scijava-jupyter-kernel aims to be a polyglot Jupyter kernel. It uses the Scijava scripting languages to execute the code in Jupyter client and it's possible to use different languages in the same notebook.

Some of the supported languages are Groovy (default), Python, Beanshell, Clojure, Java, Javascript, Ruby and Scala.

The kernel has been originally created to work with ImageJ. See here for more details.

Under the hood scijava-jupyter-kernel uses the Beaker base kernel.

A documentation in the forms of a serie of notebooks is available here.

• Install Anaconda
• Install scijava-jupyter-kernel with :

# Add the conda-forge channel
conda config --add channels conda-forge

# Create an isolated environment called `java_env` and install the kernel
conda create --name java_env scijava-jupyter-kernel

• Usage :

# Activate the `java_env` environment
source install java_env

# Check the kernel has been installed
jupyter kernelspec list

# Launch your favorite Jupyter client
jupyter notebook

# or
jupyter lab

Note : It is suggested to install the kernel in an isolated Conda environment (not in the root environment).

• Download the last released JAR file.

• Drop it in your Fiji plugins directory.

• Start Fiji and launch Analyze > Jupyter Kernel > Install Scijava Kernel.

• Set the path to your Python binary.

• Choose a log level.

• Check the kernel has been installed with : jupyter kernelspec list.

• Launch jupyter notebook or jupyter lab and select the kernel you want in the kernel list.

• CI with Travis makes sure the project builds without errors for each new commit.
• All the notebook examples are executed by the kernel during CI with nbconvert (currently not working).
• A Conda package is built for each new release.

Under Apache 2.0 license. See LICENSE.

• Hadrien Mary hadrien.mary@gmail.com