Installation#

CaliBrain can be installed from source for development or installed into a standard Python environment for running workflows. The package metadata currently declares Python >=3.6; in practice, the scientific Python dependencies used by the workflows are best managed with a recent conda or virtualenv environment.

Dependencies#

Core dependencies are declared in pyproject.toml and include:

  • numpy

  • pandas

  • scipy

  • scikit-learn

  • mne

  • POT

  • nibabel

  • pyyaml

  • matplotlib and visualization dependencies

Documentation dependencies are declared in the optional docs extra.

Source installation#

From a local checkout:

git clone https://github.com/braindatalab/CaliBrain.git
cd CaliBrain
python -m pip install -e .

For documentation builds:

python -m pip install -e ".[docs]"

Conda environment#

The repository includes environment.yml. Use it when you want conda to create the scientific Python environment:

conda env create -f environment.yml
conda activate calibrain
python -m pip install -e .

Pip requirements#

For a pip-only setup:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -r requirements.txt
python -m pip install -e .

Data directory#

Workflow examples expect precomputed forward solutions and leadfields. By default, CaliBrain uses the repository data/ directory. To use another data location, set:

export CALIBRAIN_DATA=/path/to/calibrain/data

The workflow configs also contain explicit paths for results, posterior summaries, run manifests, aggregation outputs, and calibration outputs. Inspect configs/*.py before running large experiments.

Building the documentation#

From the repository root:

cd docs
make html

The rendered site is written to docs/build/html/index.html.

Installation topics