Well, first off, I should say that I don't work with a copy of the
Barnes-Hut treecode; or any other treecode, for that matter. I use
an N-body + hydro code, where the hydro is done by SPH (which Chris
Mihos has already spoken of), and the N-body by a particle-particle /
particle-mesh scheme. It's a little bit older approach than treecodes,
but perhaps more appropriate when simulating structures which don't
occupy a large fraction of the simulation volume. But that's controversial.
Anyway, I can talk about that if anyone cares.
As for visualization, if all you're interested in is the physical
location of the particles in the simulation, then if you understand
the format of the output files, it should be easy to write a program
that can read simulation output and write files containing particle
x-y positions produced either by projecting the entire volume onto a
plane (from some given perspective/pair of angles), or by taking a
wedge out of the volume and projecting the contents of that wedge onto
a plane. With a little ingenuity, you can get fancier -- writing
velocities in the plane as well, and using whatever plotting package
you use (I use sm) to attach arrows to the particles showing
direction and magnitude of motion, or whatever.
But if you really want to get fancy, then 3-D color visualization packages
such as AVS or SGI's Explorer are the way to go. But there, you're talking
about lotsa bux.
-- Chris Metzler Department of Physics, University of Michigan 313-764-4607 (office) Randall Lab, 500 E. University 313-996-9249 (home) Ann Arbor, MI 48109-1120 USA E-MAIL: metzler@pablo.physics.lsa.umich.edu"Well, I had a dream, and in it I went to a little town, and all the girls in town were named . . .Betty." -- Laurie Anderson
"Are these net-clowns for real?" -- Serdar Argic