MASON-INTEREST-L Archives

December 2012

MASON-INTEREST-L@LISTSERV.GMU.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Sean Luke <[log in to unmask]>
Reply To:
MASON Multiagent Simulation Toolkit <[log in to unmask]>
Date:
Mon, 31 Dec 2012 13:25:45 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (27 lines)
On Dec 31, 2012, at 6:34 AM, Norman Rennhack wrote:

> Can someone give me a hint how to model and visualize a chemical, spread
> as a gradient in a 3D space? The Grid should be about 100 x 100 x 100
> and the area which gets covered with the chemical is about 30 x 30 x 30.
> ValueGrid3d was my preferred solution because it's very easy to
> calculate the concentrations of the chemicals and their changes. Maybe
> someone knows how to optimze the visualisation? For example that zero
> values do not get rendered at all?
> I can also share my sourcecode if it helps you to understand my problem!

Hi Norman.  Java3D is hideously memory-inefficient.  In particular,
ObjectGrid3D is a real problem; but large numbers of objects cause problems
for all the 3D portrayals and there's not much we can do about it.  :-(
For example, I just did some tests with PSO3D.  I can get about 20,000 
particles in 500 megs.  And if I try to rotate it, it'll bomb out on me.

So what to do?  Here's a hack.  I suggest trying a 3D mesh in the form of a stack of
ValueGrid2DPortrayal3D objects.  Dump information into 50 
ValueGrid2D objects, each 50x50 in size.  Then assign a ValueGrid2DPortrayal3D
to each one and put it in the space.  You'll need to transpose the location
of each of the grids so that they're spaced properly along the Z axis with
regard to one another, but that's just a for-loop.  Maybe we should offer this
as an option.

Sean

ATOM RSS1 RSS2