A Java/swing backend for Matplotlib using the javabridge

Many of the Python UI toolkits - wx, qt and similar, have a matplotlib backend that displays plots on screen using the toolkit. Now that the python-javabridge is bidirectional, we have all the pieces we need to use AWT and Swing as a Python toolkit, arguably the most complete, platform-independent one there is. It turns out that, with agg, it's not hard to do, so here it is.

The backend is fully-functional, probably with beta quality. To use with matplotlib.pyplot or pylab, set your backend to 'module://backend_swing':

import matplotlib
matplotlib.use('module://backend_swing')
from pylab import *
import javabridge
javabridge.start_vm()
plot([0, 1, 2, 3, 4], [0, 1, 4, 9, 16])
get_current_fig_manager().show()

First install matplotlib and the javabridge (if you have not done so already; on Linux, you might have to install libpng12-dev and libfreetype6-dev first):

pip install --user matplotlib
pip install --user git+https://github.com/CellProfiler/python-javabridge

Then you can run the demo:

PYTHONPATH=. python demo/canvas_demo.py