JupyterLab with Version Control

Requirements

  • Git and TortoiseGit (GUI)

  • Anaconda providing the Python environment

Installing GIT

  1. IconGitTGit Option
    A corporate internal software portal provides Git and TortoiseGit or a combined package of the two.

  2. Option
    Download and install Git from the project website manually.

    • Additionally, the graphical user interface TortoiseGit can be installed. However, this requires administrator privileges.

Installing the Python environment Anaconda

  1. Download the Anaconda-Package for Windows, 64 bit, from the Anaconda website

    • If the virus scanner blocks some installation scripts (*.bat files) in the following, install an older version of Anaconda, e.g. from Oct. 2019, and then update, for example via the Anaconda Navigator.

  2. Start the setup by double-clicking the file.

  3. Confirm the license agreement with mit I Agree.

  4. Install for the current user only be selecting Just Me.

  5. Keep the default destination folder for the installation and confirm with Next >

  6. Leave other options at their default values and complete the installation.

Setting up Git support for JupyterLab

  1. Download a repository of helper scripts from GitHub, for example with TortoiseGit

  2. Double click on the file Anaconda Update Script.bat and let it run. It does several things and does not require admin rights:

    • It updates the Anaconda Python distribution and creates a log file of the process next to the location of the scripts (ComputerName.log).

    • It registers the Jupyter-Notebook file extension .ipynb, so that a double click on such files starts JupyterLab and opens the notebook.

    • It installs extensions for interactive widgets (ipywidgets) in JupyterLab and installs nodeJS, which is required for this.

    • Last but not least, it installs the extension jupyterlab_git, which adds an integrated user interface for Git to JupyterLab.