Hi everyone, I'm trying to make a merger tree on Kraken using parallelHOP and the public yt installation, but am getting this error: Traceback (most recent call last): File "./merger.py", line 9, in <module> MergerTree(files, database='/lustre/scratch/oshea/PopIII_smallbox/halos.db', halo_finder_function=parallelHF) NameError: name 'parallelHF' is not defined All of the requisite Forthon and KD Tree bits seem to be there, and the only thing that I found in the yt install that might be problematic is that I can't access the Forthon directory at /lustre/scratch/proj/yt_common/trunk/src/Forthon-0.8.3/ . I don't know if that could possibly cause problems, but it made me curious. Does anyone have any tips on what's going wrong? Running the exact same script with HaloFinder instead of parallelHF works without problem (just very slowly at lower redshifts). Thanks, Brian
Hi Brian,
Traceback (most recent call last): File "./merger.py", line 9, in <module> MergerTree(files, database='/lustre/scratch/oshea/PopIII_smallbox/halos.db',
halo_finder_function=parallelHF) NameError: name 'parallelHF' is not defined
All of the requisite Forthon and KD Tree bits seem to be there, and the only thing that I found in the yt install that might be problematic is that I can't access the Forthon directory at /lustre/scratch/proj/yt_common/trunk/src/Forthon-0.8.3/ . I don't know if that
could possibly cause problems, but it made me curious.
Not being able to read that directory shouldn't change anything.
Does anyone have any tips on what's going wrong? Running the exact same script with HaloFinder instead of parallelHF works without problem (just very slowly at lower redshifts).
I'm not sure what the problem is, especially since HaloFinder works. Could you try modifying your script something like this, and see if it works? I am still thinking about how this could happen, and this might tell us something. from yt.mods import * import yt.lagos.HaloFinding as HF ... ... MergerTree(files, database='/lustre/scratch/oshea/PopIII_smallbox/halos.db', halo_finder_function=HF.parallelHF) ... ... _______________________________________________________ sskory@physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________
Making those changes seems to have it working. Looking at the differences, I think the key is that creating a merger tree with HOP only requires from yt.extensions.merger_tree import * , but that running parallelHOP apparently also requires from yt.mods import * . I didn't realize this, so when I just changed the halo_finding_function in the script, I neglected to add the yt.mods import statement. At least that's my best guess, as it seems to be working now. Thanks, Brian On Jul 20, 2010, at 2:24 PM, Stephen Skory wrote:
Hi Brian,
Traceback (most recent call last): File "./merger.py", line 9, in <module> MergerTree(files, database='/lustre/scratch/oshea/PopIII_smallbox/halos.db',
halo_finder_function=parallelHF) NameError: name 'parallelHF' is not defined
All of the requisite Forthon and KD Tree bits seem to be there, and the only thing that I found in the yt install that might be problematic is that I can't access the Forthon directory at /lustre/scratch/proj/yt_common/trunk/src/Forthon-0.8.3/ . I don't know if that
could possibly cause problems, but it made me curious.
Not being able to read that directory shouldn't change anything.
Does anyone have any tips on what's going wrong? Running the exact same script with HaloFinder instead of parallelHF works without problem (just very slowly at lower redshifts).
I'm not sure what the problem is, especially since HaloFinder works. Could you try modifying your script something like this, and see if it works? I am still thinking about how this could happen, and this might tell us something.
from yt.mods import * import yt.lagos.HaloFinding as HF
... ...
MergerTree(files, database='/lustre/scratch/oshea/PopIII_smallbox/halos.db', halo_finder_function=HF.parallelHF)
... ...
_______________________________________________________ sskory@physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________ _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Brian,
I think the key is that creating a merger tree with HOP only requires
from yt.extensions.merger_tree import * ,
but that running parallelHOP apparently also requires
from yt.mods import * .
I'm not surprised that it wants 'yt.mods' for halo finding, but I am surprised that it worked for HOP without 'yt.mods'. Thanks for tracking this down! _______________________________________________________ sskory@physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________
participants (2)
-
Brian Crosby -
Stephen Skory