pixi-kernel 0.1.0

Python Jupyter kernel using Pixi for reproducible notebooks

Pixi Kernel

Use per-directory Pixi environments to run Jupyter kernels. The idea behind this project is to allow you to capture the exact state of your environment while allowing you to make use of the rich PyPI and Conda ecosystems.

Pixi Kernel supports Python 3.8+ and Pixi 0.18+ using pyproject.toml and pixi.toml configurations.

Quick Start

1. Install this package with pip install pixi-kernel in your JupyterLab environment and restart it.
2. Create a new directory and notebook and select the Pixi kernel for the notebook.
3. Initialize a Pixi project with pixi init and pixi add ipykernel.
4. Restart the kernel and you are good to go.

See the Pixi docs for more information on how to use Pixi.

User Experience

If the Pixi Kernel is used in a project where pixi is not installed or the project does not have ipykernel installed in the environment, starting the kernel fails.

A fallback kernel is then started showing a helpful message when you try to run the notebook. This is a regular IPython kernel which allows you to run shell commands and hopefully fix the configuration of the project.

Limitations

Pixi Kernel does not support using Pixi environments other than the default one. If you have an idea how to support this, please open an issue or check the contributing guide to open a pull request.

Related

• Pyproject Local Jupyter Kernel
• Poetry-kernel