Here are two alternative ways to run a Jupyter notebook without installing jupyterlab locally. The first uses Google Drive and is less directly integrated with Github (this option is recommended), the second is less intuitive, but provides more direct interface with Github.

To run the lab online using Google Colab

This option uses proprietary software with terms of use, requires a Google account and available space on Google Drive, and can’t be integrated directly with your online Github repo, so it requires manual file management (i.e. downloading your work from Google Drive once completed and uploading it to Github for submission).

to install Python packages in Google Colab

Colab comes with pre-installed Python libraries. You can check these by entering !pip list. As you can see, this is an ordinary shell command (pip list) you would execute on command line, but to do this in Google Colab requires you to prepend the command with an exclamation mark (!). Therefore, to install a python package such as pandas, you would do !pip install pandas, etc.

to import data to your Google Colab notebook

To run the lab online using MyBinder

This option uses open source infrastructure and can integrate directly with Github while being online, however it requires your Github repo to be set to public. It also can’t save your work anywhere since each new session is temporary. A good workaround is to use git push to save your work (which means you should be pushing frequently if you don’t want to lose any progress due to having accidentally closed your web browser!).