matplotlib-users Mailing List for matplotlib

Hi,

I am trying to save figures where all the artists have animated =
True. The result I get (I have tried svg and png formats) is a set of
axes with labels, ticks, and grid, but no data points! I am using the
WxAgg backend.

Is this something to be expected, and if so, can anyone suggest a workaround?

Thanks,
Chee Sing Lee
Oregon State University
Dept. Nuclear Engineering and Radiation Health Physics

Hi,

I am trying to save figures where all the artists have animated =
True. The result I get (I have tried svg and png formats) is a set of
axes with labels, ticks, and grid, but no data points! I am using the
WxAgg backend.

Is this something to be expected, and if so, can anyone suggest a workaround?

Thanks,
Chee Sing Lee
Oregon State University
Dept. Nuclear Engineering and Radiation Health Physics

Hi guys, continuing exploring matplotlib capabilities J

 

I've applied the idea you've said later about colormap.

Now I would like on an event to refresh my view so I've written this code:

 

   .

 

   def InitCode(self):

    .

        self.norm = plt.Normalize(vmin=0, vmax=255)

        self.cmap = cm.jet

        im_full = full_sonar.imshow( z, aspect= 'auto', cmap=self.cmap,
norm=self.norm, interpolation='quadric')

        im_zoom = zoomed_sonar.imshow( z, aspect= 'auto',cmap=self.cmap,
norm=self.norm, interpolation='quadric')

 
self.fig.colorbar(im_full,cax=cax,orientation='vertical',extend='both')

        self.im_full = im_full

        self.im_zoom = im_zoom

    .

                

    def updateSonarView(self):

        self.im_full.set_cmap(self.cmap)

        self.im_zoom.set_cmap(self.cmap)

        self.im_full.set_norm(self.norm)

        self.im_zoom.set_norm(self.norm)

        self.fig.canvas.draw()

        

    def OnUpdateButton(self, event):

        self.low_clutter_val = self.clutter_low.GetPosition()

        self.high_clutter_val = self.clutter_high.GetPosition()

        self.norm = plt.Normalize(vmin=self.low_clutter_val,
vmax=self.high_clutter_val)

        self.updateSonarView()

 

But I get this error from the lib :

 

Traceback (most recent call last):

  File "./Filters\A1Filter.py", line 183, in OnUpdateButton

    self.updateSonarView()

  File "./Filters\A1Filter.py", line 157, in updateSonarView

    self.im_full.set_norm(self.norm)

  File "C:\Python25\Lib\site-packages\matplotlib\cm.py", line 129, in
set_norm

    self.changed()

  File "C:\Python25\Lib\site-packages\matplotlib\image.py", line 123, in
changed

    cm.ScalarMappable.changed(self)

  File "C:\Python25\Lib\site-packages\matplotlib\cm.py", line 174, in
changed

    self.callbacksSM.process('changed', self)

  File "C:\Python25\Lib\site-packages\matplotlib\cbook.py", line 152, in
process

    func(*args, **kwargs)

  File "C:\Python25\Lib\site-packages\matplotlib\figure.py", line 1028, in
on_changed

    cb.update_bruteforce(m)

  File "C:\Python25\Lib\site-packages\matplotlib\colorbar.py", line 649, in
update_bruteforce

    self.draw_all()

  File "C:\Python25\Lib\site-packages\matplotlib\colorbar.py", line 221, in
draw_all

    self._process_values()

  File "C:\Python25\Lib\site-packages\matplotlib\colorbar.py", line 458, in
_process_values

    b = self.norm.inverse(self._uniform_y(self.cmap.N+1))

  File "C:\Python25\Lib\site-packages\matplotlib\colors.py", line 652, in
inverse

    return vmin + val * (vmax - vmin)

  File "C:\Python25\Lib\site-packages\numpy\ma\core.py", line 1686, in
__mul__

    return multiply(self, other)

  File "C:\Python25\Lib\site-packages\numpy\ma\core.py", line 503, in
__call__

    result = self.f(d1, d2, *args, **kwargs).view(get_masked_subclass(a,b))

ValueError: shape mismatch: objects cannot be broadcast to a single shape

 

UpdateSonarview works flawlessly if I modify the colormap (self.cmap) and
call it then.

If I modify self.norm and call UpdateSonarView I get this crash.

 

Is it a bug from my bad usage of the lib? If not, I can perhaps send you a
stripped down file showing the problem?

Using matplotlib 0.98.1 under windows.

 

Laurent

Hello Jeff,
I just wanna give feedback on what got me going here:

#### data preparation
### data is loaded from a CSV file
###
lats = y # data[:,0]
## lon => x
lons = x # data[:,1]
## values => z
values = z #data[:,2]
###
lat_uniq = list(set(lats.tolist()))
nlats = len(lat_uniq)
lon_uniq = list(set(lons.tolist()))
nlons = len(lon_uniq)
color_map = plt.cm.spectral
print lats.shape, nlats, nlons
yre = lats.reshape(nlats,nlons)
xre = lons.reshape(nlats,nlons)
zre = values.reshape(nlats,nlons)
#### later in the defined map
CT = m.contourf(xre, yre, zre, cmap=color_map)

Of course, this can be simplified. But, really, in the end the solution 
was just simple and I hadn't thought of it...

If you want I can try to create a short example for the distribution.

Thanks for your help again,
Timmie

Travis Ruthenburg wrote:
> I'm new to matplotlib and was hoping someone could point me in the  
> right direction for making a waterfall plot, similar to MATLAB's  
> 'waterfall'.
> 
> I'd like to plot a series of 2-D spectra staggered by time.

Travis,

I presume you are handling the calculation of spectra, and the question 
is really about plotting offset lines.

We do not have the 3-D type plot illustrated here:
http://www.mathworks.com/access/helpdesk/help/techdoc/index.html?/access/helpdesk/help/techdoc/ref/waterfall.html&http://www.mathworks.com/cgi-bin/texis/webinator/search/

but there are a couple of ways you can plot offset lines.  One is to use 
the plot command to plot a set of lines, to which you have added the 
offsets yourself.  The other is to make a LineCollection.  I see that 
our example of the latter does not actually make use of the feature that 
  automatically applies the constant offsets, so I need to make a 
revised example, which I can't do right this minute.  In the meantime, 
the approach in line_collection.py and line_collection2.py in the mpl 
examples directory (examples/pylab_examples/ in the 0.98.x and svn 
versions) will work.

Backing off slightly, the larger points are (1) mpl is 2-D only; there 
was a 3-D module in earlier versions, but it was unmaintained and has 
been dropped; (2) 2-D plotting is very flexible; (3) often the problem 
is not one of plotting but of manipulating your input arrays, so the 
better you know numpy, the easier and more efficiently you can generate 
plots.

Eric

Hi all,

With both matplotlib versions (0.91 and 0.98), my SVGs are rendered
correctly in Inkscape (and almost well in Firefox), but they are not
rendered correctly when viewed with ImageMagick or Eye-of-Gnome.

I did screenshots here, showing this "curve cropping" issue :
http://mathieu-leplatre.info/media/matplotlib-svg/

I am running CentOS 5.2 with matplotlib compiled manually in both cases.
Cairo 1.2.4, librsvg 2.16, inkscape 0.46

Could it be related to matplotlib ? Or to operating system libraries ?

Thank you all for your support.
Mathieu.

On Tue, Jul 29, 2008 at 1:48 PM, Travis Ruthenburg <tr...@uc...> wrote:
> I'm new to matplotlib and was hoping someone could point me in the
> right direction for making a waterfall plot, similar to MATLAB's
> 'waterfall'.
>
> I'd like to plot a series of 2-D spectra staggered by time.

The specgram function is the closest equivalent:

http://matplotlib.sourceforge.net/matplotlib.pyplot.html#-specgram

I'm new to matplotlib and was hoping someone could point me in the  
right direction for making a waterfall plot, similar to MATLAB's  
'waterfall'.

I'd like to plot a series of 2-D spectra staggered by time.

Thanks,

Travis

On Tue, Jul 29, 2008 at 1:20 PM, Ben Axelrod <bax...@co...> wrote:
> Thanks for the bug fix.  Do you think this fix will make it into the 0.98.3 release?

it will.

Thanks for the bug fix.  Do you think this fix will make it into the 0.98.3 release?

-----Original Message-----
From: John Hunter [mailto:jd...@gm...]
Sent: Tuesday, July 29, 2008 1:49 PM
To: Ben Axelrod
Cc: mat...@li...
Subject: Re: [Matplotlib-users] bar plot picker with 0 height data

On Tue, Jul 29, 2008 at 12:27 PM, Ben Axelrod <bax...@co...> wrote:
> I get an exception when I set up a bar chart with a picker that has a bar
> with 0 height, and I click on the chart.

Thanks for the report -- I just committed a fix for this to svn r5922

JDH

We are in the final stages of preparing a new matplotlib release, and
a lot of work has gone into it.  If you would like to test the release
and see if it is working for you, that would be a big help

  http://matplotlib.sourceforge.net/tmp/matplotlib-0.98.3rc2.tar.gz

Unfortunately, we do not have binary builds available at this time.

Sandro, I saw that Georg released the 0.4.2 sphinx bugfix release, so
as soon as Mikhail gets that into debian you can test this release
candidate.

JDH

On Tue, Jul 29, 2008 at 12:27 PM, Ben Axelrod <bax...@co...> wrote:
> I get an exception when I set up a bar chart with a picker that has a bar
> with 0 height, and I click on the chart.

Thanks for the report -- I just committed a fix for this to svn r5922

JDH

I get an exception when I set up a bar chart with a picker that has a bar with 0 height, and I click on the chart.

Code to generate exception:

from matplotlib.pyplot import figure, show
import numpy as np

def onpick1(event):
    print 'foo'

fig = figure()
ax = fig.add_subplot(111)

xdata = [1, 2, 3]
ydata = [1, 0, 3] #change to [1, 2, 3] and code will run fine

ax.bar(xdata,
       ydata,
       picker = True)

fig.canvas.mpl_connect('pick_event', onpick1)

show()
#now click on chart

-----
Exception generated:


Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\Python24\lib\lib-tk\Tkinter.py", line 1345, in __call__
    return self.func(*args)
  File "C:\Python24\Lib\site-packages\matplotlib\backends\backend_tkagg.py", line 236, in button_press_event
    FigureCanvasBase.button_press_event(self, x, y, num, guiEvent=event)
  File "C:\Python24\Lib\site-packages\matplotlib\backend_bases.py", line 1074, in button_press_event
    self.callbacks.process(s, mouseevent)
  File "C:\Python24\Lib\site-packages\matplotlib\cbook.py", line 152, in process
    func(*args, **kwargs)
  File "C:\Python24\Lib\site-packages\matplotlib\backend_bases.py", line 983, in pick
    self.figure.pick(mouseevent)
  File "C:\Python24\Lib\site-packages\matplotlib\artist.py", line 226, in pick
    for a in self.get_children(): a.pick(mouseevent)
  File "C:\Python24\Lib\site-packages\matplotlib\axes.py", line 2306, in pick
    martist.Artist.pick(self,args[0])
  File "C:\Python24\Lib\site-packages\matplotlib\artist.py", line 226, in pick
    for a in self.get_children(): a.pick(mouseevent)
  File "C:\Python24\Lib\site-packages\matplotlib\artist.py", line 220, in pick
    inside,prop = self.contains(mouseevent)
  File "C:\Python24\Lib\site-packages\matplotlib\patches.py", line 385, in contains
    x, y = self.get_transform().inverted().transform_point(
  File "C:\Python24\Lib\site-packages\matplotlib\transforms.py", line 1840, in inverted
    return CompositeGenericTransform(self._b.inverted(), self._a.inverted())
  File "C:\Python24\Lib\site-packages\matplotlib\transforms.py", line 1338, in inverted
    self._inverted = Affine2D(inv(mtx))
  File "C:\Python24\Lib\site-packages\numpy\linalg\linalg.py", line 332, in inv
    return wrap(solve(a, identity(a.shape[0], dtype=a.dtype)))
  File "C:\Python24\Lib\site-packages\numpy\linalg\linalg.py", line 235, in solve
    raise LinAlgError, 'Singular matrix'
LinAlgError: Singular matrix

--------------
Stats:
MPL 0.98.1
Python 2.4.4
Numpy 1.1.0
I get the error on both Windows and Linux

Thanks,
-Ben

On Tue, Jul 29, 2008 at 10:55 AM, Ben Axelrod <bax...@co...> wrote:
> Text picking does not seem to work for me anymore since I upgraded from
> 0.91.4 to 0.98.1.

This is fixed in matplotlib svn and will be out in the next release.

Text picking does not seem to work for me anymore since I upgraded from 0.91.4 to 0.98.1.

These lines should set up the picker.  They come straight from "pick_event_demo.py"

    ax1.set_title('click on points, rectangles or text', picker=True)
    ax1.set_ylabel('ylabel', picker=True, bbox=dict(facecolor='red'))

    fig.canvas.mpl_connect('pick_event', onpick1)

    def onpick1(event):
        if isinstance(event.artist, Line2D):
            thisline = event.artist
            xdata = thisline.get_xdata()
            ydata = thisline.get_ydata()
            ind = event.ind
            print 'onpick1 line:', zip(npy.take(xdata, ind), npy.take(ydata, ind))
        elif isinstance(event.artist, Rectangle):
            patch = event.artist
            print 'onpick1 patch:', patch.get_path()
        elif isinstance(event.artist, Text):
            text = event.artist
            print 'onpick1 text:', text.get_text()


I apologize if this is a known issue. I tried searching the listserve, but did not find anything.
-Ben

Hi andrew, eric,

Excellent that is exactly what I was looking for :)

Laurent

> -----Message d'origine-----
> De : Eric Firing [mailto:ef...@ha...]
> Envoyé : mardi 29 juillet 2008 02:39
> À : Laurent Dufrechou
> Cc : mat...@li...
> Objet : Re: [Matplotlib-users] Colormap cluttering?
> 
> Laurent Dufrechou wrote:
> > Hello,
> >
> >
> >
> > I would like to have a cluttering functionality to colorbar.
> >
> > http://en.wikipedia.org/wiki/Clutter_(radar)
> >
> >
> >
> > Before writing it, I would like to know if there is a way to doing it
> > with matplotlib.
> >
> > What I mean by cluttering is:
> >
> >
> >
> > You’ve got a colormap associated with a graphic where value goes from
> 0
> > to 255 for example.
> >
> > Assigning a classical colormap (for example cm.jet) 0 value will be
> blue
> > and 255 one will be red.
> >
> > What I need is a low  clutter and max clutter, if I set low clutter
> to
> > 10 and ax cluter to 250 then:
> >
> > Blue will be for value from 0 to 10
> >
> > Then the colormap do his job from 10 to 250 and finally
> >
> >  From 250 to 255 colr will be set to max one = red.
> >
> >
> >
> > Is it ever done in matplotlib, if not what could be the strategy
> here…?
> >
> > I was thinking of set_over/set_under but seems not be exactly what I
> > need because I want to recreate the colormap from 10 to 250 with N
> segments.
> >
> > (moreover I don’t understand how you set the over/under value…)
> 
> If you don't really care how many colors are in the map, then for your
> 0-255 example, try this:
> 
> import numpy as np
> import matplotlib.pyplot as plt
> fakedata = np.random.rand(10,20) * 255.0
> cmap = plt.cm.jet
> norm = plt.Normalize(vmin=100, vmax=150)
> plt.imshow(fakedata, cmap=cmap, norm=norm, interpolation="nearest")
> plt.colorbar(extend="both")
> plt.show()
> 
> I made the colored range small to emphasize what you call the
> "cluttering" effect.
> 
> We are taking advantage of the default, which is that the over and
> under
> values are the top and bottom ends of the colormap.  If you want other
> colors for the ends, then after defining your cmap, use, e.g.:
> 
> cmap.set_under('w')
> cmap.set_over('k')
> 
> If you want to use a smaller number of colors in your colormap, then
> you
> need to make the colormap this way, for example:
> 
> from matplotlib.colors import LinearSegmentedColormap
> from matplotlib._cm import _jet_data
> cmap = LinearSegmentedColormap("yourname", _jet_data, N=10)
> 
> (I should add a helper function to make this more obvious and
> straightforward.)
> 
> Eric

On Mon, Jul 28, 2008 at 12:40 PM, eliben <el...@gm...> wrote:
>
> Is it only me or is the user's guide linked to from
> http://matplotlib.sourceforge.net/
> (http://matplotlib.sourceforge.net/users_guide_0.98.1.pdf) quite old ? The
> new user's guide seems to be available from
> http://matplotlib.sourceforge.net/doc/html/index.html

No, it is not just you.  We have been working on new documenttion, and
what we have is good, but I would like to see them more complete beore
we promote them.  I do have a link to them from the matplotlib website
called "new docs (beta)"

JDH

Laurent Dufrechou wrote:
> Hello,
> 
>  
> 
> I would like to have a cluttering functionality to colorbar.
> 
> http://en.wikipedia.org/wiki/Clutter_(radar)
> 
>  
> 
> Before writing it, I would like to know if there is a way to doing it 
> with matplotlib.
> 
> What I mean by cluttering is:
> 
>  
> 
> You’ve got a colormap associated with a graphic where value goes from 0 
> to 255 for example.
> 
> Assigning a classical colormap (for example cm.jet) 0 value will be blue 
> and 255 one will be red.
> 
> What I need is a low  clutter and max clutter, if I set low clutter to 
> 10 and ax cluter to 250 then:
> 
> Blue will be for value from 0 to 10
> 
> Then the colormap do his job from 10 to 250 and finally
> 
>  From 250 to 255 colr will be set to max one = red.
> 
>  
> 
> Is it ever done in matplotlib, if not what could be the strategy here…?
> 
> I was thinking of set_over/set_under but seems not be exactly what I 
> need because I want to recreate the colormap from 10 to 250 with N segments.
> 
> (moreover I don’t understand how you set the over/under value…)

If you don't really care how many colors are in the map, then for your 
0-255 example, try this:

import numpy as np
import matplotlib.pyplot as plt
fakedata = np.random.rand(10,20) * 255.0
cmap = plt.cm.jet
norm = plt.Normalize(vmin=100, vmax=150)
plt.imshow(fakedata, cmap=cmap, norm=norm, interpolation="nearest")
plt.colorbar(extend="both")
plt.show()

I made the colored range small to emphasize what you call the 
"cluttering" effect.

We are taking advantage of the default, which is that the over and under 
values are the top and bottom ends of the colormap.  If you want other 
colors for the ends, then after defining your cmap, use, e.g.:

cmap.set_under('w')
cmap.set_over('k')

If you want to use a smaller number of colors in your colormap, then you 
need to make the colormap this way, for example:

from matplotlib.colors import LinearSegmentedColormap
from matplotlib._cm import _jet_data
cmap = LinearSegmentedColormap("yourname", _jet_data, N=10)

(I should add a helper function to make this more obvious and 
straightforward.)

Eric

Laurent Dufrechou wrote:
> Hello,
> 
>  
> 
> I would like to have a cluttering functionality to colorbar.
> 
> http://en.wikipedia.org/wiki/Clutter_(radar)
> 
>  
> 
> Before writing it, I would like to know if there is a way to doing it
> with matplotlib.
> 
> What I mean by cluttering is:
> 
>  
> 
> You’ve got a colormap associated with a graphic where value goes from 0
> to 255 for example.
> 
> Assigning a classical colormap (for example cm.jet) 0 value will be blue
> and 255 one will be red.
> 
> What I need is a low  clutter and max clutter, if I set low clutter to
> 10 and ax cluter to 250 then:
> 
> Blue will be for value from 0 to 10
> 
> Then the colormap do his job from 10 to 250 and finally
> 
> From 250 to 255 colr will be set to max one = red.
> 
>  
> 
> Is it ever done in matplotlib, if not what could be the strategy here…?
> 
> I was thinking of set_over/set_under but seems not be exactly what I
> need because I want to recreate the colormap from 10 to 250 with N segments.
> 
> (moreover I don’t understand how you set the over/under value…)

Dear Laurent, if I understand you correctly, set vmin=10 and vmax=250
and it will do what you want.

-Andrew

Ben Axelrod wrote:
> I am having problems plotting anything using version 0.98.1 on Linux 
> Debian 4.  But 0.91.4 works fine. 

If you are compiling from source, the problem could be that setup.py 
does not clean out the destination directory.  For example, if you are 
installing in /usr/local/lib/python2.5/site-packages then you should 
first do "rm -r /usr/local/lib/python2.5/site-packages/matplotlib*" 
before installing another version.

Eric

>  
> 
> When I try to plot something I get this error on the console:
> 
>  
> 
> Traceback (most recent call last):
> 
>   File 
> "/usr/lib/python2.4/site-packages/matplotlib/backends/backend_gtk.py", 
> line 331, in expose_event
> 
>     self._render_figure(self._pixmap, w, h)
> 
>   File 
> "/usr/lib/python2.4/site-packages/matplotlib/backends/backend_gtkagg.py", 
> line 75, in _render_figure
> 
>     FigureCanvasAgg.draw(self)
> 
>   File 
> "/usr/lib/python2.4/site-packages/matplotlib/backends/backend_agg.py", 
> line 357, in draw
> 
>     self.renderer = self.get_renderer()
> 
>   File 
> "/usr/lib/python2.4/site-packages/matplotlib/backends/backend_agg.py", 
> line 368, in get_renderer
> 
>     self.renderer = RendererAgg(w, h, self.figure.dpi)
> 
>   File 
> "/usr/lib/python2.4/site-packages/matplotlib/backends/backend_agg.py", 
> line 116, in __init__
> 
>     self.draw_polygon = self._renderer.draw_polygon
> 
> AttributeError: draw_polygon
> 
>  
> 
>  
> 
> And here is the configuration info from when I installed from src:
> 
>  
> 
>  
> 
> BUILDING MATPLOTLIB
> 
>             matplotlib: 0.98.1
> 
>                 python: 2.4.4 (#2, Apr 15 2008, 23:43:20)  [GCC 4.1.2
> 
>                         20061115 (prerelease) (Debian 4.1.1-21)]
> 
>               platform: linux2
> 
>  
> 
> REQUIRED DEPENDENCIES
> 
>                  numpy: 1.1.0
> 
>              freetype2: 9.10.3
> 
>  
> 
> OPTIONAL BACKEND DEPENDENCIES
> 
>                 libpng: 1.2.15beta5
> 
>                Tkinter: Tkinter: 39220, Tk: 8.4, Tcl: 8.4
> 
>               wxPython: 2.8.8.1
> 
>                         * WxAgg extension not required for wxPython >= 2.8
> 
>                   Gtk+: gtk+: 2.8.20, glib: 2.12.4, pygtk: 2.8.6, pygobject:
> 
>                         [pre-pygobject]
> 
>                     Qt: Qt: 3.3.6, PyQt: 3.16
> 
>                    Qt4: no
> 
>                  Cairo: 1.2.0
> 
>  
> 
> OPTIONAL DATE/TIMEZONE DEPENDENCIES
> 
>               datetime: present, version unknown
> 
>               dateutil: present, version unknown
> 
>                   pytz: 2006p
> 
>  
> 
> OPTIONAL USETEX DEPENDENCIES
> 
>                 dvipng: 1.9
> 
>            ghostscript: 8.15.3
> 
>                  latex: 3.141592
> 
>                pdftops: 3.00
> 
>  
> 
> EXPERIMENTAL CONFIG PACKAGE DEPENDENCIES
> 
>              configobj: matplotlib will provide
> 
>       enthought.traits: no
> 
>  
> 
>  
> 
> I am guessing that I am either missing a library or need an updated 
> version.  Does anyone have any ideas?
> 
>  
> 
> Thanks,
> 
> -Ben
> 
>  
> 
> 
> ------------------------------------------------------------------------
> 
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users