School Seminar: Why Jupyter Notebooks?


For several years, 50% of TM351 teaching has been delivered through interactive coding exercises in a custom virtual machine using Jupyter Notebooks; TM112 offers Azure cloud hosted notebooks as part of an optional student exercise. In the same period, UC Berkeley has moved to exposing almost half its total undergrad population onto an introductory data science module using Jupyter notebooks, with adoption across the university campus, and Edina are currently trialling a hosted notebook service with courses from five separate subject areas at the University of Edinburgh.  Corporate adoption is on the increase with Netflix among others using notebooks across the organisation, from devops to data analysis. The UK Ministry of Justice Analytical Platform offers Jupyter notebooks as a service, as does CERN’s SWAN (Service for Web-based Analysis). Every major machine learning platform (ML is now widely available as a commodity, PAYG service) uses notebooks as a first class user environment.
So what are they, and what can they do? This talk will give a provide a brief intro, with a particular emphasis on their relevance to teaching and research.
Tony Hirst is a Senior Lecturer in Telematics (no, he has no idea what that means either) in the Open University’s School of Computing and Communications. He has worked on OU courses ranging from AI to infoskills, gaming to robotics, as well as the current data management and analysis course TM351. A regular blogger at, he’s just launched a Tracking Jupyter newsletter, because somebody had to.

Report an error on this page