Zernikeb1 subcell projection#7241
Conversation
michaeldmurphy1
force-pushed
the
zernikeb1_subcell_projection
branch
2 times, most recently
from
0581a24 to
a6e10d1
Compare
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
ZernikeB1 bases, used with Cartoon, require the parity of the component to be known as the interpolation requires this to be sufficiently accurate. When the projection is of a DataVector, it must either be a single component with the parity passed as an argument, or the underlying tmpl::list to types that make up the DataVector must be passed as a template. The GrMhd systems are updated to correctly call the projection, while all other systems are simply updated to work with the new interface. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
michaeldmurphy1
force-pushed
the
zernikeb1_subcell_projection
branch
from
a6e10d1 to
946005e
Compare
| #include "Evolution/DgSubcell/CombineVolumeGhostData.hpp" | ||
| #include "Evolution/DgSubcell/GhostData.hpp" | ||
| #include "Evolution/DgSubcell/NeighborRdmpAndVolumeData.hpp" | ||
| #include "Evolution/DgSubcell/NeighborRdmpAndVolumeData.tpp" |
There was a problem hiding this comment.
Including this in general headers mostly defeats the purpose of having a tpp file. Would explicitly instantiating the functions be feasible?
Alternatively, I might have missed something but I don't see anything obvious that requires the parity information to be constexpr, so you could make the function in the hpp something that calls compute_parity_list and passes the information to an impl that doesn't have to be templated on that tags. (The array could be converted to an initializer_list, which though sometimes painful to work with isn't templated on the size.)
| * \brief Computes the projection matrix in 1 dimension going from a DG | ||
| * mesh to a conservative finite difference subcell mesh. | ||
| * | ||
| * The parity parameter is required for ZernikeB1 bases. |
There was a problem hiding this comment.
"and is ignored for others" or something similar.
| const DirectionalIdMap<Dim, std::optional<intrp::Irregular<Dim>>>& | ||
| Neighbor_dg_to_fd_interpolants); | ||
| neighbor_dg_to_fd_interpolants, | ||
| tmpl::list<VolumeFields> /*meta*/); |
There was a problem hiding this comment.
Isn't this a list<list<tags...>>? Do you need the second list?
| DataVector even_input(num_even * num_dg_pts, 0.0); | ||
| DataVector odd_input(num_odd * num_dg_pts, 0.0); | ||
| DataVector even_output(num_even * num_ghost_pts_per_var, 0.0); | ||
| DataVector odd_output(num_odd * num_ghost_pts_per_var, 0.0); |
There was a problem hiding this comment.
Do these need to be initialized to 0, or can you leave them uninintialized?
| DataVector even_input(num_even * num_dg_pts, 0.0); | ||
| DataVector odd_input(num_odd * num_dg_pts, 0.0); | ||
| DataVector even_output(num_even * num_ghost_pts_per_var, 0.0); | ||
| DataVector odd_output(num_odd * num_ghost_pts_per_var, 0.0); |
There was a problem hiding this comment.
A possible optimization would be to allocate these as a single buffer.
| const DirectionalIdMap<Dim, std::optional<intrp::Irregular<Dim>>>& | ||
| neighbor_dg_to_fd_interpolants) { | ||
| neighbor_dg_to_fd_interpolants, | ||
| tmpl::list<VolumeFields> meta) { |
There was a problem hiding this comment.
If you're actually going to use the variable, name it something more descriptive than "meta".
| template <typename TagList, size_t Dim> | ||
| void project(const gsl::not_null<DataVector*> subcell_u, const DataVector& dg_u, | ||
| const Mesh<Dim>& dg_mesh, const Index<Dim>& subcell_extents, | ||
| tmpl::list<TagList> /*meta*/) { |
There was a problem hiding this comment.
Same question as above about lists of lists.
2 participants
Proposed changes
We need ZernikeB1 projection for cartoon subcell. We need to know the parity for each projection, which can either be passed or inferred by the type (so now the type needs to be passed).
This is achieved by either passing the parity as an argument for DataVectors, or for when the passed DataVector is multiple components (i.e. GhostData) the underlying types must be passed.
Depends on:
Upgrade instructions
Code review checklist
make docto generate the documentation locally intoBUILD_DIR/docs/html.Then open
index.html.code review guide.
bugfixornew featureif appropriate."Co-Authored-By: Claude Sonnet 4.6 noreply@anthropic.com",
"Co-Authored-by: Codex noreply@openai.com", or
"Co-Authored-By: GitHub Copilot CLI noreply@microsoft.com"
as the last line of the commit, depending on the agent.
Further comments