For more detailed instructions, see our documentation.
To install runn, we recommend using the mamba package manager:
mamba create -n runn -c conda-forge -c JoseAngelMartinB runn
git clone git@github.com:JoseAngelMartinB/runn.git
cd runn
mamba create -n runn -c conda-forge --file requirements/base.txt --file requirements/dev.txt
mamba activate runn
pip install --no-deps -e .
python -m ipykernel install --user --name runnThis will install ruun in editable mode. For more detailed instructions, see our documentation.
There are many ways to contribute to runn. Before making contributions to the runn source code, see our contribution guidelines and follow the development install instructions.
If you plan to make changes to the code then please make regular use of the following tools to verify the codebase while you work:
pre-commit: runpre-commit installin your command line to load inbuilt checks that will run every time you commit your changes. The checks are: 1. check no large files have been staged, 2. lint python files for major errors, 3. format python files to conform with the pep8 standard. You can also run these checks yourself at any time to ensure staged changes are clean by simple callingpre-commit.pytest- run the unit test suite and check test coverage.pytest -p memray -m "high_mem" --no-cov(not available on Windows) - after installing memray (mamba install memray pytest-memray), test that memory and time performance does not exceed benchmarks.
For more information, see our documentation.
If you are unable to access the online documentation, you can build the documentation locally. First, install a development environment of runn, then deploy the documentation using mike:
mike deploy develop
mike serve
Then you can view the documentation in a browser at http://localhost:8000/.
The runn package is distributed under the MIT License. See LICENSE for more information.
This package was created with Cookiecutter and the arup-group/cookiecutter-pypackage project template.