[SciPy-User] ANN: visvis

Almar Klein almar.klein@gmail....
Tue Dec 1 10:02:36 CST 2009

Hi Giovanni,

Thanks for your interesting question.

> This is interesting; but which are the differences with matplotlib/pylab?
> Can you make some examples of functions that are available with vv and maybe
> more difficult to implement with matplotlib?
> I am preparing a seminar on how to plot data with python and I am interested
> in adding references to new alternatives.

The advantage of visvis over matplotlib is that it can plot 3D data.
These can be 3D
lines and points or 3D rendering of volumetric data (for example CT
data). Also,
because it uses OpenGl, displaying images, and zooming/panning them is much
faster in visvis. Visvis is also designed to be extendable. If you
know a bit of OpenGl
you can easily create your own world-objects (I call them wobjects)
and place them
in a scene with other objects.

I tried to make visvis a relatively complete visualization library by
enabling easy
plotting of high quality, with support for labels, title, legend, etc.
However, compared
to matplotlib, visvis 2D plotting capabilities are very limited. I
plan to extend these a
bit, by for example providing support for bar plots, but I do not
intend to go much
further than that.

So you might say that visvis is more focussed at 3D visualization, and
leaves the
advanced 2D visualization to matplotlib. But there is of course some
overlap. Another
thing to note is that both visvis and matplotlib use an object
oriented structure with a
functional interface similar to Matlab, which makes the use of both
libraries very easy
and intuitive.


More information about the SciPy-User mailing list