articles. The online supplements to the editorial
> include examples.  The 3D examples related to the misc3d packages are
> also available in
> http://www.stat.uiowa.edu/~luke/R/misc3d/misc3d-pdf/.  At some point
> the code there will be added to misc3d.  It should be possible to
> adapt these ideas to other objects rendered with rgl.

Very kewl. On a Mac the greyscale plots opened in Adobe Acrobat Reader  
v8.2.2 displays properly, but the color version supp_j.pdf looks like  
a ménage à trois of three psychedelic sea urchins. I think that Adobe  
may need to do some work on their display engine for this to be a  
fully cross-platform combination. The color version of the volcano  
example is likewise carpeted with spiky artifacts.

(I have not yet tried producing plots de novo with the Mac pdf device.)

-- 
David.

> luke
>
>
> On Thu, 15 Apr 2010, baptiste auguie wrote:
>
>> I have seen pdf files with 3D objects embedded in it, using the U3D  
>> format,
>> http://en.wikipedia.org/wiki/Universal_3D
>>
>> but I don't think there's a device for this in R; in fact there may
>> not even exist a third-party post-processing route available at this
>> time to bridge the gap between rgl and this format. It sure would be
>> nice, though.
>>
>> Best,
>>
>> baptiste
>>
>> On 15 April 2010 14:12, Barry Rowlingson <b.rowlingson at lancaster.ac.uk 
>> > wrote:
>>> On Thu, Apr 15, 2010 at 1:01 PM,  <cgenolin at u-paris10.fr> wrote:
>>>> Thanks for you answer. Let me precise my question.
>>>>
>>>> In fact, I do not want to "capture" a screen, I want to save an  
>>>> object that
>>>> can be seen in 3D. With rgl, using my mouse, I can make the  
>>>> object move.
>>>> This is what I want to export: an real 3D object that my  
>>>> collaborator will
>>>> have the possibility to see in 3D.
>>>>
>>>
>>>  You mean without them having to install R and rgl and run the code
>>> that produces your graphic?
>>>
>>>  I guess you could somehow export a VRML or some other 3d file:
>>>
>>>  http://en.wikipedia.org/wiki/VRML
>>>
>>>  but I suspect of all the billions of people on the planet only  
>>> Duncan
>>> Murdoch knows enough about rgl to figure that one out...
>>>
>>>  The person at the other end would still need a VRML viewer. Just  
>>> get
>>> them to install R.
>>>
>>> Barry
>>
>
> -- 
> Luke Tierney
> Chair, Statistics and Actuarial Science


David Winsemius, MD
West Hartford, CT