Installation

DarSIA requires Python 3.12 or newer. It is not on PyPI yet, so install it from a clone of the repository.

With uv (recommended)

uv is a fast Python package/environment manager.

# install uv once
curl -LsSf https://astral.sh/uv/install.sh | sh

git clone https://github.com/pmgbergen/darsia.git
cd darsia
uv python install 3.12
uv sync --extra dev

Add --extra docs to also install the documentation toolchain (uv sync --extra dev --extra docs).

With pip

git clone https://github.com/pmgbergen/darsia.git
cd darsia
pip install -e ".[dev]"        # or ".[dev,docs]"

Optional: petsc4py

petsc4py accelerates the performance-critical linear solvers (notably the Wasserstein / Beckmann solvers). It is not installed by default.

Linux (Ubuntu/Debian):

sudo apt-get install -y libhypre-dev libmumps-seq-dev build-essential \
  gcc gfortran mpich cmake
pip install numpy mpi4py
PETSC_CONFIGURE_OPTIONS="--download-hypre --download-mumps --download-parmetis \
  --download-ml --download-metis --download-scalapack" pip install petsc petsc4py

macOS / conda:

conda install -c conda-forge petsc petsc4py

The bundled conda_env.yaml creates a complete conda environment (named darcia) with PETSc included: conda env create --file conda_env.yaml.

The graphical interface

The GUI ships with DarSIA (its Qt bindings, PySide6, are a normal dependency). Launch it with:

uv run darsia

Optionally register a desktop / Start-menu entry:

uv run darsia-install-desktop            # add
uv run darsia-install-desktop --uninstall  # remove

On Windows this needs pywin32, which is installed automatically on Windows. See The graphical interface for a full GUI tutorial.

Verifying the installation

uv run python -c "import darsia; print(darsia.__name__)"
uv run pytest tests/unit -q

Building the documentation

With the docs extra installed:

uv run sphinx-build -b html docs docs/_build/html

then open docs/_build/html/index.html. See Writing documentation for how the documentation is organised.