Introducing jupyterlab-env-kernel#

We are happy to announce the release of jupyterlab-env-kernel, a JupyterLab extension that makes it easy to create Conda environments and use them as Jupyter kernels directly from the JupyterLab interface.

Creating and managing environments often requires switching between JupyterLab and a terminal. jupyterlab-env-kernel brings this workflow into JupyterLab, making it easier to create a dedicated environment for a notebook or project.

Getting started#

Open JupyterLab and select the Conda Environments icon in the left sidebar.

From the environment manager, you can create a new environment by specifying:

  • an environment name,

  • a Python version, and

  • optional packages such as numpy, pandas, or scikit-learn.

Click Create Environment & Kernel, and the extension will create the Conda environment using mamba, install ipykernel, and automatically register the environment as a Jupyter kernel.

The new kernel is then immediately available from the notebook’s kernel selector.

Creating environments from files#

You can also create an environment from an existing environment.yml or requirements.txt file.

For example, an environment.yml might contain:

name: my-scientific-env
channels:
  - conda-forge
dependencies:
  - python=3.11
  - numpy
  - pandas
  - matplotlib
  - scikit-learn

Simply upload the file through the From File tab and create the environment. The resulting environment and Jupyter kernel are registered automatically.

Why jupyterlab-env-kernel?#

The goal is simple: make environment and kernel management a natural part of the JupyterLab workflow.

Instead of manually creating an environment, installing ipykernel, and registering a kernelspec, everything can be done from the JupyterLab interface.

Learn more#

The package is available on PyPI:

https://pypi.org/project/jupyterlab-env-kernel/

HSM News Vol 1  
On this page