On Mon, Feb 7, 2011 at 7:09 PM, Jorge Scandaliaris
<jorgesmbox...@yahoo.es>wrote:
> Jorge Scandaliaris <jorgesmbox-ml@...> writes:
>
> < snip >
> > So, I modified the lasso_demo, increasing progressively the number
> polygons
> > drawn. When I reached 100000 polygons, I was able to reproduce the
> problem.
> > It's
> > true that this a rather large number, but in my code it happens well
> below this
> > number, maybe because I assign a different color and size to each
> polygon. So
> > my
> > previous observation that setting different sizes triggered the problem
> might
> > not be the real picture. It seems as if I was hitting some sort of limit.
> >
> > Anyone would have a clue about what could be happening, or how can I try
> to
> > debug this?
> >
>
> Well, I am bit desperate. The event problem definitely is related to the
> number
> of polygons, circles, ellipses I draw, as well as to using different colors
> and
> sizes. I don't even know how to start debugging this.
>
> One step forward, though, is that I checked running the script directly
> from
> python, instead of within ipython, and it is running as it should. This is
> with
> more than 300 thousand ellipses, whereas within ipython it stops working
> with as
> little as 2 thousand ellipses.
>
> I am running both matplotlib and ipython development trees.
>
> Any expert on how events work out there? And mpl-inpython interaction?
>
> Thanks for any hints/suggestions.
>
> jorges
>
>
>
Hmm, interesting observation. There is very little in mpl that limits your
ability to produce elements for plotting (which is probably why you were
getting shrugs from the mailing list...). However, ipython has various
"tricks" for caching mpl elements, and can sometimes be a bit excessive in
memory usage. This could lead to some issues.
Now that you have identified ipython as the culprit, I would suggest taking
this to the ipython list and seeing if they can better identify the problem
for you.
Ben Root
------------------------------------------------------------------------------
The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE:
Pinpoint memory and threading errors before they happen.
Find and fix more than 250 security defects in the development cycle.
Locate bottlenecks in serial and parallel code that limit performance.
http://p.sf.net/sfu/intel-dev2devfeb
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
