The IPython Cookbook, second edition, has been published (500+ pages). Most of the contents (text and code) are now freely available on
http://ipython-books.github.io
The recipes cover a wide range of techniques and topics: code optimization, just-in-time compilation, parallel computing, graphics card programming, data analysis and visualization, statistics, machine learning, signal and image processing, dynamical systems... They can be used as a starting point for courses and exercises on Python, Jupyter, and the SciPy stack.
Cyrille