1. 07 Sep, 2018 1 commit
    • Andreas Nüßing's avatar
      [IO] add option to write out the system matrix and the rhs · 766ce3e3
      Andreas Nüßing authored
      In order to test some solvers outside of the duneuro-framework, it can be
      beneficial to just generate the matrices and the rhss with duneuro, store
      them in a file (matrix market) and read the again somewhere else. This
      commit adds this option.
      Note that this is more of a debug/experimental feature
  2. 25 May, 2018 2 commits
  3. 24 Mar, 2018 2 commits
  4. 28 Feb, 2018 2 commits
  5. 26 Feb, 2018 2 commits
  6. 24 Feb, 2018 3 commits
  7. 23 Feb, 2018 1 commit
  8. 22 Feb, 2018 1 commit
    • Andreas Nüßing's avatar
      [TransferMatrixUser] unify fitted and unfitted · 4274b4ae
      Andreas Nüßing authored
      Again, there were two basically identical versions of the transfer
      matrix user, one for fitted and the other one for unfitted solvers.
      We merge both classes into a single one which is then parameterized by
      the solver and the source model factory.
  9. 20 Feb, 2018 1 commit
    • Andreas Nüßing's avatar
      unify transfer matrix solver · bfd5f465
      Andreas Nüßing authored
      Instead of using two basically identical transfer matrix solver classes,
      we merge them into a single class that is parameterized with the solver
      and the transfer matrix rhs factory.
  10. 19 Feb, 2018 6 commits
    • Andreas Nüßing's avatar
      [Transfer] rename transfer rhs · 85bfe42b
      Andreas Nüßing authored
      As we might have different types of transfer matrix rhs in the future,
      rename the current ones to {fitted,unfitted}point...
    • Andreas Nüßing's avatar
      [Transfer] create rhs factories · cccede23
      Andreas Nüßing authored
      Instead of directly creating the transfer matrix rhs, we use a factory
      for its constrution. This is basicaly a preparation for the unified
      transfer matrix solver, als well as for different kinds of transfer
      matrix rhs (such as cem)
    • Andreas Nüßing's avatar
      [Transfer] extract rhs · 75558daf
      Andreas Nüßing authored
      We introduce an interface for the transfer matrix right hand side and
      modify the existing transfer matrix rhs (fitted and unfitted) to conform
      to this inteface
    • Andreas Nüßing's avatar
      Unify EEGForwardSolver · 5b2e1178
      Andreas Nüßing authored
      This unifies the fitted and unfitted EEGForwardSolver. Up to now, there
      were two basically identical implementations, the
      {Fitted,Unfitted}EEGForwardSolver. As they did basically the same thing, we
      merge both implementations and parameterize them via the Solver and the
      SourceModelFactory. While doing so, we move the element-search into the
      Solver class and mark several variables as const that are nerver and also
      should never be modified (mainly the VolumeConductor and the
    • Andreas Nüßing's avatar
      Merge branch 'feature/udg-subtraction' into 'master' · 41a4d6ec
      Andreas Nüßing authored
      [UDG] Subtraction
      See merge request duneuro/duneuro!51
    • Andreas Nüßing's avatar
      fix rebasing errors · 4deb2a0a
      Andreas Nüßing authored
  11. 18 Feb, 2018 5 commits
  12. 16 Feb, 2018 2 commits
  13. 15 Feb, 2018 2 commits
    • Andreas Nüßing's avatar
      [MEEGDriver] add statistics method · 844f5cc3
      Andreas Nüßing authored
      We add a new method to the MEEGDriver interface, namely `statistics`.
      The exact semantics of the method depends on the implementation, but
      it should write out statistics about the driver to the datatree.
      For the fitted driver, this currently means the volume and surface area
      of the different discrete compartments.
      For the unfitted driver, the semantics is similar as it writes out the
      volume and surface area of the different domains.
    • Andreas Nüßing's avatar
      [VolumeConductor] use always indirect mapping · c97477a5
      Andreas Nüßing authored
      In order to prepare for providing statistics about the volume conductor,
      we remove the direct mapping to be able to access element labels.
      Most of the meshes (but some legacy files) provide labeled elements and
      we thus have a direct access to these labels.
      For the legacy formats, this leads to a slight increase in time consumption
      when accessing the tensor of an element, but compared to other tasks, this
      should be negligible.
      Besides being able to access the labels, this also gets rid of the type
      erasure in the volume conductor, leading to way more readable code.
  14. 14 Feb, 2018 2 commits
  15. 09 Feb, 2018 1 commit
  16. 05 Feb, 2018 7 commits