]> https://gitweb.dealii.org/ - dealii.git/log
dealii.git
9 years agoTest umfpack/umfpack_04: Avoid printing exact number of iterations
Matthias Maier [Fri, 19 Dec 2014 11:33:45 +0000 (12:33 +0100)]
Test umfpack/umfpack_04: Avoid printing exact number of iterations

This commit introduces a new macro for tests to test an iterative solve to
convergence within a given number of iteration steps. Due to roundoff
errors the exact number doesn't have to be stable...

9 years agoTest bits/step-51: Improve output
Matthias Maier [Fri, 19 Dec 2014 11:05:36 +0000 (12:05 +0100)]
Test bits/step-51: Improve output

This allows numdiff to catch small roundoff errors.

9 years agoFix constructor of SchurMatrix for 64 bit indices.
Martin Kronbichler [Fri, 19 Dec 2014 08:27:02 +0000 (09:27 +0100)]
Fix constructor of SchurMatrix for 64 bit indices.

9 years agoFix bug in Trilinos SparseMatrix::add/copy_from
Martin Kronbichler [Thu, 18 Dec 2014 11:03:18 +0000 (12:03 +0100)]
Fix bug in Trilinos SparseMatrix::add/copy_from

We assume that the pointers to a Trilinos sparse matrix remain valid when a matrix
is copied from another one or added and the two matrices share the same sparsity pattern.
This is used in step-32. However, the fix introduced in 8d64256 could not ensure this
(while fixing a few other cases). The check if the two sparsity patterns point to the same
memory is not good (we can have the same sparsity pattern memory but different addresses).

Therefore, the add and copy_from methods need to be more careful to not set up a new
matrix structure when this is not necessary. In particular, the two methods now check
whether the column indices are the same in any case but without requiring calling
methods that change the memory layout.

9 years agoMerge pull request #358 from tamiko/dealii-8.2
Matthias Maier [Thu, 18 Dec 2014 09:17:24 +0000 (10:17 +0100)]
Merge pull request #358 from tamiko/dealii-8.2

Cherry-pick commits to the release branch

9 years agoFix previous commit 1d327fd1dcd34023a3d5bb2c0ee7524593d44015. 358/head
Wolfgang Bangerth [Thu, 18 Dec 2014 08:37:02 +0000 (02:37 -0600)]
Fix previous commit 1d327fd1dcd34023a3d5bb2c0ee7524593d44015.

Commit 1d327fd1dcd34023a3d5bb2c0ee7524593d44015 moved a function from the .cc
file into the .h file and made it inline. Subtly, it also snuck in a small
change in the code: it changed
  if (!something)
   {
     ...
   }

  if (something)
into
  if (!something)
    {
      ...
    }
  else

The problem is that the first code block in rare circumstances changed the
value of the variable something. It is remarkable that this only triggered
in a single test in the testsuite -- an indication that the testsuite is
after all reasonably good at finding even obscure problems, but also an
indication that the anisotropic refinement stuff is not particularly well
tested.

9 years agoMerge pull request #354 from tamiko/dealii-8.2
Matthias Maier [Thu, 18 Dec 2014 09:14:12 +0000 (10:14 +0100)]
Merge pull request #354 from tamiko/dealii-8.2

Remove unfinished example steps

9 years agoRemove unfinished example steps 47 and 50 354/head
Matthias Maier [Wed, 17 Dec 2014 23:11:43 +0000 (00:11 +0100)]
Remove unfinished example steps 47 and 50

9 years agoMerge pull request #353 from tamiko/dealii-8.2
Matthias Maier [Wed, 17 Dec 2014 18:01:44 +0000 (19:01 +0100)]
Merge pull request #353 from tamiko/dealii-8.2

Cherry-pick commits to the release branch

9 years agoFix copyright years in doc/users/doxygen.html 353/head
Felix Gruber [Wed, 17 Dec 2014 09:33:12 +0000 (10:33 +0100)]
Fix copyright years in doc/users/doxygen.html

The year 2013 is already included in the range 2012 - 2014.

9 years agoAdd instantations for class RungeKutta.
Bruno Turcksin [Wed, 17 Dec 2014 15:52:01 +0000 (09:52 -0600)]
Add instantations for class RungeKutta.

9 years agoMerge pull request #349 from tamiko/dealii-8.2
Matthias Maier [Wed, 17 Dec 2014 08:02:22 +0000 (09:02 +0100)]
Merge pull request #349 from tamiko/dealii-8.2

Cherry-pick commits to the release branch

9 years agofix doxygen warnings 349/head
Timo Heister [Wed, 17 Dec 2014 07:49:13 +0000 (02:49 -0500)]
fix doxygen warnings

9 years agotry to get the arpack test a bit more stable
Matthias Maier [Wed, 17 Dec 2014 00:58:28 +0000 (01:58 +0100)]
try to get the arpack test a bit more stable

Well, it turns out that depending on arpack version and CPU the result for
the first 5 eigenvalues differs. Try to make this a bit more stable by
computing 8 eigenvalues and only print the first 5.

9 years agoUpdate normalize.pl to also substitute 8.2.pre version
Matthias Maier [Wed, 17 Dec 2014 07:26:33 +0000 (08:26 +0100)]
Update normalize.pl to also substitute 8.2.pre version

9 years agoupdate AUTHORS
Matthias Maier [Tue, 16 Dec 2014 15:28:36 +0000 (16:28 +0100)]
update AUTHORS

9 years agoupdate VERSION
Matthias Maier [Tue, 16 Dec 2014 15:08:55 +0000 (16:08 +0100)]
update VERSION

9 years agoMerge pull request #344 from tamiko/fix_some_tests
Martin Kronbichler [Tue, 16 Dec 2014 14:59:03 +0000 (15:59 +0100)]
Merge pull request #344 from tamiko/fix_some_tests

Fix a bunch of tests for simserv

9 years agoMerge pull request #343 from tamiko/changes_and_doc_update
Luca Heltai [Tue, 16 Dec 2014 09:45:04 +0000 (10:45 +0100)]
Merge pull request #343 from tamiko/changes_and_doc_update

Release steps 0e 1 1a

9 years agoMerge pull request #342 from tamiko/doxygen_fixups_and_reindent
Luca Heltai [Tue, 16 Dec 2014 09:44:24 +0000 (10:44 +0100)]
Merge pull request #342 from tamiko/doxygen_fixups_and_reindent

Release steps 0a 0b 0c 0d

9 years agonormalize parameter_handler_write_json.output for different boost versions 344/head
Matthias Maier [Tue, 16 Dec 2014 07:55:56 +0000 (08:55 +0100)]
normalize parameter_handler_write_json.output for different boost versions

9 years agoUpdate codim_01.mpirun=3.debug.output
Matthias Maier [Tue, 16 Dec 2014 01:48:46 +0000 (02:48 +0100)]
Update codim_01.mpirun=3.debug.output

9 years agosort output of arpack/step-36_ar
Matthias Maier [Tue, 16 Dec 2014 01:44:37 +0000 (02:44 +0100)]
sort output of arpack/step-36_ar

9 years agoupdate changes.h and tags for release 343/head
Matthias Maier [Mon, 15 Dec 2014 23:27:06 +0000 (00:27 +0100)]
update changes.h and tags for release

9 years agoUpdate blurbb for petsc and trilinos
Matthias Maier [Mon, 15 Dec 2014 23:36:45 +0000 (00:36 +0100)]
Update blurbb for petsc and trilinos

9 years agoBugfix: Fix typo so that our step-graph gets build again... 342/head
Matthias Maier [Mon, 15 Dec 2014 22:58:31 +0000 (23:58 +0100)]
Bugfix: Fix typo so that our step-graph gets build again...

9 years agomanually fix up @ref statements
Matthias Maier [Mon, 15 Dec 2014 22:50:44 +0000 (23:50 +0100)]
manually fix up @ref statements

9 years agoRun wrapcomments.py, release step 0c
Matthias Maier [Mon, 15 Dec 2014 22:26:45 +0000 (23:26 +0100)]
Run wrapcomments.py, release step 0c

9 years agoreindent dof_handler.h, doxygen fixups
Matthias Maier [Mon, 15 Dec 2014 21:47:01 +0000 (22:47 +0100)]
reindent dof_handler.h, doxygen fixups

9 years agoMerge pull request #301 from bangerth/muparser-cleanups
Timo Heister [Sun, 14 Dec 2014 20:33:18 +0000 (15:33 -0500)]
Merge pull request #301 from bangerth/muparser-cleanups

Muparser cleanups

9 years agoMake an error message easier to read. 301/head
Wolfgang Bangerth [Sat, 13 Dec 2014 04:18:12 +0000 (22:18 -0600)]
Make an error message easier to read.

Enclose strings in <...> so that one can more easily see if there happen
to be spaces in untoward places.

9 years agoClean up documentation a bit.
Wolfgang Bangerth [Sat, 13 Dec 2014 03:50:47 +0000 (21:50 -0600)]
Clean up documentation a bit.

9 years agoReorganize code a bit.
Wolfgang Bangerth [Sat, 13 Dec 2014 03:50:36 +0000 (21:50 -0600)]
Reorganize code a bit.

Change the code so that checking whether a function parser object
has already been created now happens at the place where we use
the object, rather than as an early exit in the init_muparser()
function. I find that this makes the code easier to read.

This also allows us to remove the creation of a parser object on the
thread where the object is actually created. In many (most) cases,
we will simply create the parser later on anyway, but not necessarily.

9 years agoMerge pull request #299 from bangerth/add-test-2
Luca Heltai [Fri, 12 Dec 2014 19:56:45 +0000 (20:56 +0100)]
Merge pull request #299 from bangerth/add-test-2

Add another testcase.

9 years agoAdd another testcase. 299/head
Wolfgang Bangerth [Fri, 12 Dec 2014 19:05:30 +0000 (13:05 -0600)]
Add another testcase.

This is for the same problem as reported by Krishna Garikipati
on the mailing list this week. He verified that this used to fail
and works now, so I would like to add it to the testsuite to make
sure we don't regress in this functionality.

9 years agoMerge pull request #298 from Rombur/workstream
Wolfgang Bangerth [Fri, 12 Dec 2014 18:41:52 +0000 (12:41 -0600)]
Merge pull request #298 from Rombur/workstream

Use parallel_for instead of a pipeline in implementation3 of WorkStream.

9 years agoUse parallel_for instead of a pipeline in implementation3 of WorkStream. 298/head
Bruno Turcksin [Fri, 12 Dec 2014 18:05:37 +0000 (12:05 -0600)]
Use parallel_for instead of a pipeline in implementation3 of WorkStream.

9 years agoMerge pull request #289 from l-korous/master
Wolfgang Bangerth [Fri, 12 Dec 2014 15:14:48 +0000 (09:14 -0600)]
Merge pull request #289 from l-korous/master

Make entire library code compile on MSVC & gcc.

9 years agoSquashed changes for build working both on MSVC and other compilers. 289/head
Lukas Korous [Fri, 12 Dec 2014 15:05:07 +0000 (16:05 +0100)]
Squashed changes for build working both on MSVC and other compilers.

9 years agoMerge pull request #297 from tamiko/fix_wno-xxx_detection_for_gcc
Wolfgang Bangerth [Thu, 11 Dec 2014 20:06:21 +0000 (14:06 -0600)]
Merge pull request #297 from tamiko/fix_wno-xxx_detection_for_gcc

CMake: Correctly detect -Wno-... support for gcc

9 years agoCMake: Correctly detect -Wno-... support for gcc 297/head
Matthias Maier [Thu, 11 Dec 2014 19:49:34 +0000 (20:49 +0100)]
CMake: Correctly detect -Wno-... support for gcc

gcc does not emit a warning if an unknown -Wno-... flag is specified on the
command line, thus the ENABLE_IF_SUPPORTED macro unconditionally enabled
-Wno-... flags for gcc. Unfortunately, gcc _does_ emit a warning for the
unrecognized compiler option if another warning is emitted in the same
compilation unit. This is now fixed by always querying for the non-negated
version, i.e. -Wfoo instead of -Wno-foo.

9 years agoMerge pull request #296 from bangerth/fix-slepc-icc-problem
Luca Heltai [Thu, 11 Dec 2014 17:33:43 +0000 (18:33 +0100)]
Merge pull request #296 from bangerth/fix-slepc-icc-problem

Elide default initialization.

9 years agoElide default initialization. 296/head
Wolfgang Bangerth [Thu, 11 Dec 2014 17:12:40 +0000 (11:12 -0600)]
Elide default initialization.

Apparently, the initialization of a std::shared_ptr with NULL is not
allowed with the Intel compiler. It is also not necessary (that's what
the default constructor does anyway), so just remove the line.

9 years agoMerge pull request #295 from bangerth/code-reorg
Luca Heltai [Thu, 11 Dec 2014 15:25:58 +0000 (16:25 +0100)]
Merge pull request #295 from bangerth/code-reorg

Reorganize code in symmetric_tensor.h.

9 years agoReorganize code in symmetric_tensor.h. 295/head
Wolfgang Bangerth [Thu, 11 Dec 2014 13:09:03 +0000 (07:09 -0600)]
Reorganize code in symmetric_tensor.h.

The current implementation of SymmetricTensor<rank,dim,Number>::unrolled_to_component_indices
and its inverse function was only implemented for rank=2 and looked essentially like this:

  Assert (rank == 2, ExcNotImplemented());
  Assert (i < n_independent_components, ExcIndexRange(i, 0, n_independent_components));
  switch (dim)
    {
    case 1:
      return TableIndices<2>(0,0);
    ...

Such code cannot be generalized to rank=4 (in fact, it doesn't even compile
for rank=4) because we would have to return objects of different types
in any switch on rank. The only way around this is to use dispatch to
different functions that do the work for a particular rank.

This patch does the first part of this: set up the dispatch. Later patches
may in fact implement this function and its inverse for other ranks than 2.

9 years agoMerge pull request #294 from bangerth/add-test
Luca Heltai [Thu, 11 Dec 2014 07:53:42 +0000 (08:53 +0100)]
Merge pull request #294 from bangerth/add-test

Add a test.

9 years agoAdd a test. 294/head
Wolfgang Bangerth [Thu, 11 Dec 2014 02:46:29 +0000 (20:46 -0600)]
Add a test.

9 years agoMerge pull request #293 from tamiko/miscellaneous_fixups
Matthias Maier [Wed, 10 Dec 2014 19:24:03 +0000 (20:24 +0100)]
Merge pull request #293 from tamiko/miscellaneous_fixups

Miscellaneous fixups

9 years agoalso add the 'package' target to $ make info output 293/head
Matthias Maier [Wed, 10 Dec 2014 14:20:26 +0000 (15:20 +0100)]
also add the 'package' target to $ make info output

9 years agoadd package component to rewrite rule
Matthias Maier [Wed, 10 Dec 2014 14:13:22 +0000 (15:13 +0100)]
add package component to rewrite rule

9 years agoWhitespace and indenting fixups
Matthias Maier [Wed, 10 Dec 2014 14:10:28 +0000 (15:10 +0100)]
Whitespace and indenting fixups

9 years agoMerge pull request #291 from luca-heltai/cpack-support
Matthias Maier [Wed, 10 Dec 2014 13:10:14 +0000 (14:10 +0100)]
Merge pull request #291 from luca-heltai/cpack-support

Added cpack configuration, and Mac Bundle package.

9 years agoFixed licence, restructured setup_cpack 291/head
Luca Heltai [Wed, 10 Dec 2014 12:55:35 +0000 (13:55 +0100)]
Fixed licence, restructured setup_cpack

9 years agoMerge pull request #292 from bangerth/show-shape-functions
Luca Heltai [Wed, 10 Dec 2014 10:32:14 +0000 (11:32 +0100)]
Merge pull request #292 from bangerth/show-shape-functions

Include pictures of shape functions for FE_Q and FE_Q_Hierarchical.

9 years agoInclude pictures of shape functions for FE_Q and FE_Q_Hierarchical. 292/head
Wolfgang Bangerth [Wed, 10 Dec 2014 02:15:47 +0000 (20:15 -0600)]
Include pictures of shape functions for FE_Q and FE_Q_Hierarchical.

This patch uses pictures generated by Manuel Quezada de Luna for the shape
functions of FE_Q and FE_Q_Hierarchical.

9 years agoAdded cpack configuration, and Mac Bundle package.
Luca Heltai [Fri, 14 Nov 2014 09:37:16 +0000 (10:37 +0100)]
Added cpack configuration, and Mac Bundle package.

9 years agoMerge pull request #281 from bangerth/doc-updates-and-test-additions
Wolfgang Bangerth [Tue, 9 Dec 2014 13:12:02 +0000 (07:12 -0600)]
Merge pull request #281 from bangerth/doc-updates-and-test-additions

Doc updates and test additions

9 years agoRewrite parts of the implementation to make it easier to read. 281/head
Wolfgang Bangerth [Mon, 1 Dec 2014 16:24:37 +0000 (10:24 -0600)]
Rewrite parts of the implementation to make it easier to read.

9 years agoRewrite documentation for a function.
Wolfgang Bangerth [Mon, 1 Dec 2014 16:10:33 +0000 (10:10 -0600)]
Rewrite documentation for a function.

9 years agoMerge pull request #290 from tamiko/miscellaneous_cleanup
Luca Heltai [Tue, 9 Dec 2014 13:11:31 +0000 (14:11 +0100)]
Merge pull request #290 from tamiko/miscellaneous_cleanup

Miscellaneous cleanup

9 years agoUpdate CMake documentation 290/head
Matthias Maier [Tue, 9 Dec 2014 13:01:13 +0000 (14:01 +0100)]
Update CMake documentation

9 years agoCleanup DEAL_II_MSVC testsuite handling
Matthias Maier [Tue, 9 Dec 2014 12:57:35 +0000 (13:57 +0100)]
Cleanup DEAL_II_MSVC testsuite handling

9 years agoMerge pull request #288 from bangerth/clarify-documentation
Bruno Turcksin [Mon, 8 Dec 2014 15:31:43 +0000 (09:31 -0600)]
Merge pull request #288 from bangerth/clarify-documentation

Clarify documentation.

9 years agoClarify documentation. 288/head
Wolfgang Bangerth [Mon, 8 Dec 2014 15:27:12 +0000 (09:27 -0600)]
Clarify documentation.

9 years agoAdd tests for a function that currently does not appear to be tested.
Wolfgang Bangerth [Mon, 1 Dec 2014 16:09:52 +0000 (10:09 -0600)]
Add tests for a function that currently does not appear to be tested.

9 years agoMerge pull request #287 from davydden/refinement_optimized
Martin Kronbichler [Sat, 6 Dec 2014 15:16:25 +0000 (16:16 +0100)]
Merge pull request #287 from davydden/refinement_optimized

refine_and_coarsen_optmise

9 years agomade refine_and_coarsen_optmise easier to read and less error prone 287/head
Denis Davydov [Sat, 6 Dec 2014 12:59:23 +0000 (13:59 +0100)]
made refine_and_coarsen_optmise easier to read and less error prone

9 years agoMerge pull request #286 from kronbichler/master
Wolfgang Bangerth [Fri, 5 Dec 2014 17:14:14 +0000 (11:14 -0600)]
Merge pull request #286 from kronbichler/master

Cleanup file of news, added one contribution

9 years agoCleanup file of news, added one contribution 286/head
Martin Kronbichler [Fri, 5 Dec 2014 16:13:44 +0000 (17:13 +0100)]
Cleanup file of news, added one contribution

9 years agoMerge pull request #285 from tjhei/documentation
Wolfgang Bangerth [Thu, 4 Dec 2014 20:42:25 +0000 (14:42 -0600)]
Merge pull request #285 from tjhei/documentation

clarify documentation of ConstraintMatrix::distribute(VECTOR)

9 years agoclarify documentation of ConstraintMatrix::distribute(VECTOR) 285/head
Timo Heister [Thu, 4 Dec 2014 20:37:16 +0000 (15:37 -0500)]
clarify documentation of ConstraintMatrix::distribute(VECTOR)

This came up as a question on the mailing list so we should be more precise.

9 years agoMerge pull request #284 from bangerth/fix-cuthill-mckee-without-dofs
Martin Kronbichler [Thu, 4 Dec 2014 20:07:07 +0000 (21:07 +0100)]
Merge pull request #284 from bangerth/fix-cuthill-mckee-without-dofs

Fix cuthill mckee without dofs

9 years agoAdd a testcase. 284/head
Michał Wichrowski [Thu, 4 Dec 2014 19:58:03 +0000 (13:58 -0600)]
Add a testcase.

9 years agoDoFRenumbering::Cuthill_McKee failed in parallel when a processor had no dofs.
Wolfgang Bangerth [Thu, 4 Dec 2014 19:57:06 +0000 (13:57 -0600)]
DoFRenumbering::Cuthill_McKee failed in parallel when a processor had no dofs.

9 years agoMerge pull request #283 from bangerth/more-ms-vc-fixes
Wolfgang Bangerth [Thu, 4 Dec 2014 16:47:56 +0000 (10:47 -0600)]
Merge pull request #283 from bangerth/more-ms-vc-fixes

More ms vc fixes

9 years agoKeep the old data types in declarations and only use the workaround for MS Visual... 283/head
Wolfgang Bangerth [Thu, 4 Dec 2014 16:46:19 +0000 (10:46 -0600)]
Keep the old data types in declarations and only use the workaround for MS Visual Studio.

9 years agoPatch by Lukas Korous: Make namespace DerivativeApproximation work with MS Visual...
Lukas Korous [Wed, 3 Dec 2014 23:34:38 +0000 (00:34 +0100)]
Patch by Lukas Korous: Make namespace DerivativeApproximation work with MS Visual Studio.

9 years agoMerge pull request #282 from bangerth/patches-for-ms-vc
Wolfgang Bangerth [Wed, 3 Dec 2014 13:19:41 +0000 (07:19 -0600)]
Merge pull request #282 from bangerth/patches-for-ms-vc

Patches for Microsoft Visual Studio

9 years agoAdd a news entry. 282/head
Wolfgang Bangerth [Wed, 3 Dec 2014 13:14:36 +0000 (07:14 -0600)]
Add a news entry.

9 years agoPatch by Lukas Korous: Work around a couple of problems with Microsoft Visual Studio...
Wolfgang Bangerth [Wed, 3 Dec 2014 13:13:26 +0000 (07:13 -0600)]
Patch by Lukas Korous: Work around a couple of problems with Microsoft Visual Studio. In the process, fix a couple of oversights and typos.

9 years agoPatch by Lukas Korous: Only include <mm_malloc.h> if not on Microsoft Visual Studio.
Wolfgang Bangerth [Wed, 3 Dec 2014 13:11:25 +0000 (07:11 -0600)]
Patch by Lukas Korous: Only include <mm_malloc.h> if not on Microsoft Visual Studio.

9 years agoPatch by Lukas Korous: Only include <mm_malloc.h> if not on Microsoft Visual Studio.
Wolfgang Bangerth [Wed, 3 Dec 2014 13:10:52 +0000 (07:10 -0600)]
Patch by Lukas Korous: Only include <mm_malloc.h> if not on Microsoft Visual Studio.

9 years agoPatch by Lukas Korous: Work around a compiler problem in Microsoft Visual Studio...
Wolfgang Bangerth [Wed, 3 Dec 2014 13:09:35 +0000 (07:09 -0600)]
Patch by Lukas Korous: Work around a compiler problem in Microsoft Visual Studio by disabling a safety check (for just this compiler) that should never trigger.

9 years agoPatch by Lukas Korous: Remove the explicit instantiation of a function since this...
Wolfgang Bangerth [Wed, 3 Dec 2014 13:04:42 +0000 (07:04 -0600)]
Patch by Lukas Korous: Remove the explicit instantiation of a function since this creates problems for Microsoft Visual Studio. The instantiation is not strictly needed since the function, internal::MatrixFreeFunctions::ShapeInfo<T>::reinit() is defined in a file that is always included.

9 years agoPatch by Lukas Korous: Apply a workaround for Microsoft Visual Studio.
Wolfgang Bangerth [Wed, 3 Dec 2014 12:59:07 +0000 (06:59 -0600)]
Patch by Lukas Korous: Apply a workaround for Microsoft Visual Studio.

9 years agoPatch by Lukas Korous: Disable the testsuite for Microsoft Visual Studio because...
Wolfgang Bangerth [Wed, 3 Dec 2014 12:54:53 +0000 (06:54 -0600)]
Patch by Lukas Korous: Disable the testsuite for Microsoft Visual Studio because the testsuite uses Unix shell commands.

9 years agoMerge pull request #229 from davydden/GHEP
Toby D. Young [Tue, 2 Dec 2014 08:55:49 +0000 (09:55 +0100)]
Merge pull request #229 from davydden/GHEP

generalised hermitian

9 years agoMerge pull request #279 from tamiko/fix_periodic_bc
Matthias Maier [Mon, 1 Dec 2014 13:00:37 +0000 (14:00 +0100)]
Merge pull request #279 from tamiko/fix_periodic_bc

Fix various bugs in periodic boundary conditions

9 years agocode cleanup 279/head
Matthias Maier [Mon, 1 Dec 2014 11:38:03 +0000 (12:38 +0100)]
code cleanup

9 years agoMerge pull request #280 from bangerth/doc-update-4
Timo Heister [Mon, 1 Dec 2014 02:36:36 +0000 (21:36 -0500)]
Merge pull request #280 from bangerth/doc-update-4

Update documentation in one place.

9 years agoUpdate documentation in one place. 280/head
Wolfgang Bangerth [Mon, 1 Dec 2014 02:27:22 +0000 (20:27 -0600)]
Update documentation in one place.

9 years agomake astyle happy
Matthias Maier [Wed, 26 Nov 2014 22:34:04 +0000 (23:34 +0100)]
make astyle happy

9 years agoAvoid unnecessary inversion of a transformation matrix
Matthias Maier [Sun, 30 Nov 2014 14:38:23 +0000 (15:38 +0100)]
Avoid unnecessary inversion of a transformation matrix

9 years agoBugfix: adjust rotation for inverted matching
Matthias Maier [Wed, 26 Nov 2014 23:51:46 +0000 (00:51 +0100)]
Bugfix: adjust rotation for inverted matching

When constraining from face_1 to face_2 we have to use an (orientation,
flip, rotation) bitset that actually gives the relative orientation of
face_2 to face_1 (and not as specified face_1 to face_2).

This worked by accident because the constraining direction face_2 to face_1
is accidentally almost always used in all test cases.

9 years agoadd another version of the dof_tools_21_b
Matthias Maier [Sun, 30 Nov 2014 13:31:46 +0000 (14:31 +0100)]
add another version of the dof_tools_21_b

that also tests for correct behaviour with hanging nodes. This is done by
additionally refining the second cube once. Test that constraining face_1
-> face_2 and the opposite direction face_2 -> face_1 give the exact same
result.

Manually verified that this is indeed the case

9 years agoAlso check for reverse matching in bits/dof_tools_21_b
Matthias Maier [Sun, 30 Nov 2014 13:37:02 +0000 (14:37 +0100)]
Also check for reverse matching in bits/dof_tools_21_b

Also check for the inverse matching from face_2 to face_1 in the test
bits/dof_tools_21_b. Both directions must produce the same constraint
matrix except for the ordering of the constraints possibly being swapped.

Manually verified that this is the case

9 years agoBugfix: Conditionally use matrix in orthogonal_equality
Matthias Maier [Wed, 26 Nov 2014 21:53:04 +0000 (22:53 +0100)]
Bugfix: Conditionally use matrix in orthogonal_equality

Only apply the parameter matrix in orthogonal_equality if it is a spacedim
x spacedim matrix and can be interpreted as a rotation.

Further bugfixes and documentation updates.

9 years agoUse spacedim in compute_transformation instead of incorrect dim
Matthias Maier [Wed, 26 Nov 2014 22:30:22 +0000 (23:30 +0100)]
Use spacedim in compute_transformation instead of incorrect dim

9 years agoCode cleanup and several bugfixes
Matthias Maier [Wed, 26 Nov 2014 22:38:39 +0000 (23:38 +0100)]
Code cleanup and several bugfixes

Refactor the creation of the final (possibly rotated) interpolation matrix
in make_periodicity_constraints into its own function. Add some additional
asserts to catch corner cases.

Return early if nothing to do (fixes dof_tools_21* tests)

9 years agorestructure and clean up some code
Matthias Maier [Wed, 26 Nov 2014 22:38:17 +0000 (23:38 +0100)]
restructure and clean up some code


In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move.

Douglas Adams


Typeset in Trocchi and Trocchi Bold Sans Serif.