matplotlib-users Mailing List for matplotlib (Page 7)

I've searched and searched the online docs...please help.
DG
 

John Hunter wrote:
> On Mon, Feb 22, 2010 at 4:33 PM, John Hunter <jd...@gm...> wrote:
>> polycol = plt.hexbin(data['jetMomentum'][cut] / 1000,
>> data['deltaR'][cut],gridsize=50, norm=colors.LogNorm())
>> cb = plt.colorbar(norm=colors.LogNorm())
>>
>> but this appears to be broken:
> 
> I committed some changes to support this -- the following now works:
> 
> polycol = plt.hexbin(data['jetMomentum'][cut] / 1000,
> data['deltaR'][cut],gridsize=50, norm=colors.LogNorm())
> 
> cb = plt.colorbar(norm=polycol.norm)
> 
> Eric - I was surprised the colorbar does not use the mappable norm by
> default (if passed norm=None). Instead it uses ::
> 
> norm = colors.Normalize()
> 
> is this a feature?
Yes. You are looking at ColorbarBase, which does not have an associated 
mappable. The derived Colorbar class does grab the cmap and norm from 
the mappable used in the initialization. Is this somehow not working? 
Did you really need to specify the norm explicitly?
Eric
> 
> JDH

On Mon, Feb 22, 2010 at 4:33 PM, John Hunter <jd...@gm...> wrote:
> polycol = plt.hexbin(data['jetMomentum'][cut] / 1000,
> data['deltaR'][cut],gridsize=50, norm=colors.LogNorm())
> cb = plt.colorbar(norm=colors.LogNorm())
>
> but this appears to be broken:
I committed some changes to support this -- the following now works:
 polycol = plt.hexbin(data['jetMomentum'][cut] / 1000,
data['deltaR'][cut],gridsize=50, norm=colors.LogNorm())
 cb = plt.colorbar(norm=polycol.norm)
Eric - I was surprised the colorbar does not use the mappable norm by
default (if passed norm=None). Instead it uses ::
 norm = colors.Normalize()
is this a feature?
JDH

On Mon, Feb 22, 2010 at 3:42 PM, Jan Strube <cur...@gm...> wrote:
> Hi John,
> the attachment may not make it to the list. However, please run the modified
> test.py that I have attached.
> It requires the attached input file.
> Then change it to read the original input file.
> In my case:
> The broken case:
OK, at least now we are running the sample example :-)
The problem is that the LogFormatter has a default which is
"decadeOnly=True" and in the first case which "worked" three of the
tick locations coincidentally came down on decades (0, 1, 2 -> 1, 10,
100). In the case you were working with, only one of the ticks mapped
to the decade.
So for this case we want a locator that returns integers 0,1,2... that
will then get mapped via Eric's custom formatter to the 10^i formats.
Unfortunately, there is no easy way to set the locator for the
colorbar. An easy workaround *for this case* is to simply set the
tick locations
 cb = plt.colorbar(format=LogFormatterHB(), ticks=[0,1,2])
but in general you may not know the decade span that you need. It
does all feel a bit kludgy. The problem as you noted in one of your
earlier posts is that the data is log scaled before being passed into
the PolyCollection and the fact that it is log scaled is then lost to
the colorbar. It seems everything would fit together more naturally
if we passed in raw scalar data to the PolyCollection and set the norm
to be colors.LogNorm, and then also set norm=colors.LogNorm on the
colorbar
I tried:
 polycol = plt.hexbin(data['jetMomentum'][cut] / 1000,
data['deltaR'][cut],gridsize=50, norm=colors.LogNorm())
 cb = plt.colorbar(norm=colors.LogNorm())
but this appears to be broken:
msierig@pinchiepie:Downloads> python test.py
Traceback (most recent call last):
 File "test.py", line 29, in <module>
 cb = plt.colorbar()
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/pyplot.py",
line 1356, in colorbar
 ret = gcf().colorbar(mappable, cax = cax, ax=ax, **kw)
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/figure.py",
line 1103, in colorbar
 cb = cbar.Colorbar(cax, mappable, **kw)
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 690, in __init__
 ColorbarBase.__init__(self, ax, **kw)
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 242, in __init__
 self.draw_all()
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 260, in draw_all
 self._config_axes(X, Y)
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 332, in _config_axes
 self.update_ticks()
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 271, in update_ticks
 ticks, ticklabels, offset_string = self._ticker()
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/colorbar.py",
line 458, in _ticker
 b = np.array(locator())
 File "/home/msierig/dev/lib/python2.6/site-packages/matplotlib/ticker.py",
line 1173, in __call__
 vmin = self.axis.get_minpos()

On Mon, Feb 22, 2010 at 2:57 PM, Jan Strube <cur...@gm...> wrote:
> Hi John,
> thanks for the test case.
> Indeed I am getting different results.
> Labels that are beyond the range of the axis don't get printed.
> This is reproducible on ubuntu (test.png) and on mac os x (t.png) both have
> the svn head of mpl.
OK, I see we are getting several differences (eg the x and y viewlim
extent, pixel dimensions of saved figure) in addition to the tick
label problem you described. I modified the test script to force the
use of the Agg backend. Please run it with the verbose-debug flag and
post the image and output. Mine are attached
 > python ~/test.py --verbose-debug > debug.out

On Mon, Feb 22, 2010 at 1:44 PM, Ben Axelrod <BAx...@co...> wrote:
> I noticed that there are many modules in the current code base that are
> not listed at: http://matplotlib.sourceforge.net/modindex.html. I
> understand that a few are new files and that the documentation for these
> will be generated during the next release. But I know that most of these
> were in the last release so should have had their documentation generated.
> Am I missing something? Can the documentation for these modules be found
> somewhere else?
>
You're not missing anything in terms of the html/pdf docs. When we
converted our documentation build system to sphinx/rest, we had to convert
our docstrings to rest as well. We made good headway for a while but the
progress has stalled. To add a module to the documentation system, you need
to
 * make sure the docstrings are ReST compliant
 * add a stub in doc/api (eg see figure_api.rst
 * add your new rst file to doc/api/index.rst
We'd like to see all of the user facing modules converted, so contributions
are welcome. See also
http://matplotlib.sourceforge.net/faq/howto_faq.html?highlight=contributing#contribute-to-matplotlib-documentation
 http://matplotlib.sourceforge.net/devel/documenting_mpl.html
 http://matplotlib.sourceforge.net/sampledoc/
In the meantime, you can always use pydoc or help from the interactive
shell, eg::
 > pydoc matplotlib.patheffects
pydoc also has flags for generating html, etc.
or from the shell (ipython here)::
 In [3]: import matplotlib.patheffects
 In [4]: help matplotlib.patheffects
 ------> help(matplotlib.patheffects)
JDH

On Mon, Feb 22, 2010 at 10:11 AM, Michael Droettboom <md...@st...> wrote:
> The coordinates for Circle (and all patches) are in data coordinates.
> So the (300, 300) is relative to the values in the data itself. When
> adding a patch directly to a plot, however, the limits may not
> automatically update, so you may need to call axes.set_xlim or
> axes.set_ylim to adjust them to make the circle visible.
In this example, though, he is not adding the Circle to the Axes via
add_patch (so the transData transform is not set).
Rather, he is using it to set the clip path
 http://matplotlib.sourceforge.net/examples/api/clippath_demo.html
So I think Ryan's answer is correct.
But David, if you want the clippath to be in data coordinates, you can
set the transform yourself
 patch = patches.Circle((300,300), radius=100, transform=ax.transData)
See the transformations tutorial for more about the coordinate systems
and transformations
 http://matplotlib.sourceforge.net/users/transforms_tutorial.html
JDH

On Mon, Feb 22, 2010 at 10:01 AM, Ben Axelrod <BAx...@co...> wrote:
> John, your assesment of the problem is correct. And I believe your suggested solution is also correct. Currently, each call to a mplot3d plot method is treated independantly. They get converted into custom PolyCollections which each do the Z-order sorting.
>
> There is still an issue here however. Even if we implement the aformentioned solution, we are still only approximating a 3d library. And the result will still not be as nice as matlab. I believe that because we treat the surface as a series of 2D polygons, the intersection between two surfaces will be at the polygon edges. See the attached image for an example of what the intersection between a sphere and plane might look like.
True enough, but as your example shows it would still be a substantial
improvement over what we have now, and by getting all the faces in the
scene into a single data structure, we leave open the possibility of
doing something more sophisticated down the road (like chopping a
problematic face into multiple faces, some in front, some behind, an
intersecting object).
JDH

The coordinates for Circle (and all patches) are in data coordinates. 
So the (300, 300) is relative to the values in the data itself. When 
adding a patch directly to a plot, however, the limits may not 
automatically update, so you may need to call axes.set_xlim or 
axes.set_ylim to adjust them to make the circle visible.
Mike
Ryan May wrote:
> On Wed, Feb 17, 2010 at 6:44 PM, David Arnold <dwa...@su...> wrote:
> 
>> All,
>>
>> I'm looking at:
>>
>> http://matplotlib.sourceforge.net/examples/api/clippath_demo.html
>>
>> But I cannot figure out:
>>
>> patch=patches.Circle((300, 300), radius=100)
>>
>> Where precisely is (300,300)?
>> 
>
> I believe it's in window coordinates (pixels), with 0,0 being the lower left.
>
> Ryan
>
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

I will be sure to use "gdb" and "bt" to get that information next time I
encounter the segmentation fault or any other error that I post to
list. 
I tried going back to the numpy.dev8106 version just now in an attempt
to recreate the conditions of the error, but I was unsuccessful in
generating any segmentation faults. My guess is that I made some novice
mistake when upgrading the libraries from source. 
On Mon, 2010年02月22日 at 10:16 -0500, Michael Droettboom wrote:
> Can you provide a gdb backtrace?
> 
> Run "gdb python", then at the gdb prompt type "run -c 'import pylab; 
> pylab.clf(); pylab.plot(pylab.sin(range(101))); pylab.xlabel("Test X"); 
> pylab.ylabel("Test Y"); pylab.show()'". After it segfaults, type "bt" 
> to get a backtrace, and copy-and-paste it to this list.
> 
> Mike
> 
> Joseph D Cali wrote:
> > I am encountering the following error:
> >
> > python -c 'import pylab; pylab.clf(); pylab.plot(pylab.sin(range(101)));
> > pylab.xlabel("Test X"); pylab.ylabel("Test Y"); pylab.show()'
> >
> > /local_home/calijos/usr/lib/python2.6/site-packages/matplotlib/backends/backend_gtk.py:621: DeprecationWarning: Use the new widget gtk.Tooltip
> > self.tooltips = gtk.Tooltips()
> > Segmentation fault
> >
> > -------------------------------------------------------------
> >
> > But when I run the following command, I encounter no errors and plot is
> > as expected:
> >
> > python -c 'import pylab; pylab.clf(); pylab.plot(pylab.sin(range(100)));
> > pylab.xlabel("Test X"); pylab.ylabel("Test Y"); pylab.show()'
> >
> > I tried several other plots, any plots with less than 100 points work
> > correctly, any plots with more than 100 points segfault.
> >
> > -------------------------------------------------------------
> >
> > My current system setup is:
> >
> > System: Ubuntu 9.04 x64
> > Python: Python 2.6.2
> > Matplotlib: >0.99.1 (tied 0.99.1.1, 0.99.1.2, and latest SVN version:
> > 8126)
> > Numpy Version: 1.5.0.dev8106
> >
> > I have been a long time Matplotlib user, and have never encountered this
> > issue. It started whenever I updated numpy and scipy. However,
> > reverting to the old version does not seem to fix the situation. I also
> > cannot get numpy to crash on its own. All my software works when
> > plotting is disabled.
> >
> > Has anyone else experienced this or similar problems during an upgrade?
> >
> > Joe
> >
> >
> >
> > ------------------------------------------------------------------------------
> > SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
> > Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
> > http://p.sf.net/sfu/solaris-dev2dev
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> > 
> 
-- 
Sincerely,
Joe

John, your assesment of the problem is correct. And I believe your suggested solution is also correct. Currently, each call to a mplot3d plot method is treated independantly. They get converted into custom PolyCollections which each do the Z-order sorting.
There is still an issue here however. Even if we implement the aformentioned solution, we are still only approximating a 3d library. And the result will still not be as nice as matlab. I believe that because we treat the surface as a series of 2D polygons, the intersection between two surfaces will be at the polygon edges. See the attached image for an example of what the intersection between a sphere and plane might look like. 
As a side note, this was a major barrier to me displaying multi-colored 3d bar plots as seen here: 
http://www.benaxelrod.com/temp/bar3d-2.png. But I fixed some color parameter issues in bar3d so that I can now call bar3d only once, and pass in color arrays so that now it renders properly. I will hopefully be able to submit a patch for this soon.
-Ben
-----Original Message-----
From: John Hunter [mailto:jd...@gm...] 
Sent: Sunday, February 21, 2010 9:24 PM
To: David Arnold
Cc: Ben Axelrod; Jakub Nowacki; mat...@li...; Reinier Heeres
Subject: Re: [Matplotlib-users] mplot3d stays?
Importance: Low
On Sun, Feb 21, 2010 at 7:15 PM, David Arnold <dwa...@su...> wrote:
> Hi,
>
> What prevents me from using mplot3d in the classroom is highlighted by the following example.
I believe the problem arises because each artist (ie each polygon, line or 3d text object) is rendered separately, and so there is no way different faces from the same object to be rendered on different sides of another object in the scene.
I am no expert on the mplot3d internals or pipeline, but it seems like the solution is for each artist to transform the faces of their respective polys and place them in a Axes3D level list (or other data
structure) along with their properties (eg facecolor, alpha) and then do a zordering and clipping at the axes level rather than the artist level before rendering. One might use a custom PolyCollection for this....
For those of you with more familiarity with mplot3d internals: is this approach viable/feasible?
JDH

Hi all,
I'm having some trouble with graphs ending up wider than I'd like. I'm
using matplotlib 0.98.5.2. If I do as follows:
import pylab
pylab.clf()
figure2 = pylab.figure(1)
axessubplot2 = pylab.subplot(111)
axessubplot2.fill_between([0, 1, 2, 3, 4, 5, 6], [0, 0, 0, 0, 0, 0,
0], [1, 1, 1, 1, 1, 0, 1], color='green')
axessubplot2.fill_between([4, 5, 6], [1, 0, 1], [1, 1, 1], color='red')
figure2.savefig('graph.png', dpi=100)
then the X-axis of the graph goes up to 7 and I get a white region at
the right-hand side, even though none of my calls have provided any
values for x>6. If I for example eliminate the first point and instead
do
import pylab
pylab.clf()
figure2 = pylab.figure(1)
axessubplot2 = pylab.subplot(111)
axessubplot2.fill_between([0, 1, 2, 3, 4, 5], [0, 0, 0, 0, 0, 0], [1,
1, 1, 1, 0, 1], color='green')
axessubplot2.fill_between([3, 4, 5], [1, 0, 1], [1, 1, 1], color='red')
figure2.savefig('graph.png', dpi=100)
then all is well and I don't get such a region.
So I guess I have two questions.
1) Is this a bug? It certainly feels like one...
2) Is there a workaround / what should I do instead?
Regards,
Geoff Bache

Erik Tollerud wrote:
> I'm curious if anyone knows a good way to embed pydot
> (http://code.google.com/p/pydot/) graphs (or really, any
> graphviz-style graphs) inside matplotlib somehow. I could easily
> write out a png or something from pydot and then imshow it, but that
> seems very kludgy. Is there some way to load svg or other vector data
> into matplotlib to be shown inside a figure?
Unfortunately not. It is possible to cut-and-paste graphviz SVG into a 
matplotlib SVG using Inkscape, but that's fairly kludgy also and can't 
be automated.
It would be nice to have a graphviz importer for matplotlib (something 
that would convert a dot stream to matplotlib artists), but that's a 
mid-sized development effort... It wouldn't consider it extremely 
difficult, but it would need someone to step up to implement it.
Mike
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

The drawing between the points is done using the regular "line drawing" 
commands of the backend. So there isn't really any low-level control 
over how the line between points is drawn.
As a workaround, however, you can interpolate the data yourself and just 
pass more points to matplotlib. You may also be interested in this 
document if you're implementing new geodesics (though it doesn't solve 
the interpolation problem you describe):
http://matplotlib.sourceforge.net/devel/add_new_projection.html
Mike
T J wrote:
> Hi,
>
> When plotting,
>
> plot(x, y, marker="-")
>
> and its similar markers, what functionality in MPL is responsible for
> interpolating between the points? My naive guess is that
> interpolation is done in "display" coordinates since everything looks
> nice even when zooming in. I inquire because I'd like to make
> interpolation between two points follow some other path between the
> two points. In other words, I'd like to make a plot structure which
> will follow a different type of geodesic. Any tips or pointers in the
> right direction would be greatly appreciated.
>
> Thanks!
>
> ------------------------------------------------------------------------------
> SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
> Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
> http://p.sf.net/sfu/solaris-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

Can you provide a gdb backtrace?
Run "gdb python", then at the gdb prompt type "run -c 'import pylab; 
pylab.clf(); pylab.plot(pylab.sin(range(101))); pylab.xlabel("Test X"); 
pylab.ylabel("Test Y"); pylab.show()'". After it segfaults, type "bt" 
to get a backtrace, and copy-and-paste it to this list.
Mike
Joseph D Cali wrote:
> I am encountering the following error:
>
> python -c 'import pylab; pylab.clf(); pylab.plot(pylab.sin(range(101)));
> pylab.xlabel("Test X"); pylab.ylabel("Test Y"); pylab.show()'
>
> /local_home/calijos/usr/lib/python2.6/site-packages/matplotlib/backends/backend_gtk.py:621: DeprecationWarning: Use the new widget gtk.Tooltip
> self.tooltips = gtk.Tooltips()
> Segmentation fault
>
> -------------------------------------------------------------
>
> But when I run the following command, I encounter no errors and plot is
> as expected:
>
> python -c 'import pylab; pylab.clf(); pylab.plot(pylab.sin(range(100)));
> pylab.xlabel("Test X"); pylab.ylabel("Test Y"); pylab.show()'
>
> I tried several other plots, any plots with less than 100 points work
> correctly, any plots with more than 100 points segfault.
>
> -------------------------------------------------------------
>
> My current system setup is:
>
> System: Ubuntu 9.04 x64
> Python: Python 2.6.2
> Matplotlib: >0.99.1 (tied 0.99.1.1, 0.99.1.2, and latest SVN version:
> 8126)
> Numpy Version: 1.5.0.dev8106
>
> I have been a long time Matplotlib user, and have never encountered this
> issue. It started whenever I updated numpy and scipy. However,
> reverting to the old version does not seem to fix the situation. I also
> cannot get numpy to crash on its own. All my software works when
> plotting is disabled.
>
> Has anyone else experienced this or similar problems during an upgrade?
>
> Joe
>
>
>
> ------------------------------------------------------------------------------
> SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
> Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
> http://p.sf.net/sfu/solaris-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

thanks both!
very nice LaTeX
2010年2月22日 Darren Dale <dsd...@gm...>
> On Mon, Feb 22, 2010 at 8:36 AM, Philipp Bender <li...@ro...>
> wrote:
> > Hi,
> >
> >> How do I print as label of an axis the Theta symbol - θ?
> >
> > you can take a look in the docs, there are examples how to use LaTeX for
> the
> > labels. You get the letter in LaTeX with \theta or \Theta.
>
> But be sure to either use raw strings or escape properly: r"\theta" or
> "\\theta"
>
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

On Mon, Feb 22, 2010 at 8:36 AM, Philipp Bender <li...@ro...> wrote:
> Hi,
>
>> How do I print as label of an axis the Theta symbol - θ?
>
> you can take a look in the docs, there are examples how to use LaTeX for the
> labels. You get the letter in LaTeX with \theta or \Theta.
But be sure to either use raw strings or escape properly: r"\theta" or "\\theta"

On Mon, Feb 22, 2010 at 2:42 AM, Jan Strube <cur...@gm...> wrote:
> Hi John, Eric,
> sorry to bug again, but was either of you able to reproduce my findings that
> in svn head the tick labels don't get printed if the formatter changes them
> to be outside the range of the axis?
On svn HEAD (r8142) (and the branch) I get the attached figure with
the attached code. I'm running the code with
> python test.py
from the python shell. Are you getting a different result?
JDH

Hi,
> How do I print as label of an axis the Theta symbol - θ?
you can take a look in the docs, there are examples how to use LaTeX for the 
labels. You get the letter in LaTeX with \theta or \Theta.

On Mon, Feb 22, 2010 at 4:26 AM, Wolfgang Kerzendorf
<wke...@ms...> wrote:
> Hello,
>
> Now that I have found the awesome widgets in matplotlib I want more: dropdown menus? will that come at some stage?
I have worked on it, but not finished it. I put the code in svn under examples
 http://matplotlib.sourceforge.net/examples/widgets/menu.html
but it would be migrated to matplotlib.widgets once it is completed.
If you or someone else wants to run with it, that would be great. It
would be handy to have this functionality.
I wrote this before JJ provided his nice containers in offsetbox, and
it should probably be reworked to use the OffsetbOX
JDH

Hi all,
How do I print as label of an axis the Theta symbol - θ?
thanks in advanced

Hello,
Now that I have found the awesome widgets in matplotlib I want more: dropdown menus? will that come at some stage?
Cheers
 Wolfgang

Hi John, Eric,
sorry to bug again, but was either of you able to reproduce my findings that
in svn head the tick labels don't get printed if the formatter changes them
to be outside the range of the axis?
Cheers,
 Jan
On Fri, Feb 19, 2010 at 10:40 AM, Jan Strube <cur...@gm...> wrote:
> Hi John,
>
> I am now at r8141.
> I don't see a difference. There's a one at the bottom of the scale, that's
> all.
> I am copying and pasting the code in your other email into a python
> console.
>
> It's a bit strange, because when I put in a print statement, I see that the
> method gets called a number of times...
> [edit] I see what's going on. If I replace vv with a large constant (like
> 100), it never gets printed.
> If I replace vv with a small constant (like 1), that is below the range of
> the scale, it does get printed at each tick.[/edit]
>
> My setup is kubuntu karmic, dependencies listed below.
> I have removed my matplotlibrc, but the result is the same. I am attaching
> it nevertheless...
>
> Cheers,
> Jan
>
>
>
> ============================================================================
>
> BUILDING
> MATPLOTLIB
>
> matplotlib:
> 1.0.svn
>
> python: 2.6.4 (r264:75706, Nov 2 2009, 14:44:17)
> [GCC
>
> 4.4.1]
>
> platform:
> linux2
>
>
>
> REQUIRED
> DEPENDENCIES
>
> numpy:
> 1.3.0
>
> freetype2:
> 9.20.3
>
>
>
> OPTIONAL BACKEND
> DEPENDENCIES
>
> libpng:
> 1.2.37
>
> Tkinter:
> no
>
> * TKAgg requires
> Tkinter
>
> wxPython:
> no
>
> * wxPython not
> found
>
> pkg-config: looking for pygtk-2.0
> gtk+-2.0
>
> * Package pygtk-2.0 was not found in the
> pkg-config
> * search path. Perhaps you should add the
> directory
> * containing `pygtk-2.0.pc' to the
> PKG_CONFIG_PATH
>
> * environment variable No package 'pygtk-2.0'
> found
> * Package gtk+-2.0 was not found in the
> pkg-config
> * search path. Perhaps you should add the
> directory
> * containing `gtk+-2.0.pc' to the
> PKG_CONFIG_PATH
>
> * environment variable No package 'gtk+-2.0'
> found
> * You may need to install 'dev' package(s)
> to
> * provide header
> files.
>
> Gtk+:
> no
>
> * Could not find Gtk+ headers in any
> of
>
> * '/usr/local/include', '/usr/include',
> '.'
> Mac OS X native:
> no
>
> Qt:
> no
>
> Qt4: Qt: 4.5.2, PyQt4:
> 4.6
>
> Cairo:
> 1.8.6
>
>
>
> OPTIONAL DATE/TIMEZONE
> DEPENDENCIES
>
> datetime: present, version
> unknown
>
> dateutil:
> 1.4.1
>
> pytz:
> 2009l
>
>
>
> OPTIONAL USETEX
> DEPENDENCIES
>
> dvipng:
> 1.12
>
> ghostscript:
> 8.70
>
> latex:
> 3.1415926
>
> pdftops:
> 0.12.0
>
>
>
> [Edit setup.cfg to suppress the above
> messages]
>
>
> ============================================================================
>
>
> On Thu, Feb 18, 2010 at 10:29 PM, John Hunter <jd...@gm...> wrote:
>
>> On Thu, Feb 18, 2010 at 4:06 PM, Jan Strube <cur...@gm...> wrote:
>> > Hi John,
>> > thanks for trying this also. Yes, I think it's a bug that not the scale
>> is
>> > log, but the data is.
>> > Unfortunately, the solution really doesn't work for me.
>> > Please see the attached screenshot. (Yes, it still says log_10 entries,
>> but
>> > the code is otherwise the same)
>> > In [2]: matplotlib.__version__
>> > Out[2]: '1.0.svn'
>> > This is r8063, I think.
>> > Strange that I get different results. Could this be a backend problem? I
>> use
>> > PyQT4.
>> > I'd be happy to also update from svn if you think that helps.
>>
>> I'm running svn but not svn HEAD -- you should try updating to HEAD
>> and I will do the same later (unfortunately HEAD is broken on my work
>> machine (solaris, python2.4) because of the CXX upgrade I put in some
>> time ago. I think I am on r8083.
>>
>> I do not think this difference could be caused by a backend or GUI
>> version difference as all of the formatting logic happens in the
>> frontend. If we are on the same version of svn, we should be getting
>> the same tick labels.
>>
>> JDH
>>
>
>

On Sun, Feb 21, 2010 at 4:02 PM, Ben Axelrod <BAx...@co...> wrote:
> I am not a MPL developer,
You are now :-)
> but I am using mplot3d quite heavily right now to support 3D plots for a client of mine. I have found many bugs and
> lacking features which I require in the mplot3d library and have modified my local copy of the code significantly. I am
> eagerly awaiting Reinier's return from vacation so that I can work with him to integrate my improvements. For the most
> part, these fixes simply make the 3D plots behave more like the 2D plots. Here is a tentative list of my changes so far:
>
> * bug fix: placement of title in 3D plots to match 2D plot behavior
> * bug fix: allow facecolors and edgecolors to be specified as 'none' in 3D scatter plots to match the 2D scatter plot behavior
> * bug fix: allow all keyword arguments to be used in text3D
> * bug fix: allow an array of colors to be passed into bar3d to specify the colors on a per-bar or per-face basis
> * bug fix: allow all keyword arguments to be used in bar3d
> * bug fix: allow 3d scatter plots with 3 or 4 points with colors specified
> * new feature: new method to disable mouse rotation in 3D plots
> * new feature: new Z-order sorting heuristic to eliminate rendering issues for the common case of using bar3d to visualize a 2D histogram
> * new feature: new method text2D
> * code cleanup: warn when canvas is None which disables mouse callbacks
> * code cleanup: fully document more methods in mplot3d
I'd be happy to take a look at this patch and commit it - Reinier can
review it and make any necessary changes when he gets back.
> Although I haven't written them yet, I can probably create a couple more example codes:
> * example code: demonstrate use of transform() to do rectangle selection in 3D scatter plots
> * example code: mplot3d with wx - demonstrate turning off mouse rotations to make pan and zoom toolbar buttons work properly
>
> There are a few other bugs that I would really like fixed, but can't quite figure out right now. Hopefully Reinier will be able to shed some light on these:
> * axis label picking for 3D axes
> * how to set axis tick label properties for 3D axes
> * allow 3d boxes with transparent faces to make "wireframe" boxes
> * fix z-order sorting across multiple calls to bar3d()
>
> I should note that because of my client, I have a vested interest in seeing mplot3d (with the above bug fixes) make it
> into a stable release of MPL. But at the same time, I don't have a lot of spare time to spend on MPL development.
I see no reason why they can't make it into the (overdue, upcoming)
1.0 if you can get a patch together in the next week or two.
JDH

On Sun, Feb 21, 2010 at 7:15 PM, David Arnold <dwa...@su...> wrote:
> Hi,
>
> What prevents me from using mplot3d in the classroom is highlighted by the following example.
I believe the problem arises because each artist (ie each polygon,
line or 3d text object) is rendered separately, and so there is no way
different faces from the same object to be rendered on different sides
of another object in the scene.
I am no expert on the mplot3d internals or pipeline, but it seems like
the solution is for each artist to transform the faces of their
respective polys and place them in a Axes3D level list (or other data
structure) along with their properties (eg facecolor, alpha) and then
do a zordering and clipping at the axes level rather than the artist
level before rendering. One might use a custom PolyCollection for
this....
For those of you with more familiarity with mplot3d internals: is this
approach viable/feasible?
JDH