matplotlib-users Mailing List for matplotlib

On Mon, Feb 14, 2011 at 5:30 PM, John Hunter <jd...@gm...> wrote:
> On Sun, Feb 13, 2011 at 11:45 PM, Tom Flannaghan <tj...@ca...> wrote:
>
> > It would also be helpful if anyone has suggestions on a particular issue
> I had.
> > Currently, to plot variable-width lines (i.e. streamlines2.png) I use a
> plot
> > command for each line segment which is very slow and nasty. Is there a
> better
> > way I'm missing?
>
> You probably want to use a compound path (one object for the entire
> plot). See the tutorial at
> http://matplotlib.sourceforge.net/users/path_tutorial.html, in
> particular the compound path for the histogram example near the end,
> and let me know if you have any questions.
>
> JDH
>
>
Personally, I am more a fan of Ray's version, although Tom's version would
integrate more nicely with the current mpl codebase. There are some nice
features with Ray's version such as allowing to choose density in both x and
y directions.
I guess the major question is which style do we like better? Maybe we could
use both of these code bases to come up with a nice, generalized version?
Ben Root

On Sun, Feb 13, 2011 at 11:45 PM, Tom Flannaghan <tj...@ca...> wrote:
> It would also be helpful if anyone has suggestions on a particular issue I had.
> Currently, to plot variable-width lines (i.e. streamlines2.png) I use a plot
> command for each line segment which is very slow and nasty. Is there a better
> way I'm missing?
You probably want to use a compound path (one object for the entire
plot). See the tutorial at
http://matplotlib.sourceforge.net/users/path_tutorial.html, in
particular the compound path for the histogram example near the end,
and let me know if you have any questions.
JDH

On 2/13/11 10:45 PM, Tom Flannaghan wrote:
> Hi,
>
> I've written a script to roughly emulate the elegant streamline plots found in
> Mathematica. The code is available at
> http://www.atm.damtp.cam.ac.uk/people/tjf37/streamplot.py and example plots at
> http://www.atm.damtp.cam.ac.uk/people/tjf37/streamlines1.png and
> streamlines2.png. It's a pretty hacky script, but fast and fairly robust. If
> anyone finds this script useful and has comments/suggestions, I'm happy to do a
> bit more work.
>
> It would also be helpful if anyone has suggestions on a particular issue I had.
> Currently, to plot variable-width lines (i.e. streamlines2.png) I use a plot
> command for each line segment which is very slow and nasty. Is there a better
> way I'm missing?
>
> Tom
Tom: This is really nice! I'd like to see some version of your code 
incorporated into matplotlib. Regarding your question about 
variable-width lines, I don't know of any way to do that - but perhaps 
someone with more detailed knowledge of matplotlib internals will comment.
-Jeff
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

Thank you for your help. I upgraded to the latest development version, 
and as you said, memory use dropped a ton. I will have to test more to 
confirm that the problem is completely gone, but this appears to bring 
memory usage down to something quite manageable (at least on my 8gb box 
...).
Tom
On 02/09/2011 07:30 PM, Robert Abiad wrote:
> Tom,
>
> I just went through this, though with version 1.01 of mpl, so it may be different. You can read the
> very long thread at:
>
> http://www.mail-archive.com/mat...@li.../msg20031.html
>
> Those who maintain mpl don't think there is a memory leak. What I found was that imshow() does
> consume a lot of memory (now fixed in the development version) and that the first 2 or so uses build
> on each other, but after that it levels off giving back memory after close(). There is a
> discrepancy between what python reports it's using and what the OS reports (I had 500MB from the OS,
> but only 150MB from python). There is a chance that ipython is caching your results (try ipython
> -pylab -cs 0), but when I ran without ipython, python still had a large portion of memory.
>
> -robert
>
> On 2/9/2011 3:52 PM, Tom Dimiduk wrote:
>> I am using matplotlib pylab in association with ipython -pylab to show
>> many large (~2000x2000 or larger) images. Each time I show another
>> image it consumes more memory until eventually exhausting all system
>> memory and making my whole system unresponsive.
>>
>> The easiest way to replicate this behaviour is with
>> a = ones((3333,3333))
>> imshow(a)
>>
>> optionally
>>
>> close()
>>
>> and then
>>
>> imshow(a)
>>
>> again. I am using ipython .10.1 and matplotlib 0.99.3. Is there
>> something I should be doing differently to avoid this problem? Is it
>> fixed in a later version?
>>
>> Thanks,
>> Tom
>>
>> ------------------------------------------------------------------------------
>> 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
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
> ------------------------------------------------------------------------------
> 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
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Aman Thakral, on 2011年02月09日 15:40, wrote:
> Hi,
> 
> I was just wondering, how would I go about redrawing only a single element
> (i.e. an artist)? I know that matplotlib.artist.Artist has a draw() method
> that accepts a renderer as the argument.
> 
> I tried something along the lines of:
> 
> renderer = fig.canvas.get_renderer()
> title = ax.set_title("This is my title")
> title.draw(renderer)
> 
> But have not had any luck. I'm not sure what I'm missing. Any help would
> be greatly appreciated.
Hi Aman,
To see the affects of rendering something new, you can use:
 ax.figure.canvas.blit(ax.title.get_bbox_patch())
If you intend to change the title several times, however, the
above will just keep over-plotting on top of the previous titles.
In that case, you'll want to use the animated=True keyword and
canvas.restore_region along with the blitting -- see the
animation examples for how to do that.
best,
-- 
Paul Ivanov
314 address only used for lists, off-list direct email at:
http://pirsquared.org | GPG/PGP key id: 0x0F3E28F7 

Bartosz Telenczuk, on 2011年02月13日 23:16, wrote:
> Hi,
> 
> I am writting an application in which I update dynamically the state 
> of the figures. When I am done I call the Figure.canvas.draw function 
> to redraw the figure. However, when the window was previously closed 
> by the user the update can be skipped.
> 
> Is there a way to tell from a figure handle if the figure is shown on 
> screen or not?
> 
> Currently, I solve this problem with a callback which sets the figure 
> handle to None on figure close event. However, I guess there might be 
> a more direct way of doing that. What would you suggest?
> 
> Thanks a lot for your help.
Hi Bartosz,
I don't think there's currently a backend independent way of
doing it outside of the approach you have taken. 
Having a quick look using f = plt.gcf() before and after closing
the figure, with a WX backend, the f.canvas object changes
to an instance of <class 'wx._core._wxPyDeadObject'>, and loses
all of its usual methods. For GTK - looks like f.canvas.window
gets set to None. 
So if you're only targeting one backend, there might be some
simple way like this to find out if a figure is being shown using
from the figure handle.
best,
-- 
Paul Ivanov
314 address only used for lists, off-list direct email at:
http://pirsquared.org | GPG/PGP key id: 0x0F3E28F7 

After searching for the path in all files in the distro,
I found the problem:
 font caching of Matplotlib
After removing C:\Documents and Settings\Administrator\.matplotlib
everything works ok.
So the remaining question is,
is there an elegant way, either in py2exe or in matplotlib,
to get arounf this caching problem ?
thanks,
Stef Mientki
On 14-02-2011 21:41, Stef Mientki wrote:
> Ran into the same problem,
> are there any hints to track down the problem.
>
> I tried to remove the MatPlotLib from py2exe and added MatPlatlib manual afterwards,
> but still it point to the wrong directory.
>
> thanks,
> Stef Mientki
>
>
> On 26-12-2010 15:18, zb wrote:
>> Hi.
>>
>> First of all, I would like to thank the developers of this wonderful package.
>>
>> I started doing an app and I finally tried to compile it with py2exe. I run into some issues that I posted in the py2exe group here;
>>
>> http://sourceforge.net/mailarchive/forum.php?thread_name=ienv6p%24b42%241%40dough.gmane.org&forum_name=py2exe-users
>>
>> To make it work, I ended it up having to use an older version of matplotlib and numpy. I don't know if the developers of matplotlib are aware of these issues and I am posting here just in case they don't know.
>>
>> When I was trying to use version 1.0, I run into problems with the mpl data directory. The output of py2exe could only find the mpl-data dir at the root /dist of a drive. Then I decided to try a much earlier version of matplotlib, and I run into different set of problems. Finally, I tried matplotlib 0.99.1 and the problems went away (I didn't try version 0.99.3 because its release date was too close to the version 1.0).
>>
>> I am just trying to make the developers aware of the problems and I hope to use newer versions in the future.
>>
>> Thanks for your attention
>>

On 2/14/2011 12:41 PM, Stef Mientki wrote:
> Ran into the same problem,
> are there any hints to track down the problem.
>
> I tried to remove the MatPlotLib from py2exe and added MatPlatlib manual afterwards,
> but still it point to the wrong directory.
>
> thanks,
> Stef Mientki
>
>
> On 26-12-2010 15:18, zb wrote:
>> Hi.
>>
>> First of all, I would like to thank the developers of this wonderful package.
>>
>> I started doing an app and I finally tried to compile it with py2exe. I run into some issues that I posted in the py2exe group here;
>>
>> http://sourceforge.net/mailarchive/forum.php?thread_name=ienv6p%24b42%241%40dough.gmane.org&forum_name=py2exe-users
>>
>> To make it work, I ended it up having to use an older version of matplotlib and numpy. I don't know if the developers of matplotlib are aware of these issues and I am posting here just in case they don't know.
>>
>> When I was trying to use version 1.0, I run into problems with the mpl data directory. The output of py2exe could only find the mpl-data dir at the root /dist of a drive. Then I decided to try a much earlier version of matplotlib, and I run into different set of problems. Finally, I tried matplotlib 0.99.1 and the problems went away (I didn't try version 0.99.3 because its release date was too close to the version 1.0).
>>
>> I am just trying to make the developers aware of the problems and I hope to use newer versions in the future.
>>
>> Thanks for your attention
>>
>
I can not reproduce this problem with matplotlib 1.0.1, PyQt 4.8.3, 
py2exe-0.6.10dev, Python 2.7.1. The only thing I noticed is that some 
toolbar icons are not displayed.
Christoph

Ran into the same problem,
are there any hints to track down the problem.
I tried to remove the MatPlotLib from py2exe and added MatPlatlib manual afterwards,
but still it point to the wrong directory.
thanks,
Stef Mientki
On 26-12-2010 15:18, zb wrote:
> Hi.
>
> First of all, I would like to thank the developers of this wonderful package.
>
> I started doing an app and I finally tried to compile it with py2exe. I run into some issues that I posted in the py2exe group here;
>
> http://sourceforge.net/mailarchive/forum.php?thread_name=ienv6p%24b42%241%40dough.gmane.org&forum_name=py2exe-users
>
> To make it work, I ended it up having to use an older version of matplotlib and numpy. I don't know if the developers of matplotlib are aware of these issues and I am posting here just in case they don't know.
>
> When I was trying to use version 1.0, I run into problems with the mpl data directory. The output of py2exe could only find the mpl-data dir at the root /dist of a drive. Then I decided to try a much earlier version of matplotlib, and I run into different set of problems. Finally, I tried matplotlib 0.99.1 and the problems went away (I didn't try version 0.99.3 because its release date was too close to the version 1.0).
>
> I am just trying to make the developers aware of the problems and I hope to use newer versions in the future.
>
> Thanks for your attention
>

Have you looked into Celery? It is a queuing system with Django ORM
support. I don't have any experience with it myself, but I have heard
good things about it.
-Aman
2011年2月14日 wukan <wek...@gm...>:
> Hi ,When I use matplotlib to draw 2D graphics in django web site,I encounter
> a problem.
> when server users use matplotlib to draw graphics simultaneously will
> cause website collapse.
> when one user use matplotlib to draw graphics will have no problem.
>
> i suppose matplotlib doesnot support multithreading drawing.
> so i put the draw function in a thread . it can't work as before.
>
> so why matplotlib doesnot support multithreading drawing?
> How to solve this problem.
> Hope I can receive help from you.
>
> Regards
>
> wekay
> ------------------------------------------------------------------------------
> 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
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

Hi ,When I use matplotlib to draw 2D graphics in django web site,I encounter
a problem.
when server users use matplotlib to draw graphics simultaneously will
cause website collapse.
when one user use matplotlib to draw graphics will have no problem.
i suppose matplotlib doesnot support multithreading drawing.
so i put the draw function in a thread . it can't work as before.
so why matplotlib doesnot support multithreading drawing?
How to solve this problem.
Hope I can receive help from you.
Regards
wekay

On Mon, Feb 14, 2011 at 8:40 AM, Nils Wagner
<nw...@ia...> wrote:
> Hi all,
>
> Is it possible to apply griddata to polar coordinates or
> do I need cartesian coordinates ?
>
> http://matplotlib.sourceforge.net/api/mlab_api.html#matplotlib.mlab.griddata
You can keep the data in polar coordinates, you just need to pass in
the locations of the points in cartesian coordinates:
x = r * cos(theta)
y = r * sin(theta)
Ryan
-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma

Dear Matplotlib folks,
Those of you whose duties include teaching basic stats
might be interested in these interactive tutorial files,
designed to illustrate basic concepts.
Running the code opens up an interactive figure window,
using Matplotlib for the GUI and the plots.
When you click on a figure to add new points,
the statistical tests shown in the figure change accordingly.
http://www.dartmouth.edu/~raj/intro-stats.html
The code has lots of comments in it, which attempt to explain the
concepts as explicitly as possible. No prior knowledge of Python or
statistics is assumed. These programs require the SciPy module,
in addition to Matplotlib.
The tutorials are:
- interactive_mean_std_normal_distribution.py
- interactive_one_sample_t_test.py
- interactive_two_sample_t_test.py
- interactive_correlation_plot.py
The same webpage also contains Matlab versions of the scripts.
Please feel more than free to use any of the code for teaching,
if you find it useful.
Yours,
Rajeev Raizada
Research Assistant Professor
Neukom Institute for Computational Science
Dartmouth College
HB 6255
Hanover NH 03755
Tel: 603 646 0175
E.mail: raj...@da...
WWW: http://www.dartmouth.edu/~raj

Hi all,
Is it possible to apply griddata to polar coordinates or 
do I need cartesian coordinates ?
http://matplotlib.sourceforge.net/api/mlab_api.html#matplotlib.mlab.griddata
Nils

Feel free to 'save and run', pass along, or ignore.
This was my valentine's day present today.
I hope the bandwidth amuses more than it annoys...
Jules
#----------------------------------
# hohumheart.py
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
from matplotlib.patches import Polygon
# force square figure and square axes looks better for polar, IMO
width, height = mpl.rcParams['figure.figsize']
size = min(width, height)
# make a square figure
fig = plt.figure(figsize=(size, size))
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8], polar=True, axisbg='#ffffff')
ax.set_rmax(2.0)
ax.set_xticks([])
ax.set_yticks([])
plt.grid(False)
theta = np.linspace(0,1,100)*np.pi*2
r = 1*(1-np.cos(theta))
ncards = 5
step = 2*np.pi/ncards
for ii in range(ncards):
	tr = np.column_stack((theta+ii*step, r))
	cpatch = Polygon(tr)
	cpatch.set_facecolor('r')
	cpatch.set_edgecolor('k')
	cpatch.set_alpha(0.5)
	ax.add_patch(cpatch)
ax.set_title("I $\heartsuit$ you", fontsize=20)
plt.show()
#-----------------------

Hi,
I have some problems to plot a 3d plot_surface (and contour plot) in log scale (y and z or x,y and z).
There is nothink in the help sections of thus plot to plot them in log scale (neither in thus plot code commentary).
I tried to find a solution by myself (many try as "log=True", "xscale='log'", "scale='log'", or "set_xscale('log')"...) but it doesn't work.
I tried to use "LogFormatter" (ticker) but i failed (maybe I am doing something wrong (I am a newbe) but i tried many things).
I try to found a solution on forums and this mailing list and what i found nearest of my problem is this topic:
http://stackoverflow.com/questions/3909794/plotting-mplot3d-axes3d-xyz-surface-plot-with-log-scale‏
They said this problem had no solution. Is it true? have somebody solved this problem?
An other solution is to plot log(Data) and change ticks with "set_ticks" (for exemple) but it seems to only work with 2D plot. I tried too :(
To be unterstanding easier, this is an exemple of the code to plot in log scale:
 import matplotlib.pyplot as plt# (i am under '1.0.0')
 import numpy as np
 #Data
 x = np.logspace(0, 4, 10)
 y = np.logspace(0, 4, 10)
 Z = np.arange(100).reshape((10,10))
 #Min and Max for x,y and Z
 minx=np.min(x)
 maxx=np.max(x)
 miny=np.min(y)
 maxy=np.max(y)
 minz=np.min(Z)
 maxz=np.max(Z)
 #figure
 fig = plt.figure()
 ax = fig.gca(projection='3d')
 ax.plot_surface(X, Y, Z, rstride=8, cstride=8, alpha=0.3)
 cset = ax.contour(X, Y, Z, zdir='z', offset=minz)
 cset = ax.contour(X, Y, Z, zdir='x', offset=minx)
 cset = ax.contour(X, Y, Z, zdir='y', offset=miny)
 ax.set_xlabel("X")
 ax.set_xlim3d(minx,maxx)
 ax.set_ylabel("Y")
 ax.set_ylim3d(miny,maxy)
 ax.set_zlabel("Z")
 ax.set_zlim3d(minz,maxz)
 fig.savefig("plot00.png")
Has someone got a solution?
Thanks!
Regards,
Gaël
 		 	 		 

That solution might be fine for static plots, but I my case I prefer my
solution.
My plot is quite interactive so using your solution causes many problems
with my code.
-- 
View this message in context: http://old.nabble.com/onpick-on-a-2-y-plot-%28-via-twinx%28%29-%29-seems-to-only-allow-picking-of-second-axes%27s-artists-tp25049128p30919747.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Hi,
I've written a script to roughly emulate the elegant streamline plots found in 
Mathematica. The code is available at 
http://www.atm.damtp.cam.ac.uk/people/tjf37/streamplot.py and example plots at 
http://www.atm.damtp.cam.ac.uk/people/tjf37/streamlines1.png and 
streamlines2.png. It's a pretty hacky script, but fast and fairly robust. If 
anyone finds this script useful and has comments/suggestions, I'm happy to do a 
bit more work.
It would also be helpful if anyone has suggestions on a particular issue I had. 
Currently, to plot variable-width lines (i.e. streamlines2.png) I use a plot 
command for each line segment which is very slow and nasty. Is there a better 
way I'm missing?
Tom