0430df6 | Christophe Prud'homme | 27 January 2014, 14:11:26 UTC | up comments | 27 January 2014, 14:11:26 UTC |
055e118 | Christophe Prud'homme | 06 January 2014, 08:23:38 UTC | compilation fixes | 06 January 2014, 08:23:38 UTC |
387d176 | Christophe Prud'homme | 05 January 2014, 21:11:04 UTC | update examples | 05 January 2014, 21:11:04 UTC |
f6f082c | Christophe Prud'homme | 05 January 2014, 21:10:44 UTC | bump up to v0.97.0-alpha.1 | 05 January 2014, 21:10:44 UTC |
b4bc078 | Christophe Prud'homme | 05 January 2014, 21:09:22 UTC | split file and fix measure() (#226) | 05 January 2014, 21:09:22 UTC |
ca85b78 | Christophe Prud'homme | 05 January 2014, 21:08:55 UTC | allow geo parameters in write mode too | 05 January 2014, 21:08:55 UTC |
fabb71f | Christophe Prud'homme | 05 January 2014, 21:08:30 UTC | start working on #243 | 05 January 2014, 21:08:30 UTC |
3b1fb36 | Christophe Prud'homme | 05 January 2014, 08:38:13 UTC | create own header file for variant of integrate, see #226 | 05 January 2014, 08:38:13 UTC |
5bae3ee | Christophe Prud'homme | 04 January 2014, 13:52:58 UTC | use rank_type see #224 (@vincentchabannes) | 04 January 2014, 13:52:58 UTC |
82f6ace | Christophe Prud'homme | 04 January 2014, 13:51:53 UTC | Merge remote-tracking branch 'origin/develop' into feature/header Conflicts: feel/feeldiscr/createsubmesh.hpp | 04 January 2014, 13:51:53 UTC |
99cf3cd | Vincent Chabannes | 04 January 2014, 08:08:09 UTC | up createsubmesh : - use the point numeration of parent mesh (by default) but renunumber is possible with adding MESH_RENUMBER - add boolean option subMeshIsOnBoundaryFaces(true by default) : reduce comm for submesh<MESH_FACES> and work only if the range is define only on boundary | 04 January 2014, 08:08:09 UTC |
3ad4f06 | Vincent Chabannes | 04 January 2014, 07:48:51 UTC | fix integration on interprocess faces (only one side) | 04 January 2014, 07:48:51 UTC |
b7aceab | Vincent Chabannes | 04 January 2014, 07:47:17 UTC | move also some ghost elts if the displacement space contains this info | 04 January 2014, 07:47:17 UTC |
b42d6a3 | Vincent Chabannes | 04 January 2014, 07:45:38 UTC | add boost parameter | 04 January 2014, 07:45:38 UTC |
d21e489 | Vincent Chabannes | 04 January 2014, 07:45:09 UTC | Support extended stencil over a range of elements up #227 | 04 January 2014, 07:45:09 UTC |
5bf2a15 | Christophe Prud'homme | 03 January 2014, 11:18:56 UTC | fix compilation (bdf2->bdf + other minor changes) see #226 | 03 January 2014, 11:18:56 UTC |
41962fa | Christophe Prud'homme | 03 January 2014, 11:18:26 UTC | add dashbord script for atlas | 03 January 2014, 11:18:26 UTC |
f566b19 | Christophe Prud'homme | 03 January 2014, 11:07:04 UTC | compilation fixes (see #226) | 03 January 2014, 11:07:04 UTC |
63653f5 | Christophe Prud'homme | 03 January 2014, 10:38:23 UTC | use new Blocks interface it compiles now, it needs however to be tested | 03 January 2014, 10:38:23 UTC |
7cd91b6 | Christophe Prud'homme | 03 January 2014, 09:26:17 UTC | more compilation fixes (see #226) | 03 January 2014, 09:26:17 UTC |
29d6dc4 | Christophe Prud'homme | 03 January 2014, 09:18:28 UTC | various compilation fixes (see #226, and #88) | 03 January 2014, 09:18:28 UTC |
67e0f65 | Christophe Prud'homme | 03 January 2014, 09:09:29 UTC | Merge branch 'release/v0.96.0' into develop | 03 January 2014, 09:09:29 UTC |
d66b938 | Christophe Prud'homme | 03 January 2014, 08:58:37 UTC | various compilation fixes (see #226 and #88) | 03 January 2014, 08:58:37 UTC |
8253c57 | Christophe Prud'homme | 03 January 2014, 08:58:12 UTC | fix benchmark tilted from bit paper | 03 January 2014, 08:58:12 UTC |
91c29ba | Christophe Prud'homme | 03 January 2014, 08:30:35 UTC | fix CrouzeixRaviart interface | 03 January 2014, 08:30:35 UTC |
2872dfa | Christophe Prud'homme | 03 January 2014, 08:09:03 UTC | remove LABEL testsuite (will reduce the number of tests by half and failures by half) see #88 | 03 January 2014, 08:09:03 UTC |
e73af98 | Christophe Prud'homme | 02 January 2014, 14:03:52 UTC | fix execution in // | 02 January 2014, 14:03:52 UTC |
f0c9b5d | Christophe Prud'homme | 02 January 2014, 13:52:19 UTC | update #define (#226) | 02 January 2014, 13:52:19 UTC |
6c1dc59 | Christophe Prud'homme | 02 January 2014, 13:50:29 UTC | update(simpler and use #226) and fix execution | 02 January 2014, 13:50:29 UTC |
f54b135 | Christophe Prud'homme | 02 January 2014, 12:55:33 UTC | update | 02 January 2014, 12:55:33 UTC |
569a594 | Christophe Prud'homme | 02 January 2014, 12:55:20 UTC | fix execution of sound | 02 January 2014, 12:55:20 UTC |
05487ac | Christophe Prud'homme | 02 January 2014, 10:05:08 UTC | fix compilation (update issue #226) | 02 January 2014, 10:05:08 UTC |
75eef03 | Christophe Prud'homme | 02 January 2014, 09:02:47 UTC | fix compilation (update issue #226) | 02 January 2014, 09:02:47 UTC |
5af6e7a | Christophe Prud'homme | 30 December 2013, 21:42:22 UTC | update issue #237 - introduce bdf->priorTimes() to initialize bdf, see testsuite/feeldiscr/test_bdf2 for an example - change behavior of order change, use directly bdf order but requires initialization of priorTimes() | 30 December 2013, 21:42:22 UTC |
4af02c9 | Christophe Prud'homme | 30 December 2013, 21:40:01 UTC | fix indentation | 30 December 2013, 21:40:01 UTC |
0fc9c28 | Christophe Prud'homme | 30 December 2013, 21:39:33 UTC | @StephaneVeys update test to verify bdf implementation | 30 December 2013, 21:39:33 UTC |
e03d14b | Christophe Prud'homme | 28 December 2013, 15:55:36 UTC | a+=m was too much added more precision | 28 December 2013, 15:55:36 UTC |
01a3c96 | Christophe Prud'homme | 28 December 2013, 15:44:34 UTC | @StephaneVeys updated test_bdf2 (arguably a simpler version) | 28 December 2013, 15:44:34 UTC |
77138cc | Christophe Prud'homme | 28 December 2013, 15:44:06 UTC | call start to update bdf coefficients (update issue #237) | 28 December 2013, 15:44:06 UTC |
aff793a | Christophe Prud'homme | 28 December 2013, 15:43:05 UTC | update issue #226: add domain.hpp | 28 December 2013, 15:43:05 UTC |
1521d5d | Christophe Prud'homme | 28 December 2013, 14:13:27 UTC | Merge remote-tracking branch 'origin/develop' into feature/header Conflicts: feel/feeldiscr/functionspace.hpp feel/feeldiscr/projector.hpp | 28 December 2013, 14:13:27 UTC |
2ed3d60 | Christophe Prud'homme | 28 December 2013, 14:10:28 UTC | update issue #226: work on tutorial to reduce compilation time | 28 December 2013, 14:10:28 UTC |
f04f590 | Christophe Prud'homme | 28 December 2013, 14:09:28 UTC | fix compilation update issue #226 | 28 December 2013, 14:09:28 UTC |
cdec628 | Stephane Veys | 28 December 2013, 11:27:08 UTC | add a test using bdf | 28 December 2013, 11:27:08 UTC |
2782e72 | Vincent Chabannes | 27 December 2013, 09:58:43 UTC | fix parallelism | 27 December 2013, 09:58:43 UTC |
0e16c7a | Christophe Prud'homme | 27 December 2013, 08:23:13 UTC | update issue #226 Merge branch 'feature/header' of https://github.com/feelpp/feelpp into feature/header # Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic branch. # # Lines starting with '#' will be ignored, and an empty message aborts # the commit. | 27 December 2013, 08:23:13 UTC |
6422cc1 | Christophe Prud'homme | 27 December 2013, 08:19:54 UTC | rename header for operations(*,+,/,-) in variational formulations update issue #226 | 27 December 2013, 08:19:54 UTC |
4e361fd | Stephane Veys | 26 December 2013, 19:15:47 UTC | fix parameter broadcast when sampling WNmu is randomly choosen. However it seems there is a convergence problem with transient problems... | 26 December 2013, 19:15:47 UTC |
7ffe753 | Vincent Chabannes | 26 December 2013, 17:44:06 UTC | fix composite space with extended doftable | 26 December 2013, 17:44:06 UTC |
2aaab44 | Vincent | 26 December 2013, 11:12:17 UTC | fix clang compilation | 26 December 2013, 11:12:17 UTC |
f21c17e | Christophe Prud'homme | 26 December 2013, 10:58:27 UTC | Merge branch 'develop' into feature/header Conflicts: feel/feeldiscr/functionspace.hpp | 26 December 2013, 10:58:27 UTC |
12b4e2d | Christophe Prud'homme | 26 December 2013, 10:47:37 UTC | move BDF coefficients to bdf class and cleanup use TSBase as base class update issue #237 | 26 December 2013, 10:47:37 UTC |
1b93029 | Christophe Prud'homme | 26 December 2013, 10:38:41 UTC | start working on reorganizing headers and splitting them (update issue #226) - create one file per public feature - split gmsh.hpp, functionspace.hpp - move options to feel/options.cpp - remove some .cpp files that were dealing with program_options | 26 December 2013, 10:38:41 UTC |
5c821ab | Vincent Chabannes | 24 December 2013, 13:49:41 UTC | up rebuildDofPoints with extended doftable | 24 December 2013, 13:49:41 UTC |
b7e9593 | Vincent Chabannes | 24 December 2013, 13:38:46 UTC | use Pdh instead of Odh and use extended dof table in // | 24 December 2013, 13:38:46 UTC |
3b1e0a6 | Vincent Chabannes | 24 December 2013, 13:36:13 UTC | reactivate integrate on interprocess faces | 24 December 2013, 13:36:13 UTC |
71a6218 | Vincent Chabannes | 24 December 2013, 13:34:17 UTC | fix projector : Pattern::EXTENDED is uselessgit diff projector.hpp and also the default backend ( Backend<double>::build( BACKEND_PETSC ) is obselete I think ) | 24 December 2013, 13:34:17 UTC |
3f9a72e | Vincent Chabannes | 24 December 2013, 13:30:11 UTC | accept extended stencil with ghost elt | 24 December 2013, 13:30:11 UTC |
6f11b2d | Vincent Chabannes | 24 December 2013, 13:29:04 UTC | add boost parameter | 24 December 2013, 13:29:04 UTC |
0011e12 | Vincent Chabannes | 24 December 2013, 13:19:09 UTC | add parameters in functionspace to take into account the extended doftable | 24 December 2013, 13:19:09 UTC |
bdb36e2 | Vincent Chabannes | 24 December 2013, 13:18:08 UTC | possibility to extend the // doftable on some ghost element (which touch the interprocess faces) | 24 December 2013, 13:18:08 UTC |
57b0dd9 | Christophe Prud'homme | 23 December 2013, 12:31:57 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 23 December 2013, 12:31:57 UTC |
407b701 | Christophe Prud'homme | 23 December 2013, 12:31:40 UTC | create TSBase a base class for time stepping alorithms update issue #237 | 23 December 2013, 12:31:40 UTC |
c5d86a2 | Christophe Prud'homme | 23 December 2013, 12:28:03 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 23 December 2013, 12:28:03 UTC |
dbadc54 | Christophe Prud'homme | 21 December 2013, 06:55:25 UTC | @vincentchabannes added rank_type for rank number. make sure that it is used consistently in the code to hold the process rank of geometric entities it is stored in a uint16_type (2^[16]-1) fix issue #224 | 23 December 2013, 12:27:30 UTC |
d272912 | Alexandre Ancel | 20 December 2013, 16:25:41 UTC | CRB Added kernel init code in crb.hpp + bugfixes Start working on OpenCL kernel | 20 December 2013, 16:25:41 UTC |
334d981 | Stephane Veys | 20 December 2013, 08:19:17 UTC | use crb.dimension-max to convergence study and all proc must have the same sampling (test part) if we want to compute FEM solutions | 20 December 2013, 08:19:17 UTC |
8682692 | Alexandre Ancel | 18 December 2013, 17:09:48 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 18 December 2013, 17:09:48 UTC |
63cfc56 | Alexandre Ancel | 18 December 2013, 17:09:09 UTC | Fixed OpenCL code not meant to be used outside Harts context | 18 December 2013, 17:09:09 UTC |
71e0c46 | Stephane Priem | 18 December 2013, 16:37:06 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 18 December 2013, 16:37:06 UTC |
16ffd97 | Stephane Priem | 18 December 2013, 16:36:50 UTC | fixed macport petc | 18 December 2013, 16:36:50 UTC |
2321073 | Alexandre Ancel | 18 December 2013, 15:22:27 UTC | Added init and mem transfer code (OpenCL CRB) | 18 December 2013, 15:22:27 UTC |
fd0d50d | Vincent Chabannes | 18 December 2013, 08:12:07 UTC | add Portfile for hdf5-18 ( with +fortran as default variant) | 18 December 2013, 08:12:07 UTC |
2cc9926 | Vincent Chabannes | 18 December 2013, 08:10:59 UTC | remove doublon option | 18 December 2013, 08:10:59 UTC |
d53374a | Christophe Prud'homme | 18 December 2013, 06:05:08 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 18 December 2013, 06:05:08 UTC |
a6b12fd | Christophe Prud'homme | 18 December 2013, 06:04:45 UTC | tests for hypercube | 18 December 2013, 06:04:45 UTC |
de4a53c | Christophe Prud'homme | 18 December 2013, 06:04:29 UTC | fix for hypercube, slightly increase polynomial order to fix stiffness matrix in case of hypercube | 18 December 2013, 06:04:29 UTC |
faf1712 | Christophe Prud'homme | 18 December 2013, 06:02:36 UTC | up | 18 December 2013, 06:02:36 UTC |
aa7cdb1 | Christophe Prud'homme | 18 December 2013, 06:02:32 UTC | cleanup | 18 December 2013, 06:02:32 UTC |
0608bf0 | Alexandre Ancel | 17 December 2013, 17:20:25 UTC | CRB - Added initial entry point for online CRB step in OpenCL HARTS - Fixed cmake module if harts is not found in the system | 17 December 2013, 17:20:25 UTC |
7383252 | Stephane Veys | 17 December 2013, 12:10:12 UTC | add a bool to say if the element (parameter) is broadcasted or not | 17 December 2013, 12:10:12 UTC |
7d241d0 | Stephane Veys | 17 December 2013, 07:47:50 UTC | for eim and scm every procs continue to have the same supersampling | 17 December 2013, 07:47:50 UTC |
038479c | Christophe Prud'homme | 16 December 2013, 16:23:43 UTC | update issue #36 : test 1D mortar finite element space from P1 to P4 | 16 December 2013, 16:23:43 UTC |
d83b543 | Christophe Prud'homme | 16 December 2013, 16:23:14 UTC | update issue #36: fix local vertex dof index | 16 December 2013, 16:23:14 UTC |
d83fc3f | Christophe Prud'homme | 16 December 2013, 15:46:40 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 16 December 2013, 15:46:40 UTC |
7601a91 | Christophe Prud'homme | 16 December 2013, 15:41:46 UTC | update issue #36 implement mortar dof numbering need now to update linear and bilinear forms so that they handle change in the number of degrees of freedom | 16 December 2013, 15:41:46 UTC |
d859543 | Christophe Prud'homme | 16 December 2013, 15:39:46 UTC | added sibling meshes support for bilinear form (only element-wise case) | 16 December 2013, 15:39:46 UTC |
5ff51a4 | Stephane Veys | 16 December 2013, 14:32:22 UTC | add option to say if all procs have the same super-sampling( as before ) or not | 16 December 2013, 14:32:22 UTC |
b45b671 | Vincent Chabannes | 16 December 2013, 14:29:24 UTC | up example | 16 December 2013, 14:29:24 UTC |
f469dfc | Vincent Chabannes | 16 December 2013, 14:07:59 UTC | use submesh relation<FACES> with linear form | 16 December 2013, 14:07:59 UTC |
0cc3347 | Vincent Chabannes | 16 December 2013, 13:54:00 UTC | isRelatedTo,isSiblingOf and isParentMeshOf can be called with MeshBase* | 16 December 2013, 13:54:00 UTC |
82b2187 | Stephane Veys | 16 December 2013, 10:40:16 UTC | fix build of sampling complement in // | 16 December 2013, 10:40:16 UTC |
516f94a | Stephane Veys | 14 December 2013, 09:23:41 UTC | when super-sampling is built log-randomly, check that there is no duplicates | 14 December 2013, 09:23:41 UTC |
0f2bcce | Vincent Chabannes | 13 December 2013, 17:14:09 UTC | Merge branch 'develop' of https://github.com/feelpp/feelpp into develop | 13 December 2013, 17:14:09 UTC |
451cf04 | Vincent Chabannes | 13 December 2013, 17:14:01 UTC | fix test_operatorinterpolation + remove test_element_serialize and test_interp_twomesh in // (because lot of work to fix that) | 13 December 2013, 17:14:01 UTC |
52541b0 | Stephane Veys | 13 December 2013, 16:45:16 UTC | fix file name of super-sampling Xi | 13 December 2013, 16:45:16 UTC |
e287c4f | Stephane Veys | 13 December 2013, 16:44:32 UTC | log-equidistrubted and equidistributed super-sampling are also distributed on all procs | 13 December 2013, 16:44:32 UTC |
f4e0719 | Alexandre Ancel | 13 December 2013, 16:38:48 UTC | Removed previous reference to OpenCL with Harts in feelconfig. Harts will handle this in its own config file. | 13 December 2013, 16:38:48 UTC |
e537043 | Vincent Chabannes | 13 December 2013, 16:31:58 UTC | fix nonlinear solver use with geomap inverse : - the type of snes (trust regin) must be implement directly in code (not option) | 13 December 2013, 16:31:58 UTC |
e564393 | Vincent Chabannes | 13 December 2013, 16:31:20 UTC | fix geomapcontext when nDim!=nRealDim and nOrder > 1 | 13 December 2013, 16:31:20 UTC |