Numpy project template

Construct a numpy-based Python project from scratch for scientific computing and research.

🔧 Features

• Numpy a basic installation for the holy trifecta 🕊️ of numpy, matplotlib, scipy.
• PDM for dependency, virtualenv, and package management.
• Mkdocs material for simple, clean, automated, online code documentation.
• pre-commit with ruff integration for code linting and formatting.
• pytest with coverage for regression testing and code coverage.
• Commitizen with conventional commits for automatic versioning and changelogs.
• Copier for continuously updating project from this original template.
• Github actions for automated, build, test, and deployment.

📍 Quickstart

Have python and git installed, then:

pip install --user pdm
pdm self add copier copier-templates-extensions

cd path/to/project
pdm init --copier gh:eckelsjd/copier-numpy --trust

That's it! Follow the questionnaire and then your numpy-based scientific computing project is ready to go.

All extra arguments are passed to copier copy (you're also welcome to just use copier directly).

Note: The --trust flag enables extensions used in this template. Please see extensions.py and setup_github.py to make sure you trust this template (spoiler: these just add some global template variables and some basic git scripting).

🐍 Publishing on PyPI

Follow this tutorial to enable trusted publishing with Github actions. Then, do:

pdm bump

That's it! Your package will automatically deploy to PyPI and GitHub with a correctly-versioned vX.X.X tag.

🔄 Updating your project

copier update --trust

This will update template files in your project with the most recent releases to the copier-numpy template. See the copier docs for more info.

🔗 See similar

• Scientific Python library development
• Copier-pdm and the similar pdm-project version
• Serious scaffolding for Python projects
• LINCC framework for Python projects by the University of Washington
• copier-pylib for pure Python projects