JSPlot is a library with Matplotlib like API, except that the show() command launches a web server and you interact with the plot in the web browser.

Simple example:

>>> from jsplot import plot, show
>>> plot([1, 2, 3], [1, 2, 1])
>>> show()
Server is running at http://127.0.0.1:8000/
Quit the server with CONTROL-C.

More complicated example:

>>> from numpy import arange, pi, sin
>>> from jsplot import plot, show
>>> x = arange(0, 3*pi, 0.01)
>>> plot(x, sin(x*sin(x)), label="sin(x*sin(x))")
>>> plot(x, sin(x), label="sin(x)")
>>> show()
Server is running at http://127.0.0.1:8000/
Quit the server with CONTROL-C.

Going to the url above will show something like the following screenshot, where you can interactively zoom in/out.

https://github.com/certik/jsplot/raw/master/doc/jsplot1.png

To develop the plotserver, do:

cd plotserver
./manage.py runserver

To develop the jsplot itself, just edit files in the jsplot directory.