Template library to generate an o²S²PARC compatible interactive service with the JupyterLab environment. In addition to the different JupyterLab "flavours" offered on o²S²PARC, you can use this library to create one with your favourite software and tools.
This library is pre-configured to easily build JupyterLabs with Python, R or Julia and additional dependencies. Just choose the appropriate base_image when running the cookiecutter and follow the instructions in the README.md.
If you need to install other custom software, you can do so by selecting the custom image as base_image.
- GNU Make
- Python3
- Python3-venv (recommended to work in a virtual environment)
cookiecutterDocker(if you wish to build and test the service locally)
To install the cookiecutter:
sudo apt-get update
sudo apt-get install -y make python3-venv
python3 -m venv .venv
source .venv/bin/activate
pip install cookiecutter jinja2_timeGenerate a new Cookiecutter template layout:
python3 -m venv .venv
source .venv/bin/activate
cookiecutter git+ssh://[email protected]/ITISFoundation/cookiecutter-osparc-jupyterlab-servicegit clone git+ssh://[email protected]/ITISFoundation/cookiecutter-osparc-jupyterlab-service
cd cookiecutter-osparc-jupyterlab-service
make devenv
source .venv/bin/activate
make playThis will create different projects on the docker image bases, specified in cookiecutter.json, and will build the images (this can take some time)
make testsIn the latest version, tests only checks that cookiecutter can be run and the Docker images built. Manual testing is still required to check that the built images run as expected.
This project is licensed under the terms of the MIT License
The service code is in the folder {{cookiecutter.project_slug}}.
This cookiecutter allows users to choose a Kernel, among Python, R and Julia, where they can install packages from a requirement-like file (see env-config for details). By default, the latest kernel versions (as of April 2024) provided by the Jupyter team are used (e.g. Python 3.11). Users find in their Dockerfiles instructions on how to use older versions.
In case these options are not enough, they can build their custom image. We recommend to use one of the Jupyter Docker Stacks as base (so JupyterLab is already installed and configured) and the start-up scripts can be re-used.
By default, with this cookiecutter, users install their libraries in the main kernel (i.e. we don't add an additional kernel and environment to the ones already shipped with the base JupyterLab).
