Adam Nelson
3524728fbc
Merge remote-tracking branch 'upstream/develop' into scatt_bins
2015-10-04 13:02:12 -04:00
Sterling Harper
3327242acf
Fix Python import in test_track_output
2015-10-03 23:53:06 -04:00
Paul Romano
b4a92883f2
Allow run_tests.py to work even without MPI installed
2015-10-04 10:32:15 +07:00
Paul Romano
3ca7279bd8
Merge branch 'develop' into complex-cells
...
Conflicts:
openmc/opencg_compatible.py
openmc/surface.py
2015-10-04 10:22:48 +07:00
Adam Nelson
fdd343e8df
Fixed failing tests
2015-10-03 12:56:49 -04:00
Sterling Harper
03a8fb64e6
Address @paulromano comments on #469
2015-10-03 11:28:36 -04:00
Adam Nelson
a393378f4d
Updated tests for the revised calc_rn
2015-10-03 10:42:22 -04:00
Sterling Harper
6b92694578
Return isotropic scores to total_yn, flux_yn tests
2015-10-03 10:01:43 -04:00
Adam Nelson
147d581c7c
Merge remote-tracking branch 'upstream/develop' into scatt_bins
2015-10-03 09:29:32 -04:00
Sterling Harper
070a229a93
Fix test_score_nuscatter* inactive batches
2015-10-03 02:02:44 -04:00
Sterling Harper
8030112938
Use OrderedDict for material nuclides in PyAPI
2015-10-03 00:41:42 -04:00
Sterling Harper
d14750149a
Fix Python paths in tests
2015-10-02 23:59:41 -04:00
Paul Romano
71dfde8d12
Adopt shorthand notation for half-spaces and regions in Python API
...
Surfaces now have __neg__ and __pos__ operators, and regions have __and__,
__or__, and __invert__. This makes the syntax for complex regions much less
bulky.
2015-10-03 09:26:50 +07:00
Sterling Harper
27548d6232
Allow mpif90 in run_tests.py
2015-10-02 22:25:07 -04:00
Sterling Harper
3d5840214d
Move more tests to PythonAPI
2015-10-02 18:03:32 -04:00
Sterling Harper
15c7a1d7b6
Merge remote-tracking branch 'upstream/develop' into pyapi_tests
2015-10-02 16:35:54 -04:00
Paul Romano
6c6676cc14
Have openmc-update-inputs change 'surfaces' to 'region'
2015-09-28 19:01:10 +07:00
Paul Romano
adfa0288ce
Update all geometry.xml files to use 'region'
2015-09-28 16:54:48 +07:00
Adam Nelson
cd365eeb71
Merge branch 'nuscatter_fix' into scatt_bins
2015-09-28 04:59:46 -04:00
Adam Nelson
d719129a56
Removed explicit casting of multiplicity to real in tally.f90, removed tally module dependence on interpolation, and updated nuscatter tests to better capture nu-scattering
2015-09-28 04:59:29 -04:00
Paul Romano
98d745d9ce
Add test with complex cells
2015-09-28 14:11:18 +07:00
Adam Nelson
754c8ea57e
Merge branch 'nuscatter_fix' into scatt_bins
2015-09-26 14:07:03 -04:00
Adam Nelson
5658cb0e66
All tested and results updates
2015-09-26 14:06:45 -04:00
Adam Nelson
c1630a02db
merged w/ upstream
2015-09-26 11:39:18 -04:00
Paul Romano
dce591b037
Improve check for coincidence in sense().
...
The previous solution was to move the particle forward artificially and then
call sense() recursively. A better way is to look at the projection of the
particle's direction on the surface normal from which the sign determines which
side the particle will be on when it moves forward. This enables us to make
sense() a pure non-recursive function.
2015-09-26 19:58:16 +07:00
Paul Romano
cde3d7a00a
Make reflect(...) a bound procedure of Surface directly.
...
This obviates the need to have an implementation of reflect(...) for each
surface. As long as the surface has a defined normal vector, it can be used to
reflect a particle off the surface.
2015-09-26 19:58:16 +07:00
Paul Romano
f3f6753820
Respond to remaining comments from @wbinventor on #454
2015-09-25 14:43:32 +07:00
Sterling Harper
30425070e4
Move some tests to PythonAPI
2015-09-21 11:36:22 -04:00
Sterling Harper
5e49425714
Merge remote-tracking branch 'upstream/develop' into pyapi_tests
2015-09-21 10:26:27 -04:00
Sterling Harper
4507139173
Add PythonAPI functionality to TestHarness
2015-09-21 10:20:41 -04:00
Paul Romano
bcbb8f6a3f
Merge branch 'develop' into output-improvements
...
Conflicts:
openmc/statepoint.py
2015-09-19 10:28:47 +07:00
Sterling Harper
9c3f77fe90
Add collision estimator tests
2015-09-17 16:20:15 -04:00
Paul Romano
8b67fa7a92
Use structured array for global tallies with mean and std_dev fields
2015-09-17 07:21:04 +07:00
Paul Romano
9bd1dc0cb9
Get rid of SourceSite class and just use ndarray with compound datatype
2015-09-16 18:37:05 +07:00
Paul Romano
97a7582c44
Determine Tally.sum, sum_sq, mean, std_dev on demand. No more read_results()!
2015-09-16 18:37:02 +07:00
Adam Nelson
9dd810e7b6
Merged w/ upstream
2015-09-16 07:21:02 -04:00
Paul Romano
aac5eb6f52
Restructure openmc.particle_restart to read values on demand
2015-09-16 15:45:33 +07:00
Paul Romano
2c4f7f113e
Restructured most of openmc.statepoint.StatePoint to read values on demand from
...
properties.
2015-09-16 15:45:30 +07:00
Sterling Harper
fb518a4d7c
Make os and glob imports explicit
2015-09-14 21:42:25 -04:00
Sterling Harper
b9379e3c86
Make test cleanup source file
2015-09-14 21:05:58 -04:00
Paul Romano
4994493397
Add two test configurations to Travis pull request runs
2015-09-12 00:31:50 +08:00
Paul Romano
04aecbcb94
Remove remnants of binary files
2015-09-04 12:51:18 +07:00
Paul Romano
3c1ba76f95
Rely on FindHDF5 rather than h5fc and h5pfc scripts
...
The FindHDF5.cmake packaged with CMake is broken in a number of respects. The HL
components don't work (debian has a patch in cmake-data). It's also impossible
to prefer a parallel installation if both h5fc and h5pfc appear on your
PATH. Finally, hdf5_hl is not included in the list of libraries needed for the
Fortran_HL component. A local version of FindHDF5 is used here which fixes all
these issues. This enables one to compile OpenMC with MPI + serial HDF5 if
needed by introducing the PHDF5 preprocessor flag.
2015-09-04 12:51:18 +07:00
Paul Romano
d1139cbedc
Remove non-HDF5 configurations from tests
2015-09-04 12:51:18 +07:00
Adam Nelson
95a5271790
Added capabilities for tallying over polar and azimuthal angles, added bounds checking for these new filters and mu while doing filter match searching, updated docs and tests.
2015-08-28 20:54:44 -04:00
Adam Nelson
e6c264ac86
Merge branch 'develop' into scatt_bins
2015-08-28 19:36:15 -04:00
Adam Nelson
de58e3c76e
Added test for mu filter
2015-08-26 20:32:00 -04:00
Paul Romano
23481c6438
Update test results since modeling (n,xn) as secondary particles changes answers
2015-08-21 11:23:55 +07:00
Paul Romano
14c62e70b6
Combined main batch/particle loop structure for eigenvalue and fixed source
...
simulations in one 'simulation' mode. The fixed_source module is now gone, and
the eigenvalue module consists specifically of eigenvalue-related subroutines.
2015-08-21 11:23:50 +07:00
Paul Romano
2c8c771998
Change fixed source simulations to use source_bank rather than dummy source
...
site. This makes fixed source and eigenvalue modes more similar.
2015-08-19 11:16:37 +07:00
