matplotlib-users Mailing List for matplotlib (Page 12)
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Hello,

in my matplotlibrc, I use 'backend : WXAgg'. This works fine, since I 
use a wxPython shell as an interactive shell with pylab. However, 
'WXAgg' is not in the list of possible backends given in matplotlibrc. 
Is it save (also in the future) to use WXAgg, or should I try to switch 
to one of the other backends?

Thanks,
Lars

Hello,

I would like to have multiple image plots in a figure. Each plot should 
have its own colorbar. I tried the following:

**********************************
a = N.array(((1,2,3), (4,5,6)))

P.figure(0)
P.subplot(1,2,1)
P.imshow(a)
P.colorbar()

P.subplot(1,2,2)
P.imshow(a)
P.colorbar()
************************************

The two images display and have their own colorbar, each. However, the 
placement is not optimal. (see attached 'colorbar1.png') But when I try 
to use the "Configure subplot parameters" feature in the interactive 
figure window, only the image plots are affected and the colorbars stay 
as they are. When I hit the reset button in the "configure subplots" 
dialogue, the figure looks different from the way it looked, when it was 
generated (see attached 'colorbar2.png'). Do I have to do all the 
placement on my own by using axes.set_position or is there a more 
comfortable way using the subplot syntax?

Thanks
Lars

Using latest matplotlib (0.91.1) and have two about date formatting.
The following two questions are put here in attempt to get a plot of a month
with days 1-31 neatly fit onto the x axis.

1.  Using DateFormatter class to format dates and am confused as
to how to set the axis properties like font size and font family of the
xtick labels.  The following type of thing (leaving out the most of the
code) does not work...

        daysFmt = DateFormatter('%d')
        ax.xaxis.set_major_locator(days)
        ax.xaxis.set_major_formatter(daysFmt)

        ax.xaxis.set_minor_locator(days2)
        ax.autoscale_view()
        ax.set_xticklabels(xdates,family='serif', fontsize=8)

...because when I put that last line I do get the font size changed
but I lose the date formatting via DateFormatter.  However if I put
only the .set_major_formatter(daysFmt) line, I get the right date
formatting...but how do I change the font?

2.  I want the plot to display, for example, the month of Jan,
with *only* days 1 - 31 shown on the x axis.   But if I use the
DateFormatter  (leaving out the  .set_xticklabels line), for the date
range given (days 1-31) it includes a couple of days on either side
of the plot (the 31st from the previous month and a day or two into
the following month).  How can I get it to display only 1-31?  And if I
do, will the  points plotted on day 1 and day 31 be halfway cut off
by th edges of the plot? (and is there any fix for that?)

Any help is appreciated.

In fact, that's what I wanted to do : put the origine of my axis on le upper
left corner of my axis
with the values of y increase from top to bottom. And the tip you gave
(set_ylim) do it kindly.

Thank you again for your help.

Jaonary

On Dec 4, 2007 4:22 AM, John Hunter <jd...@gm...> wrote:

> On Dec 3, 2007 7:50 AM, Jaonary Rabarisoa <ja...@gm...> wrote:
> > Hi all,
> >
> > I'm relatively new to matplotlib and what I'll ask to you seems to be
> simple
> > but I just can't figure out how to do this. I'd like to change the
> origin of
> > my coordinate system  to the upper left of my axis when I plot my
> figures.
> > In matlab I use axis('ij') but in matplotlib I couldn't find something
> > similar. How will I do this ?
>
> I am not 100% sure what you are asking and I don't have a matlab
> install handy to test.  If you want to plot a decreasing scale on the
> y-axis, that is easy, just invert the min and max args to the ylim
>
> ax.set_ylim(ymax, ymin)
>
> If you want to do something more sophisticated, please describe in
> detail and then I may have to tell you it is not currently supported
> :-(
>
> JDH
>

Hi,
I limit the value range of my imshow using vmin and vmax. As I like the
way the contour levels are usually coloured by contour, I would like to
just restrict them to the same range but have matplotlib use the
default spacing and colouring. In other words, I want to avoid setting
levels manually. Is that possible at all? There is obviously no vmin
and vmax in contour but can I do this is some other way? Or do I have
to set up my own array for the levels to achieve this?

Cheers,
Christian

On Dec 3, 2007 7:34 PM, Barry Wark <bar...@gm...> wrote:
> sorry instances of self.plotView.image
>
> On Dec 3, 2007 3:38 PM, Chris Fonnesbeck <lis...@ma...> wrote:
>
>

I replaced the one instance of this that I found. I now get the following error:

In [7]: show()
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)

/Users/chris/<ipython console> in <module>()

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in show()
     44 def show():
     45     for manager in Gcf.get_all_fig_managers():
---> 46         manager.show()
     47
     48 def draw_if_interactive():

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in show(self)
    203                 mplBundle,
    204                 False)
--> 205         NSApplication.sharedApplication().run()
    206
    207

ValueError: NSInvalidArgumentException - NSImageCell's object value
must be an NSImage.


-- 
Christopher J. Fonnesbeck
+ Fish & Wildlife Research Institute (FWC)
+ 727.235.5570

On Dec 3, 2007 9:41 PM, John Hunter <jd...@gm...> wrote:
> On Dec 3, 2007 10:13 AM, John Hunter <jd...@gm...> wrote:
>
> > Sorry for the confusion, I was still on the *old* problem of tkagg
> > hanging w/ hist, and was wondering if it was hanging with other plot
> > commands too.  Not that you can easily test now w/ the active tcl
> > install.
>
> Chris, if you get some time to explore this further, I'd like to see
> if we can sort it out.  If you could remove the acive tcl and do a
> clean rebuild of mpl, I'd like to verify the hang still exists, and
> whether it is hist or all plot commands, and et some version info (eg
> what tcl/tk) and see a copy of the mpl build output.  Also, I'd like
> to see a run w/ --verbose-debug output.  I find it strange because I
> am on the same platform and am not having any troubles (yet) with
> tkagg.

John,

I removed ActiveTcl and the hang seems to have disappeared, both for
hist and plot.

cf

-- 
Christopher J. Fonnesbeck
+ Fish & Wildlife Research Institute (FWC)
+ 727.235.5570

On Dec 3, 2007 9:08 AM, Jos=E9 G=F3mez-Dans <jgo...@gm...> wrote:
> Hi,
> I have compiled v.0.90.1 on RHEL 5. By default, the GTKAgg backend is bei=
ng
> used (TkAgg cannot be set, as TkInter is not installed on the system, I
> think. It throws a "NO Module named Tkinter" error).
>
> At any rate, a test session is as follows:
> import pylab
> pylab.plot ( [1,2,3],[1,2,3],'-or') #Wait for a long time, up to 3-4 minu=
tes
> pylab.show() # Display the graph. Quite fast <1sec
>
> This delay is very unusual, and does not happen on my laptop. Does anyone=
 have
> any ideas?

This is not happening for me.  Does it happen every time you run mpl
or just the first time after you install.  Run some test script, eg

  > cat test.py
  from pylab import *
  plot([1,2,3])
  show()

with

  > python test.py -dGTKAgg --verbose-debug

and see if you can get some clue what mpl is doing when it hangs.
Also, post the verbose output to the list.

Thanks,
JDH

On Dec 3, 2007 7:50 AM, Jaonary Rabarisoa <ja...@gm...> wrote:
> Hi all,
>
> I'm relatively new to matplotlib and what I'll ask to you seems to be simple
> but I just can't figure out how to do this. I'd like to change the origin of
> my coordinate system  to the upper left of my axis when I plot my figures.
> In matlab I use axis('ij') but in matplotlib I couldn't find something
> similar. How will I do this ?

I am not 100% sure what you are asking and I don't have a matlab
install handy to test.  If you want to plot a decreasing scale on the
y-axis, that is easy, just invert the min and max args to the ylim

ax.set_ylim(ymax, ymin)

If you want to do something more sophisticated, please describe in
detail and then I may have to tell you it is not currently supported
:-(

JDH

On Dec 3, 2007 4:57 AM, Dirk Zickermann <dir...@go...> wrote:
> Dear users,
> I would like to generate a colored histogram using a colormap.
> via
> ...
> vals, bins, patchs = pylab.hist(hist_data ,75 ,normed=1)
> ..
> Does anybody know how to assign a colormap to a histogram to find map data
> so that one can find easily data from a 2d map in the histogram?

We need to rewrite histogram to use a PatchCollection, which would be
more efficient and would support colormapping out of the box.  In the
current implementation, you can do it by using the colormap
normalization code and mapping manually on each rectangle.  Here is an
example:

import numpy as n
from pylab import figure, show
import matplotlib.cm as cm
import matplotlib.colors as colors

fig = figure()
ax = fig.add_subplot(111)
Ntotal = 1000
N, bins, patches = ax.hist(n.random.rand(Ntotal), 20)

#I'll color code by height, but you could use any scalar


# we need to normalize the data to 0..1 for the full
# range of the colormap
fracs = N.astype(float)/N.max()
norm = colors.normalize(fracs.min(), fracs.max())

for thisfrac, thispatch in zip(fracs, patches):
    color = cm.jet(norm(thisfrac))
    thispatch.set_facecolor(color)


show()

I'll remove the exe's from SF.  It's been a while since a release, and
I guess I was a little rusty. ;)

- Charlie

On Dec 3, 2007 7:02 PM, John Hunter <jd...@gm...> wrote:
>
> On Dec 3, 2007 5:53 PM, John Hunter <jd...@gm...> wrote:
> > On Dec 3, 2007 5:49 PM, John Hunter <jd...@gm...> wrote:
> >
> > > Charlie, I don't know how you handled this last time, but is there
> > > something in setuptools you have to disable for this build?
> >
> > Well, here's a clue: matplotlib/__init__.py does not exist in this instal
>
> When I copy all the dirs from the 2.5 egg you built into
> site-packages, I get a matplotlib with rcParams and the rest of the
> __init__.py stuff.  Unfortunately, when I try to import pylab, I get
> the dreaded old msvcp71.dll error.  If I recall correctly, we used to
> hack distutils to remove the line that linked with that lib.
>
> JDH
>

On Dec 3, 2007 10:13 AM, John Hunter <jd...@gm...> wrote:

> Sorry for the confusion, I was still on the *old* problem of tkagg
> hanging w/ hist, and was wondering if it was hanging with other plot
> commands too.  Not that you can easily test now w/ the active tcl
> install.

Chris, if you get some time to explore this further, I'd like to see
if we can sort it out.  If you could remove the acive tcl and do a
clean rebuild of mpl, I'd like to verify the hang still exists, and
whether it is hist or all plot commands, and et some version info (eg
what tcl/tk) and see a copy of the mpl build output.  Also, I'd like
to see a run w/ --verbose-debug output.  I find it strange because I
am on the same platform and am not having any troubles (yet) with
tkagg.

JDH

sorry instances of self.plotView.image

On Dec 3, 2007 3:38 PM, Chris Fonnesbeck <lis...@ma...> wrote:
> OK, I'm having a bad backend day (insert joke here).
>
> Since I am on OSX, I decided to switch over to the CocoaAgg backend,
> but it fails when trying to show a histogram, yielding the following
> TypeError:
>
> In [4]: from matplotlib import use
>
> In [5]: use('CocoaAgg')
>
> In [6]: from pylab import hist, show
>
> In [7]: from numpy import random
>
> In [8]: x = random.exponential(2, 1000)
>
> In [9]: hist(x)
> Out[9]:
> (array([545, 230, 117,  59,  18,  15,  11,   1,   3,   1]),
>  array([  2.99606603e-03,   1.52977050e+00,   3.05654494e+00,
>          4.58331937e+00,   6.11009381e+00,   7.63686824e+00,
>          9.16364268e+00,   1.06904171e+01,   1.22171915e+01,
>          1.37439660e+01]),
>  <a list of 10 Patch objects>)
>
> In [10]: show()
> ---------------------------------------------------------------------------
> TypeError                                 Traceback (most recent call last)
>
> /Users/chris/Projects/Occupancy-abundance Manuscript/<ipython console>
> in <module>()
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in show()
>      44 def show():
>      45     for manager in Gcf.get_all_fig_managers():
> ---> 46         manager.show()
>      47
>      48 def draw_if_interactive():
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in show(self)
>     203                 mplBundle,
>     204                 False)
> --> 205         NSApplication.sharedApplication().run()
>     206
>     207
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in startWithBundle_(self, bundle)
>     179     def startWithBundle_(self, bundle):
>     180         #NSApplicationLoad()
> --> 181         if not
> bundle.loadNibFile_externalNameTable_withZone_('Matplotlib.nib', {},
> None):
>     182             print >>sys.stderr, 'Unable to load Matplotlib Cocoa UI!'
>     183             sys.exit()
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in awakeFromNib(self)
>      79
>      80         self.plotView.setImageFrameStyle_(NSImageFrameGroove)
> ---> 81         self.plotView.image = NSImage.alloc().initWithSize_((0,0))
>      82         self.plotView.setImage_(self.plotView.image)
>      83
>
> TypeError: cannot change a method
>
> --
> Christopher J. Fonnesbeck
> + Fish & Wildlife Research Institute (FWC)
> + 727.235.5570
>
> -------------------------------------------------------------------------
> SF.Net email is sponsored by: The Future of Linux Business White Paper
> from Novell.  From the desktop to the data center, Linux is going
> mainstream.  Let it simplify your IT future.
> http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Chris,

This error is due to more stringent error checking in pyobjc2 which
ships with Leopard. I was just preparing a patch to send to the
matplotlib dev guys, but if you replace all instances of self.image
with self._image in backend_cocoaagg.py, things should work. Because
PlotView is an NSImage subclass, it inherits a method named image.
pyobjc previously let you overwrite that method with a python instance
variable, but no longer, hence having to rename the instance variable.

barry

On Dec 3, 2007 3:38 PM, Chris Fonnesbeck <lis...@ma...> wrote:
> OK, I'm having a bad backend day (insert joke here).
>
> Since I am on OSX, I decided to switch over to the CocoaAgg backend,
> but it fails when trying to show a histogram, yielding the following
> TypeError:
>
> In [4]: from matplotlib import use
>
> In [5]: use('CocoaAgg')
>
> In [6]: from pylab import hist, show
>
> In [7]: from numpy import random
>
> In [8]: x = random.exponential(2, 1000)
>
> In [9]: hist(x)
> Out[9]:
> (array([545, 230, 117,  59,  18,  15,  11,   1,   3,   1]),
>  array([  2.99606603e-03,   1.52977050e+00,   3.05654494e+00,
>          4.58331937e+00,   6.11009381e+00,   7.63686824e+00,
>          9.16364268e+00,   1.06904171e+01,   1.22171915e+01,
>          1.37439660e+01]),
>  <a list of 10 Patch objects>)
>
> In [10]: show()
> ---------------------------------------------------------------------------
> TypeError                                 Traceback (most recent call last)
>
> /Users/chris/Projects/Occupancy-abundance Manuscript/<ipython console>
> in <module>()
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in show()
>      44 def show():
>      45     for manager in Gcf.get_all_fig_managers():
> ---> 46         manager.show()
>      47
>      48 def draw_if_interactive():
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in show(self)
>     203                 mplBundle,
>     204                 False)
> --> 205         NSApplication.sharedApplication().run()
>     206
>     207
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in startWithBundle_(self, bundle)
>     179     def startWithBundle_(self, bundle):
>     180         #NSApplicationLoad()
> --> 181         if not
> bundle.loadNibFile_externalNameTable_withZone_('Matplotlib.nib', {},
> None):
>     182             print >>sys.stderr, 'Unable to load Matplotlib Cocoa UI!'
>     183             sys.exit()
>
> /Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
> in awakeFromNib(self)
>      79
>      80         self.plotView.setImageFrameStyle_(NSImageFrameGroove)
> ---> 81         self.plotView.image = NSImage.alloc().initWithSize_((0,0))
>      82         self.plotView.setImage_(self.plotView.image)
>      83
>
> TypeError: cannot change a method
>
> --
> Christopher J. Fonnesbeck
> + Fish & Wildlife Research Institute (FWC)
> + 727.235.5570
>
> -------------------------------------------------------------------------
> SF.Net email is sponsored by: The Future of Linux Business White Paper
> from Novell.  From the desktop to the data center, Linux is going
> mainstream.  Let it simplify your IT future.
> http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

On Dec 3, 2007 5:53 PM, John Hunter <jd...@gm...> wrote:
> On Dec 3, 2007 5:49 PM, John Hunter <jd...@gm...> wrote:
>
> > Charlie, I don't know how you handled this last time, but is there
> > something in setuptools you have to disable for this build?
>
> Well, here's a clue: matplotlib/__init__.py does not exist in this instal

When I copy all the dirs from the 2.5 egg you built into
site-packages, I get a matplotlib with rcParams and the rest of the
__init__.py stuff.  Unfortunately, when I try to import pylab, I get
the dreaded old msvcp71.dll error.  If I recall correctly, we used to
hack distutils to remove the line that linked with that lib.

JDH

On Dec 3, 2007 5:49 PM, John Hunter <jd...@gm...> wrote:

> Charlie, I don't know how you handled this last time, but is there
> something in setuptools you have to disable for this build?

Well, here's a clue: matplotlib/__init__.py does not exist in this instal

JDH

On Dec 3, 2007 5:44 PM, Charlie Moad <cw...@gm...> wrote:
> I gave it a try, and I did not get this error.

I am seeing the problem on python2.5 using the binary exe installer.
It appears we are getting fouled up by setuptools special handling of
the __init__.py file, which is supposed to be empty I think to support
namespace packages.  All the stuff we are defining in
matplotlib/__init__.py, like rcParams and friends, is being lost.
Instead, a

>>> import matplotlib
>>> dir(matplotlib)

just shows the special names '__path__', '__doc__',  and '__name__',
and the  'toolkit' dir.

Charlie, I don't know how you handled this last time, but is there
something in setuptools you have to disable for this build?

JDH

I gave it a try, and I did not get this error.

- Charlie

On Dec 3, 2007 12:57 PM, J=F6rgen Stenarson <jor...@bo...> =
wrote:
> Hi,
>
> After upgrading to the new matplotlib 0.9.1 on python2.4 using the
> windows binary installer I get the following error:
>
>
> C:\>python
> Python 2.4.3 (#69, Mar 29 2006, 17:35:34) [MSC v.1310 32 bit (Intel)] on
> win32
> Type "help", "copyright", "credits" or "license" for more information.
>  >>> import pylab
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
>    File "C:\Python24\Lib\site-packages\pylab.py", line 1, in ?
>      from matplotlib.pylab import *
>    File "C:\Python24\Lib\site-packages\matplotlib\pylab.py", line 206, in=
 ?
>      from matplotlib.numerix import npyma as ma
>    File "C:\Python24\Lib\site-packages\matplotlib\numerix\__init__.py",
> line 20, in ?
>      from matplotlib import rcParams, verbose
> ImportError: cannot import name rcParams
>
> Does anyone else see the same thing?
>
>
> best regards
> J=F6rgen
>
>
>
> -------------------------------------------------------------------------
> SF.Net email is sponsored by: The Future of Linux Business White Paper
> from Novell.  From the desktop to the data center, Linux is going
> mainstream.  Let it simplify your IT future.
> http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

OK, I'm having a bad backend day (insert joke here).

Since I am on OSX, I decided to switch over to the CocoaAgg backend,
but it fails when trying to show a histogram, yielding the following
TypeError:

In [4]: from matplotlib import use

In [5]: use('CocoaAgg')

In [6]: from pylab import hist, show

In [7]: from numpy import random

In [8]: x = random.exponential(2, 1000)

In [9]: hist(x)
Out[9]:
(array([545, 230, 117,  59,  18,  15,  11,   1,   3,   1]),
 array([  2.99606603e-03,   1.52977050e+00,   3.05654494e+00,
         4.58331937e+00,   6.11009381e+00,   7.63686824e+00,
         9.16364268e+00,   1.06904171e+01,   1.22171915e+01,
         1.37439660e+01]),
 <a list of 10 Patch objects>)

In [10]: show()
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)

/Users/chris/Projects/Occupancy-abundance Manuscript/<ipython console>
in <module>()

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in show()
     44 def show():
     45     for manager in Gcf.get_all_fig_managers():
---> 46         manager.show()
     47
     48 def draw_if_interactive():

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in show(self)
    203                 mplBundle,
    204                 False)
--> 205         NSApplication.sharedApplication().run()
    206
    207

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in startWithBundle_(self, bundle)
    179     def startWithBundle_(self, bundle):
    180         #NSApplicationLoad()
--> 181         if not
bundle.loadNibFile_externalNameTable_withZone_('Matplotlib.nib', {},
None):
    182             print >>sys.stderr, 'Unable to load Matplotlib Cocoa UI!'
    183             sys.exit()

/Library/Python/2.5/site-packages/matplotlib/backends/backend_cocoaagg.py
in awakeFromNib(self)
     79
     80         self.plotView.setImageFrameStyle_(NSImageFrameGroove)
---> 81         self.plotView.image = NSImage.alloc().initWithSize_((0,0))
     82         self.plotView.setImage_(self.plotView.image)
     83

TypeError: cannot change a method

-- 
Christopher J. Fonnesbeck
+ Fish & Wildlife Research Institute (FWC)
+ 727.235.5570

Hi,

After upgrading to the new matplotlib 0.9.1 on python2.4 using the 
windows binary installer I get the following error:


C:\>python
Python 2.4.3 (#69, Mar 29 2006, 17:35:34) [MSC v.1310 32 bit (Intel)] on 
win32
Type "help", "copyright", "credits" or "license" for more information.
 >>> import pylab
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "C:\Python24\Lib\site-packages\pylab.py", line 1, in ?
     from matplotlib.pylab import *
   File "C:\Python24\Lib\site-packages\matplotlib\pylab.py", line 206, in ?
     from matplotlib.numerix import npyma as ma
   File "C:\Python24\Lib\site-packages\matplotlib\numerix\__init__.py", 
line 20, in ?
     from matplotlib import rcParams, verbose
ImportError: cannot import name rcParams

Does anyone else see the same thing?


best regards
Jörgen

On Dec 3, 2007 10:08 AM, Chris Fonnesbeck <lis...@ma...> wrote:
>
> On Dec 3, 2007 10:58 AM, John Hunter <jd...@gm...> wrote:
> > On Dec 2, 2007 9:23 PM, Chris Fonnesbeck <lis...@ma...> wrote:
> > > Running a recent build from svn on OSX 10.5, the TkAgg interface
> > > becomes unresponsive after plotting a histogram and calling show().
> > > Here is a sample from the process, in case it is useful:
> >
> > Is it only hist that is hanging?  Does plot work....  This one has me
> > a bit stumped.
>
> Same thing. Here is what I did:
>
> In [1]: from matplotlib import use

Sorry for the confusion, I was still on the *old* problem of tkagg
hanging w/ hist, and was wondering if it was hanging with other plot
commands too.  Not that you can easily test now w/ the active tcl
install.

On Dec 3, 2007 11:04 AM, Michael Droettboom <md...@st...> wrote:
> Possibly related to the _tkagg extension -- that's what defines the
> "PyAggImagePhoto" method.  If you installed ActiveState's Tcl/Tk after
> building matplotlib, that could explain this.  Try doing a clean build
> of matplotlib again (removing the build directory first).
>
> But that's only a suggestion -- I don't use Macs much.

Thanks for the suggestion. I did do a fresh svn update and build after
installing Tk/Tcl.

-- 
Christopher J. Fonnesbeck
+ Fish & Wildlife Research Institute (FWC)
+ 727.235.5570

On Dec 3, 2007 10:58 AM, John Hunter <jd...@gm...> wrote:
> On Dec 2, 2007 9:23 PM, Chris Fonnesbeck <lis...@ma...> wrote:
> > Running a recent build from svn on OSX 10.5, the TkAgg interface
> > becomes unresponsive after plotting a histogram and calling show().
> > Here is a sample from the process, in case it is useful:
>
> Is it only hist that is hanging?  Does plot work....  This one has me
> a bit stumped.

Same thing. Here is what I did:

In [1]: from matplotlib import use

In [2]: use('TkAgg')

In [3]: from numpy import random

In [4]: x = random.gamma(0.1, 1000, 1000)

In [5]: from pylab import *

In [6]: plot(x)
Out[6]: [<matplotlib.lines.Line2D instance at 0x1dda508>]

In [7]: show()
Exception in Tkinter callback
Traceback (most recent call last):
  File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib-tk/Tkinter.py",
line 1403, in __call__
    return self.func(*args)
  File "/Library/Python/2.5/site-packages/matplotlib/backends/backend_tkagg.py",
line 188, in resize
    self.show()
  File "/Library/Python/2.5/site-packages/matplotlib/backends/backend_tkagg.py",
line 192, in draw
    tkagg.blit(self._tkphoto, self.renderer._renderer, colormode=2)
  File "/Library/Python/2.5/site-packages/matplotlib/backends/tkagg.py",
line 15, in blit
    tk.call("PyAggImagePhoto", photoimage, id(aggimage), colormode, id(bbox))
TclError





-- 
Christopher J. Fonnesbeck
+ Fish & Wildlife Research Institute (FWC)
+ 727.235.5570

Possibly related to the _tkagg extension -- that's what defines the 
"PyAggImagePhoto" method.  If you installed ActiveState's Tcl/Tk after 
building matplotlib, that could explain this.  Try doing a clean build 
of matplotlib again (removing the build directory first).

But that's only a suggestion -- I don't use Macs much.

Cheers,
Mike

Chris Fonnesbeck wrote:
> On Dec 2, 2007 10:44 PM, John Hunter <jd...@gm...> wrote:
>> On Dec 2, 2007 9:23 PM, Chris Fonnesbeck <lis...@ma...> wrote:
>>> Running a recent build from svn on OSX 10.5, the TkAgg interface
>>> becomes unresponsive after plotting a histogram and calling show().
>>> Here is a sample from the process, in case it is useful:
>> I'm not seeing this with tkagg on 10.5.  hist is working fine for me
>> both in ipython and running histogram_demo.py -- what tcl/tk are you
>> building against?
> 
> I've just tried installing ActiveState's Tk/Tcl, but I get the
> following error when I try to plot anything:
> 
> Exception in Tkinter callback
> Traceback (most recent call last):
>   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib-tk/Tkinter.py",
> line 1403, in __call__
>     return self.func(*args)
>   File "/Library/Python/2.5/site-packages/matplotlib/backends/backend_tkagg.py",
> line 188, in resize
>     self.show()
>   File "/Library/Python/2.5/site-packages/matplotlib/backends/backend_tkagg.py",
> line 192, in draw
>     tkagg.blit(self._tkphoto, self.renderer._renderer, colormode=2)
>   File "/Library/Python/2.5/site-packages/matplotlib/backends/tkagg.py",
> line 15, in blit
>     tk.call("PyAggImagePhoto", photoimage, id(aggimage), colormode, id(bbox))
> TclError
> 
> 
> 
> 

-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA