[yt-users] pHOP child halos

[Brian Crosby]

Hi everyone,

I'm using Nautilus to run pHOP on a 1024^3 dataset and have encountered a strange problem.  All of the halos are found and written to the SQL database successfully, but then when pHOP attempts to go back and identify the parent/child relationships, it hangs and gives message:

Traceback (most recent call last):
  File "halofinding.py", line 14, in <module>
    MergerTree(restart_files=files, database='/gpfs/medusa/oshea/PopIII_1024/halos1024.db', halo_finder_function=parallelHF)
  File "/nics/e/sw/analysis/yt/2.1/sles11.1_gnu4.3.4/src/yt-hg/yt/analysis_modules/halo_merger_tree/merger_tree.py", line 201, in __init__
    last = self._compute_child_fraction(pair[0], pair[1], last)
  File "/nics/e/sw/analysis/yt/2.1/sles11.1_gnu4.3.4/src/yt-hg/yt/analysis_modules/halo_merger_tree/merger_tree.py", line 489, in _compute_child_fraction
    for group in h5fp:
  File "/sw/analysis/yt/2.1/sles11.1_gnu4.3.4/lib/python2.7/site-packages/h5py-1.2.0-py2.7-linux-x86_64.egg/h5py/highlevel.py", line 300, in __iter__
    return self.id.__iter__()
  File "/sw/analysis/yt/2.1/sles11.1_gnu4.3.4/lib/python2.7/site-packages/h5py-1.2.0-py2.7-linux-x86_64.egg/h5py/_sync.py", line 94, in wrap
    return func(*args, **kwds)
  File "h5g.pyx", line 449, in h5py.h5g.GroupID.__iter__ (h5py/h5g.c:3441)
  File "h5g.pyx", line 103, in h5py.h5g.GroupIter.__init__ (h5py/h5g.c:1235)
  File "/sw/analysis/yt/2.1/sles11.1_gnu4.3.4/lib/python2.7/site-packages/h5py-1.2.0-py2.7-linux-x86_64.egg/h5py/_sync.py", line 88, in wrap
    return func(*args, **kwds)
  File "h5g.pyx", line 316, in h5py.h5g.GroupID.get_num_objs (h5py/h5g.c:2469)
h5py._stub.ValueError: Invalid group id (Invalid arguments to routine: Bad value)

Does anyone have an idea what might be going wrong here, or how I should go about fixing it?

Thanks,
Brian