matplotlib-users Mailing List for matplotlib

On Friday 06 May 2005 5:58 pm, Robert Kern wrote:
> Darren Dale wrote:
> > I see that passing normed=True to the hist function is supposed scale the
> > histogram counts to represent a probability distribution. Presently, this
> > renormalizes the counts such that the most popular bin is 1, but I would
> > argue that this does not represent a probability distribution. I think it
> > would be more appropriate to renormalize such that integrating over the
> > whole distribution yields 1.
>
> Are you sure? With CVS matplotlib:
>
> In [6]:hist(y)
> Out[6]:
> (array([ 17,  44, 130, 244, 260, 176,  91,  29,   8,   1]),
>   array([-2.85420394, -2.18336129, -1.51251864, -0.841676  ,
> -0.17083335,  0.5000093 ,
>               1.17085195,  1.84169459,  2.51253724,  3.18337989]))
>
> In [7]:hist(y, normed=True)
> Out[7]:
> (array([ 0.02534126,  0.06558915,  0.19378613,  0.36372166,  0.38757226,
>   0.26235661,
>               0.13565029,  0.04322921,  0.0119253 ,  0.00149066]),
>   array([-2.85420394, -2.18336129, -1.51251864, -0.841676  ,
> -0.17083335,  0.5000093 ,
>               1.17085195,  1.84169459,  2.51253724,  3.18337989]))

It turned out that I had a funny set of data that just happened to have a max 
normed count of 1, but the distribution still summed to 1. The normed 
operation is working the way it should.

-- 
Darren S. Dale

Bard Hall
Department of Materials Science and Engineering
Cornell University
Ithaca, NY. 14850

dd...@co...

Matt Newville wrote:
> Fernando, Darren,
> 
> For me, on Mac OS X changing IPython/Shell.py to 
>      if ver[:3] >= '2.5':
> 
> does suppress the warning about the version of wx.  But 'ipython
> -pylab' still hangs on startup with a message that the program
> should be run with 'pythonw', not 'python'.  This was from a
> clean install with
>  ~> sudo pythonw setup.py install
> 
> This installs ipython into 
>  
> /System/Library/Frameworks/Python.framework/Versions/2.3/bin/ipython
> 
> which is not typically in the shells PATH. 
> 
> The shebang line for the ipython script reads
> 
> #!/System/Library/Frameworks/Python.framework/Versions/2.3/Resources/Python.app/Contents/MacOS/Python
> 
> Changing that to
> 
> #!/usr/bin/env pythonw
> 
> makes 'ipython -pylab' run fine.
> 
> Maybe you want to make Mac installs put ipython in /usr/bin and
> point to pythonw (which is admittedly just a bash wrapper around
> /System/Library/..../Python.app/Contents/MacOS/Python)??

Mmh, the problem is that I don't know how to achieve this in a clean, 
distutils-based way.  I can't go changing the shebang line by hand, since I 
don't know where things will end up, and as far as I know there is no way to 
force this kind of behavior onto distutils.  Do you know any such incantation 
to achieve the desired result?

f

Darren Dale wrote:
> I see that passing normed=True to the hist function is supposed scale the 
> histogram counts to represent a probability distribution. Presently, this 
> renormalizes the counts such that the most popular bin is 1, but I would 
> argue that this does not represent a probability distribution. I think it 
> would be more appropriate to renormalize such that integrating over the whole 
> distribution yields 1.

Are you sure? With CVS matplotlib:

In [6]:hist(y)
Out[6]:
(array([ 17,  44, 130, 244, 260, 176,  91,  29,   8,   1]),
  array([-2.85420394, -2.18336129, -1.51251864, -0.841676  , 
-0.17083335,  0.5000093 ,
              1.17085195,  1.84169459,  2.51253724,  3.18337989]))

In [7]:hist(y, normed=True)
Out[7]:
(array([ 0.02534126,  0.06558915,  0.19378613,  0.36372166,  0.38757226, 
  0.26235661,
              0.13565029,  0.04322921,  0.0119253 ,  0.00149066]),
  array([-2.85420394, -2.18336129, -1.51251864, -0.841676  , 
-0.17083335,  0.5000093 ,
              1.17085195,  1.84169459,  2.51253724,  3.18337989]))

-- 
Robert Kern
rk...@uc...

"In the fields of hell where the grass grows high
  Are the graves of dreams allowed to die."
   -- Richard Harter

The problem I have seems more fundamental than Darren's.

Even if I start pythonw,  and then do from pylab import *, I still get  
the warning messages about being unable to open the type 1 fonts,
System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/ 
site-packages/matplotlib/font_manager.py:465: UserWarning: Could not  
open font file /Users/agn/Library/Fonts/GillSanBol
   warnings.warn("Could not open font file %s"%fpath)
  etc...

and then the program crash in font_manager.py with

Traceback (most recent call last):
...............
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/axis.py", line 20, in ?
     from font_manager import FontProperties
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/font_manager.py", line 991, in ?
     fontManager = FontManager()
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/font_manager.py", line 835, in  
__init__
     rebuild()
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/font_manager.py", line 828, in  
rebuild
     self.ttfdict = createFontDict(self.ttffiles)
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/font_manager.py", line 467, in  
createFontDict
     prop = ttfFontProperty(font)
   File  
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/ 
python2.3/site-packages/matplotlib/font_manager.py", line 347, in  
ttfFontProperty
     size = str(float(font.get_fontsize()))
AttributeError: get_fontsize


George Nurser.

I see that passing normed=True to the hist function is supposed scale the 
histogram counts to represent a probability distribution. Presently, this 
renormalizes the counts such that the most popular bin is 1, but I would 
argue that this does not represent a probability distribution. I think it 
would be more appropriate to renormalize such that integrating over the whole 
distribution yields 1.

Maybe there are other schools of thought on this subject? Is there a way that 
I can generate a physicists probability distribution with the existing 
function?

Thanks,
Darren

Fernando, Darren,

For me, on Mac OS X changing IPython/Shell.py to 
     if ver[:3] >= '2.5':

does suppress the warning about the version of wx.  But 'ipython
-pylab' still hangs on startup with a message that the program
should be run with 'pythonw', not 'python'.  This was from a
clean install with
 ~> sudo pythonw setup.py install

This installs ipython into 
 
/System/Library/Frameworks/Python.framework/Versions/2.3/bin/ipython

which is not typically in the shells PATH. 

The shebang line for the ipython script reads

#!/System/Library/Frameworks/Python.framework/Versions/2.3/Resources/Python.app/Contents/MacOS/Python

Changing that to

#!/usr/bin/env pythonw

makes 'ipython -pylab' run fine.

Maybe you want to make Mac installs put ipython in /usr/bin and
point to pythonw (which is admittedly just a bash wrapper around
/System/Library/..../Python.app/Contents/MacOS/Python)??

Hope that helps,

--Matt

On Friday 06 May 2005 12:19 am, Fernando Perez wrote:
> Matt Newville wrote:
> > Darren,
> >
> >>I just upgraded wxPython a couple days ago, and got error
> >>messages about not finding wxpython 2.4 or 2.5. Can anyone say
> >>whether the wx backend and IPython are compatible with the new
> >>wxpython release?
> >
> > The wx backend from matplotlib 0.80 works fine for me with
> > wxPython 2.6.0.0 on OS X 10.3.9.
> >
> > I also get messages about not finding wxpython 2.4 or 2.5 from
> > Ipython, and on Mac OS X also have Ipython hang when started
> > with 'ipython -wthread -pylab' or 'ipython -pylab', probably
> > because it's using 'python' not 'pythonw'.  I haven't tracked it
> > down beyond that.
>
> Could you try to change in IPython/Shell.py, the line (around line 555):
>
>      if ver[:3] == '2.5':
>          import wx
>
> to
>
>      if ver[:3] >= '2.5':
>          import wx
>
> and let me know if it works again?  In that case, I'll make that change
> before the bugfix .14 release.
>

Fernando, 

Thank you, this solves it.

-- 
Darren S. Dale

Bard Hall
Department of Materials Science and Engineering
Cornell University
Ithaca, NY. 14850

dd...@co...

>>>>> "N" == N Volbers <mit...@we...> writes:

    N> Hello everyone on the list, I try to use the matplotlib class
    N> interface for my plotting application.  I have a question which
    N> I hope you can help me with. The attached file contains a test
    N> case for my problem. From the example 'embedding_in_gtk2.py' I
    N> figured the proper way to use the class interface was to

    N> 1. create a Figure() instance 2. create a FigureCanvas
    N> instance, using the Figure as argument 3. create a subplot by
    N> using FigureCanvas.subplot

I think you mean  "create a subplot by using Figure.add_subplot"

    N> This works fine; however now I was trying to set the limits of
    N> the plot using the Axes.set_xlim and Axes.set_ylim functions.

Call set_xlim and set_ylim after calling plot.  Plot automatically
calls autoscale on  the axes to incorporate the new data into its
range.  This should become an rc param and a kwarg (and this question
will soon become a FAQ!)

    N> As you can see from the test, it won't affect the plot! On the
    N> other hand, this method works fine if you use the pylab 'show'
    N> method as in 'poly_editor.py'. Is this a bug or should I use
    N> another function to set the limits?


It has nothing to do with show and everything to do that setting the
lim come after creating the plot elements

    N> BTW, the documentation recommends passing two arguments to
    N> set_xlim, set_ylim, while the examples pass a list/tuple. Are
    N> both ways correct?

Yes.

JDH

So obvious when someone tells you! Thanks for this, it is exactly what I was
looking for. 

Carl

----- Original Message ----- 
From: "John Hunter" <jdh...@ac...>
To: "Carl Mouser" <car...@bi...>
Cc: <mat...@li...>
Sent: Thursday, May 05, 2005 11:55 PM
Subject: Re: [Matplotlib-users] Deleting lines from a plot


> 
> If you know which line you want to delete, you can call
> ax.lines.remove(line) where line is a line instance.
> 
>     In [1]: line1, = plot(rand(10), 'ro')
>     In [2]: line2, = plot(rand(20), 'b--s')
>     In [3]: ax = gca()
>     In [4]: ax.lines.remove(line1)
>     In [5]: draw()
> 
> If you haven't saved the line instance, but can otherwise identify it
> (eg you know it was the first line you plotted), you can do something
> like
> 
>   ax.lines.remove(ax.lines[0])
> 
> or if you have labeled the line, you can identify and remove it with
> 
>   plot(x,y,label='my line')
>   line = [line for line in ax.lines if line.get_label()=='my line'][0]
>   ax.lines.remove(line)
> 
> JDH

Hello everyone on the list,

I try to use the matplotlib class interface for my plotting 
application.  I have a question which I hope you can help me with. The 
attached file contains a test case for my problem. From the example 
'embedding_in_gtk2.py' I figured the proper way to use the class 
interface was to

1. create a Figure() instance
2. create a FigureCanvas instance, using the Figure as argument
3. create a subplot by using FigureCanvas.subplot

This works fine; however now I was trying to set the limits of the plot 
using the Axes.set_xlim and Axes.set_ylim functions.  As you can see 
from the test, it won't affect the plot! On the other hand, this method 
works fine if you use the pylab 'show' method as in 'poly_editor.py'. Is 
this a bug or should I use another function to set the limits?

BTW, the documentation recommends passing two arguments to set_xlim, 
set_ylim, while the examples pass a list/tuple. Are both ways correct?

Thanks,

Niklas Volbers.

Matt Newville wrote:
> John,
> 
> Hmm, I just installerd ipython-0.6.13 on OS X 10.3.9.  For me it
> created ipython with

[...]

> /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site-packages/IPython/Shell.py in mainloop(self=<IPShellMatplotlibWX(Thread-1, started)>)
>     712             def OnTimer(self, event):
>     713                 update_tk(self.tk)
>     714                 self.IP.runcode()
>     715 
>     716         class App(self.wx.wxApp):
>     717             wx = self.wx
>     718             TIMEOUT = self.TIMEOUT
>     719             def OnInit(self):
>     720                 'Create the main window and insert the custom frame'
>     721                 self.agent = TimerAgent(None, self.TIMEOUT)
>     722                 self.agent.Show(self.wx.false)
>     723                 self.agent.StartWork()
>     724                 return self.wx.true
>     725         
>     726         self.app = App(redirect=False)
> --> 727         self.wx_mainloop(self.app)
>         self.wx_mainloop = None
>         self.app = <IPython.Shell.App; proxy of C++ wxPyApp instance at _05a3a8e0_p_wxPyApp>
>     728         self.join()
>     729 
>     730 # A set of matplotlib public IPython shell classes, for single-threaded
>     731 # (Tk* and FLTK* backends) and multithreaded (GTK* and WX* backends) use.
>     732 class IPShellMatplotlib(IPShell):
>     733     """Subclass IPShell with MatplotlibShell as the internal shell.
>     734 
>     735     Single-threaded class, meant for the Tk* and FLTK* backends.
>     736 
>     737     Having this on a separate class simplifies the external driver code."""
>     738     
>     739     def __init__(self,argv=None,user_ns=None,debug=1):
>     740         IPShell.__init__(self,argv,user_ns,debug,shell_class=MatplotlibShell)
>     741 
>     742 class IPShellMatplotlibGTK(IPShellGTK):
> 
> TypeError: 'NoneType' object is not callable

Could you try the change I suggested in my last email?  That might fix this 
crash as well, as it will prevent the None object from being returned for wx 
2.6.  Let me know if it helps,

f

John,

Hmm, I just installerd ipython-0.6.13 on OS X 10.3.9.  For me it
created ipython with

#!/System/Library/Frameworks/Python.framework/Versions/2.3/Resources/Python.app/Content
s/MacOS/Python

Changing that to 
#!/usr/bin/env pythonw

caused ipython to crash, with IPython_crash_report.txt below.

--Matt

***************************************************************************

IPython post-mortem report

IPython version: 0.6.13 

Platform info  : os.name -> posix, sys.platform -> darwin

***************************************************************************

Current user configuration structure:

{'Version': 0,
 'alias': [''],
 'args': [],
 'autocall': 1,
 'autoindent': 1,
 'automagic': 1,
 'banner': 1,
 'c': '',
 'cache_size': 1000,
 'classic': 0,
 'color_info': 1,
 'colors': 'Linux',
 'confirm_exit': 1,
 'debug': 0,
 'deep_reload': 0,
 'editor': 'vi',
 'execfile': [],
 'execute': [''],
 'gthread': 0,
 'help': 0,
 'ignore': 0,
 'import_all': [],
 'import_mod': [],
 'import_some': [[]],
 'include': [],
 'ipythondir': '/Users/newville/.ipython',
 'log': 0,
 'logfile': '',
 'logplay': '',
 'magic_docstrings': 0,
 'messages': 1,
 'multi_line_specials': 1,
 'nosep': 0,
 'opts': Struct({'pylab': 1}),
 'pdb': 0,
 'pprint': 1,
 'profile': '',
 'prompt_in1': 'In [\\#]:',
 'prompt_in2': '   .\\D.:',
 'prompt_out': 'Out[\\#]:',
 'prompts_pad_left': 1,
 'pylab': 1,
 'quick': 0,
 'rcfile': 'ipythonrc',
 'readline': 1,
 'readline_merge_completions': 1,
 'readline_omit__names': 0,
 'readline_parse_and_bind': ['tab: complete',
                             '"\\C-l": possible-completions',
                             'set show-all-if-ambiguous on',
                             '"\\C-o": tab-insert',
                             '"\\M-i": "    "',
                             '"\\M-o": "\\d\\d\\d\\d"',
                             '"\\M-I": "\\d\\d\\d\\d"',
                             '"\\C-r": reverse-search-history',
                             '"\\C-s": forward-search-history',
                             '"\\C-p": history-search-backward',
                             '"\\C-n": history-search-forward'],
 'readline_remove_delims': '\'"-/~',
 'screen_length': -2,
 'separate_in': '\n',
 'separate_out': '',
 'separate_out2': '',
 'system_verbose': 0,
 'tk': 0,
 'upgrade': 0,
 'wthread': 0,
 'xmode': 'Context'}

***************************************************************************

Crash traceback:

---------------------------------------------------------------------------
exceptions.TypeErrorPython 2.3: /System/Library/Frameworks/Python.framework/Versions/2.3/Resources/Python.app/Contents/MacOS/Python
                                                   Thu May  5 23:13:31 2005
A problem occured executing Python code.  Here is the sequence of function
calls leading up to the error, with the most recent (innermost) call last.

/Users/newville/bin/ipython 
     -2 import IPython
     -1 
      0 IPython.Shell.start().mainloop()

/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site-packages/IPython/Shell.py in mainloop(self=<IPShellMatplotlibWX(Thread-1, started)>)
    712             def OnTimer(self, event):
    713                 update_tk(self.tk)
    714                 self.IP.runcode()
    715 
    716         class App(self.wx.wxApp):
    717             wx = self.wx
    718             TIMEOUT = self.TIMEOUT
    719             def OnInit(self):
    720                 'Create the main window and insert the custom frame'
    721                 self.agent = TimerAgent(None, self.TIMEOUT)
    722                 self.agent.Show(self.wx.false)
    723                 self.agent.StartWork()
    724                 return self.wx.true
    725         
    726         self.app = App(redirect=False)
--> 727         self.wx_mainloop(self.app)
        self.wx_mainloop = None
        self.app = <IPython.Shell.App; proxy of C++ wxPyApp instance at _05a3a8e0_p_wxPyApp>
    728         self.join()
    729 
    730 # A set of matplotlib public IPython shell classes, for single-threaded
    731 # (Tk* and FLTK* backends) and multithreaded (GTK* and WX* backends) use.
    732 class IPShellMatplotlib(IPShell):
    733     """Subclass IPShell with MatplotlibShell as the internal shell.
    734 
    735     Single-threaded class, meant for the Tk* and FLTK* backends.
    736 
    737     Having this on a separate class simplifies the external driver code."""
    738     
    739     def __init__(self,argv=None,user_ns=None,debug=1):
    740         IPShell.__init__(self,argv,user_ns,debug,shell_class=MatplotlibShell)
    741 
    742 class IPShellMatplotlibGTK(IPShellGTK):

TypeError: 'NoneType' object is not callable

***************************************************************************

History of session input:

*** Last line of input (may not be in above history):

Matt Newville wrote:
> Darren,
> 
> 
>>I just upgraded wxPython a couple days ago, and got error
>>messages about not finding wxpython 2.4 or 2.5. Can anyone say
>>whether the wx backend and IPython are compatible with the new
>>wxpython release?
> 
> 
> The wx backend from matplotlib 0.80 works fine for me with
> wxPython 2.6.0.0 on OS X 10.3.9.
> 
> I also get messages about not finding wxpython 2.4 or 2.5 from
> Ipython, and on Mac OS X also have Ipython hang when started
> with 'ipython -wthread -pylab' or 'ipython -pylab', probably
> because it's using 'python' not 'pythonw'.  I haven't tracked it
> down beyond that.

Could you try to change in IPython/Shell.py, the line (around line 555):

     if ver[:3] == '2.5':
         import wx

to

     if ver[:3] >= '2.5':
         import wx

and let me know if it works again?  In that case, I'll make that change before 
the bugfix .14 release.

Cheers,

f

>>>>> "Matt" == Matt Newville <new...@ca...> writes:

    Matt> The wx backend from matplotlib 0.80 works fine for me with
    Matt> wxPython 2.6.0.0 on OS X 10.3.9.

    Matt> I also get messages about not finding wxpython 2.4 or 2.5
    Matt> from Ipython, and on Mac OS X also have Ipython hang when
    Matt> started with 'ipython -wthread -pylab' or 'ipython -pylab',
    Matt> probably because it's using 'python' not 'pythonw'.  I
    Matt> haven't tracked it down beyond that.

ipython 0.6.13 on OSX uses 

  #!/usr/bin/env pythonw

I don't get the warning when using ipython -wthread on OSX with
wxpython 2.5.4.1 or linux with wxpython 2.5.3.2.  Both are using
ipython 0.6.13.

JDH

Darren,

> I just upgraded wxPython a couple days ago, and got error
> messages about not finding wxpython 2.4 or 2.5. Can anyone say
> whether the wx backend and IPython are compatible with the new
> wxpython release?

The wx backend from matplotlib 0.80 works fine for me with
wxPython 2.6.0.0 on OS X 10.3.9.

I also get messages about not finding wxpython 2.4 or 2.5 from
Ipython, and on Mac OS X also have Ipython hang when started
with 'ipython -wthread -pylab' or 'ipython -pylab', probably
because it's using 'python' not 'pythonw'.  I haven't tracked it
down beyond that.

--Matt