248ef8ef97
more tests
...
- invalid files
- scalar/array size,center,exponent for add_primitive
- rotation invariance for add primitive
2020-08-26 23:56:20 +02:00
229f6139c8
better user experience
...
- no meaningless warnings regarding floating point precision
- meaningful error message for invalid header
2020-08-26 23:54:56 +02:00
b9f1421c6b
[skip ci] removed unnecessary line continuations "\"
2020-08-26 17:27:08 -04:00
3f24d16603
renamed VTK.write() to VTK.to_file()
2020-08-26 16:15:15 -04:00
da46e5ea9a
cannot use np.linalg.norm when dealing with exponent triple
2020-08-26 15:49:25 -04:00
06b524d13e
added 'iso' as possible lattice (with no symmetry)
2020-08-26 21:39:20 +02:00
e3f310fa16
[skip ci] updated version information after successful test of v3.0.0-alpha-27-g68c2908b
2020-08-26 15:05:04 +02:00
68c2908ba8
Merge branch 'rename-hdf5-output' into development
2020-08-26 12:34:56 +02:00
1a9904402a
Merge remote-tracking branch 'origin/development' into geom-out-of-place
2020-08-26 06:22:24 +02:00
7dba9518cc
adjust to new Geom behavior
2020-08-25 23:43:59 +02:00
cbf402c1a4
[skip ci] updated version information after successful test of v3.0.0-alpha-22-gcdc9f078
2020-08-25 23:06:46 +02:00
fee21cbd9c
more sensible tests
2020-08-25 17:17:49 +02:00
cdc9f07838
Merge branch 'increase-geom-test-coverage' into 'development'
...
Increase geom test coverage
See merge request damask/DAMASK!204
2020-08-25 14:40:42 +02:00
4fe4190ee0
mapping renamed
...
now we have two mappings, 'phase' and 'homogenization'. They `unwind`
the definition in 'materialpoint' in 'material.yaml'. Since we
never introduced the inverse mapping, I have removed the cellResults
subfolder.
2020-08-25 11:02:51 +02:00
bf401e56cf
avoid name clash with damask.Geom
2020-08-25 09:56:24 +02:00
800dac5d01
correct type handling
...
no reason to calculate fill twice
2020-08-25 08:47:24 +02:00
96d4975fc4
using numpy functionality
2020-08-25 08:34:04 +02:00
51da632108
aiming at testing each individual statement
2020-08-25 07:49:56 +02:00
2d98325fa4
increasing test coverage
2020-08-25 01:21:17 +02:00
453f5a14c5
duplicated test (w/o wait for parallel out)
2020-08-25 00:53:05 +02:00
2751cdb6af
support comments also in vtr
2020-08-25 00:40:14 +02:00
9a2ac31545
explicit is better then implicit
2020-08-24 23:35:46 +02:00
0ceba2a6d3
use central functionality
2020-08-24 23:28:26 +02:00
71e08ea66a
small fixes:
...
- numpy has multiple float variants
- start renaming microstructure -> materialpoint where it does not hurt
- no need for type check if comparing against set of valid directions
(or I miss some corner cases)
2020-08-24 23:23:47 +02:00
2560f014a3
Merge remote-tracking branch 'origin/increase-geom-test-coverage' into geom-out-of-place
2020-08-24 17:17:31 -04:00
55443a3ff5
reflect change of variable names in file layout version
2020-08-24 22:54:09 +02:00
8ef473b71c
new names in output
...
- constituent -> phase
- materialpoint -> homogenization
this gives a one-to-one correspondence to the material.yaml definition
next steps
1) adjust damask.Result python class
2) implement materialpoint output (homogeneous over the whole domain)
for homogenized quantities (currently T,F,P,phi)
2020-08-24 19:22:53 +02:00
56157af7db
updated script for building fedora (rpm) package
2020-08-24 19:06:05 +02:00
279d43ccc6
include version info + date in comments (Table class)
2020-08-24 19:04:59 +02:00
9d505c8518
better readable
...
- more explicit reporting
- always relative import for DAMASK
2020-08-24 09:58:24 +02:00
b393da4955
relative imports
...
... it's all about damask.XXX
2020-08-24 07:28:10 +02:00
0d3ef29554
more logical encapsulation
...
- getting data out of VTK is a functionality of the VTK class. Need to
discuss whether the VTK class should also provide grid and size (works
only for rectilinear grid, so hiding VTK type is not possible
anymore), but the current situation requires the Geom class to rely on
'internals' of the VTK class
- header is specific to *.geom format, not to Geom objects.
2020-08-24 06:46:22 +02:00
507a165733
functionality for getting data
...
- handy for Geom class, but could be of general interest
- add/set/get comments: follows same functions as for Geom class
2020-08-24 06:40:36 +02:00
3738002e4a
forgot the new reference files...
2020-08-23 22:27:53 -04:00
446ac03b07
All geom methods are now out-of-place, i.e. return an updated duplicate (to allow for daisy chaining).
...
* Added comments when methods acted.
* Added diff method
* Added flip method
* Fixed add_primitive inversion bug (again...)
* Fixed cell centering bug in add_primitive
* Added missing tests
2020-08-23 19:43:46 -04:00
22da3af92e
cleaning (prospector complaint)
2020-08-24 00:44:18 +02:00
44015082a3
dead code (copy and paste error)
2020-08-24 00:34:07 +02:00
c7cbd961a1
more systematic monkeypatching
...
allows comparison of reference files (log/history does not contain
information about damask.version and datetime.datetime.now)
2020-08-24 00:31:38 +02:00
9867830d68
store history of geometry modifications
2020-08-23 23:52:48 +02:00
a1c78b778a
WIP: track history in VTK objects
2020-08-23 23:22:53 +02:00
a814db5b64
PRIVATE: Master includes yaml changes
2020-08-23 19:31:30 +02:00
7b50a3b364
material.config is not used anymore
2020-08-23 15:20:11 +02:00
e007aed407
updates in PRIVATE
...
- lightweigth tarball: no examples, no Makefile (explicit cmake)
- improved build instructions for ubuntu/debian package
2020-08-23 13:28:15 +02:00
d65371d42f
removed unused and untested functionality
...
PEP 20: Explicit is better than implicit.
2020-08-23 12:55:55 +02:00
6f79573140
more test coverage
2020-08-23 11:05:56 +02:00
07af7b2f98
geom is written to vtr exlcusively, not to general vtk
...
base tests on vtr
2020-08-23 10:50:15 +02:00
64e7582e8e
consistent interface to ndimage filters
...
Boolean 'periodic' sets mode to 'wrap' or 'nearest'
2020-08-23 10:02:22 +02:00
975db01f31
microstructure -> materialpoint
...
at least for new functionality
2020-08-23 09:38:08 +02:00
5fb2d30ee4
fast and for arbitrary type
2020-08-23 09:02:17 +02:00
b731b1e768
Geom.to_vtk improvements
...
- should be integer
- should have a test
2020-08-23 08:49:49 +02:00
Martin Diehl