Mathematica 9 is now available
Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2010

[Date Index] [Thread Index] [Author Index]

Search the Archive

ListPlot3D[] - rather slow?

  • To: mathgroup at smc.vnet.net
  • Subject: [mg113410] ListPlot3D[] - rather slow?
  • From: Crni Gorac <cgorac at gmail.com>
  • Date: Wed, 27 Oct 2010 05:17:33 -0400 (EDT)

I have kind of finite difference solver, typically used over mesh of
250x250 size, producing scalar values for each mesh point (also -
typically values produced are rather smooth).  So I'm trying to
visualize results using ListPlot3D[], basically passing only the
resulting 2D list of results as ListPlot3D[] argument.  However, the
resulting visualization is very slow to build, and interact with (am
using Mathematica 7.0.0 on Linux, on a 2.4 GHz Core 2 Duo machine,
with NVIDIA Quadro FX 770M graphics) - each redraw takes couples
second, and soon after Mathematica just start eating my swap memory
and brings the whole system very slow.  Now, I have some experience
with using OpenGL and that kind of stuff, and I know several tens of
thousands of polygons is not something particularly problematic for
above graphics hardware.  So - I'm wondering why ListPlot3D[] is so
slow, and are there any options of this function that could be used to
make the resulting visualization more responsive (so far, I've only
tried setting PerformanceGoal option of this function to "Speed", but
then resulting visualization looks like crap, and also I've tried to
set PlotStyle option to "None" and this kind of works, but still it
would be much more helpful to be able to see actual surface generated,
instead of lines mesh...  Also - is there any other method that could
be recommended for this type of visualization (I guess lots of people
are doing same kind of work, and would be interested in alike kind of
visualization)?

Thanks.


  • Prev by Date: Re: fyi, A nice package for making plot ticks
  • Next by Date: Re: Minimize - What am I doing worng?
  • Previous by thread: Re: how to read Wolfram technology conference 2010
  • Next by thread: Mathlink Issue