Commit Graph

230 Commits

Author SHA1 Message Date
Martin Diehl b11ca71786 no need to pass determinant if not needed 2022-09-10 10:31:18 +02:00
Martin Diehl 61dc14e0b9 Merge remote-tracking branch 'origin/development' into implicit-none-external 2022-06-27 15:13:20 +02:00
Daniel Otto de Mentock de89d0fe38 Merge branch 'Marc-Precision-handling' into 'development'
Marc precision handling

See merge request damask/DAMASK!607
2022-06-27 08:37:41 +00:00
Martin Diehl 3b57934d6e Marc precision handling 2022-06-27 08:37:41 +00:00
Martin Diehl b8c3d75700 [skip sc] enforce interfaces (quick fix: declare as external) 2022-06-21 22:59:01 +02:00
Martin Diehl 0b8473d7ec Merge remote-tracking branch 'origin/development' into polishing 2022-06-15 20:13:39 +02:00
Martin Diehl 0d63284c01 increase test coverage 2022-06-12 16:16:25 +02:00
Martin Diehl f60c950a7c numerically better normalization
slightly more expensive, but guarantees a det(R) = 1.0
2022-06-12 10:25:14 +02:00
Martin Diehl 78838c2f75 consistent variants including space 2022-06-08 23:27:09 +02:00
Martin Diehl b376b10b7a classic Intel gives FPE 2022-05-08 17:47:20 +02:00
Martin Diehl 8f9fbb30e5 use fused multiply-add where possible
only possible for Intel compiler
2022-05-08 15:48:13 +02:00
Martin Diehl 2741fbd977 whitespace adjustments
- 1 empty line between dummy arguments/return value
- 2 empty lines between variable definition and actual code
2022-05-08 09:27:21 +02:00
Martin Diehl c66e2336c2 some versions of ifort have problems with 'do concurrent' 2022-02-14 15:08:35 +01:00
Martin Diehl bd657e6c62 fixed style and corrected copy and paste errors 2022-02-09 12:29:55 +01:00
Martin Diehl a673abb413 indicate that this mapping should be used only for C 2022-01-30 06:24:50 +01:00
Martin Diehl 774ec4e8c4 not used 2022-01-29 14:47:14 +01:00
Martin Diehl dd0f2cfa3c tau is often easier to use than pi
https://tauday.com/tau-manifesto
2022-01-29 14:46:09 +01:00
Martin Diehl c5fd6a3677 random seed should contain some entropy
setting all elements to the same value does not make sense. Moreover,
negative values are allowed
2022-01-18 10:47:26 +01:00
Martin Diehl 2f74e0d070 avoid failing self test
increase number of samples to have less corner cases.
Needs to be allocatable to avoid stack/heap issue on ifort
2022-01-01 11:39:02 +01:00
Martin Diehl fb51e3c4cd functions have no side-effects, hence 'pure' 2021-12-29 07:19:26 +01:00
Martin Diehl f833d348e0 testing random sampling 2021-12-17 08:01:15 +01:00
Martin Diehl f40d731fe1 use the Box-Muller transform instead of random sampling
still needs testing.
2021-12-16 21:22:44 +01:00
Sharan Roongta 7e7098baf7 Merge branch 'rotate-Voigt' into 'development'
Rotate voigt

See merge request damask/DAMASK!463
2021-11-29 16:47:39 +00:00
Philip Eisenlohr a531b7ccae transitioned remaining real exponents to int 2021-11-28 12:46:26 -05:00
Martin Diehl 1fbeabc949 Merge remote-tracking branch 'origin/development' into rotate-Voigt 2021-11-27 19:55:02 +01:00
Martin Diehl bfc6b69ee2 integer exponents: faster and shorter 2021-11-25 20:52:22 +01:00
Philip Eisenlohr da23c916ca polish 2021-11-21 15:49:04 -05:00
Martin Diehl 28ff027b38 exploit symmetry for stress calculation 2021-11-20 22:36:01 +01:00
Martin Diehl a857285e34 backward transformation for Voigt 2021-11-20 22:09:01 +01:00
Martin Diehl dfe6d0a195 more support for Voigt notation 2021-11-18 21:06:38 +01:00
Philip Eisenlohr da9fdf53d2 consistent indentation and line-spacings in reporting 2021-11-15 12:35:44 -05:00
Vitesh 4a8c836117 Merge branch 'store-sim-setup' into 'development'
store information on slip and twin systems

See merge request damask/DAMASK!422
2021-08-03 13:16:59 +00:00
Martin Diehl 044a048944 taking care of corner cases (e.g. restart)
adjusting tests to take care of new 'setup' group
2021-08-01 22:46:11 +02:00
Martin Diehl af714cfd5b broken Intel compiler (do concurrent)
errors occur under unclear conditions, better avoid at the moment:

https://community.intel.com/t5/Intel-Fortran-Compiler/do-concurrent-broken-with-openmp/m-p/1301528#M156942
https://community.intel.com/t5/Intel-Fortran-Compiler/Bug-with-do-concurrent-and-openmp/m-p/1173473m
2021-07-27 16:44:13 +02:00
Martin Diehl c7ed0e7934 Merge remote-tracking branch 'origin/development' into do-concurrent 2021-07-26 20:20:52 +02:00
Martin Diehl e56f2e09a4 trivial and never used
even the more commoly used fucntions for 3x3 matrices are not all in use
2021-07-26 10:22:53 +02:00
Martin Diehl d19ab4c4f6 co concurrent possible here
note: do concurrent cannot be used for the double loops for 66 to 3333!
2021-07-25 14:37:50 +02:00
Martin Diehl 31a4000655 no need for arbitrary dimension
not sure if correct, not used at all (even identity4th for 3 dim is not
used, but now at least tested)
2021-07-25 13:32:11 +02:00
Martin Diehl 3bb5ae3d9e concurrent/parallel execution is possible here 2021-07-25 13:20:39 +02:00
Martin Diehl 4fab078d18 multinomial tested+simplified 2021-07-22 15:31:30 +02:00
Martin Diehl 013c4c0a92 common variable names 2021-07-16 10:31:40 +02:00
Martin Diehl 72ab936ec3 cleaning 2021-05-23 00:12:55 +02:00
Martin Diehl 2ca3a824a4 better fail immediately 2021-02-25 23:11:36 +01:00
Martin Diehl 6f65de27fc not used
was only used for reporting (see v.2.0.0)
2021-01-07 14:59:12 +01:00
Martin Diehl 80c8c9e51e fixing use for new Intel Fortran
visibility of use associated entities can now be limited
2020-11-28 07:23:55 +01:00
Martin Diehl 3d7aceb99b avoid underflow/division by zero 2020-11-11 20:57:17 +01:00
Martin Diehl cd7ada0da9 use Fortran internals for error handling 2020-11-11 20:09:14 +01:00
Martin Diehl 41fbc58c1b standard conforming name 2020-09-22 13:09:12 +02:00
Martin Diehl 63f9078f04 don' hardcode output unit
replaced remaining write(6,..), flush(6) with write(OUTPUT_UNIT,...),
flush(OUTPUT_UNIT)
2020-09-22 12:15:36 +02:00
Martin Diehl 797073b015 print choses output unit automatically, no need to rely on 6
also shorter...
2020-09-17 19:28:41 +02:00