Usage

RAMSES-CPP is designed to be plug-and-play with the existing ecosystem of the original Fortran RAMSES code.

Running a Simulation

The simulation is configured using standard Fortran Namelist (.nml) files.

To run a simulation, execute the appropriate dimensional executable (e.g., ramses_3d) and pass the path to your namelist file:

cd build
./ramses_3d path/to/your/namelist.nml

Example: 3D Sedov Blast

A standard Sedov 3D blast wave test is included in the namelist directory:

./ramses_3d ../namelist/sedov3d.nml

Parallel Execution (MPI)

If you compiled RAMSES-CPP with MPI support, you can execute it across multiple processors using mpirun or mpiexec:

mpirun -np 4 ./ramses_3d ../namelist/sedov3d.nml

The code will automatically perform domain decomposition using a Hilbert curve and distribute the workload across the available ranks.

Output and Visualization

Simulations output their state to directories named output_XXXXX/ (where XXXXX is the output step number).

Because RAMSES-CPP maintains strict binary compatibility with legacy RAMSES data formats, it generates the expected amr, hydro, and info files.

You can use existing Python visualization tools to parse and plot these results seamlessly:

# Example using the included Python plotting script for a hydro test
python3 ../tests/hydro/implosion/plot-implosion.py

Automated Testing

The repository includes a comprehensive test suite to verify the physics and AMR logic.

Note: You must include ./tests/visu/ in your PYTHONPATH for the plotting and verification scripts to function.

To run the tests:

export PYTHONPATH=$PYTHONPATH:$(pwd)/tests/visu
cd tests
./run_test_suite.sh -t hydro  # Run hydrodynamics tests
./run_test_suite.sh -t mhd    # Run MHD tests

Test Configuration (`config.txt`)

Each individual test directory contains a config.txt file that defines the build-time requirements for that benchmark. The test runner (run_test_suite.sh) automatically parses these flags and propagates them as CMake options:

NDIM=N: Compiles for N-dimensions.
SOLVER=mhd: Automatically enables RAMSES_USE_MHD.
RT=1: Automatically enables RAMSES_USE_RT.
NENER=N: Sets the number of non-thermal energy variables.

The test suite performs a clean build for every test to ensure there is no configuration leakage between benchmarks.

Verification Tool

The repository includes a verify_ref utility to compare C++ snapshots against reference Fortran results to ensure bit-perfect structural parity.

./verify_ref path/to/fortran_amr path/to/cpp_amr