ETH Computational Competencies: JupyterHub
The newly established JupyterHub service at ETH brings JupyterNotebooks to everyone involved in teaching and learning at ETH. Easily reachable by one click from your course page in Moodle, a JupyterNotebook opens on the server - ready to go with no installation required. Write your text and formulas, write and run your code, and include your results and plots, all in one document, the JupyterNotebook.
What are JupyterNotebooks?
JupyterNotebooks are interactive documents which combine text, executable code, and visualizations. They can be accessed, run, and modified in a web browser without cumbersome installation of interpreters, compilers or similar. For example, they can be used to analyse, plot, and document time series (example simple time series), to compute, animate, and document the path of a bouncing ball (example bouncing ball), or to model and display chemical structures (external pageexample molecule structures by Greg Landrum), amongst a variety of others (external pagegallery of JupyterNotebooks). The Notebooks serve several needs:
- computational: from loading and plotting your data to challenging coding needs (in Python, R, Julia, Octave, or OpenModelica)
- text and documentation: write your documents with markdown, including mathematical formulas and symbols using LaTeX
- visualizations and demonstrations: animate figures to illustrate circumstances which can be difficult to grasp
JupyterNotebooks for teaching at ETH
With this broad range of functionality, JupyterNotebooks and the JupyterHub can enrich and facilitate teaching in any discipline. No matter if computational needs are most important, or visualizations, or getting started with the first steps in computing.
Using the ETH JupyterHub, teachers can provide students with interactive exercises and multimedia handouts. And a JupyterNotebook can also serve as a pre-defined coding environment, ready to use. Students will be able to open, modify, save and run JupyterNotebooks on the Hub without having to install anything on their computer. This applies to both pre-defined Notebooks or their own work, in the framework of a course.
A JupyterNotebook can serve as an interactive textbook, accompanying your lecture. Teachers may create static HTML versions of JupyterNotebooks with external pageJupyterBook and upload them to Moodle courses. And with the Binder service, JupyterNotebooks can be published in an executable environment, accessible for anyone.
The JupyterHub at ETH is reachable through a plug-in from Moodle course without further registration needs. In Moodle, an assignment can be created in the form of a downloadable JupyterNotebook. Students will be able to submit their JupyterNotebooks as solutions to assignments. In addition to the standard Moodle functionality of grading and feedback, grading is supported by nbgrader which allows to prepare JupyterNotebooks with automatic grading capabilities.
Getting started
Please contact us at LET via to open the JupyterHub for your course and for any questions regarding JupyterNotebooks and the JupyterHub. We will happily provide technical and didactical support to teachers. And we welcome and encourage everyone who is ready to embark on an exciting journey with us using the JupyterHub.
Do you need funding to incorporate a computational application with JupyterHub/JupyterNotebooks in one of your lectures? Apply now for a grant from the Innovedum Fund.