Hi everybody,

A quick followup:  at Sam Skillman's suggestion, I removed the .harrays and .yt files that were already existing in the DD0050 directory.  After removing these, everything worked fine.

Thanks for the quick turnaround!

--Brian

On Sun, Apr 18, 2010 at 2:05 PM, Brian O'Shea <bwoshea@gmail.com> wrote:
Dear yt-users,

I'm having a strange problem with yt (svn trunk, r1693).  If I run this script:

from yt.mods import *
fn = "DD0050/DD0050"
pf = load(fn)
dd = pf.h.all_data()

I get the following error:

yt         INFO       2010-04-18 13:41:48,183 Getting the binary hierarchy
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/mnt/home/oshea/yt-x86_64/src/yt-trunk-svn/yt/lagos/OutputTypes.py", line 150, in _get_hierarchy
    self.__hierarchy = self._hierarchy_class(self, data_style=self.data_style)
  File "/mnt/home/oshea/yt-x86_64/src/yt-trunk-svn/yt/lagos/HierarchyType.py", line 374, in __init__
    AMRHierarchy.__init__(self, pf, data_style)
  File "/mnt/home/oshea/yt-x86_64/src/yt-trunk-svn/yt/lagos/HierarchyType.py", line 55, in __init__
    self._parse_hierarchy()
  File "/mnt/home/oshea/yt-x86_64/src/yt-trunk-svn/yt/lagos/HierarchyType.py", line 435, in _parse_hierarchy
    if self._parse_binary_hierarchy(): return
  File "/mnt/home/oshea/yt-x86_64/src/yt-trunk-svn/yt/lagos/HierarchyType.py", line 506, in _parse_binary_hierarchy
    self.grid_dimensions[:] = f["/ActiveDimensions"][:]
  File "/mnt/home/oshea/yt-x86_64/lib/python2.6/site-packages/h5py-1.2.0-py2.6-linux-x86_64.egg/h5py/highlevel.py", line 961, in __getitem__
    basetype = self.id.dtype
  File "h5d.pyx", line 109, in h5py.h5d.DatasetID.dtype.__get__ (h5py/h5d.c:1423)
  File "h5t.pyx", line 364, in h5py.h5t.TypeID.dtype.__get__ (h5py/h5t.c:2919)
  File "h5t.pyx", line 367, in h5py.h5t.TypeID.py_dtype (h5py/h5t.c:2976)
TypeError: No NumPy equivalent for TypeIntegerID exists

I installed this version of yt on MSU's supercomputer using the install script at http://svn.enzotools.org/yt/trunk/doc/install_script.sh (the supercomputer uses gcc v4.1.2), and as far as I can tell everything relevant, including hdf5 and h5py, installed successfully.  The same simple set of commands works fine on my laptop with the same dataset.  I'm pointing my LD_LIBRARY_PATH environmental variable to the correct directory, so I don't think it's that.  Any ideas about what's going on here?

Some more system details:  The output of uname -a for this system is is "Linux dev-amd09 2.6.16.60-0.21-smp #1 SMP Tue May 6 12:41:02 UTC 2008 x86_64 x86_64 x86_64 GNU/Linux".  The system uses gcc v4.1.2 and I'm installing yt _without_ wxPython.

Thanks,
Brian