matplotlib-users Mailing List for matplotlib

Hello,
I started off with the Polygon plotting example of the basemap toolkit as I
want to visualize Radar data with GoogleEarth.
I have a polygon shapefile and another data file in which I have one value
for each of the polygons.
I do the basemap.read_shapefile and the axes.add_patch stuff once in the
beginning, keeping a list of the polygon-instances.
Then, for each data file I set the face color of the polygons according to
their respective value and call pyplot.savefig() to save to a png file. This
way I have to do the time and memory consuming task of reading the shapefile
and producing the polygon patches (~40000) only once at the beginning.
That worked splendidly.
Now I wanted some data values to be transparent (so that only rainfall is
actually drawn). So I thought about using set_visible(None) on the
respective polygons.
That seems to work as well, but now I have the problem, that wherever a
polygon becomes invisible, the patch from the previous run remains visible.
So apparently, the figure's canvas (or whatever object is responsible for
the final image pixels) is not initialized between subsequent calls to
savefig.
I searched a bit through the code of figure.py, backend_bases.py,
backend_agg.py but didn't find any method that looked like it would erase
the canvas.
If I do clf() then savefig() always saves the same image over and over
again. I suppose it's because no new data is drawn to the canvas in this
case.
Any ideas, how to solve this?
Thank you for your help
Thomas

Hi,
I'm dealing with a problem how to redraw just label ticks in one certain subplot with in matplotlib. If I change description of axis in the subplot (by .set_yticks() and .set_yticklabels()) I have to redraw whole figure (figure.canvas.draw()) to see the changes.
But I need to redraw either just the one subplot with axis' description or just the descriptions. 
Is it possible to use there something like "blit" techniques? 
Thanks for any advice.
Czenek

On Mon, Oct 27, 2008 at 6:48 PM, Eric Firing <ef...@ha...> wrote:
>> It appears that the figure gets *bigger* as I make a and b *smaller* !??!
>>
>
> You will need to be more explicit about what you are doing to reach this
> conclusion.
What he is probably seeing comes from the fact that some of the figure
elements (line width, font size) are in physical dimensions. As you
make the figure size smaller with figsize, these dimensions are
unchanged, and so they look bigger in proportion to things that scale
proportionately with the figsize, eg the axes area. There is not
built-in scaling of these physical dimensions with figsize.
JDH