Generation of smoothing lengths from coordinates of tracer particles
Hello,
I am a student researcher working on the Chimera project. Part of my work involves creating density maps from tracer particles that accurately represent what we get directly interpolated to the grid from chimera code. If I can do this I know I can make accurate element abundance maps for up to 150 species. We are limited by computing resources to creating maps for only 14 species directly.
Here is a video of the best conservative density plot I have been able to come up with thus far. The smoothing lengths for the tracers are found in a round-about way right now that is less than ideal. Is there a way in yt to create smoothing lengths using just particle coordinates like this? (IE the complex nearest neighbor math I would rather not code myself if I don’t have to)
Thanks,
Amos M
Hi Amos,
(Cool movie! :)
On Fri, Aug 8, 2014 at 11:04 AM, Amos Manneschmidt amannes1@vols.utk.edu wrote:
Hello,
I am a student researcher working on the Chimera project. http://chimerasn.org/ Part of my work involves creating density maps from tracer particles that accurately represent what we get directly interpolated to the grid from chimera code. If I can do this I know I can make accurate element abundance maps for up to 150 species. We are limited by computing resources to creating maps for only 14 species directly.
Here is a video of the best conservative density plot https://www.youtube.com/watch?v=WESKqLNi_jg I have been able to come up with thus far. The smoothing lengths for the tracers are found in a round-about way right now that is less than ideal. Is there a way in yt to create smoothing lengths using just particle coordinates like this? (IE the complex nearest neighbor math I would rather not code myself if I don’t have to)
Yes, you can definitely do this -- it requires delving somewhat into the Cython code. On my todo list is coming up with an easy way to get back "distance to Nth nearest neighbor." If that's what you're looking for, it would be pretty straightforward for me to implement that in the current version. I can try to do that today and write back to the list, but my time is somewhat fractured this afternoon. If something more complex is needed, we can certainly do that. Right now the way all of the smoothing, FOF-ing, etc etc, happens enables a lot of different correlations to be plugged in, so we should be able to find something you're looking for.
-Matt
Thanks,
Amos M
yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Matt,
Distance to Nth nearest neighbor output would be fantastic. No rush however, I won’t be able to do much with it until Tuesday at the earliest anyway.
Thanks a bunch,
Amos
On August 8, 2014 at 4:07:58 PM, Matthew Turk (matthewturk@gmail.com) wrote:
Hi Amos,
(Cool movie! :)
On Fri, Aug 8, 2014 at 11:04 AM, Amos Manneschmidt amannes1@vols.utk.edu wrote: Hello,
I am a student researcher working on the Chimera project. Part of my work involves creating density maps from tracer particles that accurately represent what we get directly interpolated to the grid from chimera code. If I can do this I know I can make accurate element abundance maps for up to 150 species. We are limited by computing resources to creating maps for only 14 species directly.
Here is a video of the best conservative density plot I have been able to come up with thus far. The smoothing lengths for the tracers are found in a round-about way right now that is less than ideal. Is there a way in yt to create smoothing lengths using just particle coordinates like this? (IE the complex nearest neighbor math I would rather not code myself if I don’t have to)
Yes, you can definitely do this -- it requires delving somewhat into the Cython code. On my todo list is coming up with an easy way to get back "distance to Nth nearest neighbor." If that's what you're looking for, it would be pretty straightforward for me to implement that in the current version. I can try to do that today and write back to the list, but my time is somewhat fractured this afternoon. If something more complex is needed, we can certainly do that. Right now the way all of the smoothing, FOF-ing, etc etc, happens enables a lot of different correlations to be plugged in, so we should be able to find something you're looking for.
-Matt
Thanks,
Amos M
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
participants (2)
-
Amos Manneschmidt -
Matthew Turk