MEDCoupling

MEDCoupling is a robust and versatile C++/Python library for advanced mesh and field manipulation. It is part of the SALOME platform and co-developed by CEA and EDF.

Whether you're performing mesh intersections, transferring fields between meshes, or analyzing simulation results, MEDCoupling offers the building blocks you need for scientific computing workflows.

---

Table of Contents

• About MEDCoupling
• Getting Started
  • Installation
  • Usage Highlights
  • Resources
• Developer Guide
  • Building from Source
  • Contributing
  • Roadmap

---

About MEDCoupling

MEDCoupling is a powerful library designed for the manipulation and analysis of structured and unstructured meshes, as well as associated field data. It supports:

• Mesh intersection, node merging, extraction, and more
• Field creation and transformation
• Data transfer between non-conforming meshes
• I/O in the .med (MEDFile) format

The library handles complex geometries with hybrid cell types (tetrahedra, hexahedra, higher-order elements, etc.), and supports fields defined on nodes, cells, or Gauss points.

---

Getting Started

Installation

Precompiled binaries for multiple platforms (Windows, Linux) are available on the SALOME platform website.

For more flexibility, you can also build MEDCoupling from source—see the Developer Guide section below.

Minimal standalone Installation

First you need to clone configuration and medcoupling repositories:

 git clone https://github.com/SalomePlatform/configuration.git
 git clone https://github.com/SalomePlatform/medcoupling.git

Then you can specify a minimal cmake configuration:

 mkdir build
 cd build
 cmake -DCONFIGURATION_ROOT_DIR=$PWD/../configuration  -DMEDCOUPLING_MICROMED=ON -DMEDCOUPLING_ENABLE_PYTHON=OFF -DMEDCOUPLING_ENABLE_RENUMBER=OFF -DMEDCOUPLING_BUILD_DOC=OFF -DMEDCOUPLING_USE_MPI=OFF -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL_PREFIX=$PWD/../install $PWD/../medcoupling
 make -j4
 make install

Remarks: - this installation should work with libcppunit installed in your system ([sudo] apt-get install libcppunit-dev ); - you may need to specify CPPUNIT_ROOT_DIR.

If you have specific needs you can use the following cmake flags:

• MEDCOUPLING_ENABLE_PYTHON=ON : to enable PYTHON wrappers (dependencies: swig & numpy & scipy)
• MEDCOUPLING_MICROMED=OFF : to enable medfichier (dependencies: hdf5 & medfile)
• MEDCOUPLING_USE_MPI=ON : to enable parallel interpolator (dependency: MPI)
• MEDCOUPLING_ENABLE_PARTITIONER=ON : to enable partitioner (dependencies: metis | parmetis | scotch | ptscotch )
• MEDCOUPLING_ENABLE_RENUMBER=ON : to enable renumbering (dependencies: boost | scotch)

Finally you can run tests: make test

Usage Highlights

MEDCoupling provides a rich set of tools for building, analyzing, and processing meshes and fields in scientific computing workflows:

🔨 Mesh Creation & Editing

• Create structured and unstructured meshes programmatically.
• Build hybrid meshes with mixed element types (e.g., triangles and quads).
• Modify connectivity or extract sub-meshes.

🔁 Mesh Operations

• Merge or aggregate meshes, with optional node deduplication.
• Intersect meshes to compute shared domains or regions (in 2d).
• Compute cell barycenter on any cell type.

🌡️ Field Definition & Transformation

• Define scalar/vector fields on nodes, cells, or integration points.
• Compute measure fields (cell volumes) on any cell type.
• Use constants, expressions, or interpolated data.
• Transfer fields between non-matching meshes (remap).
• Combine or transform fields using arithmetic or geometric logic.

💾 Input/Output & Interoperability

• Read/write .med files (MEDFile format), compatible with SALOME, Code_Saturne, etc.
• Support for time-dependent data and multi-step simulations.
• Export data for post-processing in ParaView or SALOME.
• Used a main ICoCo mesh and field transfer format for code coupling.

🐍 Python Integration

• Access full functionality via Python bindings.
• Interact easily with NumPy arrays.
• Script mesh operations and field processing in notebooks or automation pipelines.

Resources

• 📘 Tutorials: MEDCoupling Tutorials on GitHub
• 📚 Documentation: Official MEDCoupling User & API Docs

  Note: Despite the title, this is the main user and API reference documentation.

---

Developer Guide

Building from Source

The recommended way to build MEDCoupling is using SALOME’s official package manager, SAT, which handles all dependencies (e.g., HDF5, MEDFile) and supports both native and custom configurations.

---

Contributing

We welcome contributions!

To ensure consistency, please install and use pre-commit:

pip install pre-commit
cd $PATH_TO_MEDCOUPLING/
pre-commit install

This will automatically run code formatters and linters on modified files before each commit. If hooks fail, there are two possibilities:

• It was a formatting issue — the formatter applied changes. You need to git add the file(s) and retry the commit.
• It was a linter error — the linter will tell you what to fix. Apply the corrections, then git add and commit again.

If you're in a hurry or blocked by linter issues, you can:

• Skip a specific linter hook:

  SKIP=linter-name git commit -m "wip: commit bypassing linter"

• Skip all hooks (not recommended, as formatting won't be applied):

  git commit -n -m "wip: commit skipping all pre-commit hooks"

---

Roadmap

We're actively improving MEDCoupling. Upcoming features and goals include:

🔧 Usability Improvements

• Unified and searchable documentation (via Sphinx + Breathe)
• Executable notebooks replacing static .rst tutorials
• Modernized CMake targets
• Native Python builds
• spack and conan package support

🛠️ Developer Experience

• Clearer modular architecture (medcoupling_tools)
• Modern C++ tooling (clang-tidy, clang-format, etc.)
• GitHub-based CI/CD workflows

---

For more information about SALOME and the broader ecosystem, visit salome-platform.org. We look forward to your feedback and contributions!