1. 04 Jun, 2020 2 commits
  2. 03 Jun, 2020 1 commit
  3. 28 May, 2020 1 commit
  4. 12 May, 2020 1 commit
  5. 11 May, 2020 1 commit
    • Nicolas CHEVAUGEON's avatar
      modified: xcrack-test/geometricderivative/geomderivRhs.h · 056f1998
      Nicolas CHEVAUGEON authored
          addede header for
      
          	modified:   xcrack-test/geometricderivative/geomderivRhs.h
      
      [xcrack-test] some #include added in mains
      
       due to clean up of the header of xcrack some #include were missing in :
            xcrack-test/front/main.cc
            xcrack-test/geometricderivative/main.cc
            xcrack-test/geometricderivative/geomdervRhs.h
      
      This 3 files have also been clag-formated.
      056f1998
  6. 10 May, 2020 3 commits
    • Nicolas CHEVAUGEON's avatar
      [xfem-seq-test] two level-set test updated · 06e40c16
      Nicolas CHEVAUGEON authored
        xfem-seq-test/ls_propag/main.cc and xfem-seq-test/lset_transform/main.cc are updated to use cut function from xcut/xLegacySimpleCut/xMeshCut.h/cc
      06e40c16
    • Nicolas CHEVAUGEON's avatar
      [xfem-seq-test] updated for interface change · dc154024
      Nicolas CHEVAUGEON authored
      updated for the interface changes in xScalarFunctionDerivDiscXFEMOctree and xPhysSurfOctree
      xfem-seq-test/inclusion_xfem_high_order/main.cc updated to take into account interface change in
      xScalarFunctionDerivDiscXFEMOctree and xPhysSurfOcree
      
      ref for xfem-seq-test/hexQuadLevelset have been reordered
      
      modified files :
       xfem-seq-test/hexQuadLevelset/reference/S_0001_R_0000_TWO.pos
       xfem-seq-test/hexQuadLevelset/reference/S_0001_R_0000_TWO_SIMPLEX.pos
       xfem-seq-test/hole_xfem_high_order/main.cc
       xfem-seq-test/inclusion_xfem_high_order/main.cc
      dc154024
    • Nicolas CHEVAUGEON's avatar
      [xtest] some reference are updated. · 77e8a5c5
      Nicolas CHEVAUGEON authored
      77e8a5c5
  7. 07 May, 2020 1 commit
    • Nicolas CHEVAUGEON's avatar
      [xfem-seq-test] change from xfile commit... · df7037b6
      Nicolas CHEVAUGEON authored
      [xfem-seq-test] change from xfile commit 0ef7f77b90c548e059a94923cd0a7d7626727771 forced an update of some ref
      
      As often, some results are same between the 2 releases, up to a reordering.
      I verifyed manually the where needed and updated them.
      df7037b6
  8. 05 May, 2020 1 commit
  9. 04 May, 2020 1 commit
  10. 16 Apr, 2020 1 commit
  11. 10 Apr, 2020 1 commit
  12. 08 Apr, 2020 2 commits
    • Nicolas CHEVAUGEON's avatar
      [Xtest] Implement the changes from Xfiles's commit · 4cc1c1d7
      Nicolas CHEVAUGEON authored
            Changes from Xfile commit
            f2906dd071066c2b8bb022578526ee32949cfa40
            are propagatted in Xtest.
      
            Some reference have been changed when only the ordering of the data
            were different.
      
      MODIFIED FILES :
      	modified:   xcrack-test/postpro_crack/reference/crack.msh
      	modified:   xcrack-test/postpro_crack/reference/crack.vtu
      	modified:   xdoublecut-seq-test/simple_use_xphysurfCutDom/main.h
      	modified:   xfem-seq-test/ls_propag/main.cc
      	modified:   xinterface-xoctree-test/assembly/main.cc
      	modified:   xinterface-xoctree-test/assembly/main.h
      	modified:   xtls-test/WeigthedDirectionalAverageConvergence3D/main.h
      4cc1c1d7
    • Nicolas CHEVAUGEON's avatar
      [xtest] xMesh::partition_tag removed follow up · 5b865799
      Nicolas CHEVAUGEON authored
           This is a followup of  commit
           8be52876c164c6da54ba492e1ea248e407a10c13 of Xfiles.
           It updates the changes in xMesh into test files related to the
           partition_tag.
           After the previous update, some test were still failing and the
           follwing additional changes were done :
      
           - [xfem-para-test/complex2d/] ref changed. The ref for this test
             case stongly depends on  the memory usage. The matrix ordering
            depends on the relative address of the entity.
             So that the produced results matrix where not the same than in
             the previous reference.
             Actually the numeric value where identical but not the ordering.
             To verify that the new matrices  were the same, I loaded them
             in a linear algebra tool and checked that
             the eigen value where indeed the same.
      
           - [xdoublecut-seq-test/hammerhead] The way hammerhead was
             previoulsy set up, it could crash if the ordering of the nodes in
              the memory was not the same as the ordering of the nodes in the
             file that stored the ls value at nodes or the svls value at nodes.
             For the svls value, since a reader already exist in the library,
             I modifyed it directly in Xfiles.
             for the ls it is modifyied directly in the main.
      
       Changes to be committed:
      
      	modified:   xdoublecut-seq-test/hammerhead/main.cc
      	modified:   xfem-para-test/complex2d/main.cc
      	modified:   xfem-seq-test/functional_xRefCutToAOMD_cutMeshByRef_3D_Recursif/main.cc
      	modified:   xfem-seq-test/highOrderQuadHex/main.cc
      	modified:   xinterface-xoctree-test/anisotropicHoleXfem/main.cc
      	modified:   xinterface-xoctree-test/assembly/main.cc
      	modified:   xinterface-xoctree-test/functional_HangingL2Projection_2D/main.cc
      	modified:   xinterface-xoctree-test/functional_HangingL2Projection_3D/main.cc
      	modified:   xinterface-xoctree-test/hole_xfem_octree/main.cc
      	modified:   xinterface-xoctree-test/hole_xfem_octree_no_hanging/main.cc
      	modified:   xinterface-xoctree-test/hole_xfem_octree_no_hanging3D/main.cc
      	modified:   xinterface-xoctree-test/inclusion_xfem_octree/main.cc
      	modified:   xinterface-xoctree-test/thin_layer/main.cc
      5b865799
  13. 30 Mar, 2020 1 commit
  14. 11 Mar, 2020 1 commit
    • Nicolas CHEVAUGEON's avatar
      [removing tags ] in the process of cleaning up tags · f891e064
      Nicolas CHEVAUGEON authored
          this commit apply some slight change in xtest  due to the changes in the tag
          handling in xMesh for data related to the octree->aomd interface
          (or more generally the data used to handle meshes with hanging entities )
          it also remove some left meshDataIdLookup for string that correspond to
          tag already defined in xMesh.
      f891e064
  15. 05 Mar, 2020 2 commits
  16. 04 Mar, 2020 1 commit
  17. 03 Mar, 2020 1 commit
  18. 27 Feb, 2020 1 commit
    • Alexis SALZMAN's avatar
      [xfastmarching-test] illustrate fmeik new usage · df34464e
      Alexis SALZMAN authored
      To illustrate new change in fmeik family this commit add
      example of call to fmeik with different container then
      entitistorage.
      TLSFastMarchingTool test is still using entitistorage.
      But now in DistStart2DComplexMeshDefinition we use
      a xAttachedDataManagerAOMDGeneric and in
      DistStart3DComplexMeshDefinitionWithModes we use
      a xUnorderedMapDataManager.
      Both work well and give same results ....
      
      As suggested by N.C. we can now compare performance of
      exlibris container by simply change the datamanagerFM_t
      type in those tests...
      df34464e
  19. 25 Feb, 2020 1 commit
    • Alexis SALZMAN's avatar
      [xfastmarching-test] reactivating DistStart3DComplexMeshDefinitionWithModes · bf1d2194
      Alexis SALZMAN authored
      This test is re-activate by adding a call to a script that generate
      mesh. In data folder, the call, done by proc 0, launch genmeshtest
      if devbranch.msh is not present. Unfortunately for unforeseen reasons
      launching this script on tnode 10 from application via a system cmd lead
      gmsh to crash. This crash appears also if gmsh, with same arguments as genmesh,
      is launched by hand on a terminal. If genmeshtest is launched by hand on
      a terminal gmsh works fine !?. Don't have time to figure out why gmsh
      behave like this ....
      
      This way the msh file (rather large) is not tracked in Xtest. If not present,
      application is expected to generate it, on the fly, the first time it run.
      Then for every following run msh file is used directly.
      
      Compare to previous situation where no mesh lead to ignore totally this
      test in ctest statistic (i.e. the run and the ndiff pass without error),
      now if mesh not present it stop. And if mesh is present it run normally
      and check what it has to.
      
      A .gitignore file is added to force user not to track devbranch.msh.
      TO DISCUSS: maybe we don't care storing such large files in Xtest.
      In this case all those scripts have to be removed.
      bf1d2194
  20. 24 Feb, 2020 2 commits
  21. 10 Feb, 2020 2 commits
  22. 05 Feb, 2020 4 commits
    • Alexis SALZMAN's avatar
      [xfem-para-test] add functional_xValueLinearCombinationParam test · 4306c07d
      Alexis SALZMAN authored
      This test case is testing the xValueLinearCombinationParam concept
      introduced in Xfiles commit 135c3fc
      
      It's 2 cubes related by some linear relation.
      
      First step create only xValueLinearCombination (nested)
      Then a mix of nested xValueLinearCombination and xValueLinearCombinationParam
      is tested and compare with first step.
      Then only xValueLinearCombinationParam (nested) are tested and compare with first step.
      
      In step 2 parametric coefficients are change by hand to check
      differences with step 1
      In step 3  parametric coefficients are change by hand and with a
      computation (flux assembly). Again it shows different with step 1.
      
      In this test case all values are in the same double_manager but some
      un-committed test prove that values (parametric or master) can comes from one or many
      other double_manager ....
      4306c07d
    • Alexis SALZMAN's avatar
      [xfem-seq-test] clone_doubleManager use xCloneTemplateValue<double> · eb039d50
      Alexis SALZMAN authored
      use explicit xCloneTemplateValue<double> in replacement of the type that
      has been removed by xFiles commit 7b0a56e
      eb039d50
    • Alexis SALZMAN's avatar
      [xfem-para-test] change complex2d with USE_ZMUMPS · e243573f
      Alexis SALZMAN authored
      following Xfile comit 0a09622 now complex mumps solver instance is
      created only when USE_ZMUMPS is available
      e243573f
    • Grégory LEGRAIN's avatar
      Revert unwanted change · 88b810d6
      Grégory LEGRAIN authored
      88b810d6
  23. 04 Feb, 2020 1 commit
  24. 03 Feb, 2020 1 commit
  25. 27 Jan, 2020 1 commit
    • Grégory LEGRAIN's avatar
      Update testcases according to template xField · 2d863e94
      Grégory LEGRAIN authored
      Following commit 3cc692f488c2220e9d49abbde4eb9b62bad11eb8 in Xfiles.
      * Update testcases
      * Add new testcase complex2d in xfem-para-test
      
      This testcase highlights the use of a complex-valued field. For the
      moment, only the matrix is tested in the reference. When zmumps will be
      available, the reference will be updated with solutions.
      2d863e94
  26. 20 Dec, 2019 1 commit
    • Nicolas CHEVAUGEON's avatar
      removing the deps of xfem on xcut · aaf3bf31
      Nicolas CHEVAUGEON authored
        some test case have been updated so that it use constructor of enriched
        function that do not directly refer to xPhysSurfxxx
         xValueCreatorRampedHeaviside is moved to xtls, so that this test must
         follow :  Moving xfem-seq-test/mechanics_ramp_heaviside to xtls-test
       + Some dependencies on xTensor library were missing in some CMakeLists.txt
      
      Squashed commit of the following:
      
      commit e0c65ead57b2a86e2c2ee30249267f0e78452c15
      Author: chevaugeon <nicolas.chevaugeon@ec-nantes.fr>
      Date:   Fri Dec 20 15:32:00 2019 +0100
       removing the deps of xfem on xcut
          some test case have been updtated so that it use constructor of enriched
          function that do not directly refer to xPhysSurfxxx
          Some dependencies on xTensor library were missing in some CMakeLists.txt
      
      commit 683f7b82f50df7e1b5743ac039581b30b21eed1e
      Author: chevaugeon <nicolas.chevaugeon@ec-nantes.fr>
      Date:   Fri Dec 20 15:28:57 2019 +0100
      
          Moving xfem-seq-test/mechanics_ramp_heaviside to xtls-test
      
          I'm in the process of removing the deps of xfem on xcut and
          xValueCreatorRampedHeaviside is moved to xtls, so that this test must
          follow
      aaf3bf31
  27. 11 Dec, 2019 1 commit
  28. 10 Dec, 2019 1 commit
  29. 09 Dec, 2019 1 commit
    • Nicolas CHEVAUGEON's avatar
      removing dependecies of xfem on xexport · ddb1e384
      Nicolas CHEVAUGEON authored
      +xtls-test/CMakeLists.txt modifyed so that the tests that don't use Taucs still compiles when USE_TAUCS = NO
      
      Squashed commit of the following:
      
      commit 056bf5ef0af02d6cd24acc0219934b5779467b05
      Author: Nicolas CHEVAUGEON <nchevaug@titan.ec-nantes.fr>
      Date:   Mon Dec 9 10:08:01 2019 +0100
      
          Finishing the move of Export Algo to xexport (to remove the cyclic dep between xfem and xexport)
      
          +xtls-test/CMakeLists.txt modifyed so that the tests that don't use Taucs still compiles when USE_TAUCS = NO
      
      commit db3ffdf3b906b8dbd445273b4005f4532f100c6c
      Author: chevaugeon <nicolas.chevaugeon@ec-nantes.fr>
      Date:   Sun Dec 8 09:12:43 2019 +0100
      
          removing dependecies of xfem on xexport
      ddb1e384
  30. 27 Nov, 2019 1 commit