ngsPETSc.utils.fenicsx
This module contains all the functions related to wrapping NGSolve meshes to FEniCSx We adopt the same docstring conventiona as the FEniCSx project, since this part of the package will only be used in combination with FEniCSx.
Attributes
Classes
This class is used to wrap a Netgen geometric model to a DOLFINx mesh. |
Functions
|
Convenience wrapper to extract elements from a NetGen mesh based on topological dimension |
|
Trim zeros from a two dimensional numpy array along axis 1 |
|
Extract element tags from a Netgen mesh (on a given MPI rank) and distribute them onto |
Module Contents
- ngsPETSc.utils.fenicsx._ngs_to_cells
- ngsPETSc.utils.fenicsx._dim_to_element_wrapper(ngmesh: Any) dict[int, Any]
Convenience wrapper to extract elements from a NetGen mesh based on topological dimension
- ngsPETSc.utils.fenicsx._numpy_trim(array: numpy.typing.ArrayLike, dtype: numpy.typing.DTypeLike) numpy.ndarray
Trim zeros from a two dimensional numpy array along axis 1 and return it as a specified dtype.
- Args:
array: The input array to trim. dtype: The desired output dtype.
- Returns:
The trimmed array.
- class ngsPETSc.utils.fenicsx.GeometricModel(geo, comm: mpi4py.MPI.Comm, comm_rank: int = 0)
This class is used to wrap a Netgen geometric model to a DOLFINx mesh. Args:
geo: The Netgen model comm: The MPI communicator to use for mesh creation
- geo
- comm
- comm_rank = 0
- model_to_mesh(hmax: float, gdim: int = 2, partitioner: Callable[[mpi4py.MPI.Comm, int, int, dolfinx.cpp.graph.AdjacencyList_int32], dolfinx.cpp.graph.AdjacencyList_int32] = dolfinx.mesh.create_cell_partitioner(dolfinx.mesh.GhostMode.shared_facet), transform: Any = None, routine: Any = None, meshing_options: dict[str, Any] | None = None) tuple[dolfinx.mesh.Mesh, tuple[dolfinx.mesh.MeshTags | None, dolfinx.mesh.MeshTags | None], dict[tuple[int, str], tuple[int, Ellipsis]]]
Given a NetGen model, take all physical entities of the highest topological dimension and create the corresponding linear DOLFINx mesh.
- Args:
hmax: The maximum diameter of the elements in the triangulation model: The NetGen model gdim: Geometrical dimension of the mesh partitioner: Function that computes the parallel
distribution of cells across MPI ranks
transform: PETSc DMPLEX Transformation to be applied to the mesh routine: Function to be applied to the mesh after generation
takes as plan the mesh and the NetGen model and returns the same objects after the routine has been applied.
- Returns:
A DOLFINx mesh for the given NetGen model. It also extracts cell tags, facet tags and a mapping from the NetGen label to the corresponding integer marker(s).
- extract_regions()
Extract regions from the Netgen mesh.
- extract_linear_mesh(gdim: int, partitioner=dolfinx.mesh.create_cell_partitioner(dolfinx.mesh.GhostMode.shared_facet)) tuple[dolfinx.mesh.MeshTags, dolfinx.mesh.MeshTags]
Extract a DOLFINx mesh (and correpsonding cell and facet tags) from the Netgen mesh.
- Args:
gdim: Geometric dimension of the mesh partitioner: Function that computes the parallel distribution of cells across MPI ranks
- Note:
This function updates the self._mesh to be in sync with the NetGen model.
- Returns:
The cell and facet tags of the DOLFINx mesh.
- curveField(order: int, permutation_tol: float = 1e-08)
This method returns a curved mesh as a Firedrake function.
- Parameters:
order – the order of the curved mesh.
permutation_tol – tolerance used to construct the permutation of the reference element.
- refineMarkedElements(dim: int, elements: numpy.typing.NDArray[numpy.int32], netgen_flags: dict | None = None)
Refine mesh based on marked elements.
- ngsPETSc.utils.fenicsx.extract_element_tags(comm_rank: int, ngmesh, dolfinx_mesh: dolfinx.mesh.Mesh, dim: int) dolfinx.mesh.MeshTags
Extract element tags from a Netgen mesh (on a given MPI rank) and distribute them onto the corresponding DOLFINx mesh.
- Args:
comm_rank: The MPI rank to extract the element from. ngmesh: The Netgen mesh object. dolfinx_mesh: The DOLFINx mesh to which the facet tags will be distributed to. dim: The topological dimension of the entities to extract.