matplotlib-users Mailing List for matplotlib (Page 3)

I have an application with two matplotlib figures, both using the gtkAgg
backend. One or the other is displayed on the screen depending on the
GUI state. Initially after creating the two figures, I can interact with
both of them just fine, but after switching between figures two or three
times the figure becomes sluggish in handling pans and zooms, and I
start getting errors like this while panning:

Traceback (most recent call last):
  File
"/usr/lib/pymodules/python2.6/matplotlib/backends/backend_gtk.py", line
216, in button_release_event
    FigureCanvasBase.button_release_event(self, x, y, event.button,
guiEvent=event)
  File "/usr/lib/pymodules/python2.6/matplotlib/backend_bases.py", line
1219, in button_release_event
    self.callbacks.process(s, event)
  File "/usr/lib/pymodules/python2.6/matplotlib/cbook.py", line 165, in
process
    func(*args, **kwargs)
  File "/usr/lib/pymodules/python2.6/matplotlib/backend_bases.py", line
2016, in release_pan
    a.end_pan()
  File "/usr/lib/pymodules/python2.6/matplotlib/axes.py", line 2458, in
end_pan
    del self._pan_start
AttributeError: _pan_start

I tried to create a minimum working example but so far have been
unsuccessful; the problem did not recur in the simpler program I created.

Can anyone suggest to me where I would look to find what could cause
this error and the (presumably related) performance degradation?

On Friday, December 17, 2010, Nat Echols <nat...@gm...> wrote:
> I updated the version of matplotlib distributed with our group's software to 1.0.0 a week and a half ago (we also use wxPython 2.8.11.0 and Python 2.7).  Since then, multiple users (all on Mac, so far, although we also distribute a Linux version) have been reporting this error:
>
> RuntimeError : Could not open facefile /usr/local/bin/phenix-dev-610/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/mpl-data/fonts/ttf/VeraMono.ttf; Cannot_Open_Resource
>
> The full traceback is appended below.  Googling for this turned up multiple messages relating to older versions of matplotlib, but the only solution I could find was to remove the fontCache file in ~/.matplotlib (and it appears that in some cases removing the entire directory is necessary).  The problem is that this appears to be necessary every time a user upgrades to a new version of our code and removes an older installation, and since we distribute nightly builds, it has become a daily ritual explaining how to fix the problem.  I would really appreciate it if someone can suggest a more elegant solution than making my code nuke ~/.matplotlib.  Alternately, if there is a way to completely prevent the creation or use of ~/.matplotlib, this would make my life much easier; so far this is the only Python module that I've seen that insists on writing to the home directory and crashes if it doesn't get its way.
>
> Or do I need to file a bug ticket?
>
> thanks,
> Nat
>
> Traceback (most recent call last):
>  File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Runtime.py", line 155, in OnUpdate
>    self.propagate_event(event, "OnUpdate")
>  File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Runtime.py", line 211, in propagate_event
>    fn(event)
>  File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py", line 517, in OnUpdate
>    self.stats_plot.update_stats(callback.data)
>  File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py", line 760, in update_stats
>    stats.bs_ave)
>  File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py", line 802, in set_plots
>    self.canvas.draw()
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_wxagg.py", line 59, in draw
>    FigureCanvasAgg.draw(self)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 394, in draw
>    self.figure.draw(self.renderer)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py", line 55, in draw_wrapper
>    draw(artist, renderer, *args, **kwargs)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/figure.py", line 798, in draw
>    func(*args)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py", line 55, in draw_wrapper
>    draw(artist, renderer, *args, **kwargs)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axes.py", line 1934, in draw
>    a.draw(renderer)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py", line 55, in draw_wrapper
>    draw(artist, renderer, *args, **kwargs)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axis.py", line 1017, in draw
>    tick.draw(renderer)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py", line 55, in draw_wrapper
>    draw(artist, renderer, *args, **kwargs)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axis.py", line 234, in draw
>    self.label1.draw(renderer)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py", line 55, in draw_wrapper
>    draw(artist, renderer, *args, **kwargs)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/text.py", line 524, in draw
>    bbox, info = self._get_layout(renderer)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/text.py", line 298, in _get_layout
>    ismath=False)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 180, in get_text_width_height_descent
>    font = self._get_agg_font(prop)
>  File "/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 221, in _get_agg_font
>    font = FT2Font(str(fname))
> RuntimeError: Could not open facefile /usr/local/bin/phenix-dev-610/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/mpl-data/fonts/ttf/VeraMono.ttf; Cannot_Open_Resource
>
>

I think this issue has been totally fixed in the development branch
and I am pretty sure it is in the maintenance branch as well.  Have
you tried the latest svn source in the v1_0_maint branch?  It should
also have many, many other fixes.

Ben Root

On 12/17/2010 6:28 PM, Christoph Gohlke wrote:
> http://www.lfd.uci.edu/~gohlke/pythonlibs/#matplotlib

Wow!
That's quite an offering.

Alan Isaac

On Dec 16, 2010 at 09:24 PM -0800, Robert Field wrote:
> That's what I thought at first too, but imagemagick/graphicsmagick aren't 
> able to do the work. I've found something else to use in the meantime. 
> Thanks,

You could save an additional pdf version with your matplotlib code and 
convert that with imagemagick.

Unofficial matplotlib binaries for Python 2.7 for Windows are available 
at <http://www.lfd.uci.edu/~gohlke/pythonlibs/#matplotlib>.

Christoph

On 12/11/2010 11:55 PM, Abdul Rasheed wrote:
> hi all
> i am new to networkx and Matplotlib. I downloaded networkx1.3 under
> python 2.7.1 for windows xp. when i tried to install matplotlib, it
> requires python 2.6 only. What shall i do in this situation? kindly
> suggest me to have successful installation of matplotlib. Awaiting for
> your replies. thanking you in advance.
>
> */With Warm Wishes and Regards / *
> ____
> A. Abdul Rasheed, M.C.A., M.E., Ph.D.,
> Assistant Professor,
> Department of Computer Applications,
> Valliammai Engineering College,
> SRM Nagar, Kattankulathur - 603 203.
> Kancheepuram District. Tamil Nadu. INDIA.
> Contact: 91 - 44 - 27454784 Ext: 451 (O) / 996 23 000 55
>
>

Ah, imagemagick/graphicsmagick couldn't successfully parse svg, so no conversion was possible. I wound up looking at batik, a java-based solution, which is not fast, but has the benefit of actually working.  

--Rob

On Dec 17, 2010, at 3:11 PM, Alan G Isaac wrote:

> On 12/17/2010 12:24 AM, Robert Field wrote:
>> imagemagick/graphicsmagick aren't able to do the work. I've found something else to use in the meantime.
> 
> It would not be off topic to share your solution with the list.
> 
> Alan Isaac
> 

Sorry for the long delay,

simplest case would be if you can install networkx for py2.6 and the
matplotlib you already have on that py2.6 too.

Other, painful option is to compile matplotlib on WIndows yourself
(but I cannot help with this).

I don't know networkx but I strongly believe that py2.6 is recent enough.

Friedrich

2010/12/12 Abdul Rasheed <pr...@li...>:
> hi all
>   i am new to networkx and Matplotlib. I downloaded networkx1.3 under python
> 2.7.1 for windows xp. when i tried to install matplotlib, it requires python
> 2.6 only. What shall i do in this situation? kindly suggest me to have
> successful installation of matplotlib. Awaiting for your replies. thanking
> you in advance.

On 12/17/2010 12:24 AM, Robert Field wrote:
> imagemagick/graphicsmagick aren't able to do the work. I've found something else to use in the meantime.

It would not be off topic to share your solution with the list.

Alan Isaac

2010/12/15 John Floan <joh...@nt...>:
> I have tried to install matplotlib to our IBM cluster P5(aix5)
> and get this error message (see below):
> The python script output xlC_r xlC_r (2 times) (see end of this email).
> Is that right?
> John
>
> building 'matplotlib.ft2font' extension
> xlC_r xlC_r
> -bI:/usr/local/python/python-2.6.3/lib/python2.6/config/python.exp
> build/temp.aix-5.3-2.6/src/ft2font.o
> build/temp.aix-5.3-2.6/src/mplutils.o
> build/temp.aix-5.3-2.6/CXX/cxx_extensions.o
> build/temp.aix-5.3-2.6/CXX/IndirectPythonInterface.o
> build/temp.aix-5.3-2.6/CXX/cxxsupport.o
> build/temp.aix-5.3-2.6/CXX/cxxextensions.o -L/opt/freeware/lib
> -L/usr/local/lib -lfreetype -lz -lstdc++ -lm -o
> build/lib.aix-5.3-2.6/matplotlib/ft2font.so
> /usr/vacpp/bin/xlC_r: 1501-228 (W) input file xlC_r not found
> error: command 'xlC_r' failed with exit status 252

I think this is a bug in the layer beneath mpl, so neither user nor
matplotlib error.  I'm not at all an IBM cluster expert, I'm not even
a user.

All,
Can anyone tell if he should upgrade maybe distutils or setuptools?  I
don't remember which did what ...

John,
Did you compile numpy yourself?  I strongly believe you did.  In that
case, did the linkage instructions work for that properly?  If yes,
then I might be proven wrong.

Did the commands for compilation (those which creates e.g.
cxx_extensions.o) use xlC_r too?

Friedrich

Hi:

I want to add a legend to a stem plot with two plots. The basic code is:

###########################
from pylab import *

x = [1,2,3,4,5]
y1 = [1, 2, 3, 4, 5]
y2 = [5, 4, 3, 2, 1]

subplot(211)
plot(x, y1, 'rx-')
plot(x, y2, 'bx-')
legend(('a', 'b'))

subplot(212)
stem(x, y1, 'r')
stem(x,y2, 'b')
legend(('a', 'b'))

show()
###########################

The left subplot is the same plot using plot instead of stem. You can
see the result here: http://imagebin.org/128376 .

I expect the legend for the stem case to look similar to the plot
case, however, for the stem, I get the two labels of the legend
associated with the first stem command: 'a' linked to the marker and
'b' linked to the stem line.

Is there any way to fix this? Is this a bug?

Alejandro.

I updated the version of matplotlib distributed with our group's software to
1.0.0 a week and a half ago (we also use wxPython 2.8.11.0 and Python 2.7).
 Since then, multiple users (all on Mac, so far, although we also distribute
a Linux version) have been reporting this error:

RuntimeError : Could not open facefile
/usr/local/bin/phenix-dev-610/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/mpl-data/fonts/ttf/VeraMono.ttf;
Cannot_Open_Resource

The full traceback is appended below.  Googling for this turned up multiple
messages relating to older versions of matplotlib, but the only solution I
could find was to remove the fontCache file in ~/.matplotlib (and it appears
that in some cases removing the entire directory is necessary).  The problem
is that this appears to be necessary every time a user upgrades to a new
version of our code and removes an older installation, and since we
distribute nightly builds, it has become a daily ritual explaining how to
fix the problem.  I would really appreciate it if someone can suggest a more
elegant solution than making my code nuke ~/.matplotlib.  Alternately, if
there is a way to completely prevent the creation or use of ~/.matplotlib,
this would make my life much easier; so far this is the only Python module
that I've seen that insists on writing to the home directory and crashes if
it doesn't get its way.

Or do I need to file a bug ticket?

thanks,
Nat

Traceback (most recent call last):
 File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Runtime.py", line 155, in
OnUpdate
   self.propagate_event(event, "OnUpdate")
 File "/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Runtime.py", line 211, in
propagate_event
   fn(event)
 File
"/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py",
line 517, in OnUpdate
   self.stats_plot.update_stats(callback.data)
 File
"/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py",
line 760, in update_stats
   stats.bs_ave)
 File
"/usr/local/bin/phenix-dev-613/phenix/wxGUI2/Programs/Refine/Output.py",
line 802, in set_plots
   self.canvas.draw()
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_wxagg.py",
line 59, in draw
   FigureCanvasAgg.draw(self)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py",
line 394, in draw
   self.figure.draw(self.renderer)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py",
line 55, in draw_wrapper
   draw(artist, renderer, *args, **kwargs)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/figure.py",
line 798, in draw
   func(*args)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py",
line 55, in draw_wrapper
   draw(artist, renderer, *args, **kwargs)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axes.py",
line 1934, in draw
   a.draw(renderer)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py",
line 55, in draw_wrapper
   draw(artist, renderer, *args, **kwargs)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axis.py",
line 1017, in draw
   tick.draw(renderer)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py",
line 55, in draw_wrapper
   draw(artist, renderer, *args, **kwargs)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/axis.py",
line 234, in draw
   self.label1.draw(renderer)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/artist.py",
line 55, in draw_wrapper
   draw(artist, renderer, *args, **kwargs)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/text.py",
line 524, in draw
   bbox, info = self._get_layout(renderer)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/text.py",
line 298, in _get_layout
   ismath=False)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py",
line 180, in get_text_width_height_descent
   font = self._get_agg_font(prop)
 File
"/usr/local/bin/phenix-dev-613/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py",
line 221, in _get_agg_font
   font = FT2Font(str(fname))
RuntimeError: Could not open facefile
/usr/local/bin/phenix-dev-610/build/mac-intel-osx-x86_64/base/Python.framework/Versions/Current/lib/python2.7/site-packages/matplotlib/mpl-data/fonts/ttf/VeraMono.ttf;
Cannot_Open_Resource

That's what I thought at first too, but imagemagick/graphicsmagick aren't able to do the work. I've found something else to use in the meantime.  Thanks,
Rob

On Dec 16, 2010, at 8:59 PM, Benjamin Root wrote:

> On Friday, December 10, 2010, Robert Field <Rob...@no...> wrote:
>> Newbie here, and trying to wade through this stuff, and it's not coming too quickly.  I'm just trying to take svg data I already have and turn it around into png/pdf/jpg files.  Surely this is not terribly difficult. Any help appreciated!
>> ------------------------------------------------------------------------------
>> Lotusphere 2011
>> Register now for Lotusphere 2011 and learn how
>> to connect the dots, take your collaborative environment
>> to the next level, and enter the era of Social Business.
>> http://p.sf.net/sfu/lotusphere-d2d
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>> 
> Matplotlib is probably not what you want.  For rasterized conversions,
> there is ImageMagick.  Most Linux distros come with 'convert' that you
> can use at the command line. Inkscape is good for visual editing of
> vector data.
> 
> HTH!
> Ben Root

On Friday, December 10, 2010, Robert Field <Rob...@no...> wrote:
> Newbie here, and trying to wade through this stuff, and it's not coming too quickly.  I'm just trying to take svg data I already have and turn it around into png/pdf/jpg files.  Surely this is not terribly difficult. Any help appreciated!
> ------------------------------------------------------------------------------
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
Matplotlib is probably not what you want.  For rasterized conversions,
there is ImageMagick.  Most Linux distros come with 'convert' that you
can use at the command line. Inkscape is good for visual editing of
vector data.

HTH!
Ben Root

On Monday, December 13, 2010, usherbsallen <ste...@us...> wrote:
>
> Hi,
> I am using imshow to illustrate the content of arrays. But I am having
> problems when arrays are too large. There seem to be a limitation that is
> not due to the computer memory size. My data are unsigned integer 1 byte and
> I am limited to array of about 12000 x 5000.
> When trying with a 16000 x 5000 array, my 4 GB computer got very slow and
> never perform the creation of the image. After 10 minutes, I have to kill
> the process. On a 16 GB computer, I've got a segmentation fault message. The
> size limitation are approximately the same on both computers.
>
> I found a solution by calling twice the imshow function with only part of
> the array each time. But I think there might be a better solution.
>
> My feeling is that the limitation is not due to matplotlib or python but
> rather from gtk or other part of the imaging tools.
>
> Steve
> --
> View this message in context: http://old.nabble.com/memory-problem-in-imshow-tp30446084p30446084.html
> Sent from the matplotlib - users mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

What version of mpl are you using?  There was a bug fixed over the
summer regarding seg faults and memory usage.

Ben Root

Sorry for the delay.
This is the script is used (modified so that it include the data).

	



Le 15 déc. 10 à 16:50, Benjamin Root a écrit :

> On Wed, Dec 15, 2010 at 9:46 AM, Benoist Laurent <be...@ib...>  
> wrote:
> Hi all,
>
> I'm still a bit stuck with this probleme of polar annotation.
> Let me present the problem in a different way.
>
> I've got the center of my circle, its radius and even some points on  
> the circle.
> Actually, I'd like to annotate these points (red crosses in the  
> joined picture).
> How would you do that?
>
> My best try gave me the green numbers.
>
> Thanks in advance,
> Ben
>
>
>
> Can you include the source code (if it is simple) that you used to  
> generate this example, and we could probably help you out.
>
> Ben Root
>

On Wed, Dec 15, 2010 at 08:02:42PM -0600, Benjamin Root wrote:

> Assuming you are doing a pcolor on some 2-D numpy array called 'z':
> 
> ax.set_xlim(xmax=z.shape[1])
> ax.set_ylim(ymax=z.shape[0])

Thanks.  I also found (after a bit of digging) that I could just use:

   axis ('off')

Cheers, Ben

Skip Montanaro, on 2010-12-15 12:49,  wrote:
> > Skip,You can call figlegend() and build a legend for the figure,
> > irrespectively of any axes.
> 
> Thanks.  Sounds like exactly what I need.

Hi Skip,
I just wanted to chime in and give you a concrete example of how
you can get all of the line instances from one axes and include
them in the legend of another, or in a figlegend:

  ax = plt.subplot(1,2,1) 
  ax2 = plt.subplot(1,2,2) 
  ax.plot([0,1], label='ax1') 
  ax2.plot([4,3], 'r--',label='ax2') 
  lines =ax.get_lines()
  lines.extend(ax2.get_lines())
  labels = [l.get_label() for l in lines] 
  leg = ax.legend(lines, labels)
  # or, alternatively...
  leg2 = plt.figlegend(lines,labels,loc='center')


this is a slight revision to what was previously discussed here:
<http://old.nabble.com/aligning-multiple-legends-td29638901.html>

-- 
Paul Ivanov
314 address only used for lists,  off-list direct email at:
http://pirsquared.org | GPG/PGP key id: 0x0F3E28F7 

I figured it out, I think. I read the docs (
http://matplotlib.sourceforge.net/api/figure_api.html#matplotlib.figure.Figure.add_axes)
on add_axes and found the part about adding a label to a new axes to force
matplotlib to add a new axes.

So I changed this line from:

subplot = self.figure.add_subplot(n_before + 1, 1, n_before + 1)

to

subplot = self.figure.add_subplot(n_before + 1, 1, n_before + 1, label='%s'
% self.c)

and everything works as advertised.

On Thu, Dec 16, 2010 at 10:48 AM, Åke Kullenberg
<ake...@gm...>wrote:

> I am using matplotlib in wxpython to dynamically add subplots to a figure
> and I am using the geometry_change() method to manage how they are shown.
> Basically I want all subplot to share the space evenly in one column. With
> the simple test code I have pasted in below I have run into some weird
> behavior though.
>
> Two cases, one good and one weird. (launch the wxpython app and do as
> follows to replicate)
>
> Good behavior:
> 1. Right-click in an empty frame.
> 2. Click the 'add subplot' menu item to add a subplot.
> 3. Repeat 1 & 2 to add a second subplot.
> 4. Right-click on top of the LOWER subplot, and click the 'delete' menu
> item to delete that particular subplot.
> 5. Right-click and click the 'add subplot' menu item to add a second
> subplot again.
>
> You will have some debug stuff in the console, but this run does exactly
> what you would expect it to. You first end up with two subplots sharing the
> column 50-50, you delete the LOWER one, and you add a second one again to
> end up in the same situation, i.e., two subplots sharing the colum 50-50.
>
> Weird behavior:
> 1. Right-click in an empty frame.
> 2. Click the 'add subplot' menu item to add a subplot.
> 3. Repeat 1 & 2 to add a second subplot.
> 4. Right-click on top of the UPPER subplot, and click the 'delete' menu
> item to delete that particular subplot.
> 5. Right-click and click the 'add subplot' menu item to add a second
> subplot again.
>
> This time you would of course expect the same behavior as above, but after
> 5 you won't end up with two subplots sharing the column 50-50. Instead you
> have one subplot (which is the most recent one you added) in upper half, and
> the lower half is blank.
>
> Looking at the console after 5 it says this:
>
> add ----------------------------------------
> changing geometry for subplot 0 to (211)
> adding subplot(212)
> count=2 c=3 n_before=1 n_after=1
> --------------------------------------------
>
> So it has changed the geometry as you would expect. It has also added the
> new second subplot as you would expect too. But somehow the axes count
> before and after is the same (n_before and n_after). In other words it looks
> like it hasn't added the subplot with 212, it looks like 211 has been used
> instead.
>
> Am I doing something stupid, or is this a bug os some sort?
>
>
> Comment regarding the code. The Frame class has two attributes outside of
> the usual figures etc, count & c. Both are simple counters. But while count
> is increased/decreased as subplots are added/subtracted, c is only increased
> as subplots are added it is never decreased. I then use c for making a
> simple plot in every subplot. I made it that way so I could tell the
> subplots apart.
>
> Code starts here:
>
> import wx
> import matplotlib
> matplotlib.use('wxagg')
> from matplotlib.figure import Figure
> from matplotlib.backends.backend_wxagg import FigureCanvasWxAgg as
> FigureCanvas
>
> class Frame(wx.Frame):
> def __init__(self, title):
> wx.Frame.__init__(self, None, title=title, pos=(150,150), size=(800,800))
>  self.panel = wx.Panel(self)
> self.figure = Figure()
> self.canvas = FigureCanvas(self, -1, self.figure)
>  self.figure.canvas.mpl_connect('button_press_event', self.OnEvent)
> self.figure.canvas.draw()
>  self.Bind(wx.EVT_SIZE, self.OnSize)
> self.count = 0
> self.c = 0
>  def OnEvent(self, event):
> self.clicked_axes = event.inaxes
> if event.button == 3:
>  id_add_subplot = wx.NewId()
> id_reset = wx.NewId()
> self.Bind(wx.EVT_MENU, self.OnAddSubplot, id=id_add_subplot)
>  self.Bind(wx.EVT_MENU, self.OnReset, id=id_reset)
> self.menu = wx.Menu()
>  self.menu.Append(id_add_subplot, "Add subplot")
> self.menu.Append(id_reset, "Reset")
>  if self.clicked_axes is not None:
> id_delete_subplot = wx.NewId()
> self.Bind(wx.EVT_MENU, self.OnDeleteSubplot, id=id_delete_subplot)
>  self.menu.Append(id_delete_subplot, "Delete")
> self.PopupMenu(self.menu)
>  self.menu.Destroy()
> def OnDeleteSubplot(self, event):
> self.count -= 1
>  self.figure.delaxes(self.clicked_axes)
> n = len(self.figure.axes)
> print ''
>  print 'delete -------------------------------------'
> for i in xrange(n):
>  print 'changing geometry for subplot %s to (%s%s%s)' % (i, n, 1, i + 1)
> self.figure.axes[i].change_geometry(n, 1, i + 1)
>  self.figure.axes[i].update_params()
> self.figure.subplots_adjust()
> self.figure.canvas.draw()
>  print 'count=%s c=%s n=%s' % (self.count, self.c, n)
> print '--------------------------------------------'
>  def OnAddSubplot(self, event):
> self.count += 1
> self.c += 1
>  n_before = len(self.figure.axes)
> print ''
> print 'add ----------------------------------------'
>  for i in xrange(n_before):
> print 'changing geometry for subplot %s to (%s%s%s)' % (i, n_before + 1, 1,
> i + 1)
>  self.figure.axes[i].change_geometry(n_before + 1, 1, i + 1)
> self.figure.axes[i].update_params()
>  print 'adding subplot(%s%s%s)' % (n_before + 1, 1, n_before + 1)
> subplot = self.figure.add_subplot(n_before + 1, 1, n_before + 1)
>  n_after = len(self.figure.axes)
> subplot.plot([self.c,self.c,self.c])
> self.figure.subplots_adjust()
>  self.figure.canvas.draw()
> print 'count=%s c=%s n_before=%s n_after=%s' % (self.count, self.c,
> n_before, n_after)
>  print '--------------------------------------------'
> def OnReset(self, event):
>  self.count = 0
> self.figure.clear()
> self.figure.canvas.draw()
>  def OnSize(self, event):
> pixels = tuple(event.GetSize())
> self.canvas.SetSize(pixels)
>  self.figure.set_size_inches(float(pixels[0])/self.figure.get_dpi(),
> float(pixels[1])/self.figure.get_dpi())
> if __name__ == "__main__":
> app = wx.PySimpleApp()
> frame = Frame('Hello')
>  frame.Show()
> app.MainLoop()
>
>

I am using matplotlib in wxpython to dynamically add subplots to a figure
and I am using the geometry_change() method to manage how they are shown.
Basically I want all subplot to share the space evenly in one column. With
the simple test code I have pasted in below I have run into some weird
behavior though.

Two cases, one good and one weird. (launch the wxpython app and do as
follows to replicate)

Good behavior:
1. Right-click in an empty frame.
2. Click the 'add subplot' menu item to add a subplot.
3. Repeat 1 & 2 to add a second subplot.
4. Right-click on top of the LOWER subplot, and click the 'delete' menu item
to delete that particular subplot.
5. Right-click and click the 'add subplot' menu item to add a second subplot
again.

You will have some debug stuff in the console, but this run does exactly
what you would expect it to. You first end up with two subplots sharing the
column 50-50, you delete the LOWER one, and you add a second one again to
end up in the same situation, i.e., two subplots sharing the colum 50-50.

Weird behavior:
1. Right-click in an empty frame.
2. Click the 'add subplot' menu item to add a subplot.
3. Repeat 1 & 2 to add a second subplot.
4. Right-click on top of the UPPER subplot, and click the 'delete' menu item
to delete that particular subplot.
5. Right-click and click the 'add subplot' menu item to add a second subplot
again.

This time you would of course expect the same behavior as above, but after 5
you won't end up with two subplots sharing the column 50-50. Instead you
have one subplot (which is the most recent one you added) in upper half, and
the lower half is blank.

Looking at the console after 5 it says this:

add ----------------------------------------
changing geometry for subplot 0 to (211)
adding subplot(212)
count=2 c=3 n_before=1 n_after=1
--------------------------------------------

So it has changed the geometry as you would expect. It has also added the
new second subplot as you would expect too. But somehow the axes count
before and after is the same (n_before and n_after). In other words it looks
like it hasn't added the subplot with 212, it looks like 211 has been used
instead.

Am I doing something stupid, or is this a bug os some sort?


Comment regarding the code. The Frame class has two attributes outside of
the usual figures etc, count & c. Both are simple counters. But while count
is increased/decreased as subplots are added/subtracted, c is only increased
as subplots are added it is never decreased. I then use c for making a
simple plot in every subplot. I made it that way so I could tell the
subplots apart.

Code starts here:

import wx
import matplotlib
matplotlib.use('wxagg')
from matplotlib.figure import Figure
from matplotlib.backends.backend_wxagg import FigureCanvasWxAgg as
FigureCanvas

class Frame(wx.Frame):
def __init__(self, title):
wx.Frame.__init__(self, None, title=title, pos=(150,150), size=(800,800))
self.panel = wx.Panel(self)
self.figure = Figure()
self.canvas = FigureCanvas(self, -1, self.figure)
self.figure.canvas.mpl_connect('button_press_event', self.OnEvent)
self.figure.canvas.draw()
self.Bind(wx.EVT_SIZE, self.OnSize)
self.count = 0
self.c = 0
def OnEvent(self, event):
self.clicked_axes = event.inaxes
if event.button == 3:
id_add_subplot = wx.NewId()
id_reset = wx.NewId()
self.Bind(wx.EVT_MENU, self.OnAddSubplot, id=id_add_subplot)
self.Bind(wx.EVT_MENU, self.OnReset, id=id_reset)
self.menu = wx.Menu()
self.menu.Append(id_add_subplot, "Add subplot")
self.menu.Append(id_reset, "Reset")
if self.clicked_axes is not None:
id_delete_subplot = wx.NewId()
self.Bind(wx.EVT_MENU, self.OnDeleteSubplot, id=id_delete_subplot)
self.menu.Append(id_delete_subplot, "Delete")
self.PopupMenu(self.menu)
self.menu.Destroy()
def OnDeleteSubplot(self, event):
self.count -= 1
self.figure.delaxes(self.clicked_axes)
n = len(self.figure.axes)
print ''
print 'delete -------------------------------------'
for i in xrange(n):
print 'changing geometry for subplot %s to (%s%s%s)' % (i, n, 1, i + 1)
self.figure.axes[i].change_geometry(n, 1, i + 1)
self.figure.axes[i].update_params()
self.figure.subplots_adjust()
self.figure.canvas.draw()
print 'count=%s c=%s n=%s' % (self.count, self.c, n)
print '--------------------------------------------'
def OnAddSubplot(self, event):
self.count += 1
self.c += 1
n_before = len(self.figure.axes)
print ''
print 'add ----------------------------------------'
for i in xrange(n_before):
print 'changing geometry for subplot %s to (%s%s%s)' % (i, n_before + 1, 1,
i + 1)
self.figure.axes[i].change_geometry(n_before + 1, 1, i + 1)
self.figure.axes[i].update_params()
print 'adding subplot(%s%s%s)' % (n_before + 1, 1, n_before + 1)
subplot = self.figure.add_subplot(n_before + 1, 1, n_before + 1)
n_after = len(self.figure.axes)
subplot.plot([self.c,self.c,self.c])
self.figure.subplots_adjust()
self.figure.canvas.draw()
print 'count=%s c=%s n_before=%s n_after=%s' % (self.count, self.c,
n_before, n_after)
print '--------------------------------------------'
def OnReset(self, event):
self.count = 0
self.figure.clear()
self.figure.canvas.draw()
def OnSize(self, event):
pixels = tuple(event.GetSize())
self.canvas.SetSize(pixels)
self.figure.set_size_inches(float(pixels[0])/self.figure.get_dpi(),
float(pixels[1])/self.figure.get_dpi())
if __name__ == "__main__":
app = wx.PySimpleApp()
frame = Frame('Hello')
frame.Show()
app.MainLoop()

On Wed, Dec 15, 2010 at 4:08 PM, Ben Elliston <bj...@ai...> wrote:

> I am trying to plot an 840 x 680 element matrix using pcolor.
> matplotlib has decided that the xrange should be 900 and the yrange
> 700.  Is there a way to tell matplotlib to always use x/yranges that
> always match the dimensions of the matrix?
>
> Thanks, Ben
>
>
Assuming you are doing a pcolor on some 2-D numpy array called 'z':

ax.set_xlim(xmax=z.shape[1])
ax.set_ylim(ymax=z.shape[0])

That should make everything look right.  You could also consider using
imshow() which will automatically trim the limits for you.  imshow() and
pcolor() are slightly different, but can often serve similar purposes.

I hope that helps!

Ben Root

I am trying to plot an 840 x 680 element matrix using pcolor.
matplotlib has decided that the xrange should be 900 and the yrange
700.  Is there a way to tell matplotlib to always use x/yranges that
always match the dimensions of the matrix?

Thanks, Ben

hello,

When I save as an "eps" a figure created by matplotlib I face the problem
that the inclined lines in the plot appear to be jagged when I open the
"eps" (please see figure below). This problem doesn't appear when I save the
figure as a pdf or png. Do you you know if there is a simple solution to
this problem that I can implement in my example code (shown below)?
I would greatly appreciate any advice. 
#----example code
from numpy import *
from pylab import *
x=linspace(-1,1,100)
y=x
figure(1)
p1,=plot(x,y,lw=3)
savefig("figeps.eps")
#---jagged line in plot
http://old.nabble.com/file/p30465591/jagged.png 

-- 
View this message in context: http://old.nabble.com/jagged-line-in-eps-from-matplitlib-tp30465591p30465591.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

On Mon, Dec 13, 2010 at 9:10 PM, Xunchen Liu <xun...@gm...> wrote:

> Hello,
>
> I'm plotting some experimental data and found my x axis variable are
> displayed like 1, 2, 3, +1000. It seems depend on how you set a stopper or
> something.
> I'm wondering how to make it display just 1001, 1002, 1003?
>
> thanks!
>

The behavior you are seeing is called "tick offset" and is controlled by the
axis tick formatter.  This can be controlled in different ways, but here is
one approach:


import numpy as np
import matplotlib.pyplot as plt

x = np.arange(1000, 1010)
y = np.random.random((10,))

fig = plt.figure()
ax = fig.gca()

ax.plot(x, y)

# Get the formatter for the major ticks of the x-axis
# and set the 'useOffset' attribute to False.
ax.get_xaxis().get_major_formatter().set_useOffset(False)

plt.show()


I hope that helps!
Ben Root

On Wed, Dec 15, 2010 at 9:46 AM, Benoist Laurent <be...@ib...> wrote:

> Hi all,
>
> I'm still a bit stuck with this probleme of polar annotation.
> Let me present the problem in a different way.
>
> I've got the center of my circle, its radius and even some points on the
> circle.
> Actually, I'd like to annotate these points (red crosses in the joined
> picture).
> How would you do that?
>
> My best try gave me the green numbers.
>
> Thanks in advance,
> Ben
>
>
>
Can you include the source code (if it is simple) that you used to generate
this example, and we could probably help you out.

Ben Root

Hi all,

I'm still a bit stuck with this probleme of polar annotation.
Let me present the problem in a different way.

I've got the center of my circle, its radius and even some points on  
the circle.
Actually, I'd like to annotate these points (red crosses in the joined  
picture).
How would you do that?

My best try gave me the green numbers.

Thanks in advance,
Ben








Le 26 nov. 10 à 14:41, Benoist Laurent a écrit :

> Thank you for your answer.
> I read the text and annotate manual pages.
>
> I don't understand how the "polar"  xycoords/textcoords works.
> I guess I should use this but its not clear to me.
>
>
>
>
> Le 26 nov. 10 à 14:13, Alan G Isaac a écrit :
>
>> On 11/26/2010 8:12 AM, Benoist Laurent wrote:
>>> How can I manage the annotation of points forming a circle if I have
>>> their coordinates?
>>
>> http://matplotlib.sourceforge.net/api/pyplot_api.html#matplotlib.pyplot.annotate
>>
>> http://matplotlib.sourceforge.net/api/pyplot_api.html#matplotlib.pyplot.text
>>
>> http://matplotlib.sourceforge.net/examples/api/joinstyle.html
>>
>> hth,
>> Alan Isaac
>>
>> ------------------------------------------------------------------------------
>> Increase Visibility of Your 3D Game App & Earn a Chance To Win $500!
>> Tap into the largest installed PC base & get more eyes on your game  
>> by
>> optimizing for Intel(R) Graphics Technology. Get started today with
>> the
>> Intel(R) Software Partner Program. Five $500 cash prizes are up for
>> grabs.
>> http://p.sf.net/sfu/intelisp-dev2dev
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>
>
> ------------------------------------------------------------------------------
> Increase Visibility of Your 3D Game App & Earn a Chance To Win $500!
> Tap into the largest installed PC base & get more eyes on your game by
> optimizing for Intel(R) Graphics Technology. Get started today with  
> the
> Intel(R) Software Partner Program. Five $500 cash prizes are up for  
> grabs.
> http://p.sf.net/sfu/intelisp-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>