matplotlib-users Mailing List for matplotlib

I eventually got it to mostly work using:
³sudo port install py26-matplotlib @1.0.1 +gtk2²

It doesn't match the correct egg format it seems, but it does work.


On 8/10/11 11:01 AM, "David Just" <Jus...@ma...> wrote:

> Hello,
> 
> I¹ve been trying to install matplotlib correctly for the past couple of days.
> 
> I¹m using a macports python2.6 built with +universal on.
> 
> If I install with the the downloadable .egg file I end up with a message that
> states that _path.so has problems. (built for 32bit I guess)
> ImportError: 
> dlopen(/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2
> .6/site-packages/matplotlib-1.0.1_r0-py2.6.egg/matplotlib/_path.so, 2): no
> suitable image found.  Did find:
>     
> /opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site
> -packages/matplotlib-1.0.1_r0-py2.6.egg/matplotlib/_path.so: no matching
> architecture in universal wrapper
> 
> 
> If I install with easy_install-2.6 matplotlib  from source I get the
> following:
>     ....
>     Best match: matplotlib 0.91.1
>     ....
>     src/_image.cpp:952: error: Œpng_infopp_NULL¹ was not declared in this
> scope    
>     error: Setup script exited with error: command '/usr/bin/gcc-4.2' failed
> with exit status 1
> 
> If I install with port install matplotlib it fails because the install in not
> the correct path structure in the site-packages folder.
> 
> 
> I¹ve tried searching the web and have found some indications that I need to
> install 32bit python 2.6  but cannot find how to do that, and all of my
> production machinese have 64bit python installed on them.
> 
> 
> Any help at this point would be appreciated.
> Thanks. 
> 
> 
> 
> 

I have an array of images stored as an array of numpy arrays.   I need to be
able to efficiently scroll through that set of images.   My first attempt at
doing this goes something like this:

--init--

        self.ax = pyplot.imshow(imgdta[0], interpolation='spline36',
cmap=cm.gray, picker=True)      # draw the plot @UndefinedVariable
        pyplot.axes().set_axis_off()
        self.fig = self.ax.get_figure()
        self.canvas = FigureCanvasGTKAgg(self.fig)

--onscroll--
        self.ax.set_array(imdta[n]) # 0 < n < num_images
        self.canvas.draw()


This method of changing the image data does not seem to be very preferment.
It takes ~.25 seconds to go from one image to the next.   Can anybody
suggest a faster way?  This also ends up in a canvas that¹s much larger than
I need, is there a better way to define my view area?


Thank you,
Dave. 

On Wed, Aug 10, 2011 at 1:36 PM, Jay Wineinger <jay...@gm...>wrote:

> Hi everyone,
>
> I'm trying to construct a radar plot for 5 variables.  I've found the
> example code (radar_chart.py) for doing this with matplotlib but I'm having
> trouble adjusting one aspect of the rendering.  That sample file produces
> radar plots with  circular frames and ticks (see
> http://wineinger.me/plotting/radar_chart1.png).  That sample code does
> seem to have an option to use a polygon frame rather than a circle (the
> default), but this doesn't seem to do much except erase part of the outer
> tick lines (http://wineinger.me/plotting/polygon.png vs
> http://wineinger.me/plotting/circle.png).
>

Originally, this example did produce the polygon frame correctly.
Improvements to matplotlib have, however, broken this example. If you plot
this on a non-white background, you'll see that there is indeed a polygon
patch for the axes. Nevertheless the border (i.e. "spine") of the plot is
circular because of improvements that separated the spine from the patch.
The example code only overrides PolarAxes._gen_axes_patch, but should also
override PolarAxes._gen_axes_spines to fix this issue. Unfortunately, I
don't have time to work on this.


> Plus, I'm looking for the tick lines to parallel the frame, so having a
> polygon frame with circular ticks would not be useful to me.
>

I thought the grid was correctly rendered as polygons instead of circles
(when choosing a polygon frame). I don't, however, see anything in the
example code to change the grid lines, so my memory might be wrong.

Sorry I couldn't be more help.
-Tony

Hi Ben,

I will try to do it. Once it is done I let you know.

Cheers,
marc

On Wed, Aug 10, 2011 at 7:41 PM, Benjamin Root <ben...@ou...> wrote:
>
>
> On Wed, Aug 10, 2011 at 12:32 PM, Marc Magrans de Abril
> <mar...@gm...> wrote:
>>
>> Dear colleagues,
>>
>> I'm trying to display an histogram with logarithmic bins using the
>> logarithmic scale in the x-axis. When I execute the attached python
>> script it crashes giving a "Bus Error" message (see also the attached
>> core dump).
>>
>> If I remove the line "plt.xscale('log')" the script goes up to the end.
>>
>> I'm running the program on MacOS 10.6.8, Python 2.5, numpy 1.2.1, and
>> matplotlib 0.98.5.2.
>>
>> Anyone knows the reason of the crash?
>>
>> Thanks,
>> marc
>>
>
> If I remember correctly, there was a log-scale bug in plt.hist() that got
> fixed a little while back.  Your version of numpy and matplotlib is a little
> old.  Is it possible to update those two packages on your system?
>
> Ben Root
>
>

Hi Ben,

I will try to do it. Once it is done I let you know.

Cheers,
marc


On Wed, Aug 10, 2011 at 7:41 PM, Benjamin Root <ben...@ou...> wrote:
>
>
> On Wed, Aug 10, 2011 at 12:32 PM, Marc Magrans de Abril
> <mar...@gm...> wrote:
>>
>> Dear colleagues,
>>
>> I'm trying to display an histogram with logarithmic bins using the
>> logarithmic scale in the x-axis. When I execute the attached python
>> script it crashes giving a "Bus Error" message (see also the attached
>> core dump).
>>
>> If I remove the line "plt.xscale('log')" the script goes up to the end.
>>
>> I'm running the program on MacOS 10.6.8, Python 2.5, numpy 1.2.1, and
>> matplotlib 0.98.5.2.
>>
>> Anyone knows the reason of the crash?
>>
>> Thanks,
>> marc
>>
>
> If I remember correctly, there was a log-scale bug in plt.hist() that got
> fixed a little while back.  Your version of numpy and matplotlib is a little
> old.  Is it possible to update those two packages on your system?
>
> Ben Root
>
>

On Wed, Aug 10, 2011 at 12:32 PM, Marc Magrans de Abril <
mar...@gm...> wrote:

> Dear colleagues,
>
> I'm trying to display an histogram with logarithmic bins using the
> logarithmic scale in the x-axis. When I execute the attached python
> script it crashes giving a "Bus Error" message (see also the attached
> core dump).
>
> If I remove the line "plt.xscale('log')" the script goes up to the end.
>
> I'm running the program on MacOS 10.6.8, Python 2.5, numpy 1.2.1, and
> matplotlib 0.98.5.2.
>
> Anyone knows the reason of the crash?
>
> Thanks,
> marc
>
>
If I remember correctly, there was a log-scale bug in plt.hist() that got
fixed a little while back.  Your version of numpy and matplotlib is a little
old.  Is it possible to update those two packages on your system?

Ben Root

Hi everyone,

I'm trying to construct a radar plot for 5 variables.  I've found the
example code (radar_chart.py) for doing this with matplotlib but I'm having
trouble adjusting one aspect of the rendering.  That sample file produces
radar plots with  circular frames and ticks (see
http://wineinger.me/plotting/radar_chart1.png).  That sample code does seem
to have an option to use a polygon frame rather than a circle (the default),
but this doesn't seem to do much except erase part of the outer tick lines (
http://wineinger.me/plotting/polygon.png vs
http://wineinger.me/plotting/circle.png).  Plus, I'm looking for the tick
lines to parallel the frame, so having a polygon frame with circular ticks
would not be useful to me.

Here is an example of what I'd like to produce:
http://www.corda.com/docsource/doc7/Manuals/project_graphics/design_graphs_radar_transparency.gif.
 Can anyone provide some sample code to get me started or at least some
direction on how to accomplish this?  I'm new to matplotlib though I do have
a fair bit of python experience.

thanks!
Jay

Dear colleagues,

I'm trying to display an histogram with logarithmic bins using the
logarithmic scale in the x-axis. When I execute the attached python
script it crashes giving a "Bus Error" message (see also the attached
core dump).

If I remove the line "plt.xscale('log')" the script goes up to the end.

I'm running the program on MacOS 10.6.8, Python 2.5, numpy 1.2.1, and
matplotlib 0.98.5.2.

Anyone knows the reason of the crash?

Thanks,
marc

Dear colleagues,

I'm trying to display an histogram with logarithmic bins using the
logarithmic scale in the x-axis. When I execute the attached python
script it crashes giving a "Bus Error" message (see also the attached
core dump).

If I remove the line "plt.xscale('log')" the script goes up to the end.

I'm running the program on MacOS 10.6.8, Python 2.5, numpy 1.2.1, and
matplotlib 0.98.5.2.

Anyone knows the reason of the crash?

Thanks,
marc

Hello,

I¹ve been trying to install matplotlib correctly for the past couple of
days.

I¹m using a macports python2.6 built with +universal on.

If I install with the the downloadable .egg file I end up with a message
that states that _path.so has problems. (built for 32bit I guess)
ImportError: 
dlopen(/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/pytho
n2.6/site-packages/matplotlib-1.0.1_r0-py2.6.egg/matplotlib/_path.so, 2): no
suitable image found.  Did find:
    
/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/si
te-packages/matplotlib-1.0.1_r0-py2.6.egg/matplotlib/_path.so: no matching
architecture in universal wrapper


If I install with easy_install-2.6 matplotlib  from source I get the
following:
    ....
    Best match: matplotlib 0.91.1
    ....
    src/_image.cpp:952: error: Œpng_infopp_NULL¹ was not declared in this
scope    
    error: Setup script exited with error: command '/usr/bin/gcc-4.2' failed
with exit status 1

If I install with port install matplotlib it fails because the install in
not the correct path structure in the site-packages folder.


I¹ve tried searching the web and have found some indications that I need to
install 32bit python 2.6  but cannot find how to do that, and all of my
production machinese have 64bit python installed on them.


Any help at this point would be appreciated.
Thanks. 

Does anyone maintain a PPA with a python-matplotlib-nox suitable for Ubuntu
11.04 amd64 server i.e. built without X or GUI libraries?

On a fresh Ubuntu server install, python-numpy has relatively few dependencies,
but python-matplotlib pulls in everything the full GUI needs.

$ sudo aptitude install python-matplotlib -s
The following NEW packages will be installed:

  blt{a} cpp{a} cpp-4.5{a} defoma{a} fontconfig{a} hicolor-icon-theme{a}
  libatk1.0-0{a} libatk1.0-data{a} libavahi-client3{a}
  libavahi-common-data{a} libavahi-common3{a} libblas3gf{a}
  libcloog-ppl0{a} libcups2{a} libdatrie1{a} libelfg0{a} libfontenc1{a}
  libgdk-pixbuf2.0-0{a} libgfortran3{a} libgl1-mesa-dri{a}
  libgl1-mesa-glx{a} libglade2-0{a} libgmpxx4ldbl{a} libgtk2.0-0{a}
  libgtk2.0-bin{a} libgtk2.0-common{a} libice6{a} libjasper1{a}
  libjpeg62{a} liblapack3gf{a} libmpc2{a} libmpfr4{a} libpango1.0-0{a}
  libppl-c2{a} libppl7{a} libsm6{a} libthai-data{a} libthai0{a}
  libtiff4{a} libutempter0{a} libxaw7{a} libxcb-atom1{a}
  libxcb-shape0{a} libxcomposite1{a} libxcursor1{a} libxdamage1{a}
  libxfixes3{a} libxfont1{a} libxft2{a} libxi6{a} libxinerama1{a}
  libxmu6{a} libxpm4{a} libxrandr2{a} libxss1{a} libxt6{a} libxtst6{a}
  libxv1{a} libxxf86dga1{a} libxxf86vm1{a} python-dateutil{a}
  python-glade2{a} python-gtk2{a} python-matplotlib
  python-matplotlib-data{a} python-numpy{a} python-pyparsing{a}
  python-tk{a} python-tz{a} shared-mime-info{a} tcl8.5{a} tk8.5{a}
  ttf-lyx{a} x-ttcidfont-conf{a} x11-common{a} x11-utils{a} xbitmaps{a}
  xfonts-encodings{a} xfonts-utils{a} xterm{a}

0 packages upgraded, 80 newly installed, 0 to remove and 0 not upgraded.
Need to get 33.4 MB of archives. After unpacking 117 MB will be used.

Jeff

Hello,
I begin with matplotlib and I draw volume is a 3D view. It runs acoorectly
but some behavior are not understood. Here is my source code:
		fig = plt.figure()		
		ax = Axes3D(fig)
		print "long  =",self.onevolume['long']
		print "lat   =",self.onevolume['lat']
		print "upper =",self.onevolume['upper']
		print "lower =",self.onevolume['lower']
	
ax.plot(self.onevolume['long'],self.onevolume['lat'],self.onevolume['upper'])
	
ax.plot(self.onevolume['long'],self.onevolume['lat'],self.onevolume['lower'])
		for i in range(len(self.onevolume)):
			ax.plot([self.onevolume['long'][i],self.onevolume['long'][i]],
				[self.onevolume['lat'][i],self.onevolume['lat'][i]],
				[self.onevolume['lower'][i],self.onevolume['upper'][i]])			
		ax.set_xlabel('Longitude (deg)')
		ax.set_title('Volumic definition of  ' +  volume)
		print "axis=",ax.axis()
		canvas = FigureCanvas(fig)  # a gtk.DrawingArea
Here is the trace on my terminal:
long  = [-5.13888884 -4.04805565 -4.05249977 -3.75972223 -3.92777777
-4.0666666
 -4.9000001  -5.13888884]
lat   = [-49.71111298 -49.4030571  -49.3955574  -49.4375     -50.34027863
 -50.33333206 -49.76666641 -49.71111298]
upper = [999 999 999 999 999 999 999 999]
lower = [245 245 245 245 245 245 245 245]
axis= (-0.095000000000000001, 0.089999999999999997, -0.095000000000000001,
0.089999999999999997)
Then I dont'understand the results given by function axis() regarding the
input values.
Finally, label is not written on x axis, why ?
Here is my figure: http://old.nabble.com/file/p32231816/volume.jpeg 
Than you for all suggestions

-- 
View this message in context: http://old.nabble.com/first-steps-with-matplotlib-tp32231816p32231816.html
Sent from the matplotlib - users mailing list archive at Nabble.com.