]>
https://gitweb.dealii.org/ - dealii.git/log
Martin Kronbichler [Tue, 17 Aug 2021 10:01:44 +0000 (12:01 +0200)]
Merge pull request #12612 from drwells/fix-mass-diagonal
Make MassOperator's diagonals true diagonals, not lumped ones.
Wolfgang Bangerth [Tue, 17 Aug 2021 03:05:15 +0000 (21:05 -0600)]
Merge pull request #12596 from marcfehling/fix-load
Load: align dealii and p4est meshes before repartitioning.
David Wells [Sat, 14 Aug 2021 21:39:02 +0000 (17:39 -0400)]
Explicitly use mass lumping in VectorTools::project_matrix_free().
This restores the behavior present before the last commit and its generally the
right choice.
David Wells [Sat, 31 Jul 2021 16:14:17 +0000 (12:14 -0400)]
Implement explicit mass lumping in MassOperator.
David Wells [Sat, 31 Jul 2021 14:31:56 +0000 (10:31 -0400)]
Modify a test to check the diagonal values (not lumped).
David Wells [Thu, 29 Jul 2021 21:40:18 +0000 (17:40 -0400)]
Make MassOperator's diagonals true diagonals, not lumped ones.
The previous implementation used row sums (i.e., mass lumping) instead of
computing the diagonal. This should be changed for two reasons:
1. It's not really the diagonal (and its inconsistent with LaplaceOperator).
2. Lumped mass matrices are only positive for either low-order elements (where
the basis functions are all positive) or very well-behaved elements (like
FE_Q). In particular, the values on the diagonal are either zero or negative
for TET10 if we use lumping, which isn't going to work.
Jean-Paul Pelteret [Mon, 16 Aug 2021 19:11:29 +0000 (21:11 +0200)]
Merge pull request #12658 from drwells/fix-issue-template
Make it an issue template.
Daniel Arndt [Mon, 16 Aug 2021 14:33:57 +0000 (10:33 -0400)]
Merge pull request #12656 from bangerth/increment
Peter Munch [Mon, 16 Aug 2021 14:33:11 +0000 (16:33 +0200)]
Merge pull request #12641 from peterrum/dataout_nonconti
Add assert to DataOut to force users to use block vectors for non-contiguous indexsets
Jean-Paul Pelteret [Mon, 16 Aug 2021 13:21:53 +0000 (15:21 +0200)]
Merge pull request #12587 from bangerth/explicit
Be more explicit about types and taking addresses.
Jean-Paul Pelteret [Mon, 16 Aug 2021 13:19:49 +0000 (15:19 +0200)]
Merge pull request #12622 from bangerth/doc
Add another example to the ScalarFunctionFromFunctionObject documentation.
David Wells [Mon, 16 Aug 2021 11:40:27 +0000 (07:40 -0400)]
Merge pull request #12657 from bangerth/types
Use std::size_t for index types.
David Wells [Mon, 16 Aug 2021 11:34:37 +0000 (07:34 -0400)]
Make it an issue template.
My bad, this doesn't make sense as a PR template.
David Wells [Mon, 16 Aug 2021 01:39:39 +0000 (21:39 -0400)]
Merge pull request #12655 from bangerth/doc-2
Improve error message.
Wolfgang Bangerth [Sun, 15 Aug 2021 23:47:04 +0000 (17:47 -0600)]
Use std::size_t for index types.
Wolfgang Bangerth [Sun, 15 Aug 2021 23:37:08 +0000 (17:37 -0600)]
Use 'auto' in one place.
Wolfgang Bangerth [Sun, 15 Aug 2021 23:12:46 +0000 (17:12 -0600)]
Change post- to pre-increment statements in foor loops (tests/).
Wolfgang Bangerth [Sun, 15 Aug 2021 23:12:40 +0000 (17:12 -0600)]
Change post- to pre-increment statements in foor loops (examples/).
Wolfgang Bangerth [Sun, 15 Aug 2021 23:12:34 +0000 (17:12 -0600)]
Change post- to pre-increment statements in foor loops (include/ and source/).
Wolfgang Bangerth [Sun, 15 Aug 2021 22:52:00 +0000 (16:52 -0600)]
Improve error message.
Wolfgang Bangerth [Sun, 15 Aug 2021 22:18:41 +0000 (16:18 -0600)]
Merge pull request #12643 from zjiaqi2018/changelog_feiv_AD
Changelog: FEInterfaceValues for AD
Wolfgang Bangerth [Sun, 15 Aug 2021 22:17:31 +0000 (16:17 -0600)]
Merge pull request #12652 from drwells/little-doc-fixes
Fix two small documentation issues.
Wolfgang Bangerth [Sun, 15 Aug 2021 22:15:46 +0000 (16:15 -0600)]
Merge pull request #12654 from drwells/add-pull-request-template
Add a pull request template.
David Wells [Sun, 15 Aug 2021 19:16:55 +0000 (15:16 -0400)]
Add a pull request template.
David Wells [Sun, 15 Aug 2021 17:11:06 +0000 (13:11 -0400)]
Fix two small documentation issues.
Its not worth opening two PRs.
Timo Heister [Sat, 14 Aug 2021 20:34:24 +0000 (16:34 -0400)]
Merge pull request #12644 from jppelteret/fe_interface_values_04
Categorise functions in FEInterfaceValuesViews
Jiaqi Zhang [Sat, 14 Aug 2021 20:32:09 +0000 (20:32 +0000)]
changelog for feiv AD
Jean-Paul Pelteret [Sat, 14 Aug 2021 08:43:14 +0000 (10:43 +0200)]
Reorganise sections and order
Peter Munch [Sat, 14 Aug 2021 07:29:40 +0000 (09:29 +0200)]
Merge pull request #12650 from konsim83/documentation/NewClangFormat
Fix clang-format version to 11.1 in coding guidelines
Konrad Simon [Fri, 13 Aug 2021 18:53:02 +0000 (20:53 +0200)]
fix clang-format version to 11.1
David Wells [Fri, 13 Aug 2021 11:57:44 +0000 (07:57 -0400)]
Merge pull request #12648 from peterrum/fe_wedge_pyramid_degree
Add comments to wedge and pyramid FEs regarding degree
Peter Munch [Fri, 13 Aug 2021 05:33:24 +0000 (07:33 +0200)]
Merge pull request #12645 from tjhei/mg_transfer_warn
add Assert to MGTransfer*
Peter Munch [Thu, 12 Aug 2021 07:43:22 +0000 (09:43 +0200)]
Add comments to wedge and pyramid FEs regarding degree
Peter Munch [Thu, 12 Aug 2021 17:16:21 +0000 (19:16 +0200)]
Merge pull request #12649 from drwells/remove-old-namespace-comment
Remove documentation of the old Simplex namespace.
Timo Heister [Wed, 11 Aug 2021 20:22:21 +0000 (16:22 -0400)]
add Assert to MGTransfer*
The error message when you forget to call distribute_mg_dofs() is not
helpful. Instead, add an Assert.
David Wells [Thu, 12 Aug 2021 14:32:13 +0000 (10:32 -0400)]
Remove documentation of the old Simplex namespace.
Jean-Paul Pelteret [Wed, 11 Aug 2021 18:35:38 +0000 (20:35 +0200)]
Categorise functions in FEInterfaceValuesViews
Daniel Arndt [Wed, 11 Aug 2021 13:58:15 +0000 (09:58 -0400)]
Merge pull request #12642 from peterrum/fe_eval_pre
Replace some post-increments by pre-increments in fe_evaluation.h
Peter Munch [Wed, 11 Aug 2021 06:33:26 +0000 (08:33 +0200)]
Merge pull request #12638 from gfcas/integrate_difference
Integrate difference
Peter Munch [Wed, 11 Aug 2021 06:27:58 +0000 (08:27 +0200)]
Replace some post-increments by pre-increments in fe_evaluation.h
Jean-Paul Pelteret [Wed, 11 Aug 2021 04:47:12 +0000 (06:47 +0200)]
Merge pull request #12603 from zjiaqi2018/feinterface_AD
FEInterfaceValues for AD
Jean-Paul Pelteret [Wed, 11 Aug 2021 03:05:10 +0000 (05:05 +0200)]
Merge pull request #12636 from gfcas/step_40_timer
step-40 disable timer summary output
Jean-Paul Pelteret [Wed, 11 Aug 2021 02:51:15 +0000 (04:51 +0200)]
Improve some assertions
Peter Munch [Tue, 10 Aug 2021 19:18:02 +0000 (21:18 +0200)]
Add assert to DataOut to force users to use block vectors for non-contiguous indexsets
Peter Munch [Tue, 10 Aug 2021 18:13:48 +0000 (20:13 +0200)]
Merge pull request #12639 from gfcas/vector_tools_interpolate
Replace AssertDimension
Peter Munch [Tue, 10 Aug 2021 13:54:21 +0000 (15:54 +0200)]
Merge pull request #12625 from bergbauer/integrate_hessians_on_cells
Integrate hessians on cells
Daniel Arndt [Tue, 10 Aug 2021 13:51:57 +0000 (09:51 -0400)]
Merge pull request #12637 from gfcas/typos
Fabian Castelli [Tue, 10 Aug 2021 13:41:45 +0000 (15:41 +0200)]
Remove unused n_q_points variable
Fabian Castelli [Sun, 8 Aug 2021 09:12:46 +0000 (11:12 +0200)]
Range based quadrature loop
Martin Kronbichler [Tue, 10 Aug 2021 12:48:17 +0000 (14:48 +0200)]
Merge pull request #12635 from peterrum/gc_prolongate_and_add
Override MGTransferGlobalCoarsening::prolongate_and_add()
Fabian Castelli [Tue, 10 Aug 2021 12:43:22 +0000 (14:43 +0200)]
Update include/deal.II/numerics/vector_tools_interpolate.templates.h
Co-authored-by: David Wells <drwells@email.unc.edu>
Fabian Castelli [Sun, 8 Aug 2021 17:13:18 +0000 (19:13 +0200)]
Replace AssertDimension
Fabian Castelli [Sun, 8 Aug 2021 07:31:37 +0000 (09:31 +0200)]
Use AssertDimension macro
Fabian Castelli [Sun, 8 Aug 2021 07:18:51 +0000 (09:18 +0200)]
Mapping argument for do_integrate_difference
Fabian Castelli [Sun, 8 Aug 2021 17:45:20 +0000 (19:45 +0200)]
Another typo
Fabian Castelli [Sat, 7 Aug 2021 17:10:07 +0000 (19:10 +0200)]
Comma spacing
Fabian Castelli [Sat, 7 Aug 2021 06:36:14 +0000 (08:36 +0200)]
step-40 disable timer summary output
Maximilian Bergbauer [Thu, 5 Aug 2021 15:40:53 +0000 (17:40 +0200)]
Integrate hessians on cells
Peter Munch [Tue, 10 Aug 2021 09:14:18 +0000 (11:14 +0200)]
Override MGTransferGlobalCoarsening::prolongate_and_add()
Martin Kronbichler [Tue, 10 Aug 2021 09:11:12 +0000 (11:11 +0200)]
Merge pull request #12632 from peterrum/step-47
Step-47: unify code
Martin Kronbichler [Tue, 10 Aug 2021 09:10:31 +0000 (11:10 +0200)]
Merge pull request #12593 from bergbauer/evaluate_and_integrate_hessians
Evaluate and integrate hessians
Peter Munch [Sat, 7 Aug 2021 18:47:42 +0000 (20:47 +0200)]
Step-47: unify code
Jiaqi Zhang [Mon, 9 Aug 2021 19:47:40 +0000 (19:47 +0000)]
add get_function_*
Maximilian Bergbauer [Thu, 17 Jun 2021 11:50:11 +0000 (13:50 +0200)]
Evaluate and integrate hessians on faces
Jean-Paul Pelteret [Sun, 8 Aug 2021 05:21:39 +0000 (07:21 +0200)]
Merge pull request #12631 from peterrum/feeval_doc
Small modifications of FEEvaluation documentation
Peter Munch [Sat, 7 Aug 2021 18:40:52 +0000 (20:40 +0200)]
Small modifications of FEEvaluation documentation
Jean-Paul Pelteret [Sat, 7 Aug 2021 10:04:51 +0000 (12:04 +0200)]
Merge pull request #12630 from jppelteret/fe_values_functions_01
Add face and cell queries to FEFaceValuesBase and FEInterfaceValues
Peter Munch [Sat, 7 Aug 2021 07:14:52 +0000 (09:14 +0200)]
Merge pull request #12614 from marcfehling/hp-indexset
Standardized output for many `mpi/hp_unify_dof_indices_X` tests.
Peter Munch [Sat, 7 Aug 2021 07:09:13 +0000 (09:09 +0200)]
Merge pull request #12615 from peterrum/rpe_test
Extend test
Peter Munch [Sat, 7 Aug 2021 07:08:47 +0000 (09:08 +0200)]
Merge pull request #12616 from peterrum/ConstraintTypes
Switch from unsigned int to ConstraintTypes
Peter Munch [Sat, 7 Aug 2021 07:02:23 +0000 (09:02 +0200)]
Merge pull request #12627 from masterleinad/update_cuda_output
Update test output for CUDA
Jean-Paul Pelteret [Fri, 6 Aug 2021 18:22:03 +0000 (20:22 +0200)]
Add a changelog entry
Jean-Paul Pelteret [Fri, 6 Aug 2021 18:01:29 +0000 (20:01 +0200)]
Update a test
Jean-Paul Pelteret [Fri, 6 Aug 2021 17:07:36 +0000 (19:07 +0200)]
Add FEInterfaceValues::get_cell() and FEInterfaceValues::get_face_number()
Jean-Paul Pelteret [Fri, 6 Aug 2021 16:57:46 +0000 (18:57 +0200)]
Add FEFaceValuesBase::get_face_number()
Jean-Paul Pelteret [Fri, 6 Aug 2021 18:25:22 +0000 (20:25 +0200)]
Merge pull request #12628 from jppelteret/fe_interface_values_03
Add dof_indices() to FEInterfaceValues
Peter Munch [Fri, 6 Aug 2021 17:31:33 +0000 (19:31 +0200)]
Merge pull request #12629 from peterrum/signals_residual_step_fix
Fix signal residual_step
Peter Munch [Fri, 6 Aug 2021 15:28:44 +0000 (17:28 +0200)]
Fix signal residual_step
Jean-Paul Pelteret [Thu, 5 Aug 2021 17:57:34 +0000 (19:57 +0200)]
Add dof_indices() to FEInterfaceValues
Daniel Arndt [Thu, 5 Aug 2021 18:05:34 +0000 (18:05 +0000)]
Update test putput for CUDA
Timo Heister [Thu, 5 Aug 2021 16:32:00 +0000 (12:32 -0400)]
Merge pull request #12411 from jppelteret/fe_interface_values_02
FEInterfaceValues: Make class interfaces consistent with FEValues and its views
David Wells [Thu, 5 Aug 2021 14:45:09 +0000 (10:45 -0400)]
Merge pull request #12623 from marcfehling/identity-constraints
Added diagnostic functions for `AffineConstraints`.
Marc Fehling [Thu, 5 Aug 2021 04:06:33 +0000 (22:06 -0600)]
Merge pull request #12618 from peterrum/TriangulationDescription_examples
Extend documentation of TriangulationDescription
Marc Fehling [Wed, 4 Aug 2021 23:42:05 +0000 (17:42 -0600)]
Added tests for AffineConstraints diagnostics.
Marc Fehling [Wed, 4 Aug 2021 23:18:09 +0000 (17:18 -0600)]
Added diagnostic functions for AffineConstraints.
Wolfgang Bangerth [Wed, 4 Aug 2021 23:03:00 +0000 (17:03 -0600)]
Add another example to the ScalarFunctionFromFunctionObject documentation.
Jean-Paul Pelteret [Sun, 1 Aug 2021 08:01:49 +0000 (10:01 +0200)]
Add a changelog entry
Jean-Paul Pelteret [Sat, 31 Jul 2021 10:46:39 +0000 (12:46 +0200)]
Update tests
Jean-Paul Pelteret [Sat, 31 Jul 2021 10:46:10 +0000 (12:46 +0200)]
Update examples
Peter Munch [Sun, 1 Aug 2021 06:09:15 +0000 (08:09 +0200)]
Extend documention of TriangulationDescription
Peter Munch [Tue, 3 Aug 2021 19:10:01 +0000 (21:10 +0200)]
Continue
Peter Munch [Tue, 3 Aug 2021 18:36:18 +0000 (20:36 +0200)]
Work on GPU
Peter Munch [Tue, 3 Aug 2021 08:12:26 +0000 (10:12 +0200)]
Rename
Peter Munch [Tue, 3 Aug 2021 07:53:31 +0000 (09:53 +0200)]
Introduce enum class
Martin Kronbichler [Mon, 2 Aug 2021 19:56:02 +0000 (21:56 +0200)]
Merge pull request #12610 from drwells/add-more-wv-rules
Add more Witherden-Vincent rules
David Wells [Mon, 2 Aug 2021 17:07:39 +0000 (13:07 -0400)]
Merge pull request #12619 from drwells/fix-parenthesis
Fix a parenthesis.
David Wells [Mon, 2 Aug 2021 11:59:07 +0000 (07:59 -0400)]
Fix a parenthesis.
I see a bunch of warnings of the form
/Users/runner/work/dealii/dealii/include/deal.II/matrix_free/tools.h:497:35:
error: taking the absolute value of unsigned type 'bool' has no effect [-Werror,-Wabsolute-value]
854
std::abs(values_dofs[j][0] &&
on the CI tester.
David Wells [Thu, 29 Jul 2021 16:21:54 +0000 (12:21 -0400)]
Add a lot more Witherden-Vincent rules.
This includes most of the even-degree rules.
For posterity: here's the script I used to print quadpy rules rather than
copying numbers individually:
import quadpy as qp
def format_16(x):
return '{:.16e}'.format(float(x))
def print_scheme(scheme):
sym_data = scheme.symmetry_data
print("// WV-{}, {}D".format(scheme.degree, scheme.dim))
for key in sym_data.keys():
if key == 'centroid' or key == 's4':
print("b_point_permutations.push_back({centroid});")
print("b_weights.push_back({});".format(format_16(sym_data[key][0][0])))
elif key == 'd3_aa' or key == 's31':
weights = [format_16(w) for w in sym_data[key][0]]
points = [format_16(p) for p in sym_data[key][1]]
assert len(weights) == len(points)
for i in range(len(weights)):
print("process_point_1({}, {});".format(points[i], weights[i]))
elif key == 'd3_ab' or key == 's211':
weights = [format_16(w) for w in sym_data[key][0]]
points1 = [format_16(p) for p in sym_data[key][1]]
points2 = [format_16(p) for p in sym_data[key][2]]
assert len(weights) == len(points1)
assert len(points2) == len(points1)
for i in range(len(weights)):
print(("process_point_3({},"
"\n {},"
"\n {});")
.format(points1[i], points2[i], weights[i]))
elif key == 's22':
assert scheme.dim == 3
weights = [format_16(w) for w in sym_data[key][0]]
points = [format_16(p) for p in sym_data[key][1]]
assert len(weights) == len(points)
for i in range(len(weights)):
print(("process_point_2({},"
"\n {});")
.format(points[i], weights[i]))
else:
assert False
Martin Kronbichler [Sun, 1 Aug 2021 11:40:40 +0000 (13:40 +0200)]
Merge pull request #12613 from drwells/fix-structdim-comments
Fix two comments in TriaAccessor.
Marc Fehling [Sat, 31 Jul 2021 21:06:00 +0000 (15:06 -0600)]
Merge pull request #12559 from tamiko/modernize_step_69
Step-69: Use SolutionTransfer instead of low-level boost archive
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.