matplotlib-users Mailing List for matplotlib (Page 10)

Jose,
Attached is an example that should at least point you in the right 
direction. I will also add it to the examples directory in the 
distribution.
Eric
Jose Gomez-Dans wrote:
> Hi,
> I am using Matplotlib to produce colormaps which I use with other
> programs. I would like to produce a PNG file with the used colormap
> (so that I can overlay). Rather than doing an imshow(<something>)
> followed by colorbar, is there an easy way I could pass the colorbar
> my cmap instance to have it plotted on its own? In essence, what I
> want to do is to do a savefig, and get something along the lines of
> what is shown here: <http://www.igidl.ul.pt/colorscale.gif>. I will
> need to change the scale (i.e.,, the numbers :D) with each run.
> 
> So far, I have "cut+pasted" from a whole image, which is hardly convenient! :(
> 
> Cheers,
> Jose

Xavier Gnata <gn...@ob...> writes:
> I do not know if we should post bug reports against matplotlib svn. 
Posting bug reports is likely to be helpful, but I suspect the
developers' list might be more appropriate for bugs in the svn version.
For bugs in released versions, I think John has told people to file a
bug in the Sourceforge tracker and also send a message to the mailing
list.
> Anyway, imshow is now fully broken this way :
[...]
> --> 200 w, h, d = renderer.get_text_width_height_descent(
> ValueError: need more than 2 values to unpack
It's not in imshow really, but in the usetex branch of
get_text_width_height_descent in the agg backend. I fixed the immediate
problem, though baseline alignment is unlikely to work with usetex as of
now.
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Hi,
I do not know if we should post bug reports against matplotlib svn. 
Please tell us.
Anyway, imshow is now fully broken this way :
imshow(ones((100,100)))
/usr/lib/python2.4/site-packages/matplotlib/pylab.py in imshow(*args, 
**kwargs)
 1960 try:
 1961 ret = gca().imshow(*args, **kwargs)
-> 1962 draw_if_interactive()
 1963 except:
 1964 hold(b)
/usr/lib/python2.4/site-packages/matplotlib/backends/backend_tkagg.py in 
draw_if_interactive()
 56 figManager = Gcf.get_active()
 57 if figManager is not None:
---> 58 figManager.show()
 59
 60
/usr/lib/python2.4/site-packages/matplotlib/backends/backend_tkagg.py in 
show(self)
 350 if sys.platform=='win32' : self.window.update()
 351 else:
--> 352 self.canvas.draw()
 353 self._shown = True
 354
/usr/lib/python2.4/site-packages/matplotlib/backends/backend_tkagg.py in 
draw(self)
 189
 190 def draw(self):
--> 191 FigureCanvasAgg.draw(self)
 192 tkagg.blit(self._tkphoto, self.renderer._renderer, 
colormode=2)
 193 self._master.update_idletasks()
/usr/lib/python2.4/site-packages/matplotlib/backends/backend_agg.py in 
draw(self)
 381
 382 self.renderer = self.get_renderer()
--> 383 self.figure.draw(self.renderer)
 384
 385 def get_renderer(self):
/usr/lib/python2.4/site-packages/matplotlib/figure.py in draw(self, 
renderer)
 610
 611 # render the axes
--> 612 for a in self.axes: a.draw(renderer)
 613
 614 # render the figure text
/usr/lib/python2.4/site-packages/matplotlib/axes.py in draw(self, 
renderer, inframe)
 1336
 1337 for zorder, i, a in dsu:
-> 1338 a.draw(renderer)
 1339
 1340 self.transData.thaw() # release the lazy objects
/usr/lib/python2.4/site-packages/matplotlib/axis.py in draw(self, 
renderer, *args, **kwargs)
 588 tick.set_label1(label)
 589 tick.set_label2(label)
--> 590 tick.draw(renderer)
 591 if tick.label1On and tick.label1.get_visible():
 592 extent = tick.label1.get_window_extent(renderer)
/usr/lib/python2.4/site-packages/matplotlib/axis.py in draw(self, renderer)
 168 if self.tick2On: self.tick2line.draw(renderer)
 169
--> 170 if self.label1On: self.label1.draw(renderer)
 171 if self.label2On: self.label2.draw(renderer)
 172
/usr/lib/python2.4/site-packages/matplotlib/text.py in draw(self, renderer)
 773 def draw(self, renderer):
 774 self.update_coords(renderer)
--> 775 Text.draw(self, renderer)
 776 if self.get_dashlength() > 0.0:
 777 self.dashline.draw(renderer)
/usr/lib/python2.4/site-packages/matplotlib/text.py in draw(self, renderer)
 315 angle = self.get_rotation()
 316
--> 317 bbox, info = self._get_layout(renderer)
 318 trans = self.get_transform()
 319 if rcParams['text.usetex']:
/usr/lib/python2.4/site-packages/matplotlib/text.py in _get_layout(self, 
renderer)
 198 baseline = None
 199 for line in lines:
--> 200 w, h, d = renderer.get_text_width_height_descent(
 201 line, self._fontproperties, 
ismath=self.is_math_text(line))
 202 if baseline is None:
ValueError: need more than 2 values to unpack
I have tried both Tk and GTKagg backends with the same result.
Xavier.
-- 
############################################
Xavier Gnata
CRAL - Observatoire de Lyon
9, avenue Charles André
69561 Saint Genis Laval cedex
Phone: +33 4 78 86 85 28
Fax: +33 4 78 86 83 86
E-mail: gn...@ob...
############################################ 

Hi,
I am using Matplotlib to produce colormaps which I use with other
programs. I would like to produce a PNG file with the used colormap
(so that I can overlay). Rather than doing an imshow(<something>)
followed by colorbar, is there an easy way I could pass the colorbar
my cmap instance to have it plotted on its own? In essence, what I
want to do is to do a savefig, and get something along the lines of
what is shown here: <http://www.igidl.ul.pt/colorscale.gif>. I will
need to change the scale (i.e.,, the numbers :D) with each run.
So far, I have "cut+pasted" from a whole image, which is hardly convenient! :(
Cheers,
Jose

> Alan G Isaac <ai...@am...> writes:=20
>> I recall some discussion of problem importing mpl EPS files=20
>> into Microsoft products in the past, and I think I recall=20
>> that there was a fix.=20
On 2007年9月01日, Jouni K. Sepp=E4nen apparently wrote:
> Perhaps you are thinking of one of these posts, which suggest using the=
=20
> svg backend and converting to emf via Inkscape or Visio:=20
> http://article.gmane.org/gmane.comp.python.matplotlib.general/9241=20
> http://article.gmane.org/gmane.comp.python.matplotlib.general/9250=20
No, but thanks.
I thought I recalled that someone found a glitch in the mpl=20
EPS Prolog, which when fixed removed the problem. Perhaps=20
this is not in 0.9.1? Let's see: OK, I guess what I had in
mind was the message below, which is not related...
The problem seems to be the construction of dates by=20
repeated use of ``glyphshow``. Apparenly the MS interpreter =20
does not support ``glyphshow``???
By the way, this usage is apparently seeking a compressed=20
rendering: is that right? Wouldn't using ``xshow`` be more=20
efficient, not to mention more obvious to humans?
Cheers,
Alan Isaac
######################################################################
Date: 2007年6月04日 15:49:20 -0500
From: Dave Baum <Dav...@mo...>
Subject: [Matplotlib-users] EPS images and MS Word
To: mat...@li...
I have been looking at the problem of figures saved as EPS images not =20
printing correctly under from MS Word (an invalidrestore error is =20
generated). I believe that matplotlib is incorrectly creating the =20
EPS file. According to the DSC Spec (http://partners.adobe.com/=20
public/developer/en/ps/5001.DSC_Spec.pdf), the Prolog section should =20
only contain procset definitions (which should also be enclosed in %%=20
BeginResource / %%EndResource pairs). It looks like the generated =20
EPS files contain an ordinary dictionary definition as well as font =20
information in the Prolog. I believe these commands should more =20
properly be moved to the Script section of the document (perhaps =20
within the %%BeginSetup/%%EndSetup sub-section).
I manually edited an EPS file to make the prolog empty, and moved the =20
mpldict definition to the body of the file, and it then worked fine =20
with MS Word.
Dave Baum
######################################################################

Alan G Isaac wrote:
> On 2007年8月31日, Eric Firing apparently wrote:
>> marker='None' is allowed, as is ' ' and ''. Do you need None? 
> 
> 
> I recently returned to some old figures where I did things 
> like this:
> 
> ls1 = dict(linestyle='-', marker=None, color=(0,0,1), linewidth=2, zorder=1000)
> fig_ax.plot_date(dates, data, **ls1)
> 
> This now fails with a KeyError.
> 
> However changing to marker='' works fine.
> I find that a bit less intuitve than marker=None,
> but I understand you to say that this should mean "use the default".
Yes, I think that quite some time ago we discussed this and decided to 
use the string "None" as well as empty strings, and to reserve the 
Python None for "use the default". I believe it was originally in the 
context of color specifications, and we then extended it to be a general 
convention. What is inconsistent is that in some cases one can 
explicitly use, e.g., marker=None for the default and in other cases, 
such as the plot command, one cannot. It is a question of whether the 
kwargs are being passed directly to the initializer or whether they are 
being applied after initialization via setters. I don't think it would 
be difficult to remove this inconsistency, and maybe it is worth doing. 
 I'm not inspired to do it immediately, though. Certainly the question 
should be addresses if the API is redone using traits.
Eric

Alan G Isaac <ai...@am...> writes:
> I recall some discussion of problem importing mpl EPS files
> into Microsoft products in the past, and I think I recall
> that there was a fix. 
Perhaps you are thinking of one of these posts, which suggest using the
svg backend and converting to emf via Inkscape or Visio:
http://article.gmane.org/gmane.comp.python.matplotlib.general/9241
http://article.gmane.org/gmane.comp.python.matplotlib.general/9250
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

On 2007年8月31日, Eric Firing apparently wrote:
> marker='None' is allowed, as is ' ' and ''. Do you need None? 
I recently returned to some old figures where I did things 
like this:
ls1 = dict(linestyle='-', marker=None, color=(0,0,1), linewidth=2, zorder=1000)
fig_ax.plot_date(dates, data, **ls1)
This now fails with a KeyError.
However changing to marker='' works fine.
I find that a bit less intuitve than marker=None,
but I understand you to say that this should mean "use the default".
Thank you,
Alan Isaac

> Alan G Isaac wrote:
>> I meant to be copying an annotation example from 
>> http://matplotlib.sourceforge.net/examples/annotation_demo.bak.py 
On 2007年8月31日, Eric Firing apparently wrote:
> The signature of the Annotation.__init__ does not match what you were 
> trying to feed it. 
OK, I can see that at 
http://matplotlib.sourceforge.net/matplotlib.text.html
So then the examples *are* wrong at
http://matplotlib.sourceforge.net/examples/annotation_demo.bak.py
Right?
Your example is sending me down the right track.
But I am still getting an odd result with
a = mpl.text.Annotation('text', (0.5,0.5), xycoords="axes fraction", fontsize=20)
fig_ax.add_artist(a)
Shouldn't this put the text smack in the middle of the figure?
OK, I give up for now. Time for some sleep.
Thanks!
Alan

I recall some discussion of problem importing mpl EPS files
into Microsoft products in the past, and I think I recall
that there was a fix. I'm using 0.9.1 and the dates in
my dateplots do not show up when imported into PowerPoint.
(An annotation below the dates does show up.)
Any clues?
Thank you,
Alan Isaac

Alan G Isaac wrote:
> I meant to be copying an annotation example from
> http://matplotlib.sourceforge.net/examples/annotation_demo.bak.py
I don't see much resemblance to that example; and I haven't tried that 
example, but I did verify that the examples/annotation_demo.py in svn 
works as expected.
> but it is not working. Am I just too bleary eyed because it 
> is later here, or is there a problem with the example?
Time for bed, rest those bleary eyes...
The signature of the Annotation.__init__ does not match what you were 
trying to feed it. Here is a modification that does match, but that may 
not be anything like what you were trying to do:
import pylab
import matplotlib as mpl
test = pylab.figure()
test_ax = test.gca()
test_ax.plot([1,2,3])
a = mpl.text.Annotation(
'F: a figure title (points)',
(-10, -10),
xycoords='figure points',
horizontalalignment='right',
verticalalignment='top',
fontsize=20)
test_ax.add_artist(a)
test.savefig('temp.eps')
------------------
Eric
> 
> Thank you,
> Alan Isaac
> 
> 
> %%%%%%%%%%%% Illustrate Annotation Problem %%%%%%%%%%%%%%%%%%%%%
> import pylab
> import matplotlib as mpl
> test = pylab.figure()
> test_ax = test.gca()
> test_ax.plot([1,2,3])
> 
> #the following line fails with "ValueError: too many values to unpack" 
> a = mpl.text.Annotation(
> test,
> 'F: a figure title (points)',
> loc=(-10, -10),
> coords='figure points',
> horizontalalignment='right',
> verticalalignment='top',
> fontsize=20)
> 
> test_ax.add_artist(f)
> test.savefig(r'c:\temp\temp.eps')

Alan G Isaac wrote:
> Line2D documentation reads:
> 
> marker: [ '+' | ',' | '.' | '1' | '2' | '3' | '4'
> 
> 1. Minor documentation bug: missing end bracket
Sort of. Actually, what is also missing is a very long list of possible 
markers, as given in the docstring for the set_marker method. Some 
docstring modifications and consolidations are needed.
> 2. Why is marker=None no longer allowed?
marker='None' is allowed, as is ' ' and ''. Do you need None? The idea 
is to distinguish between 'None' as in 'no marker--don't draw anything', 
and None as the default for a kwarg, meaning use the rcParams value. In 
plot, if you trace through the chain of half a dozen or so functions, 
you find that the kwargs are handled using setters, not at the stage of 
initialization of the Line2D instance, and setters generally don't 
accept None.
Eric
> 
> Cheers,
> Alan Isaac
> 
> 
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems? Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

I meant to be copying an annotation example from
http://matplotlib.sourceforge.net/examples/annotation_demo.bak.py
but it is not working. Am I just too bleary eyed because it 
is later here, or is there a problem with the example?
Thank you,
Alan Isaac
%%%%%%%%%%%% Illustrate Annotation Problem %%%%%%%%%%%%%%%%%%%%%
import pylab
import matplotlib as mpl
test = pylab.figure()
test_ax = test.gca()
test_ax.plot([1,2,3])
#the following line fails with "ValueError: too many values to unpack" 
a = mpl.text.Annotation(
test,
'F: a figure title (points)',
loc=(-10, -10),
coords='figure points',
horizontalalignment='right',
verticalalignment='top',
fontsize=20)
test_ax.add_artist(f)
test.savefig(r'c:\temp\temp.eps')

Line2D documentation reads:
 marker: [ '+' | ',' | '.' | '1' | '2' | '3' | '4'
1. Minor documentation bug: missing end bracket
2. Why is marker=None no longer allowed?
Cheers,
Alan Isaac

Jordan Dawe wrote:
> I've been trying to plot a pcolor over a contourf with a masked array in 
> the pcolor so that parts of the contour will show through underneath, 
> but whenever I try to do this the pcolor wipes out the contourf. I can 
> do this fine with a contourf over another contourf, but I'm plotting 
> model topography, and I would really prefer to leave the discretization 
> visible instead of showing contourf's interpolation. Any way to get a 
> pcolor to plot over a contourf without wiping out the contourf beneath it?
> 
> Jordan
Jordan,
I'm not sure I understand what you are trying to do; please provide a 
simple, self-contained script that illustrates the problem. (Is the 
contour supposed to be seen through holes in the pcolor where it is 
masked? Or are you talking about using transparency?)
Eric