[GRASS-dev] matplotlib as a wxGUI dep [was: [SoC] Weekly report #3 - GRASS Interactive Scatter Plot Tool]

Michael Barton Michael.Barton at asu.edu
Sat Jul 13 18:39:03 PDT 2013


I agree with all. wxPlot is OK but is limited in the kinds of graphs that can be produced. MatPlotLib could do something that looks like d.histogram but much nicer. It can also make polar coordinate graphs, line plots, and many other things.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity 
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 	480-965-6262 (SHESC), 480-965-8130/727-9746 (CSDC)
fax:          480-965-7671 (SHESC),  480-727-0709 (CSDC)
www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu












On Jul 13, 2013, at 6:14 PM, Hamish <hamish_b at yahoo.com>
 wrote:

> Michael wrote:
>>>> Are you planning to use wxPlot for the scatterplot display? This
>>>> provides a consistent graphical interface with the other plotting
>>>> modules.
> 
> Stepan:
>>> currently matplotlib [1] is used. The library provides events
>>> [2], which allows you to find out where the user clicked. With this
>>> information you are able to keep track of selected regions in a
>>> scatter plot.
> 
> Michael:
>> OK. I originally wanted to use matplotlib because it is a much
>> richer environment, but it was nixed by the dev team because it added
>> a new dependency. If we start using it, it will make numerous,
>> powerful analytical functions accessible.
> 
> 
> Hi,
> 
> fwiw I wouldn't mind matplotlib being around, since I already know the
> Matlab plotting commands very well, and matplotlib is "not dissimilar".
> But I've been quiet on pushing that since I'm sure there are other things
> (java, C++, ruby, whatever) that other grass devs know well, but may not
> be generally in line with the project's needs.  So "+0" from me.
> 
> I don't know much about PyPlot or how the wx profile tool is done, so I
> can't really comment on the pros and cons. Just that
> trunk/scripts/i.spectral/i.spectral.py would be a good test case
> (currently uses gnuplot or d.linegraph).
> 
> Another idea is to replace the d.linegraph module with a python wrapper
> function, and d.histogram too.
> 
> 
> I would add that adding a dependency for the wxGUI is not adding the
> dependency to GRASS, since GRASS can be built and run without the wxGUI.
> 
> 
> 2c,
> Hamish



More information about the grass-dev mailing list