This post started as a personal log to help me remember how to perform the installation, until I thought it could help other people. Or help myself in a future in which I totally forget how to do it.
Working with a virtual environment is always a good idea. The pythono virtual environment wrapper is particularly helpful.
mkvirtualenv mayavi-plots
Download sip:
https://riverbankcomputing.com/software/sip/download
Download PyQt:
http://pyqt.sourceforge.net/Docs/PyQt4/installation.html#downloading-pyqt4
Go into the directory where you have downloaded the sip and PyQt:
tar xvf sip-4.19.8.tar.gz
cd sip-4.19.8
python configure.py
make
sudo make install
Note that the python that runs configure.py is that of the virtual environment), as we can see from the next command:
which python
/home/ctroupin/Software/PythonEnvs/mayavi/bin/python
Process in a similar way for PyQt (the compilation can take a few minutes):
tar xvf PyQt4_gpl_x11-4.12.1.tar.gz
cd PyQt4_gpl_x11-4.12.1
python configure-ng.py
make
make install
Both packages will be installed in the mayavi-plots virtalenv.
With the version installed with pip,
pip install mayavi
I got the issue related to undefined ‘magnification’ attribute of a ‘WindowToImageFilter’, so I made the installation using setup.py:
pip_uninstall mayavi
git clone git@github.com:enthought/mayavi.git
python setup.py install
Jupyter notebook 5.0 has an iopub.data_rate set to 10^7
Solution: in bash, run
jupyter-notebook --NotebookApp.iopub_data_rate_limit=1.0e10