Setup and tools

Project setup and dependency management

  • All tool configuration and dependency information for AEIC code is held in the top-level pyproject.toml file.

  • AEIC uses Python 3.13.

  • Dependencies are managed using the uv tool.

Local Development

If you intend to develop the source code of AEIC, you should create a fork, clone the fork locally, and install AEIC in development mode using the uv tool (see below). For example:

  1. Fork the main Git repository.

  2. Clone the forked repository locally:

    git clone git@github.com:{YourName}/AEIC.git

  3. In your Python environment, install in development mode:

    cd AEIC
uv sync

You should now be able to import AEIC as you would with any standard library within a Python session managed by uv.

uv

The uv tool tool manages dependencies within a virtual environment. Installation instructions are here.

The most basic functionality needed is to run uv sync in the top-level directory. This creates a virtual environment, installs all of the dependencies, pins the Python version (based on the mandatory requires-python field in the pyproject.toml), and automatically does an editable install of the local package (just like you had done pip install -e .), which is almost always what you want for local development.

pre-commit

The pre-commit tool helps to manage Git hooks, in particular, it lets you set up scripts for checking whether changes you’re tracking in Git are OK, before you commit them. This lets you catch things like large file commits before they get into the repository history. It also means fewer “Ruff fixes” commits, since you can include running ruff as a pre-commit hook, so that you never end up committing code that ruff doesn’t like.

  1. Run pip install --user pre-commit or uv run pip install --user pre-commit (this installs the pre-commit executable in ~/.local/bin, so make sure you have that on your path).

  2. Run pre-commit install at the top of your working copy of the repository. This sets up the necessary Git hooks to run the pre-commit tool.

  3. The pre-commit hooks will run and will prevent you from committing until you make your changes “nice”.

Testing

Automated tests are in the tests directory and use the Pytest test runner. GitHub Actions are set up to run all tests when pull requests are created.