Hi,<br />
<br />
I tried the dev module, but it still didn't work. I should note that I'm using the common install.<br />
<br />
I also tried a script that does a few projections and profiles without slices, and it also crashed with the same error. It looks like only a few nodes are crashing, while the others run along fine for a while.<br />
<br />
I've attached the full output, which lists the modules I have loaded near the top.<br />
<br />
Thanks,<br />
<br />
-- Greg<br />
<br />
> Hi Greg,<br />
><br />
> Thanks for writing. This looks familiar to me, and I think it is<br />
> fixed in the current development branch. Can you use the dev module<br />
> for yt instead of stable?<br />
><br />
> One sidenote is that often slicing in parallel is not a net gain; I<br />
> have found very few datasets where it is in fact substantially faster,<br />
> most of them gigantic AMR runs with >1e5 grids. I would recommend<br />
> just doing slices in serial.<br />
><br />
> Best,<br />
><br />
> Matt<br />
><br />
> On Fri, Jul 29, 2011 at 11:15 AM, <meecegre@msu.edu> wrote:<br />
>> Hi everyone,<br />
>><br />
>> I've come across an error while trying to use yt for parallel analysis on<br />
>> Kraken. Wondering if its something people have seen before.<br />
>><br />
>> I have my simple script that does slices of a few different quanties. When I<br />
>> ran it with 12 tasks on Kraken, it worked fine for the first four datasets,<br />
>> but crashed during the fifth one. I tried a simpler script that just does a<br />
>> slice of density for that dataset, and it crashed again with the same error:<br />
>><br />
>> IndexError: arrays used as indices must be of integer (or boolean) type<br />
>><br />
>> When I ran this code again with only 1 task (still on Kraken, everything<br />
>> else the same), it worked fine. More of the output, as well as my job script<br />
>> and plotting script, are below.<br />
>><br />
>> Since it works in serial , and works for some data sets, the problem seems<br />
>> to be something to do with parallelism. Is there any sort of simple<br />
>> explanation or something I'm doing wrong?<br />
>><br />
>> Just wondering if this is something people have come across before.<br />
>><br />
>> --Greg<br />
>><br />
>> (Last few lines of output)<br />
>> mg, mc, mv, pos = self.find_max_cell_location(field, finest_levels)<br />
>> mg, mc, mv, pos = self.find_max_cell_location(field, finest_levels)<br />
>> File<br />
>> "/lustre/scratch/proj/sw/yt/current/lib/python2.7/site-packages/yt-2.1stable-py2.7-linux-x86_64.egg/yt/data_objects/object_finding_mixin.py",<br />
>> line 69, in find_max_cell_location<br />
>> File<br />
>> "/lustre/scratch/proj/sw/yt/current/lib/python2.7/site-packages/yt-2.1stable-py2.7-linux-x86_64.egg/yt/data_objects/object_finding_mixin.py",<br />
>> line 69, in find_max_cell_location<br />
>> File<br />
>> "/lustre/scratch/proj/sw/yt/current/lib/python2.7/site-packages/yt-2.1stable-py2.7-linux-x86_64.egg/yt/data_objects/object_finding_mixin.py",<br />
>> line 69, in find_max_cell_location<br />
>> max_grid = self.grids[mg]<br />
>> IndexError: arrays used as indices must be of integer (or boolean) type<br />
>> max_grid = self.grids[mg]<br />
>> max_grid = self.grids[mg]<br />
>> IndexError: arrays used as indices must be of integer (or boolean) type<br />
>> IndexError: arrays used as indices must be of integer (or boolean) type<br />
>><br />
>> (Job script)<br />
>> #!/bin/bash<br />
>> #PBS -A TG-AST090040<br />
>> #PBS -l size=12,walltime=2:00:00<br />
>> #PBS -j oe<br />
>> #PBS -N slicer<br />
>><br />
>> cd /lustre/scratch/oshea/nested_ics_real1/analysis<br />
>><br />
>> ls<br />
>> pwd<br />
>><br />
>> export MPLCONFIGDIR=${PBS_O_WORKDIR}/.matplotlib/<br />
>> [ ! -d ${MPLCONFIGDIR} ] && mkdir ${MPLCONFIGDIR}<br />
>><br />
>> module unload PrgEnv-pgi<br />
>> module load PrgEnv-gnu<br />
>> module load yt<br />
>><br />
>> aprun -n 12 python ./simple_slice.py --parallel<br />
>><br />
>><br />
>> (simple_slice.py)<br />
>> from yt.mods import *<br />
>><br />
>> DD = 40<br />
>><br />
>> fileName = "../DD%s/DD%s" % ((str(DD)).zfill(4), (str(DD)).zfill(4))<br />
>><br />
>> pf = load(fileName)<br />
>> pc = PlotCollection(pf)<br />
>><br />
>> pc.add_slice("Density", 0)<br />
>><br />
>> pc.save("./%s" % str(DD).zfill(4))<br />
>><br />
>><br />
>><br />
>> _______________________________________________<br />
>> yt-users mailing list<br />
>> yt-users@lists.spacepope.org<br />
>> <a target="_blank" href="http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org">http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org</a><br />
>><br />
>><br />
> _______________________________________________<br />
> yt-users mailing list<br />
> yt-users@lists.spacepope.org<br />
> <a target="_blank" href="http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org">http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org</a><br />
><br />