For Setup:

python -m venv .venv source .venv/bin/activate # On Windows use `.venv\Scripts\Activate.ps1` pip install uv uv sync --active

For data, you may download our already preprocessed data from google drive which includes graphs for 100 popular repositories from GitHub. These repositories were chosen for their purely Python codebases and popularity. You should extract the zip file inside ./codevoyant_output. You may also analyze any repoository and generate graphs:

uv run analyze.py --repo <path_to_repo> --output <output_directory> 

To run the Streamlit app:

uv run streamlit run src/ui/app.py

To run benchmarking, you should make sure you download the data as mentioned above. Then run:

uv run python -m src.benchmark.run

After finishing the quick start, install development dependencies:

uv pip install ruff mypy types-networkx

Add extensions:

1. Ruff
2. MyPy

You'll have format on save and type checking based on .vscode/settings.json.

You can specifically run mypy and ruff via:

mypy . ruff format .