Last active
February 11, 2026 14:19
-
-
Save jorgensd/459a4ae463abe76bd4867f42a0c16c04 to your computer and use it in GitHub Desktop.
MED format converter
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Copyright 2026 Jørgen S. Dokken and Simula Research Laboratory | |
| Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: | |
| The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. | |
| THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # READ MED files into DOLFINx with cell and facet tags | |
| # Author: Jørgen S. Dokken and Simula Research Laboratory | |
| # SPDX-License-Identifier: MIT | |
| import h5py | |
| from pathlib import Path | |
| import dolfinx | |
| import numpy as np | |
| import basix.ufl | |
| from mpi4py import MPI | |
| from dolfinx.cpp.mesh import create_mesh, create_cell_partitioner | |
| # MAP: type to (celltype, degree) | |
| _salome_to_dolfinx = { | |
| "SE2": (dolfinx.mesh.CellType.interval, 1), | |
| "TR3": (dolfinx.mesh.CellType.triangle, 1), | |
| "QU4": (dolfinx.mesh.CellType.quadrilateral, 1), | |
| "TE4": (dolfinx.mesh.CellType.tetrahedron, 1), | |
| "PY5": (dolfinx.mesh.CellType.pyramid, 1), | |
| "HE8": (dolfinx.mesh.CellType.hexahedron, 1), | |
| } | |
| def compute_local_range(comm: MPI.Intracomm, N: np.int64): | |
| """ | |
| Divide a set of `N` objects into `M` partitions, where `M` is | |
| the size of the MPI communicator `comm`. | |
| NOTE: If N is not divisible by the number of ranks, the first `r` | |
| processes gets an extra value | |
| Returns the local range of values | |
| """ | |
| rank = comm.rank | |
| size = comm.size | |
| n = N // size | |
| r = N % size | |
| # First r processes has one extra value | |
| if rank < r: | |
| return [rank * (n + 1), (rank + 1) * (n + 1)] | |
| else: | |
| return [rank * n + r, (rank + 1) * n + r] | |
| def read_med_mesh(filename: str | Path, comm: MPI.Intracomm): | |
| facet_data = None | |
| if comm.size == 1: | |
| h5_kwargs={} | |
| else: | |
| h5_kwargs= {"driver" : "mpio", | |
| "comm": MPI.COMM_WORLD} | |
| with h5py.File(filename, "r", **h5_kwargs) as h5file: | |
| if "ENS_MAA" not in h5file.keys(): | |
| raise ValueError( | |
| f"Could not find a set of meshes (Ensemble de Maillages) in {filename}." | |
| ) | |
| mesh_set = h5file["ENS_MAA"] | |
| meshes = mesh_set.keys() | |
| if len(meshes) > 1: | |
| raise RuntimeError(f"Multiple meshes found in {filename}") | |
| mesh = mesh_set[meshes.__iter__().__next__()] | |
| time_steps = mesh.keys() | |
| if len(time_steps) == 1: | |
| mesh = mesh[time_steps.__iter__().__next__()] | |
| # Nodes must be re-ordered according to NUM | |
| nodes = mesh["NOE"] | |
| attrs = mesh.parent.attrs | |
| if "DIM" in attrs: | |
| gdim = attrs["DIM"] | |
| elif "ESP" in attrs: | |
| gdim = attrs["ESP"] | |
| elif "ESPACE" in attrs: | |
| gdim = attrs["ESPACE"] | |
| else: | |
| gdim = 3 # Fallback assumption | |
| # Read all nodes on rank 0 as they must be reordered | |
| num_nodes = nodes["NUM"].size | |
| coords = nodes["COO"] | |
| in_dim = int(coords.size / num_nodes) | |
| assert coords.size % num_nodes == 0 | |
| if comm.rank == 0: | |
| coordinates = coords[:].reshape(in_dim, -1).T | |
| node_ordering = nodes["NUM"][:] | |
| assert min(node_ordering) == 1 | |
| else: | |
| coordinates = np.zeros((0, in_dim), dtype=np.float64) | |
| node_ordering = np.zeros(0, dtype=np.int64) | |
| # node_marker = nodes["FAM"] | |
| if len(node_ordering) != coordinates.shape[0]: | |
| raise RuntimeError("Not all coordinates have global index") | |
| coordinates = coordinates[node_ordering[:] - 1].astype(np.float64).copy() | |
| # Create a sorter to map Global ID -> Local Index (0, 1, 2...) | |
| # This allows us to look up where ID '105' is located in the array | |
| # Find cell types of highest topological dimension | |
| elements = mesh["MAI"] | |
| cell_types = [] | |
| # Sort to have same accessing on each process to align | |
| # coordinate elements | |
| element_keys = sorted(list(elements.keys())) | |
| for el in element_keys: | |
| dx_type = _salome_to_dolfinx[el] | |
| cell_types.append((el, dx_type)) | |
| def tdim(ct: tuple[str, tuple[dolfinx.mesh.CellType, int]]) -> int: | |
| return dolfinx.mesh.cell_dim(ct[1][0]) | |
| max_tdim = tdim(max(cell_types, key=tdim)) | |
| cells = filter(lambda el: tdim(el) == max_tdim, cell_types) | |
| facets = list(filter(lambda el: tdim(el) == max_tdim - 1, cell_types)) | |
| topologies = [] | |
| top_markers = [] | |
| coord_els = [] | |
| for salome_cell, (ct, degree) in cells: | |
| # Ignore NUM key, we do not use it | |
| top = elements[salome_cell] | |
| num_cells = top["NUM"].size | |
| local_cell_range = compute_local_range(comm, num_cells) | |
| # Extract data and permute (assume VTK ordering) | |
| num_dofs = int(top["NOD"].size / num_cells) | |
| assert np.isclose(top["NOD"].size % num_cells, 0) | |
| conn_data = np.zeros((local_cell_range[1]-local_cell_range[0], num_dofs), dtype=np.int64) | |
| for i in range(num_dofs): | |
| conn_data[:,i] = top["NOD"][local_cell_range[0]+i*num_cells:local_cell_range[1]+i*num_cells] | |
| permutation = np.argsort( | |
| dolfinx.cpp.io.perm_vtk(ct, conn_data.shape[1]) | |
| ) | |
| conn_flattened = (conn_data[:, permutation] - 1).flatten() | |
| # Store data | |
| top_markers.append(top["FAM"][slice(*local_cell_range)]) | |
| topologies.append(conn_flattened) | |
| coord_els.append( | |
| dolfinx.fem.coordinate_element( | |
| basix.ufl.element( | |
| "Lagrange", | |
| dolfinx.mesh.to_string(ct), | |
| degree, | |
| shape=(int(gdim),), | |
| ).basix_element | |
| )._cpp_object | |
| ) | |
| assert len(facets) < 2 | |
| if len(facets) == 1: | |
| salome_facet, (ft, _) = facets[0] | |
| ftop = elements[salome_facet] | |
| num_facets = ftop["NUM"].size | |
| # Extract data and permute (assume VTK ordering) | |
| num_facet_dofs = int(ftop["NOD"].size / num_facets) | |
| assert np.isclose(ftop["NOD"].size % num_facets, 0) | |
| local_facet_range = compute_local_range(comm, num_facets) | |
| conn_data = np.zeros((local_facet_range[1]-local_facet_range[0], num_facet_dofs), dtype=np.int64) | |
| for i in range(num_facet_dofs): | |
| conn_data[:,i] = ftop["NOD"][local_facet_range[0]+i*num_facets:local_facet_range[1]+i*num_facets] | |
| permutation = np.argsort( | |
| dolfinx.cpp.io.perm_vtk(ft, conn_data.shape[1]) | |
| ) | |
| facet_conn = conn_data[:, permutation] - 1 | |
| facet_values = ftop["FAM"][slice(*local_facet_range)] | |
| facet_data = (facet_conn, facet_values) | |
| else: | |
| raise RuntimeError( | |
| f"We do not support time dependent meshes from {filename}" | |
| ) | |
| if comm.size == 1: | |
| assert coordinates.shape[0] == len(np.unique(np.hstack(topologies))) | |
| assert min(np.unique(np.hstack(topologies))) == 0 | |
| assert ( | |
| max(np.unique(np.hstack(topologies))) | |
| == len(np.unique(np.hstack(topologies))) - 1 | |
| ) | |
| part = create_cell_partitioner(dolfinx.mesh.GhostMode.none) | |
| _cpp_mesh = create_mesh(comm, topologies, coord_els, coordinates, part, 2) | |
| mesh = dolfinx.mesh.Mesh(_cpp_mesh, None) | |
| if len(topologies) == 1: | |
| cell_data, cell_values = dolfinx.io.distribute_entity_data( | |
| mesh, | |
| mesh.topology.dim, | |
| topologies[0].reshape(-1, mesh.geometry.cmap.dim), | |
| top_markers[0], | |
| ) | |
| ct = dolfinx.mesh.meshtags_from_entities( | |
| mesh, | |
| mesh.topology.dim, | |
| dolfinx.graph.adjacencylist(cell_data), | |
| cell_values.astype(np.int32), | |
| ) | |
| ct.name = "CellTags" | |
| if facet_data is not None: | |
| fdata, fvalues = dolfinx.io.distribute_entity_data( | |
| mesh, mesh.topology.dim - 1, *facet_data | |
| ) | |
| mesh.topology.create_connectivity(mesh.topology.dim - 1, mesh.topology.dim) | |
| ft = dolfinx.mesh.meshtags_from_entities( | |
| mesh, | |
| mesh.topology.dim - 1, | |
| dolfinx.graph.adjacencylist(fdata), | |
| fvalues.astype(np.int32), | |
| ) | |
| ft.name = "FacetTags" | |
| return mesh, ct, ft | |
| return mesh, ct, None | |
| else: | |
| return mesh, None, None | |
| if __name__ == "__main__": | |
| filename = Path("Mesh_2tetraRegions.med") | |
| mesh, ct, ft = read_med_mesh(filename, MPI.COMM_WORLD) | |
| dolfinx.io.vtkhdf.write_mesh("test_med.vtkhdf", mesh) | |
| with dolfinx.io.XDMFFile(mesh.comm, "test_med.xdmf", "w") as xdmf: | |
| xdmf.write_mesh(mesh) | |
| if ct is not None: | |
| xdmf.write_meshtags(ct, mesh.geometry) | |
| if ft is not None: | |
| xdmf.write_meshtags(ft, mesh.geometry) | |
| with dolfinx.io.XDMFFile(mesh.comm, "test_med.xdmf", "r") as xdmf: | |
| mesh = xdmf.read_mesh(name="mesh") | |
| ct = xdmf.read_meshtags(mesh, name="CellTags") | |
| mesh.topology.create_connectivity(mesh.topology.dim - 1, mesh.topology.dim) | |
| ft = xdmf.read_meshtags(mesh, name="FacetTags") |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment