Fork me on GitHub
Back to main page | License and credit


Jupyter/JupyterLab should be installed on your computer if you have installed Anaconda. but please also verify your Jupyter installation.

If you instead installed Miniconda, you can install JupyterLab by:

$ conda install -c conda-forge jupyterlab

If you are not using conda, you can install JupyterLab using pip:

pip install jupyterlab

Regardless of how you installed JupyterLab, please also verify your installation.

Installation of extensions on Linux, macOS, and Windows

Here you will find how to install optional JupyterLab extensions which will be demonstrated during the workshop.

You don’t have nodejs/npm?: it is easy to install the extensions if installed Python through conda, but if you installed JupyterLab through a virtual environment, it is not easy to follow this section because nodejs or npm are non-Python programs that are easy to install through Conda, but not other ways. In this case, it’s OK to skip this section: you won’t be able to do a few exercises, but you won’t miss much else.

You are not required to install these extensions, but if you wish to you can follow the instructions below.

JupyterLab extensions are npm packages (standard package format for Javascript), and to install the extensions you need to install the nodejs package:

$ conda install nodejs

If you get error: CondaVerificationError use:

$ conda install -c conda-forge nodejs

JupyterLab extension manager

To manage JupyterLab extensions, it’s easiest to use the built-in extension manager. Once it is installed, you can easily activate or deactivate other extensions. Some extensions require the installation of a companion Python package, while others only require you to install the extension via the extension manager.

This is documented in the JupyterLab documentation.

You see the extension manager in the left sidebar. If you use JupyterLab 2.1 or newer (released 7 april 2020), then it is already enabled.

If you use jupyterlab less than 2.1, then you need to enable the extension manager. It needs to be enabled from the Settings→Enable Extension Manager option. (For even older versions, it might not be available by default, in this case, consider upgrading).


Installing widgets via the ipywidgets package is not required. Widgets will be demonstrated by the instructor and typing along is optional.

ipywidgets can be installed via conda:

$ conda install ipywidgets

or via pip (which needs an extra activation step):

$ pip install ipywidgets
$ jupyter nbextension enable --py widgetsnbextension

and then activate it in JupyterLab by installing the @jupyter-widgets/jupyterlab-manager extension via the extension manager.

Git extension

jupyterlab-git is a JupyterLab extension for version control using Git.
Install it via:

$ conda install -c conda-forge jupyterlab-git


$ pip install jupyterlab-git

and then activate it by finding and installing jupyterlab-git via the extension manager.

jupyterlab-git will also install the nbdime package as one of its dependencies. nbdime provides tools for diffing and merging Jupyter notebooks.

GitHub extension

@jupyterlab/github is a JupyterLab extension for accessing GitHub repositories.

First install it by:

$ pip install jupyterlab_github

and then activate it by finding and installing @jupyterlab/github via the extension manager.

Other languages

Python will be the main language used during the CodeRefinery workshop, but if you use other programming languages the chances are high that someone has written a Jupyter kernel for that language!

A full list of available Jupyter kernels can be found here. Find your favorite language in the list and click the link for installation and usage instructions.

How to verify the installation

To see whether JupyterLab is working as expected, type the command

$ jupyter-lab

and see if it opens up a new tab in your browser showing the JupyterLab interface.

Please also verify which Python version JupyterLab sees. Ideally it can see Python 3.

On Windows, the JupyterLab App can also be launched by clicking on the JupyterLab icon in the Anaconda menu.