[gpaw-users] UnitCellFilter calculation crashes

[Honkala Karoliina]

Dear all,

I'm running a UnitCellFilter calculation for bulk CuO in plane wave basis
with gpaw version 0.9.1.10452 downloaded yesterday from trunk together with
ASE 3.7.0.
The calculation crashes after converging electronic structure for the first
time. The key information in the error message is

File "/common/usrlocal/gpaw/gpaw-0.91rev10452-ase3.7.1-intel/lib64/python2.6/site-packages/gpaw/paw.py", line 360, in initialize
    self.initialize(atoms)
File "/common/usrlocal/gpaw/gpaw-0.91rev10452-ase3.7.1-intel/lib64/python2.6/site-packages/gpaw/paw.py", line 360, in initialize
    magmom_av[:, 2] = magmom_a
ValueError:     magmom_av[:, 2] = magmom_a
operands could not be broadcast together with shapes (10) (8)
ValueError: operands could not be broadcast together with shapes (10) (8)
GPAW CLEANUP (node 0): <type 'exceptions.ValueError'> occurred.  Calling MPI_Abort!
GPAW CLEANUP (node 3): <type 'exceptions.ValueError'> occurred.  Calling MPI_Abort!

During electronic structure interations  Open MPI also complains that
"An MPI process has executed an operation involving a call to the
"fork()" ..." but my understanding is (I might be wrong as I'm not so familiar with this)
that this has not anything to do with the error that crashes my calculation.

I use 4 cores in the calculation and the number of k-points is 4, this is spin parallel calculation and
the number of bands is 52, which is divisible by 4.

Any ideas what I could do to get this job to run ?

Best regards,
                   Karoliina
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://listserv.fysik.dtu.dk/pipermail/gpaw-users/attachments/20130627/cf61e654/attachment.html