[IAEP] Working math graphs/plots

[Gary C Martin]

I've been poking about for something like gnuplot, something python  
friendly, with an idea to wrap it up into a simple Sugar activity to  
allow basic (and advanced, but that would be not primary to the UI)  
graph/plot experimentation. Now most of the things I've looked as so  
far hit some road bump, usually needing full compile environment or  
some weird dependancy I couldn't sort out, but yesterday I tried  
matplotlib:

	http://matplotlib.sourceforge.net/

All you need is an internet connection, about 5Mb space, and not to be  
scared of Terminal. Here's how you download and install it (tested on  
an XO running 8.2-767):

	sudo yum install python-matplotlib

Python can now be used to generate all sorts of graphs, the quick way  
to start experimenting is via ipython:

	ipython -pylab

And then try this for starters:

	x = randn(1000)
	hist(x, 100)

You'll be presented with the graph (10,000 gausian random numbers in a  
histogram with 100 bins) in full screen (alt-esc or use frame icon to  
close). Take a look at their gallery to see a wide range of much more  
exciting examples:

	http://matplotlib.sourceforge.net/gallery.html

I'll probably experiment with this for a few weeks and then start  
thinking about a simple/clean Sugar UI to expose the needed features.  
I'm posting to IAEP incase anyone wants to chime in with some use  
cases for our target demographics.

--Gary

P.S. I'm open to other (working) graph tool suggestions, if I've  
missed something obvious.