Thank you very much for your willingness to contribute to QUEENS! We strongly believe in the synergy effect of developing and using QUEENS together as a community.
We invite you to share your methodological contributions to both deterministic and probabilistic models and analyses, such as:
- Parameter studies and identification
- Sensitivity analysis
- Surrogate modeling
- Uncertainty quantification
- Bayesian inverse analysis
- Optimization
In addition to methodological contributions, we also greatly appreciate infrastructure contributions that help ensure QUEENS runs smoothly and efficiently. These include, but are not limited to, bug fixes and improvements related to
- Code quality
- Performance
- User interface
- Benchmarking
- Testing
- Tutorials
- Documentation
We welcome all types of code contributions, irrespective of size and complexity.
Note: If you're unsure whether your contribution fits within the QUEENS framework, don't hesitate to ask the community by starting a discussion or by opening an issue 😊
Issues are generally used to remind or inform yourself or others about certain things in the software. We use them to report bugs, start a feature request, or plan tasks. In case you have a general question, please refer to GitHub Discussions.
To create an issue, select one of our templates and provide a detailed description. We use labels to organize our issues, so please label issues with the mandatory labels
status:labeltopic:labeltype:label
More labels can of course be assigned if they contribute to categorizing the issue.
Before you open a new issue, please check within the existing issues if your bug has already been reported. Opening an issue is a valid contribution on its own and does not mean you have to solve them yourself.
Install QUEENS as described in the README.md and run:
pip install -e .[develop]
or to do a safe develop install use:
pip install -e .[safe_develop]
To help you write style-compliant code, we use the pre-commit package to manage all our git hooks automatically. Please run:
pre-commit install --install-hooks --overwrite
pre-commit install --hook-type commit-msg
QUEENS code follows the PEP 8 style. Non-compliant code will lead to failing CI pipelines and will therefore not be merged. The code checks are conducted with Pylint, isort, and Black. Compliance with Google style docstrings is checked with ruff. Complete and meaningful docstrings are required as they are used to generate the documentation.
Like every codebase, QUEENS follows some project-specific coding conventions. Below is a list of common ones:
- Use
pathlib.Pathobjects instead of strings to handle paths and directories. - If relative paths within the QUEENS source are needed, use the relative_path_from_queens_source function.
- Decorate the init method of QUEENS objects with the
log_init_argsdecorator from src/queens/utils/logger_settings.py. This automatically logs the arguments passed to the init. - We only allow disabling pylint warnings for specific lines, not for entire files. If you disable warnings, please use the long pylint description, not just the code.
Please provide meaningful commit messages based on the Conventional Commits guidelines. These are verified by the commit-msg hook (managed by commitizen).
New code must be tested. Please also make sure that all existing tests pass by running pytest in
your source directory. For further information, see our testing README.md.
Please use the available pull request template and fill out all sections of the template. When you have submitted a pull request and the CI pipeline passes, it will be reviewed. Once your pull request is approved, there is a 24h waiting time (business days only) until the branch is merged into the main branch by the QUEENS maintainers. This ensures that the community has a chance to have a final look over the changes. This rule can be circumvented if and only if:
- All the active maintainers approve the pull request.
- The pull request is labeled as a quickfix by one of the maintainers and approved. Examples for this are one-liners, typos or urgent fixes that are time-critical.