Open ranocha opened 2 years ago
Bumping to add that we should do the same for DGMultiMesh
constructors. For example, cells_per_dimension
is a keyword argument here
DGMultiMesh(dg::DGMulti; cells_per_dimension,
coordinates_min=(-1.0, -1.0), coordinates_max=(1.0, 1.0),
is_on_boundary=nothing,
periodicity=ntuple(_ -> false, NDIMS))
but not for the StructuredMesh
equivalent
DGMultiMesh(dg::DGMulti{NDIMS}, cells_per_dimension, mapping;
is_on_boundary=nothing,
periodicity=ntuple(_ -> false, NDIMS), kwargs...) where {NDIMS}
Yes, it would definitely be great to unify these :+1:
I agree, I also do not like that for P4est
meshes we explicitly supply the dimensionality as a type (not sure if this is the correct Julian term), i.e., P4estMesh{2}
while for TreeMesh
, StructuredMesh
and DGMultiMesh
the dimensionality is inferred from the arguments while again for the unstructured 2d mesh the constructor reads UnstructuredMesh2D
.
It may be nice to unify the signatures of some mesh constructors or provide additional convenience constructors. This can be nice when testing different meshes, e.g., for #1191. Right now, we have
For example, it wold be nice to just replace
TreeMesh
byStructuredMesh
orP4estMesh
and get something that just works.