Development

How to set up a development environment, run tests, and contribute to pylcm.

Setup¶

pylcm uses pixi for dependency management. Python 3.14+ is required.

# Clone the repository
git clone https://github.com/OpenSourceEconomics/pylcm.git
cd pylcm

The first pixi run command will install dependencies automatically, but you can run pixi install explicitly if you prefer.

Install pre-commit hooks (requires prek):

pixi global install prek
prek install

Running Tests¶

# All tests
pixi run tests

# Tests with coverage
pixi run tests-with-cov

# Specific test file
pixi run pytest tests/test_specific_module.py

# Specific test function
pixi run pytest tests/test_specific_module.py::test_function_name

Code Quality¶

# Type checking (ty, not mypy)
pixi run -e tests-cpu ty

# Run all pre-commit hooks
prek run --all-files

Building Documentation¶

# Build docs
pixi run build-docs

# Live preview (watches for changes)
pixi run view-docs

Conventions¶

Code Style¶

• Ruff for linting and formatting (configured in pyproject.toml)

• All functions require type annotations

• Google-style docstrings in imperative mood (“Return”, not “Returns”)

• Use # ty: ignore[error-code] for type suppression (never # type: ignore)

• Never use from __future__ import annotations

Testing¶

• Plain pytest functions — no test classes

• Use @pytest.mark.parametrize for test variations

Naming¶

• func for callable abbreviations (never fn)

• state_names / action_names (not states_names)

• arg_names (not argument_names)

Plotting¶

• Always use plotly (plotly.graph_objects, plotly.subplots.make_subplots), never matplotlib

Docstrings¶

• MyST syntax (single backticks for code, $...$ for math), not reStructuredText

• Inline field docstrings (PEP 257) for dataclass attributes

• Docstring types must match annotations (Mapping not “Dict” when annotation is Mapping)

Contributing¶

Report bugs and suggest features on the GitHub issue tracker.