Jupyter Notebook
Hints and tips for running Jupyter Notebook sessions
When running Jupyter Notebook sessions, Python can behave differently than when run from a script on in a terminal session.
Display Issues
When trying to plot something I get the error:
ImportError: Cannot load backend 'qtagg' which requires the 'qt' interactive framework, as 'headless' is currently running
This occurs because the Jupyter session has no access to a desktop environment with which to open a separate window. There are several solutions to this issue, depending on what your requirements are.
- For displaying plots inline in the notebook, either install the Python module 'matplotlib-inline' and then import matplotlib modules with 'import matplotlib-inline'
or
Put the following in a cell and execute it:%matplotlib inline
or specify the backend in your code:import matplotlib
matplotlib.use('module://matplotlib_inline.backend_inline')
- For writing output to files without displaying use:
matplotlib.use('Agg')