Installation

Requirements

VANE supports Python 3.10 to 3.13 on Linux, macOS and Windows. Its runtime dependencies are NumPy, SciPy, pandas, Matplotlib, Plotly and Typer; pip installs them automatically.

Post-processing existing .lin files needs nothing else. Running linearization sweeps (The OpenFAST linearization sweep) also needs an OpenFAST executable whose release reads the input decks, and, for closed-loop settle runs of a floating turbine, a controller library built for the host.

VANE is distributed through its GitHub repository, <https://github.com/SMI-Lab-Inha/VANE>. A virtual environment is recommended.

Install the release wheel

Download vane-0.1.0-py3-none-any.whl from the v0.1.0 release, check it against SHA256SUMS.txt from the same page, and install it:

$ python -m pip install vane-0.1.0-py3-none-any.whl

Install from the repository

Install the tagged release directly from GitHub:

$ python -m pip install "git+https://github.com/SMI-Lab-Inha/VANE.git@v0.1.0"

or clone the repository and install it, in editable mode with the development tools for work on VANE itself:

$ git clone https://github.com/SMI-Lab-Inha/VANE.git
$ cd VANE
$ python -m pip install .
$ python -m pip install -e ".[dev]"   # for development

Optional extras

excel

openpyxl, needed only to write .xlsx tables. CSV, JSON and .mat export need no extra.

pdf

weasyprint, needed only to render the HTML study report to PDF.

docs

The Sphinx toolchain used to build this manual.

dev

ruff, mypy, pytest and its plugins, pre-commit and build.

$ python -m pip install ".[excel,pdf]"

Check the installation

$ python -c "import vane; print(vane.__version__)"
$ vane --version
$ python -m pytest -q

Tests that need OpenFAST, the OpenFAST r-test decks or other local reference data skip when that data is not available; the validation record (Validation) lists the environment variables that enable them.

Note

On Windows, some OpenBLAS 0.3.34 builds have crashed with an access violation in numpy.linalg.eig and scipy.linalg.eig when eigenvectors of matrices of order 200 or more were requested, which VANE’s eigensolver does for full turbine models. If you meet this crash, use NumPy and SciPy builds linked against a different OpenBLAS release (0.3.33 is unaffected).