]> https://gitweb.dealii.org/ - dealii.git/log
dealii.git
7 years agoGrammar: clarify obtain vs. attain. 4542/head
David Wells [Sat, 24 Jun 2017 14:35:27 +0000 (10:35 -0400)]
Grammar: clarify obtain vs. attain.

Attain and obtain are nearly, but not quite, synonyms: 'attain' implies
achievement or accomplishment while 'obtain' just implies
ownership. This patch fixes our errant usage of 'attain' (one comment
also had a logic error, which I fixed).

7 years agoMerge pull request #4534 from bangerth/dofhandler-cleanup-3
Martin Kronbichler [Thu, 22 Jun 2017 12:44:55 +0000 (14:44 +0200)]
Merge pull request #4534 from bangerth/dofhandler-cleanup-3

Deprecate (hp::)DoFHandler::invalid_dof_index.

7 years agoMerge branch 'master' into dofhandler-cleanup-3 4534/head
Wolfgang Bangerth [Thu, 22 Jun 2017 02:07:02 +0000 (21:07 -0500)]
Merge branch 'master' into dofhandler-cleanup-3

7 years agoMerge pull request #4536 from bangerth/dofhandler-cleanup-7
Daniel Arndt [Wed, 21 Jun 2017 15:35:14 +0000 (17:35 +0200)]
Merge pull request #4536 from bangerth/dofhandler-cleanup-7

Use a cheaper call.

7 years agoDeprecate (hp::)DoFHandler::invalid_dof_index.
Wolfgang Bangerth [Tue, 20 Jun 2017 08:17:04 +0000 (02:17 -0600)]
Deprecate (hp::)DoFHandler::invalid_dof_index.

Also replace all uses by numbers::invalid_dof_index. I can't see a reason to
have both of these numbers, in particular because the only logical choice is
to have them have the same value.

7 years agoUse a cheaper call. 4536/head
Wolfgang Bangerth [Wed, 21 Jun 2017 14:01:23 +0000 (08:01 -0600)]
Use a cheaper call.

The functions in question take an IndexSet argument that can either mean
something, or if it is empty, is ignored. In the places where an empty
IndexSet is passed, it actually has size zero, not just zero elements,
but in the existing code, we test whether 'indices.n_elements()==0'.
It turns out that 'n_elements()' is a pretty expensive function to call.
What we really wanted, of course, was 'indices.size()==0', and this
patch makes that change.

I ran all multigrid tests (the functions in question are multigrid
related), and they all succeed.

7 years agoMerge pull request #4532 from bangerth/dofhandler-cleanup-6
David Wells [Wed, 21 Jun 2017 12:53:28 +0000 (08:53 -0400)]
Merge pull request #4532 from bangerth/dofhandler-cleanup-6

Check whether a function call makes sense at the place where the function is called

7 years agoRename a variable for greater clarity. 4532/head
Wolfgang Bangerth [Wed, 21 Jun 2017 09:38:16 +0000 (03:38 -0600)]
Rename a variable for greater clarity.

7 years agoAvoid a global communication by using a function that's already there.
Wolfgang Bangerth [Wed, 21 Jun 2017 09:35:44 +0000 (03:35 -0600)]
Avoid a global communication by using a function that's already there.

7 years agoMerge pull request #4533 from bangerth/dofhandler-cleanup-4
Martin Kronbichler [Wed, 21 Jun 2017 07:29:34 +0000 (09:29 +0200)]
Merge pull request #4533 from bangerth/dofhandler-cleanup-4

Remove hp::DoFHandler::MGVertexDoFs.

7 years agoMerge pull request #4528 from bangerth/dofhandler-cleanup-2
Martin Kronbichler [Wed, 21 Jun 2017 07:19:53 +0000 (09:19 +0200)]
Merge pull request #4528 from bangerth/dofhandler-cleanup-2

Simplify the logic of CellDataTransferBuffer::(un)pack_data().

7 years agoMerge pull request #4535 from drwells/compile-transfinite-1d
Martin Kronbichler [Wed, 21 Jun 2017 07:03:47 +0000 (09:03 +0200)]
Merge pull request #4535 from drwells/compile-transfinite-1d

Compile TransfiniteInterpolationManifold<1> in the library.

7 years agoCompile TransfiniteInterpolationManifold<1> in the library. 4535/head
David Wells [Wed, 21 Jun 2017 02:15:25 +0000 (22:15 -0400)]
Compile TransfiniteInterpolationManifold<1> in the library.

In addition, disable usage of the class by throwing an exception.

This class is not as useful in 1D as it is in higher spatial dimensions, but
compiling it in the library improves generic programming (i.e., one can write a
solver that runs in 1D and 2D and on a 2D execution path can use
TransfiniteInterpolationManifold).

7 years agoCheck whether a function call makes sense at the place where the function is called.
Wolfgang Bangerth [Tue, 20 Jun 2017 13:22:52 +0000 (07:22 -0600)]
Check whether a function call makes sense at the place where the function is called.

This is instead of returning early in the function. Check that the arguments are indeed
valid within the function.

7 years agoSimplify the logic of CellDataTransferBuffer::(un)pack_data(). 4528/head
Wolfgang Bangerth [Mon, 19 Jun 2017 11:46:40 +0000 (05:46 -0600)]
Simplify the logic of CellDataTransferBuffer::(un)pack_data().

Specifically, instead of serializing by hand with memcpy etc, use the BOOST
facilities for serialization. This works for 2 out of 3 members of this class
without problem, and the last one is easy to deal with as well.

Secondly, serialize into a gzip compressed stream to reduce the amount of data to
transfer. This reduces the size of messages significantly; for example, in the
last cycle of step-40, we get the following data -- old packet size on the left,
new (compressed) data size on the right:
    6400   1694
    8440   2200
    8440   2412
    6400   1925
    7828   2042
    6060   1589
    1504    583
    1504    520
    6060   1849
    7828   2280
    6060   1947
    7828   2375
    1504    612
      76    102
      76    104
    6400   2025
    8440   2537

I did also try to run things with BOOST's bzip2 compressor instead of gzip,
given that bzip2 often does a better job than gzip, but apparently that
isn't the case here as I get the following:
    6400   1912
    8440   2381
    8440   2712
    6400   2230
    7828   2197
    1504    681
    6060   2177
    6060   1803
    1504    556
    7828   2678
    6060   2285
    7828   2738
    1504    720
      76    122
      76    123
    6400   2392
    8440   2896

7 years agoRemove hp::DoFHandler::MGVertexDoFs. 4533/head
Wolfgang Bangerth [Tue, 20 Jun 2017 09:21:29 +0000 (03:21 -0600)]
Remove hp::DoFHandler::MGVertexDoFs.

The hp::DoFHandler class does not support multilevel DoFs, and so having a
remnant part of the infrastructure for them makes no sense. Removing this
part from the class, however, requires a couple of specializations of
functions in the DoFAccessor class that may have previously compiled,
but would likely have led to segmentation faults if anyone tried to use them.

7 years agoMerge pull request #4526 from bangerth/dofhandler-cleanup
David Wells [Tue, 20 Jun 2017 22:03:59 +0000 (18:03 -0400)]
Merge pull request #4526 from bangerth/dofhandler-cleanup

Clean up some DoFHandler things.

7 years agoMerge pull request #4531 from bangerth/fix-typos
Bruno Turcksin [Tue, 20 Jun 2017 13:40:04 +0000 (09:40 -0400)]
Merge pull request #4531 from bangerth/fix-typos

Fix a couple of typos.

7 years agoFix a couple of typos. 4531/head
Wolfgang Bangerth [Tue, 20 Jun 2017 09:19:24 +0000 (03:19 -0600)]
Fix a couple of typos.

7 years agoMerge pull request #4529 from bangerth/add-static-asserts
Daniel Arndt [Tue, 20 Jun 2017 09:18:01 +0000 (11:18 +0200)]
Merge pull request #4529 from bangerth/add-static-asserts

Add a couple of static asserts to SolutionTransfer.

7 years agoGuard static_asserts against MS VC. 4529/head
Wolfgang Bangerth [Tue, 20 Jun 2017 02:27:49 +0000 (20:27 -0600)]
Guard static_asserts against MS VC.

7 years agoMerge pull request #4530 from drwells/fix-parameterhandler-tests
Wolfgang Bangerth [Mon, 19 Jun 2017 23:36:15 +0000 (18:36 -0500)]
Merge pull request #4530 from drwells/fix-parameterhandler-tests

Fix two test outputs.

7 years agoFix two test outputs. 4530/head
David Wells [Mon, 19 Jun 2017 20:41:53 +0000 (16:41 -0400)]
Fix two test outputs.

Commit 852db02945e4eedf251d528d701b0c2b10a368c2 removed the redundant
"list" (i.e., previously we printed "List list of ...") from
List::description. This commit updates the output files in the
corresponding tests.

7 years agoUse constant strings in two places.
David Wells [Mon, 19 Jun 2017 20:41:34 +0000 (16:41 -0400)]
Use constant strings in two places.

7 years agoAdd a couple of static asserts to SolutionTransfer.
Wolfgang Bangerth [Mon, 19 Jun 2017 11:58:14 +0000 (05:58 -0600)]
Add a couple of static asserts to SolutionTransfer.

This class (in both of its incarnations) takes both a dimension and a DoFHandler
as template argument. The dimension of the DoFHandler better match the explicitly
given dimension. The added static_assert checks that and prevents instantiation
of the class if that is not the case.

7 years agoReturn an object by value, rather than filling a reference argument. 4526/head
Wolfgang Bangerth [Mon, 19 Jun 2017 06:06:00 +0000 (00:06 -0600)]
Return an object by value, rather than filling a reference argument.

In times of move assignment operators, there really is no reason any more not
to do this.

7 years agoMerge pull request #4525 from drwells/ignore-no-virtual-dtor-pragma
Wolfgang Bangerth [Mon, 19 Jun 2017 05:16:47 +0000 (23:16 -0600)]
Merge pull request #4525 from drwells/ignore-no-virtual-dtor-pragma

Add a pragma for ignoring nonvirtual destructors.

7 years agoMerge pull request #4476 from tamiko/add_clang_format_file
Wolfgang Bangerth [Mon, 19 Jun 2017 05:16:06 +0000 (23:16 -0600)]
Merge pull request #4476 from tamiko/add_clang_format_file

Add a .clang-format file

7 years agoMerge pull request #4519 from luca-heltai/key-value-separator-in-map-pattern
Wolfgang Bangerth [Mon, 19 Jun 2017 05:15:11 +0000 (23:15 -0600)]
Merge pull request #4519 from luca-heltai/key-value-separator-in-map-pattern

Key Value separator in Patterns::Map

7 years agoMerge pull request #4524 from drwells/use-d-linear-shape-function
Wolfgang Bangerth [Mon, 19 Jun 2017 03:52:53 +0000 (21:52 -0600)]
Merge pull request #4524 from drwells/use-d-linear-shape-function

Use GeometryInfo<dim>::d_linear_shape_function in two places.

7 years agoRename a member variable to better reflect its purpose.
Wolfgang Bangerth [Sun, 18 Jun 2017 02:55:49 +0000 (20:55 -0600)]
Rename a member variable to better reflect its purpose.

7 years agoBreak comments.
Wolfgang Bangerth [Sun, 18 Jun 2017 02:46:20 +0000 (20:46 -0600)]
Break comments.

7 years agoMove the CellDataTransferBuffer class out of an enclosing (useless) class.
Wolfgang Bangerth [Sun, 18 Jun 2017 02:34:16 +0000 (20:34 -0600)]
Move the CellDataTransferBuffer class out of an enclosing (useless) class.

7 years agoRename the cellinfo structure to CellDataTransferBuffer.
Wolfgang Bangerth [Sun, 18 Jun 2017 02:30:55 +0000 (20:30 -0600)]
Rename the cellinfo structure to CellDataTransferBuffer.

Because that is what the class really represents. Also update the documentation.

7 years agoRemove a commented out piece of code.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:43:50 +0000 (19:43 -0600)]
Remove a commented out piece of code.

The code suggests that one could update the cell DoF cache, but the cache only
exists for non-MG DoF indices, whereas the function where the comment resides
deals with MG level DoFs.

7 years agoNon-functional source cleanups at the textual level.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:38:50 +0000 (19:38 -0600)]
Non-functional source cleanups at the textual level.

7 years agoRemove a wrong comment. Add some code documentation.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:37:59 +0000 (19:37 -0600)]
Remove a wrong comment. Add some code documentation.

7 years agoMake control flow more obvious.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:18:55 +0000 (19:18 -0600)]
Make control flow more obvious.

7 years agoReplace an explicit by an implicit loop.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:17:16 +0000 (19:17 -0600)]
Replace an explicit by an implicit loop.

7 years agoNon-functional source cleanups at the textual level.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:15:09 +0000 (19:15 -0600)]
Non-functional source cleanups at the textual level.

7 years agoDo not check that 'new' returned a non-NULL pointer.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:08:56 +0000 (19:08 -0600)]
Do not check that 'new' returned a non-NULL pointer.

'new' hasn't done that since C++98 but instead throws a std::bad_alloc itself. Consequently,
the assertion would never have failed.

7 years agoAdd documentation.
Wolfgang Bangerth [Sun, 18 Jun 2017 01:05:27 +0000 (19:05 -0600)]
Add documentation.

7 years agoExplicitly delete copy constructor and operator.
Wolfgang Bangerth [Sun, 18 Jun 2017 00:58:24 +0000 (18:58 -0600)]
Explicitly delete copy constructor and operator.

7 years agoUse std::unique_ptr instead of std::shared_ptr for DoFHandler::policy.
Wolfgang Bangerth [Sun, 18 Jun 2017 00:55:13 +0000 (18:55 -0600)]
Use std::unique_ptr instead of std::shared_ptr for DoFHandler::policy.

This is possible because DoFHandler has no copy constructor or operator, and consequently
there is no reason to share a policy.

7 years agoAdjust serialization output for use of std::unique_ptr.
Wolfgang Bangerth [Mon, 19 Jun 2017 01:46:59 +0000 (19:46 -0600)]
Adjust serialization output for use of std::unique_ptr.

7 years agoUse std::unique_ptrs instead of raw pointers for level/face objects in class DoFHandler.
Wolfgang Bangerth [Sun, 18 Jun 2017 00:52:15 +0000 (18:52 -0600)]
Use std::unique_ptrs instead of raw pointers for level/face objects in class DoFHandler.

7 years agoAdd a pragma for ignoring nonvirtual destructors. 4525/head
David Wells [Mon, 19 Jun 2017 01:10:27 +0000 (21:10 -0400)]
Add a pragma for ignoring nonvirtual destructors.

An OpenCascade header triggers this warning, so lets ignore it in external
libraries.

7 years agoUse GeometryInfo<dim>::d_linear_shape_function in two places. 4524/head
David Wells [Mon, 19 Jun 2017 00:23:23 +0000 (20:23 -0400)]
Use GeometryInfo<dim>::d_linear_shape_function in two places.

7 years agoMerge pull request #4523 from masterleinad/fix_uninitialized_var_dof_info
Timo Heister [Sun, 18 Jun 2017 18:18:21 +0000 (20:18 +0200)]
Merge pull request #4523 from masterleinad/fix_uninitialized_var_dof_info

Fix uninitialized member variables in DoFInfo

7 years agoFix uninitialized member variables in DoFInfo 4523/head
Daniel Arndt [Sun, 18 Jun 2017 15:16:42 +0000 (17:16 +0200)]
Fix uninitialized member variables in DoFInfo

7 years agoMerge pull request #4522 from asartori86/fix_assert
Denis Davydov [Sun, 18 Jun 2017 05:45:37 +0000 (07:45 +0200)]
Merge pull request #4522 from asartori86/fix_assert

Added missing template

7 years agoRemove an empty 'private:' section.
Wolfgang Bangerth [Sun, 18 Jun 2017 00:39:47 +0000 (18:39 -0600)]
Remove an empty 'private:' section.

7 years agoadded missing spacedim template 4522/head
alberto sartori [Sat, 17 Jun 2017 21:27:14 +0000 (23:27 +0200)]
added missing spacedim template

7 years agoMerge pull request #4505 from bangerth/strict-enums
Daniel Arndt [Fri, 16 Jun 2017 22:51:40 +0000 (00:51 +0200)]
Merge pull request #4505 from bangerth/strict-enums

Use only enum values, don't look at bits.

7 years agoMerge pull request #4521 from drwells/fix-trilinos-quick-test
Wolfgang Bangerth [Fri, 16 Jun 2017 15:37:02 +0000 (09:37 -0600)]
Merge pull request #4521 from drwells/fix-trilinos-quick-test

Fix the Trilinos quick test.

7 years agoFix the Trilinos quick test. 4521/head
David Wells [Fri, 16 Jun 2017 14:17:13 +0000 (10:17 -0400)]
Fix the Trilinos quick test.

It looks like we did not update this constructor when we removed the serial
vector (the parallel vector's constructor needs an IndexSet, not an integer).

7 years agoMerge pull request #4520 from drwells/vertex-inlining
Martin Kronbichler [Thu, 15 Jun 2017 14:26:40 +0000 (16:26 +0200)]
Merge pull request #4520 from drwells/vertex-inlining

Mark the vertex lookup functions as inline.

7 years agoFixed typos. 4519/head
Luca Heltai [Thu, 15 Jun 2017 07:40:44 +0000 (09:40 +0200)]
Fixed typos.

7 years agoMark the vertex lookup functions as inline. 4520/head
David Wells [Mon, 12 Jun 2017 15:13:23 +0000 (11:13 -0400)]
Mark the vertex lookup functions as inline.

This lowers the number of calls to vertex() by a lot (down from 229 million to
46 million in a benchmark) and also provides about a 5% speed improvement.

For the sake of posterity, here is the benchmark used here and also in
b584c58d95 and cbb473c59d:

#include <deal.II/base/geometry_info.h>

#include <deal.II/grid/grid_in.h>
#include <deal.II/grid/manifold_lib.h>
#include <deal.II/grid/tria.h>

#include <chrono>
#include <fstream>
#include <iostream>

using namespace dealii;

int main()
{
  constexpr int dim = 2;
  PolarManifold<dim> polar_manifold;
  TransfiniteInterpolationManifold<dim> inner_manifold;
  constexpr types::manifold_id polar_manifold_id = 1;
  constexpr types::manifold_id inner_manifold_id = 2;

  Triangulation<dim> triangulation;
  GridIn<dim> grid_in;
  std::ifstream input_grid("triangulation.msh");
  grid_in.attach_triangulation(triangulation);
  grid_in.read_msh(input_grid);
  triangulation.set_all_manifold_ids(inner_manifold_id);
  for (auto cell : triangulation.active_cell_iterators())
    {
      if (cell->at_boundary())
        {
          cell->set_all_manifold_ids(polar_manifold_id);
        }
    }
  triangulation.set_manifold(polar_manifold_id, polar_manifold);
  inner_manifold.initialize(triangulation);
  triangulation.set_manifold(inner_manifold_id, inner_manifold);

#ifdef DEBUG
  constexpr unsigned int n_refinements = 6;
#else
  constexpr unsigned int n_refinements = 8;
#endif

  for (unsigned int refinement_n = 0; refinement_n < n_refinements; ++refinement_n)
    {
      const auto t0 = std::chrono::high_resolution_clock::now();
      triangulation.refine_global(1);
      const auto t1 = std::chrono::high_resolution_clock::now();
      std::cout << "Number of active cells: "
                << triangulation.n_active_cells()
                << '\n';
      std::cout << "Refinement time: "
                << std::chrono::duration_cast<std::chrono::milliseconds>(t1 - t0).count()
                << '\n';
    }
}

and here is triangulation.msh:

$NOD
25
1  1 0 0
2  0.70710678118654757274 0.70710678118654746172 0
3  6.1232339957367660359e-17 1 0
4  -0.70710678118654746172 0.70710678118654757274 0
5  -1 1.2246467991473532072e-16 0
6  -0.70710678118654768376 -0.70710678118654746172 0
7  -1.8369701987210296875e-16 -1 0
8  0.70710678118654735069 -0.70710678118654768376 0
9  0.5 0 0
10  0.35355339059327378637 0.35355339059327373086 0
11  3.0616169978683830179e-17 0.5 0
12  -0.35355339059327373086 0.35355339059327378637 0
13  -0.5 6.1232339957367660359e-17 0
14  -0.35355339059327384188 -0.35355339059327373086 0
15  -9.1848509936051484375e-17 -0.5 0
16  0.35355339059327367535 -0.35355339059327384188 0
17  -0.22097086912079610954 -0.22097086912079610954 0
18  0.22097086912079610954 -0.22097086912079610954 0
19  -0.22097086912079610954 0.22097086912079610954 0
20  0.22097086912079610954 0.22097086912079610954 0
21  0 -0.22097086912079610954 0
22  -0.22097086912079610954 0 0
23  0.22097086912079610954 0 0
24  0 0.22097086912079610954 0
25  0 0 0
$ENDNOD
$ELM
20
1 3 0 0 4 10 9 1 2
2 3 0 0 4 11 10 2 3
3 3 0 0 4 12 11 3 4
4 3 0 0 4 12 4 5 13
5 3 0 0 4 14 13 5 6
6 3 0 0 4 14 6 7 15
7 3 0 0 4 15 7 8 16
8 3 0 0 4 16 8 1 9
9 3 0 0 4 17 14 15 21
10 3 0 0 4 21 15 16 18
11 3 0 0 4 17 22 13 14
12 3 0 0 4 19 12 13 22
13 3 0 0 4 17 21 25 22
14 3 0 0 4 21 18 23 25
15 3 0 0 4 19 22 25 24
16 3 0 0 4 24 25 23 20
17 3 0 0 4 18 16 9 23
18 3 0 0 4 20 23 9 10
19 3 0 0 4 19 24 11 12
20 3 0 0 4 24 20 10 11
$ENDELM

7 years agoMerge pull request #4515 from jppelteret/fix_tw_solverbase_ad_4443
Bruno Turcksin [Wed, 14 Jun 2017 20:24:52 +0000 (16:24 -0400)]
Merge pull request #4515 from jppelteret/fix_tw_solverbase_ad_4443

Propogate settings set in TrilinosWrappers::Solver<XYZ> to base class.

7 years agoMerge pull request #4516 from drwells/improve-vertex-caching
Martin Kronbichler [Wed, 14 Jun 2017 18:32:11 +0000 (20:32 +0200)]
Merge pull request #4516 from drwells/improve-vertex-caching

Improve vertex caching in a few places.

7 years agoMerge pull request #4517 from masterleinad/fixup_petsc_tests
Jean-Paul Pelteret [Wed, 14 Jun 2017 16:12:16 +0000 (18:12 +0200)]
Merge pull request #4517 from masterleinad/fixup_petsc_tests

Fix up PETSc tests

7 years agoImproved code in Patterns::List
Luca Heltai [Wed, 14 Jun 2017 16:07:03 +0000 (18:07 +0200)]
Improved code in Patterns::List

7 years agoAdded new parameter to Patterns::Map
Luca Heltai [Wed, 14 Jun 2017 15:54:36 +0000 (17:54 +0200)]
Added new parameter to Patterns::Map

7 years agoMerge pull request #4518 from luca-heltai/string-delimiter-in-split-string
Luca Heltai [Wed, 14 Jun 2017 16:09:37 +0000 (18:09 +0200)]
Merge pull request #4518 from luca-heltai/string-delimiter-in-split-string

Added split_string_list with string as separator.

7 years agoImprove vertex caching in a few places. 4516/head
David Wells [Mon, 12 Jun 2017 15:13:04 +0000 (11:13 -0400)]
Improve vertex caching in a few places.

Calling cell->vertex() requires several index lookups before indexing into the
global vertex array so this is relatively expensive. This patch lowers, in a
TransfiniteInterpolationManifold benchmark, the number of vertex lookups from
about 361 million to 229 million and lowers the total wall time by about 5%.

7 years agoAdded changelog. 4518/head
Luca Heltai [Wed, 14 Jun 2017 14:06:17 +0000 (16:06 +0200)]
Added changelog.

7 years agoUtilities::split_string_list take a string instead of char
Luca Heltai [Wed, 14 Jun 2017 13:50:05 +0000 (15:50 +0200)]
Utilities::split_string_list take a string instead of char

7 years agoFix up PETSc tests 4517/head
Daniel Arndt [Wed, 14 Jun 2017 13:40:39 +0000 (15:40 +0200)]
Fix up PETSc tests

7 years agoPropogate settings set in TrilinosWrappers::Solver<XYZ> to base class. 4515/head
Jean-Paul Pelteret [Wed, 14 Jun 2017 13:00:27 +0000 (15:00 +0200)]
Propogate settings set in TrilinosWrappers::Solver<XYZ> to base class.

Also allow direct setting of AdditionalData in
TrilinosWrappers::SolverBase.

Fixes #4443.
Supersedes #4512.

7 years agoMerge pull request #4514 from kronbichler/fix_derivative_form_test
Martin Kronbichler [Wed, 14 Jun 2017 08:04:00 +0000 (10:04 +0200)]
Merge pull request #4514 from kronbichler/fix_derivative_form_test

Fix derivative form test by outputting numbers with spaces

7 years agoFix derivative form test by outputting numbers with spaces 4514/head
Martin Kronbichler [Wed, 14 Jun 2017 06:43:52 +0000 (08:43 +0200)]
Fix derivative form test by outputting numbers with spaces

7 years agoMerge pull request #4513 from masterleinad/fix_parameter_handler_test
Martin Kronbichler [Tue, 13 Jun 2017 19:54:00 +0000 (21:54 +0200)]
Merge pull request #4513 from masterleinad/fix_parameter_handler_test

Fix parameter_handler_double_02

7 years agoMerge pull request #4509 from masterleinad/deprecations_1
Wolfgang Bangerth [Tue, 13 Jun 2017 15:31:59 +0000 (09:31 -0600)]
Merge pull request #4509 from masterleinad/deprecations_1

Remove deprecated member functions in vector classes

7 years agoFix parameter_handler_double_02 4513/head
Daniel Arndt [Tue, 13 Jun 2017 15:05:57 +0000 (17:05 +0200)]
Fix parameter_handler_double_02

7 years agoMerge pull request #4510 from masterleinad/deprecations_2
David Wells [Tue, 13 Jun 2017 12:44:40 +0000 (08:44 -0400)]
Merge pull request #4510 from masterleinad/deprecations_2

Clean up AdditionalData in SolverCG and SolverGMRES

7 years agoMerge pull request #4511 from drwells/inline-line-to-cell-vertices
Martin Kronbichler [Tue, 13 Jun 2017 12:41:58 +0000 (14:41 +0200)]
Merge pull request #4511 from drwells/inline-line-to-cell-vertices

Inline GeometryInfo<dim>::line_to_cell_vertices.

7 years agoInline GeometryInfo<dim>::line_to_cell_vertices. 4511/head
David Wells [Mon, 12 Jun 2017 15:13:37 +0000 (11:13 -0400)]
Inline GeometryInfo<dim>::line_to_cell_vertices.

This provides about a 5% speed boost in TransfiniteInterpolationManifold since
the compiler can rewrite the loop over the line vertices.

7 years agoMerge pull request #4504 from bangerth/save-flags
Daniel Arndt [Tue, 13 Jun 2017 10:40:46 +0000 (12:40 +0200)]
Merge pull request #4504 from bangerth/save-flags

Save the state of the stream we're going to write into.

7 years agoCleanup AdditionalData in SolverCG and SolverGMRES 4510/head
Daniel Arndt [Sat, 10 Jun 2017 09:48:35 +0000 (11:48 +0200)]
Cleanup AdditionalData in SolverCG and SolverGMRES

7 years agoMerge pull request #4506 from kronbichler/fix_block_relaxation
Daniel Arndt [Mon, 12 Jun 2017 22:24:14 +0000 (00:24 +0200)]
Merge pull request #4506 from kronbichler/fix_block_relaxation

Implement vmult() method in RelaxationBlock{Jacobi,SOR,SSOR}

7 years agoRemove more deprecated member methods in vector classes 4509/head
Daniel Arndt [Mon, 12 Jun 2017 20:33:06 +0000 (22:33 +0200)]
Remove more deprecated member methods in vector classes

7 years agoRemove deprecated member functions in vector classes
Daniel Arndt [Wed, 7 Jun 2017 15:53:11 +0000 (17:53 +0200)]
Remove deprecated member functions in vector classes

7 years agoMerge pull request #4508 from luca-heltai/unique-ptr-parameter-handler
Wolfgang Bangerth [Mon, 12 Jun 2017 13:40:14 +0000 (07:40 -0600)]
Merge pull request #4508 from luca-heltai/unique-ptr-parameter-handler

Changed clone and create in Patterns to return std::unique_ptr

7 years agoMake clone and crete return std::unique_ptr in Patterns. 4508/head
Luca Heltai [Mon, 12 Jun 2017 10:48:33 +0000 (12:48 +0200)]
Make clone and crete return std::unique_ptr in Patterns.

7 years agoMerge pull request #4507 from luca-heltai/double-to-number
Luca Heltai [Mon, 12 Jun 2017 08:44:17 +0000 (10:44 +0200)]
Merge pull request #4507 from luca-heltai/double-to-number

Fixed DerivativeForm.

7 years agoFixed Determinant in DerivativeForm. 4507/head
Luca Heltai [Mon, 12 Jun 2017 08:08:00 +0000 (10:08 +0200)]
Fixed Determinant in DerivativeForm.

7 years agoImplement vmult() method in RelaxationBlock{Jacobi,SOR,SSOR} 4506/head
Martin Kronbichler [Sun, 11 Jun 2017 20:53:48 +0000 (22:53 +0200)]
Implement vmult() method in RelaxationBlock{Jacobi,SOR,SSOR}

7 years agoMerge pull request #4494 from kronbichler/multigrid_performance_improvement
Martin Kronbichler [Sat, 10 Jun 2017 20:32:23 +0000 (22:32 +0200)]
Merge pull request #4494 from kronbichler/multigrid_performance_improvement

Use shortcut in multigrid smoothing in case we want to zero out vector

7 years agoSimplify step() in PreconditionChebyshev. 4494/head
Martin Kronbichler [Sat, 10 Jun 2017 10:25:11 +0000 (12:25 +0200)]
Simplify step() in PreconditionChebyshev.

7 years agoAdd changelog. 4505/head
Wolfgang Bangerth [Sat, 10 Jun 2017 00:59:58 +0000 (18:59 -0600)]
Add changelog.

7 years agoUse only enum values, don't look at bits.
Wolfgang Bangerth [Sat, 10 Jun 2017 00:57:31 +0000 (18:57 -0600)]
Use only enum values, don't look at bits.

This fixes an (undocumented) oddity in ParameterHandler::print_parameters(): We
were at times looking at individual bits instead of just the declared values
of ParameterHandler::OutputStyle. This presumably allowed for calling that
function with a combination of the OutputStyle flags, for reasons that no
longer seem particularly relevant nor obvious.

This patch removes this possibility from the current implementation of
the function, but retains it for the (deprecated) function
ParameterHandler::print_parameters_section().

7 years agoAdd changelog entry. 4504/head
Wolfgang Bangerth [Sat, 10 Jun 2017 00:28:59 +0000 (18:28 -0600)]
Add changelog entry.

7 years agoAdd test.
Wolfgang Bangerth [Sat, 10 Jun 2017 00:28:45 +0000 (18:28 -0600)]
Add test.

7 years agoSave the state of the stream we're going to write into.
Wolfgang Bangerth [Sat, 10 Jun 2017 00:28:33 +0000 (18:28 -0600)]
Save the state of the stream we're going to write into.

7 years agoMerge pull request #4495 from bangerth/const-prm-print-parameters
David Wells [Fri, 9 Jun 2017 19:31:05 +0000 (15:31 -0400)]
Merge pull request #4495 from bangerth/const-prm-print-parameters

Make ParameterHandler::print_parameters() const.

7 years agoAdd changelog
Martin Kronbichler [Fri, 9 Jun 2017 16:23:08 +0000 (18:23 +0200)]
Add changelog

7 years agoAugment documentation
Martin Kronbichler [Fri, 9 Jun 2017 06:19:01 +0000 (08:19 +0200)]
Augment documentation

7 years agoUse shortcut in multigrid smoothing in case we want to zero out vector
Martin Kronbichler [Tue, 6 Jun 2017 21:50:38 +0000 (23:50 +0200)]
Use shortcut in multigrid smoothing in case we want to zero out vector

7 years agoMerge pull request #4502 from tjhei/document_cell_id
Martin Kronbichler [Fri, 9 Jun 2017 10:58:38 +0000 (12:58 +0200)]
Merge pull request #4502 from tjhei/document_cell_id

document CellId better


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.