Installation

gridfm-trustlayer needs Python 3.10, 3.11 or 3.12. It runs on CPU only and has no PyTorch or GPU dependency.

We recommend a fresh virtual environment:

python -m venv .venv
source .venv/bin/activate      # on Windows: .venv\Scripts\activate
python -m pip install --upgrade pip

Install the package

python -m pip install gridfm-trustlayer

The development version, from the main branch:

python -m pip install "git+https://github.com/gridfm/gridfm-trustlayer"

Check that the command-line tool is available:

gridfm_trustlayer --help

Optional extras

Some features need extra packages. Install them by adding the extra's name in brackets:

Extra	Adds	You need it to
datakit	gridfm-datakit	calibrate line_loading: trustlayer computes line loading with datakit's branch-flow routine. Not needed for Vm and Va alone.
graphkit	gridfm-graphkit and gridfm-datakit	the same, in an environment where you also run models with graphkit, which needs PyTorch (install it first, following graphkit's guide).
interactive	ipywidgets	use the config builder, a Jupyter form that writes the config file for you.

For example, both at once:

python -m pip install "gridfm-trustlayer[datakit,interactive]"

Using the GridFM tools in one environment

If you also generate data with gridfm-datakit or run models with gridfm-graphkit, you can install trustlayer into the same environment. The graphkit extra sets the minimum versions trustlayer works with. See the datakit and graphkit installation guides for their own requirements.

Install for development

Clone the repository and install it in editable mode with the development and test tools:

git clone https://github.com/gridfm/gridfm-trustlayer.git
cd gridfm-trustlayer
python -m pip install -e ".[dev,test]"
pre-commit install

pre-commit run --all-files then runs the same checks as the CI: ruff, mypy and the test suite. To build this documentation locally, run mkdocs serve and open http://127.0.0.1:8000.

Next step

Follow the Quick start to run the full workflow on your own data.