matplotlib-users Mailing List for matplotlib

I'm having some problems understanding the difference between pylab.xticks()
and pylab.yticks()
Consider the following:

> import pylab as P
> import numpy as N
>
> data = N.random.random((10, 10))
> P.matshow(data)
> P.xticks([0, 1, 2], ['1', '2', '3'])
> P.show()


Why does this work, but if I change P.xticks to P.yticks, it doesn't?

The error message that I get back doesn't give me any insight as to what
could cause the difference between xticks and yticks.

 File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/matplotlib/figure.py",
line 612, in draw
    for a in self.axes: a.draw(renderer)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/matplotlib/axes.py",
line 1287, in draw
    self.transData.freeze()  # eval the lazy objects
ValueError: Domain error on eval_scalars in Transformation::freeze


 matplotlib.__version__
Out[2]: '0.90.1'

I am still in the learning phase, so any insight as to what's going on is
appreciated.

Best,
    Jan

Dear  all,
thanks for your help. this is what I was looking for!
Dirk


2007/9/26, Eric Firing <ef...@ha...>:
>
> David Huard wrote:
> > Hi Dirk,
> >
> > If you haven't already done so, look at the numpy.ma <http://numpy.ma/>
> > module. It provides a masked array object that deals gracefully with
> > missing values. To the best of my knowledge, most matplotlib functions
> > understand masked arrays and deal with it accordingly, exception made of
> > those requiring a full matrix (such as contour). Take a look at
>
> contour handles masked arrays correctly, as far as I know; contourf has
> some bugs in its masked array handling, but depending on the type and
> distribution of voids, it may still be good enough.
>
> pcolor and image have no problems with masked arrays.
>
> Eric
>
> > examples/image_masked.py. Also, in the Basemap toolkit, there is at
> > least one example showing how to plot a masked array on a map.
> >
> > Cheers,
> >
> > David
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Hello.

I've been having trouble getting Unicode characters to render.  I just
get a box in the title of my figure, rather than the character I need.

Here is my code:
#!/usr/bin/env python
from pylab import *
plot([1,2,3,4])
title(u"\u0251")
savefig("test.eps")
savefig("test.png")
show()

That character is LATIN SMALL LETTER ALPHA.  It's used in the
International Phonetic Alphabet.

I'm on Linux and I'm using matplotlib 0.90.1-2 (debian package
version).  I have a few TTF fonts in my system that contain that
glyph.  One is 'Arial Unicode MS', which I copied from my windows
machine.

As you can see, I will need to generate an EPS that renders the
character...  That EPS file will be imported into MS Word on a Windows
PC and printed.

I will happily use any solution that allows me to use that character
in the final product... :)  It doesn't have to be unicode..

I believe that my fonts are configured correctly on this Linux
system--I can use the Arial Unicode MS font in Open Office.  However,
I'm not sure that MPL is finding them.

When I point the TTFPATH environment variable a directory that only
contains ARIALUNI.TTF, I get gibberish for all characters in my
figure.

When I use ~/.matplotlib/matplotlibrc to list Arial Unicode MS as the
only font in the san-serif family, I don't observe any change in the
text in the figure.

...I did successfully instantiate an FT2FONT object out of my
ARIALUNI.TTF file, but, I didn't know what to do with it at that
point.

Help?

Cheers,
--Dave Loyall
Omaha, Nebraska, USA

On 9/27/07, Charles Seaton <cs...@st...> wrote:

> I am having the same problem as Eugen, and the suggested solution of using
> a.xaxis.get_major_locator().refresh()
> to force the creation of the full set of ticklabels doesn't seem to work for
> me.

matplotlib creates a prototypical tick (the prototick) and then
creates new ones on as as needed basis, copying properties from the
prototick.  Of course, position is one of the properties that cannot
be copied, which is why you are having trouble in your example.
Fortunately, there is an easy solution.

Call ax.xaxis.get_major_ticks() and access the label attribute:

for tick in ax.xaxis.get_major_ticks():
   label = tick.label1

Axis.get_major_ticks  will force a call to the locator and update the
tick list.  The Axes methods like get_xticklabels are just working on
the existing tick list rather than calling the get_major_ticks method
which is why you are not getting the full list.  This is a bug.  I
just made changes in svn so that all the accessor methods
(ax.get_xticklines, ax.get_yticklabels, and friends) all trigger a
call to axis.get_major_ticks rather so they should give the same
results going forward.


JDH


FYI, the Tick attributes are:

      tick1line  : a Line2D instance
      tick2line  : a Line2D instance
      gridline   : a Line2D instance
      label1     : a Text instance
      label2     : a Text instance
      gridOn     : a boolean which determines whether to draw the tickline
      tick1On    : a boolean which determines whether to draw the 1st tickline
      tick2On    : a boolean which determines whether to draw the 2nd tickline
      label1On   : a boolean which determines whether to draw tick label
      label2On   : a boolean which determines whether to draw tick label

On Friday 28 September 2007 07:36:23 am John Hunter wrote:
> On 9/27/07, Darren Dale <dd...@co...> wrote:
> > Hi Eric, John,
> >
> > Have either of you been following this thread?
>
> I am now :-)
>
> As Eric suggests, None is overloaded vis-a-vis color handling, because
> for mpl properties it generally means do the default as defined by rc.
>  For colors people often want to use None for "no color" which is why
> we added support for the string "None".  Does this work in your use
> case Tom?

The above exchange was off-list, we're back on now. I think that would be what 
Tom is looking for, but it doesnt work:

In [1]: plot([1,2])
Out[1]: [<matplotlib.lines.Line2D instance at 0x710cb0>]

In [2]: savefig('dsd.png', facecolor='None', edgecolor='None')
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)

/home/darren/<ipython console> in <module>()

/usr/lib64/python2.5/site-packages/matplotlib/pyplot.py in savefig(*args, 
**kwargs)
    272 def savefig(*args, **kwargs):
    273     fig = gcf()
--> 274     return fig.savefig(*args, **kwargs)
    275 if Figure.savefig.__doc__ is not None:
    276     savefig.__doc__ = dedent(Figure.savefig.__doc__)

/usr/lib64/python2.5/site-packages/matplotlib/figure.py in savefig(self, 
*args, **kwargs)
    768                 kwargs[key] = rcParams['savefig.%s'%key]
    769
--> 770         self.canvas.print_figure(*args, **kwargs)
    771
    772     def colorbar(self, mappable, cax=None, **kw):

/usr/lib64/python2.5/site-packages/matplotlib/backends/backend_qt4agg.py in 
print_figure(self, *args, **kwargs)
    153         self.update(l, self.renderer.height-t, w, h)
    154
    155     def print_figure(self, *args, **kwargs):
--> 156         FigureCanvasAgg.print_figure(self, *args, **kwargs)
    157         self.draw()

/usr/lib64/python2.5/site-packages/matplotlib/backend_bases.py in 
print_figure(self, filename, dpi, facecolor, edgecolor, orientation, format, 
**kwargs)
   1194                 edgecolor=edgecolor,
   1195                 orientation=orientation,
-> 1196                 **kwargs)
   1197         finally:
   1198             self.figure.dpi.set(origDPI)

/usr/lib64/python2.5/site-packages/matplotlib/backends/backend_agg.py in 
print_png(self, filename, *args, **kwargs)
    415
    416     def print_png(self, filename, *args, **kwargs):
--> 417         self.draw()
    418         self.get_renderer()._renderer.write_png(str(filename))
    419

/usr/lib64/python2.5/site-packages/matplotlib/backends/backend_qt4agg.py in 
draw(self)
    140         if DEBUG: print "FigureCanvasQtAgg.draw", self
    141         self.replot = True
--> 142         FigureCanvasAgg.draw(self)
    143         self.update()
    144

/usr/lib64/python2.5/site-packages/matplotlib/backends/backend_agg.py in 
draw(self)
    377
    378         self.renderer = self.get_renderer()
--> 379         self.figure.draw(self.renderer)
    380
    381     def get_renderer(self):

/usr/lib64/python2.5/site-packages/matplotlib/figure.py in draw(self, 
renderer)
    586         self.transFigure.freeze()  # eval the lazy objects
    587
--> 588         if self.frameon: self.figurePatch.draw(renderer)
    589
    590         for p in self.patches: p.draw(renderer)

/usr/lib64/python2.5/site-packages/matplotlib/patches.py in draw(self, 
renderer)
    198         #renderer.open_group('patch')
    199         gc = renderer.new_gc()
--> 200         gc.set_foreground(self._edgecolor)
    201         gc.set_linewidth(self._linewidth)
    202         gc.set_alpha(self._alpha)

/usr/lib64/python2.5/site-packages/matplotlib/backend_bases.py in 
set_foreground(self, fg, isRGB)
    617             self._rgb = fg
    618         else:
--> 619             self._rgb = colors.colorConverter.to_rgb(fg)
    620
    621     def set_graylevel(self, frac):

/usr/lib64/python2.5/site-packages/matplotlib/colors.py in to_rgb(self, arg)
    277
    278         except (KeyError, ValueError, TypeError), exc:
--> 279             raise ValueError('to_rgb: Invalid rgb arg "%s"\n%s' % 
(str(arg), exc))
    280             # Error messages could be improved by handling TypeError
    281             # separately; but this should be rare and not too hard

ValueError: to_rgb: Invalid rgb arg "None"
invalid literal for float(): None

After thinking it over, I did not go for
Robert or David's cool numpy tricks, but
I'll append a simple object in case someone
else wants to do more.

Cheers,
Alan Isaac

class EmpiricalCDF(object):
    '''Empirical cdf.
    First point will be (xmin,0).
    Last point will be (xmax,1).
    '''
    def __init__(self, data, sortdata=True):
        if sortdata:
            data = N.sort(data)
        self.data = data
        self.nobs = len(data)
    def gen_xp(self):
        data, nobs = self.data, self.nobs
        prob = N.linspace(0, 1, nobs+1)
        xsteps = ( data[(idx)//2] for idx in xrange(2*nobs) )
        psteps = ( prob[(idx+1)//2] for idx in xrange(2*nobs) )
        return xsteps, psteps
    def get_steps(self):
        '''Return: 2-tuple of arrays,
        the data values and corresponding cumulative 
        probabilities.
        '''
        xsteps, psteps = self.gen_xp()
        return N.fromiter(xsteps,'f'), N.fromiter(psteps,'f')

Hal Huntley wrote:
> Thanks to Eric Firing and Christopher Barker for input on trying to
> resolve the problem.  Christopher said:
> %%%
> You might try just:
> 
> easy_install numpy
> 
> easy_install matplotlib.
> %%%
> 
> 
> I did that and now the problem moved and I get:
> ===
>>>> from pylab import *
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/pylab.py", line 1, in ?
>     from matplotlib.pylab import *
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/pylab.py", line 222, in ?
>     new_figure_manager, draw_if_interactive, show = pylab_setup()
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/__init__.py", line 24, in pylab_setup
>     globals(),locals(),[backend_name])
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gtkagg.py", line 10, in ?
>     from matplotlib.backends.backend_gtk import gtk, FigureManagerGTK,
>     FigureCanvasGTK,\
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gtk.py", line 21, in ?
>     from matplotlib.backends.backend_gdk import RendererGDK, FigureCanvasGDK
>   File
>   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gdk.py", line 35, in ?
>     from matplotlib.backends._ns_backend_gdk import pixbuf_get_pixels_array
> ImportError: No module named _ns_backend_gdkd4
> ===

Is it a rotten egg?

> 
> Googling around has indicated that the X11 stuff wasn't available when the
> numpy and matplotlib were made. The gtk files and "-devel" seem to be there
> when I do an "rpm -qa".  I went and got a new numpy from source and did
> "python setup.py install".  It is interesting that when I get in to python
> now, I can do:
> 
>>>> import gtk
>>>> import numpy
>>>> import matplotlib
> 
> and they all just return the prompt, indicating, I thought, that the
> programs were installed ok.  
> 
> Here is the naive question ->  Is there something wrong, then with a
> "from pylab import *"? The user is trying to do that. 

No, that should work fine.  I don't know how the egg-based 
matplotlib/pylab should work with the installed-from-source numpy, 
though.  If the versions are compatible, then I expect it would work.

You gave a traceback resulting from "from pylab import *" after 
installing the two eggs, correct?  If so, what was the result after you 
installed numpy from source?  I would expect no difference, because the 
problem reported in the traceback is a missing matplotlib module, not a 
missing numpy module.

Or did you mean that you installed matplotlib from source?  That would 
make more sense, and certainly should work if all the right header files 
and libraries are present.

If you installed matplotlib from source, what do you get from
ls /usr/lib/python2.4/site-packages/matplotlib/backends/*.so

Another diagnostic would be to delete the build directory from the 
matplotlib source tree (assuming you are now building matplotlib from 
source), and save the output from the "python setup.py build" command. 
This should make it clear whether the necessary headers really were found.

Eric

Looking back through the archives, I found this discussion of manipulating
ticklabels.


Jouni K. Sepp=C3=A4nen wrote:
>=20
> Eugen Wintersberger <eug...@jk...>
> writes:
>=20
>> The first line is the printing of the ticklabel list before the
>> pylab.show() command. The second after pylab.show(). Why the list
>> contains 1 entry before and 7 (as is should) after pylab.show()?
>=20
> I think matplotlib is deferring the creation of the tick labels because
> for all it knows, you might be going to plot something that will cause
> the axes to be rescaled and change the tick locations and labels.
>=20
>> I would like to access the labels before the final show command in a
>> script. But how is this possible with this behavior?=20
>=20
> You could set the tick positions and labels explicitly, or if you like
> the default tick locator, add the following before getting the ticklabel
> objects:
>=20
> a.xaxis.get_major_locator().refresh()
>=20

I am having the same problem as Eugen, and the suggested solution of using=
=20
a.xaxis.get_major_locator().refresh()
to force the creation of the full set of ticklabels doesn't seem to work fo=
r
me.

Sample script:

from pylab import *
#ion()
a =3Daxes([0.2,0.2,0.7,0.7])
t =3D arange(0.0, 2.0, 0.01)
s =3D sin(2*pi*t)
a.plot(t, s)
a.grid(True)
# matlab handle graphics style
a.xaxis.get_major_locator().refresh()
xticklabels =3D getp(a, 'xticklabels')
setp(xticklabels, 'color', 'r', fontsize=3D'medium')
setp(xticklabels, dashdirection=3D0, dashlength=3D10, dashrotation=3D90)
#savefig('axprops_demo')
show()

even with a.xaxis.get_major_locator().refresh(), only the first ticklabel
has its position adjusted.

replacing=20
setp(xticklabels, dashdirection=3D0, dashlength=3D10, dashrotation=3D90)
with code modifying the position
for i in range(len(xticklabels)):
   xtp =3D getp(xticklabels[i],'position')
   setp(xticklabels[i],position=3D(xtp[0],-0.05))

also only effects the first ticklabel.

changes to color and fontsize of the first ticklabel do carry over the full
set of ticklabels, but TextWithDash properties do not carry over.=20

Uncommenting ion() at the beginning of the script generates the desired
image with offset ticklabels, but requires using interactive mode.

Any help either getting=20
a.xaxis.get_major_locator().refresh()
to work properly, or to get TextWithDash properties to carry over from the
initial single ticklabel to the full set of ticklabels would be greatly
appreciated.=20

thanks,

Charles Seaton
Research Associate
STC-CMOP
www.stccmop.org

--=20
View this message in context: http://www.nabble.com/Problem-with-tick-label=
s-in-scripts-tf4260235.html#a12932137
Sent from the matplotlib - users mailing list archive at Nabble.com.

Thanks to Eric Firing and Christopher Barker for input on trying to
resolve the problem.  Christopher said:
%%%
You might try just:

easy_install numpy

easy_install matplotlib.
%%%


I did that and now the problem moved and I get:
===
>>> from pylab import *
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/pylab.py", line 1, in ?
    from matplotlib.pylab import *
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/pylab.py", line 222, in ?
    new_figure_manager, draw_if_interactive, show = pylab_setup()
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/__init__.py", line 24, in pylab_setup
    globals(),locals(),[backend_name])
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gtkagg.py", line 10, in ?
    from matplotlib.backends.backend_gtk import gtk, FigureManagerGTK,
    FigureCanvasGTK,\
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gtk.py", line 21, in ?
    from matplotlib.backends.backend_gdk import RendererGDK, FigureCanvasGDK
  File
  "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-linux-i686.egg/matplotlib/backends/backend_gdk.py", line 35, in ?
    from matplotlib.backends._ns_backend_gdk import pixbuf_get_pixels_array
ImportError: No module named _ns_backend_gdk
===

Googling around has indicated that the X11 stuff wasn't available when the
numpy and matplotlib were made. The gtk files and "-devel" seem to be there
when I do an "rpm -qa".  I went and got a new numpy from source and did
"python setup.py install".  It is interesting that when I get in to python
now, I can do:

>>> import gtk
>>> import numpy
>>> import matplotlib

and they all just return the prompt, indicating, I thought, that the
programs were installed ok.  

Here is the naive question ->  Is there something wrong, then with a
"from pylab import *"? The user is trying to do that. 

Thanks for any assistance that can be given.

Hal
=======================


> Hal Huntley wrote:

> > I'm a sys admin trying to install the matplotlib python package for
> > someone on our staff. I'm not overly familar with python.

> > We are trying to use matplotlib with pylab.  

> > uname -a
> > Linux andromeda 2.6.13-1.1532_FC4smp #1 SMP Thu Oct 20 01:51:51 EDT 2005
> > i686 i686 i386 GNU/Linux

> > The python version we have is the default that we got with the Fedora
> > Core 4 package.

> > python -V
> > Python 2.4.1

> > I've installed using /usr/bin/easy_install:
> > matplotlib-0.90.1-py2.4-win32.egg  numpy-1.0.3.1-py2.4-win32.egg
> 
> I've never understood eggs and easy_install, but it certainly doesn't 
> look right that you are installing things labelled "win32" on a Linux 
> box!  I'm surprised you got that far.
> 
> For linux, I like to install from source, since the distro packages for 
> numpy and mpl are usually too old.  If you get the tarballs, build and 
> install numpy first, then mpl.  You don't need or want Numeric or 
> numarray.  The build process may fail.  If it does, it is because you 
> need to install distro packages with headers for various libraries, and 
> you should be able to see from the failures which headers are missing.
> 
> Maybe someone else can provide more detailed instructions for Fedora C4.
> 
> Eric
> 
> 
> 
> > 
> > If we get in to python and give the command:
> >>>> from pylab import *
> > 
> > We get:
> > ==
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/pylab.py", line 1, in ?
> >     from matplotlib.pylab import *
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/matplotlib/pylab.py", line 201, in ?
> >     import mlab  #so I can override hist, psd, etc...
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/matplotlib/mlab.py", line 64, in ?
> >     import nxutils
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/matplotlib/nxutils.py", line 17, in ?
> >     from matplotlib._ns_nxutils import *
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/matplotlib/_ns_nxutils.py", line 7, in ?
> >     __bootstrap__()
> >   File
> >   "/usr/lib/python2.4/site-packages/matplotlib-0.90.1-py2.4-win32.egg/matplotlib/_ns_nxutils.py", line 5, in __bootstrap__
> >     del __bootstrap__, __loader__
> > NameError: global name '__loader__' is not defined
> > ==
> > 
> > I've read that numeric may need to be installed, so I got "Numeric-24.2"
> > and did:
> >     python setup.py build
> >     python setup.py install
> > 
> > That didn't help solve the issue.
> > 
> > Is there some other package that needs to be installed?  I checked the FAQ
> > and this issue does not seem to be there.
> > 
> > Regards,
> > 
> > Hal Huntley
> > SRI International

On 9/27/07, Tom Johnson <tj...@gm...> wrote:
> However, it will be a little while before I can provide a status
> update----I have a presentation very soon and do not have the time to
> regenerate all my images so that the background matches my
> presentation background.  Since my setup had been creating
> 'transparent' facecolors, I did not bother setting facecolor at all
> and kept the default. :(

Assuming that GPL 8.60 fixes the problem, do you anticipate that mpl
will move to support None as a color?  I realize this is a nontrivial
change to the color functionality...but there is a major portability
benefit.  With None as an option, I would probably always set the axis
background color, figure facecolor, and figure edgecolor to None.
Then, my figures would work no matter which theme I selected for my
presentation....and it avoids the situation I am in now, which
requires significant time to regenerate images.


>
> Thanks!
>

On 9/27/07, Darren Dale <dd...@co...> wrote:
> On Thursday 27 September 2007 01:28:46 am Tom Johnson wrote:
> > On 9/26/07, Darren Dale <dd...@co...> wrote:
> > > I used your script to create the eps file, and created the attached
> > > postscript
> > > (you need an \end{document} in your latex code).
> >
> > Whoops!
> >
> >
> > Do you see anything wrong
> >
> > > with the resulting postscript? It looks fine to me.
> >
> > Indeed.  I didn't realize this before, but the problem is actually with the
> > pdf.  I have attached it.  Can you confirm that your pdf looks like mine?
>
> No, it does not look like yours. See attached.


Interesting.


>
> > Mine looks like this no matter which viewer I use (acrobat, evince, xpdf).
> > This makes me wonder if it is 1) the eps file or 2) the compilation
> > process.
>
> It is probably a problem with either ghostscript or pdftops.
>
> > Actually, the problem exists as early as the dvi file.
>
> The dvi looks fine here, and so does my pdf. It is often the case that
> problems with usetex are solved by updating the external dependencies. I am
> using:
>
> GPL Ghostscript 8.60
> pdftops version 3.00
> pdfeTeX 3.141592-1.30.5-2.2 (tetex-3.0_p1)
>
>
>


Hmm...I have:

ESP Ghostscript 8.15.04 (2007-03-14)
pdftops version 3.01    (coming from libpoppler1 version 0.5.4-0ubuntu8)
pdfeTeX 3.141592-1.21a-2.2 (tetex-3.0.dfsg.3-4)

The CUPS page (http://www.cups.org/espgs/index.php) indicates that ESP
8.15.04 and GPL 8.57 have merged into 8.60.  This is probably the
solution.  So I will give that a try as a first fix and report back to
the list.

However, it will be a little while before I can provide a status
update----I have a presentation very soon and do not have the time to
regenerate all my images so that the background matches my
presentation background.  Since my setup had been creating
'transparent' facecolors, I did not bother setting facecolor at all
and kept the default. :(

Thanks!

On Thu, 27 Sep 2007, David Huard apparently wrote:
> scipy/sandbox/dhuard/stats.py 

Well, I also did not think of a double application of 
argsort to rank the observations. Nice.  OK, I've got plenty 
to work with now.

Thanks,
Alan

On Wed, 26 Sep 2007, Robert Kern apparently wrote:
> n = len(x)
> x2 = np.repeat(x, 2)
> y2 = np.hstack([0.0, np.repeat(np.arange(1,n) / float(n), 2), 1.0])
> pylab.plot(x2, y2)

OK, that's pretty slick.
I did not think about ``repeat``.
Thanks,
Alan

Hi Alan,

There is an empiricalcdf function in scipy/sandbox/dhuard/stats.py
It's not fancy but it might do what you want.

David

2007/9/26, Alan Isaac <ai...@am...>:
>
> Is there a standard function or practice for
> plotting the CDF of a series?  (I am aware
> of the output of hist.)
>
> Thank you,
> Alan Isaac
>
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

On Thursday 27 September 2007 09:34:28 am Darren Dale wrote:
> On Thursday 27 September 2007 01:28:46 am Tom Johnson wrote:
> > On 9/26/07, Darren Dale <dd...@co...> wrote:
> > > I used your script to create the eps file, and created the attached
> > > postscript
> > > (you need an \end{document} in your latex code).
> >
> > Whoops!
> >
> >
> > Do you see anything wrong
> >
> > > with the resulting postscript? It looks fine to me.
> >
> > Indeed.  I didn't realize this before, but the problem is actually with
> > the pdf.  I have attached it.  Can you confirm that your pdf looks like
> > mine?
>
> No, it does not look like yours. See attached.
>
> > Mine looks like this no matter which viewer I use (acrobat, evince,
> > xpdf). This makes me wonder if it is 1) the eps file or 2) the
> > compilation process.
>
> It is probably a problem with either ghostscript or pdftops.
>
> > Actually, the problem exists as early as the dvi file.
>
> The dvi looks fine here, and so does my pdf. It is often the case that
> problems with usetex are solved by updating the external dependencies. I am
> using:
>
> GPL Ghostscript 8.60
> pdftops version 3.00
> pdfeTeX 3.141592-1.30.5-2.2 (tetex-3.0_p1)

On my machine, pdftops is provided by poppler-0.6, not xpdf.

Darren

On Thursday 27 September 2007 01:28:46 am Tom Johnson wrote:
> On 9/26/07, Darren Dale <dd...@co...> wrote:
> > I used your script to create the eps file, and created the attached
> > postscript
> > (you need an \end{document} in your latex code).
>
> Whoops!
>
>
> Do you see anything wrong
>
> > with the resulting postscript? It looks fine to me.
>
> Indeed.  I didn't realize this before, but the problem is actually with the
> pdf.  I have attached it.  Can you confirm that your pdf looks like mine?

No, it does not look like yours. See attached.

> Mine looks like this no matter which viewer I use (acrobat, evince, xpdf).
> This makes me wonder if it is 1) the eps file or 2) the compilation
> process.

It is probably a problem with either ghostscript or pdftops.

> Actually, the problem exists as early as the dvi file.

The dvi looks fine here, and so does my pdf. It is often the case that 
problems with usetex are solved by updating the external dependencies. I am 
using:

GPL Ghostscript 8.60
pdftops version 3.00
pdfeTeX 3.141592-1.30.5-2.2 (tetex-3.0_p1)

Thanks much! But, it didn't help. The second figure got still really
slow and closes right away after it finishes drawing not giving enough
time to me to see the figure. Then, it crashed for the third one. Any
idea?

Young Jin

-----Original Message-----
From: Wolfgang Kerzendorf [mailto:wke...@go...]=20
Sent: Wednesday, September 26, 2007 9:03 PM
To: Lee, Young-Jin
Cc: mat...@li...
Subject: Re: [Matplotlib-users] multiple figures interactively

This might help you it destroys the whole window:
pylab.get_current_fig_manager().destroy()
and then you open it again with show
Lee, Young-Jin wrote:
>
> Hi,
>
> I 'm writing a python program that draws figures one by one=20
> interactively with the user's input in dos mode. Basically, I give the

> program a decision after each figure and then it draws the next one.=20
> After the first figure, it got very much slowed down for the second=20
> one and crashed for the third one. I feel like it has some memory=20
> issues as I keep using 'show' after I close each. I used 'clf()'=20
> before I draw a new one, but it doesn't seem to help. Any idea?
Thanks.
>
> Young Jin
>
>
------------------------------------------------------------------------
>
>
------------------------------------------------------------------------
-
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>
------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>  =20

Andrew Straw wrote:
> I came across this piece by Adam Hupp on programming.reddit.com just 
> now. It looks interesting:
>
> http://hupp.org/adam/weblog/2007/09/03/etframes-applying-the-ideas-of-edward-tufte-to-matplotlib/ 
>
>   
This is indeed really interesting. Thanks for this link,

David

This might help you it destroys the whole window:
pylab.get_current_fig_manager().destroy()
and then you open it again with show
Lee, Young-Jin wrote:
>
> Hi,
>
> I ’m writing a python program that draws figures one by one 
> interactively with the user’s input in dos mode. Basically, I give the 
> program a decision after each figure and then it draws the next one. 
> After the first figure, it got very much slowed down for the second 
> one and crashed for the third one. I feel like it has some memory 
> issues as I keep using ‘show’ after I close each. I used ‘clf()’ 
> before I draw a new one, but it doesn’t seem to help. Any idea? Thanks.
>
> Young Jin
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> ------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>   

Alan Isaac wrote:
> Is there a standard function or practice for
> plotting the CDF of a series?  (I am aware
> of the output of hist.)

import numpy as np
from matplotlib import pylab

x = ... # whatever
n = len(x)
x2 = np.repeat(x, 2)
y2 = np.hstack([0.0, np.repeat(np.arange(1,n) / float(n), 2), 1.0])
pylab.plot(x2, y2)

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco

On Sep 26, 2007, at 5:09 PM, Alan Isaac wrote:

> Is there a standard function or practice for
> plotting the CDF of a series?  (I am aware
> of the output of hist.)

Without really knowing what CDF is (I am assuming
it is Cumulative Density Fucntion or something similar).
I would suggest taking a look at numpy and the histogram
and cumsum functions therein.

Cheers
   Tommy

Tom Johnson wrote:
> How can I keep a 1:1 aspect ratio after adding a colorbar? 

After adding the colorbar, try

axis('scaled')
or
axis('image')
or
axis('equal')

Or for full control use the set_aspect() method of the Axes object for 
your plot.  There is also an axis() method corresponding to the pylab 
axis() function if you are using an OO style.

help(axis) will explain the various axis() options.

Eric
> 
> ratio='exact' seems to include the colorbar in the calculations.  For 
> example, if I have a circle and color it....I want it to still look like 
> a circle (rather than an ellipse) after adding the colorbar.
> 
> ~t
> 
> 
> ------------------------------------------------------------------------
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Is there a standard function or practice for
plotting the CDF of a series?  (I am aware
of the output of hist.)

Thank you,
Alan Isaac

On 9/26/07, Simon Forman <saj...@gm...> wrote:
> Hello,
>
> I just installed Matplotlib (and NumPy) on a windows XP machine, and
> I'm getting the following traceback when I try to use the TkAgg
> backend.
>
> Python 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.1310 32 bit (Intel)] on
> win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import matplotlib
> >>> matplotlib.use('TkAgg')
> >>> from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File "C:\Python25\Lib\site-packages\matplotlib\backends\backend_tkagg.py",
> line 8, in <module>
>     import tkagg                 # Paint image to Tk photo blitter extension
>   File "C:\Python25\lib\site-packages\matplotlib\backends\tkagg.py",
> line 1, in <module>
>     import _tkagg
> ImportError: DLL load failed: The specified module could not be found.
> >>>
>
> I found one old (2004) post
> http://mail.python.org/pipermail/python-list/2004-April/258963.html
> that seems to indicate that some sort version mismatch of the Tk/Tcl
> libraries may be to blame.  But I don't know how to diagnose that or
> what to do about it.
>
> FWIW, I installed using matplotlib-0.90.1.win32-py2.5.exe and
> numpy-1.0.3.1.win32-py2.5.exe.
>
> Thanks in advance for any help you can give me.
> Sincerely,
> ~Simon
>

Please disregard.  I was missing MSVCP71.dll.

Sorry for the 'noise',
Sincerely,
~Simon

On Wednesday 26 September 2007 03:30:41 pm Tom Johnson wrote:
> Any comments on this?  The behavior is, at best, inconsistent.  At worst,
> the current behavior is incorrect, as it is not possible to have a white
> facecolor when using usetex/xpdf.

I used your script to create the eps file, and created the attached postscript 
(you need an \end{document} in your latex code). Do you see anything wrong 
with the resulting postscript? It looks fine to me.

> On 9/23/07, Tom Johnson <tj...@gm...> wrote:
> > On 9/23/07, Tom Johnson <tj...@gm...> wrote:
> > > Also, it seems like there needs to be an extra keyword or option.
> > > Suppose someone wanted a white facecolor in the (usetex=True) EPS file.
> > >  It doesn't seem like this is currently possible.  It would be nice if
> > > I could specify:
> > >
> > > savefig(file, facecolor=None, edgecolor=None)
> >
> > It would also be nice if we could specify the axis background color to be
> > None as well (again, perhaps only useful when creating an EPS)...when
> > using the plot command.

That is probably a nontrivial change to mpl's existing color support.

Darren