matplotlib-users Mailing List for matplotlib

>>>>> "Ryan" == Ryan Krauss <rya...@gm...> writes:

    Ryan> I seem to be o.k. now.  I tried setup.py clean.  That
    Ryan> doesn't remove the build dir?

No, it only rebuilds what is needed, but if the recursive dependency
analysis is broken, you can get into an inconsistent state.  This
happens most commonly when headers are modified, because this doesn't
trigger rebuilds of the *.cxx files that include them.  In this case,
I rebuilt the swig runtime headers with a more recent version of SWIG
to help Charlie try and get python2.5 builds going.

Note to other svn users: flush your build subdir before trying to
rebuild after a recent svn update.

See also http://matplotlib.sourceforge.net/faq.html#FAILURE

JDH

>>>>> "Ryan" == Ryan Krauss <rya...@gm...> writes:

    Ryan> I just upgraded matplotlib, numpy, scipy, and ipython to
    Ryan> currect svn.  I am getting a segfault with pylab:

Try rm -rf the build subdir and recompile -- I just updated the SWIG
srcs which may be causing your problem.

JDH

    Ryan> Python 2.4.2 (#2, Sep 30 2005, 21:19:01) [GCC 4.0.2 20050808
    Ryan> (prerelease) (Ubuntu 4.0.1-4ubuntu8)] on linux2 Type "help",
    Ryan> "copyright", "credits" or "license" for more information.
    >>>> import pylab
    Ryan> loaded rc file /home/ryan/.matplotlib/matplotlibrc
    Ryan> matplotlib version 0.87.5 verbose.level debug-annoying
    Ryan> interactive is False platform is linux2 loaded modules:
    Ryan> ['pylab', '_bisect', '__future__', 'copy_reg',
    Ryan> 'sre_compile', 'distutils', 'itertools', '_sre', '__main__',
    Ryan> 'popen2', 'site', '__builtin__', 'datetime', 'distutils.re',
    Ryan> 'matplotlib.re', 'matplotlib.tempfile', 'encodings',
    Ryan> 'pytz.datetime', 'shutil', 'distutils.string', 'dateutil',
    Ryan> 'matplotlib.datetime', 'posixpath', '_random', 'tempfile',
    Ryan> 'errno', 'matplotlib.warnings', 'binascii',
    Ryan> 'encodings.codecs', 'sre_constants', 're', 'matplotlib.md5',
    Ryan> 'os.path', 'pytz.sys', '_codecs', 'distutils.sysconfig',
    Ryan> 'encodings.exceptions', 'pytz.sets', 'math', 'fcntl',
    Ryan> 'stat', 'zipimport', 'string', 'warnings',
    Ryan> 'encodings.types', 'UserDict', 'encodings.utf_8',
    Ryan> 'matplotlib', 'distutils.os', 'sys', 'pytz.tzinfo', 'pytz',
    Ryan> 'matplotlib.__future__', 'codecs', 'matplotlib.sys',
    Ryan> 'readline', 'matplotlib.pytz', 'types', 'md5',
    Ryan> 'matplotlib.dateutil', 'matplotlib.os', 'thread', 'sre',
    Ryan> 'bisect', 'matplotlib.distutils', 'signal',
    Ryan> 'distutils.errors', 'random', 'linecache',
    Ryan> 'matplotlib.shutil', 'posix', 'encodings.aliases', 'sets',
    Ryan> 'exceptions', 'sre_parse', 'pytz.bisect', 'distutils.sys',
    Ryan> 'os', 'strop'] numerix numpy 1.0rc1.dev3190 font search path
    Ryan> ['/usr/lib/python2.4/site-packages/matplotlib/mpl-data']
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoIt.ttf
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmsy10.ttf
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoBd.ttf
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraBI.ttf
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmex10.ttf
    Ryan> trying fontname
    Ryan> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/Vera.ttf
    Ryan> $HOME=/home/ryan CONFIGDIR=/home/ryan/.matplotlib loaded
    Ryan> ttfcache file /home/ryan/.matplotlib/ttffont.cache
    Ryan> Segmentation fault


    Ryan> Help me please.  I am running Ubuntu Breezy

    Ryan> Ryan

    Ryan> -------------------------------------------------------------------------
    Ryan> Take Surveys. Earn Cash. Influence the Future of IT Join
    Ryan> SourceForge.net's Techsay panel and you'll get the chance to
    Ryan> share your opinions on IT & business topics through brief
    Ryan> surveys -- and earn cash
    Ryan> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
    Ryan> _______________________________________________
    Ryan> Matplotlib-users mailing list
    Ryan> Mat...@li...
    Ryan> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

If I change numerix to numarray, this goes away.  Is anyone else out
there successfully running svn mpl with:

In [2]: numpy.__version__
Out[2]: '1.0rc1.dev3190'

In [3]: matplotlib.__version__
Out[3]: '0.87.5'

Thanks,

Ryan

On 9/19/06, Ryan Krauss <rya...@gm...> wrote:
> I just upgraded matplotlib, numpy, scipy, and ipython to currect svn.
> I am getting a segfault with pylab:
>
> Python 2.4.2 (#2, Sep 30 2005, 21:19:01)
> [GCC 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu8)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import pylab
> loaded rc file /home/ryan/.matplotlib/matplotlibrc
> matplotlib version 0.87.5
> verbose.level debug-annoying
> interactive is False
> platform is linux2
> loaded modules: ['pylab', '_bisect', '__future__', 'copy_reg',
> 'sre_compile', 'distutils', 'itertools', '_sre', '__main__', 'popen2',
> 'site', '__builtin__', 'datetime', 'distutils.re', 'matplotlib.re',
> 'matplotlib.tempfile', 'encodings', 'pytz.datetime', 'shutil',
> 'distutils.string', 'dateutil', 'matplotlib.datetime', 'posixpath',
> '_random', 'tempfile', 'errno', 'matplotlib.warnings', 'binascii',
> 'encodings.codecs', 'sre_constants', 're', 'matplotlib.md5',
> 'os.path', 'pytz.sys', '_codecs', 'distutils.sysconfig',
> 'encodings.exceptions', 'pytz.sets', 'math', 'fcntl', 'stat',
> 'zipimport', 'string', 'warnings', 'encodings.types', 'UserDict',
> 'encodings.utf_8', 'matplotlib', 'distutils.os', 'sys', 'pytz.tzinfo',
> 'pytz', 'matplotlib.__future__', 'codecs', 'matplotlib.sys',
> 'readline', 'matplotlib.pytz', 'types', 'md5', 'matplotlib.dateutil',
> 'matplotlib.os', 'thread', 'sre', 'bisect', 'matplotlib.distutils',
> 'signal', 'distutils.errors', 'random', 'linecache',
> 'matplotlib.shutil', 'posix', 'encodings.aliases', 'sets',
> 'exceptions', 'sre_parse', 'pytz.bisect', 'distutils.sys', 'os',
> 'strop']
> numerix numpy 1.0rc1.dev3190
> font search path ['/usr/lib/python2.4/site-packages/matplotlib/mpl-data']
> trying fontname
> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoIt.ttf
> trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmsy10.ttf
> trying fontname
> /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoBd.ttf
> trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraBI.ttf
> trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmex10.ttf
> trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/Vera.ttf
> $HOME=/home/ryan
> CONFIGDIR=/home/ryan/.matplotlib
> loaded ttfcache file /home/ryan/.matplotlib/ttffont.cache
> Segmentation fault
>
>
> Help me please.  I am running Ubuntu Breezy
>
> Ryan
>

I just upgraded matplotlib, numpy, scipy, and ipython to currect svn.
I am getting a segfault with pylab:

Python 2.4.2 (#2, Sep 30 2005, 21:19:01)
[GCC 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu8)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import pylab
loaded rc file /home/ryan/.matplotlib/matplotlibrc
matplotlib version 0.87.5
verbose.level debug-annoying
interactive is False
platform is linux2
loaded modules: ['pylab', '_bisect', '__future__', 'copy_reg',
'sre_compile', 'distutils', 'itertools', '_sre', '__main__', 'popen2',
'site', '__builtin__', 'datetime', 'distutils.re', 'matplotlib.re',
'matplotlib.tempfile', 'encodings', 'pytz.datetime', 'shutil',
'distutils.string', 'dateutil', 'matplotlib.datetime', 'posixpath',
'_random', 'tempfile', 'errno', 'matplotlib.warnings', 'binascii',
'encodings.codecs', 'sre_constants', 're', 'matplotlib.md5',
'os.path', 'pytz.sys', '_codecs', 'distutils.sysconfig',
'encodings.exceptions', 'pytz.sets', 'math', 'fcntl', 'stat',
'zipimport', 'string', 'warnings', 'encodings.types', 'UserDict',
'encodings.utf_8', 'matplotlib', 'distutils.os', 'sys', 'pytz.tzinfo',
'pytz', 'matplotlib.__future__', 'codecs', 'matplotlib.sys',
'readline', 'matplotlib.pytz', 'types', 'md5', 'matplotlib.dateutil',
'matplotlib.os', 'thread', 'sre', 'bisect', 'matplotlib.distutils',
'signal', 'distutils.errors', 'random', 'linecache',
'matplotlib.shutil', 'posix', 'encodings.aliases', 'sets',
'exceptions', 'sre_parse', 'pytz.bisect', 'distutils.sys', 'os',
'strop']
numerix numpy 1.0rc1.dev3190
font search path ['/usr/lib/python2.4/site-packages/matplotlib/mpl-data']
trying fontname
/usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoIt.ttf
trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmsy10.ttf
trying fontname
/usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraMoBd.ttf
trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/VeraBI.ttf
trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/cmex10.ttf
trying fontname /usr/lib/python2.4/site-packages/matplotlib/mpl-data/Vera.ttf
$HOME=/home/ryan
CONFIGDIR=/home/ryan/.matplotlib
loaded ttfcache file /home/ryan/.matplotlib/ttffont.cache
Segmentation fault


Help me please.  I am running Ubuntu Breezy

Ryan

On 09/19/06 15:22, John Hunter wrote:
>>"Kenny" == Kenny Ortmann <ya...@em...> writes:
> 
>     Kenny> I'm writing a program with a graph.  You have to select
>     Kenny> segments of the graph, which requires 2 points.
>
>     Kenny> working with widgets is new to me and im used to working
>     Kenny> with while loops and what not.  I've been just checking to
>     Kenny> see if i have 0 or 1 points in the on_point wxmpl code, and
>     Kenny> then if i have 1 and the next point is good opening a
>     Kenny> dialog to ask and then resetting number of pionts to 0.
> 
>     Kenny> anyone have any better ideas of how to do this? the amount
>     Kenny> of "global"/"self.points" variables is driving me nuts
 >
> If you want to select a horizontal or vertical region of the graph,
> eg, xmin, xmax, the SpanSelector widget is your friend. 

Although there isn't a wxmpl event for measuring only horizontal or vertical 
regions, you can use the wxmpl.EVT_SELECTION event to pick a rectangular 
region.  For an example of how that works, you can look at 
`demos/picking_points.py' in the wxmpl source distribution.

Without knowing more about the application it's hard to recommend what to do 
to improve things.  Perhaps you could add a button to the wxPython application 
that enables selection, so that the user clicks it and then picks the area of 
the plot to apply the function to?  You could also subclass wxmpl.PlotPanel, 
add some basic support for the two-point selections, and then reuse that code 
in your application.

> I believe recent versions of the wxmpl code worth with mpl events and
> widgets, but I haven't tested this.

Although it is possible to modify `wxmpl.py' to emit mpl events, I'm afraid 
that wxmpl still doesn't support mpl events or widgets out-of-the-box. 
There's still too much that can break if the two event-handling systems are 
incorrectly mixed in an application.

Ken

>>>>> "Kenny" == Kenny Ortmann <ya...@em...> writes:

    Kenny> I'm writing a program with a graph.  You have to select
    Kenny> segments of the graph, which requires 2 points.

    Kenny> working with widgets is new to me and im used to working
    Kenny> with while loops and what not.  I've been just checking to
    Kenny> see if i have 0 or 1 points in the on_point wxmpl code, and
    Kenny> then if i have 1 and the next point is good opening a
    Kenny> dialog to ask and then resetting number of pionts to 0.

    Kenny> anyone have any better ideas of how to do this? the amount
    Kenny> of "global"/"self.points" variables is driving me nuts

If you want to select a horizontal or vertical region of the graph,
eg, xmin, xmax, the SpanSelector widget is your friend.  I believe
recent versions of the wxmpl code worth with mpl events and widgets,
but I haven't tested this.


#!/usr/bin/env python
"""
The SpanSelector is a mouse widget to select a vmin/vmax
range.  When you left click drag in the axes, a rectangle shows the
selected region.  When you release, the rectangle disappears and a
callback is called with min/max.
"""
import pylab
from matplotlib.widgets import SpanSelector

fig = pylab.figure(figsize=(8,6))
ax = fig.add_subplot(211, axisbg='#FFFFCC')

x,y = 4*(pylab.rand(2,100)-.5)
ax.plot(x,y,'o')
ax.set_xlim(-2,2)
ax.set_ylim(-2,2)
ax.set_title('Press left mouse button and drag to test')


def onselect(vmin, vmax):
    print vmin, vmax

# set useblit True on gtkagg for enhanced performance
span = SpanSelector(ax, onselect, 'horizontal', useblit=False,
                    rectprops=dict(alpha=0.5, facecolor='red') )

ax2 = fig.add_subplot(212)
ax2.plot([1,2,3])

span2 = SpanSelector(ax2, onselect, 'vertical')

pylab.show()

I'm writing a program with a graph.  You have to select segments of the
graph, which requires 2 points.

is there a way to write this with a while loop, where

while(morepoints == "yes")
{
    while(waiting on point 1)
    {
        check if point 1 is a good point
        if point 1 is good
            break from while
    }
    while(waiting on point 2)
    {
        check if point 2 is a good point
        if point 2 is good
            break from while
    }
    ask user if they want more points
    if yes break from while
}

working with widgets is new to me and im used to working with while loops
and what not.  I've been just checking to see if i have 0 or 1 points in
the on_point wxmpl code, and then if i have 1 and the next point is good
opening a dialog to ask and then resetting number of pionts to 0.

anyone have any better ideas of how to do this? the amount of
"global"/"self.points" variables is driving me nuts

hi all,
I have a problem with datestr2num. All my dates are sorted in %d/%m/%Y format,
but datestr2num seems to take days for months before the 12th day:
100>datas['Dates']
Sortie[100]:
recarray([05/01/2006, 06/01/2006, 07/01/2006, 08/01/2006, 09/01/2006,
       10/01/2006, 11/01/2006, 12/01/2006, 13/01/2006, 14/01/2006,
       15/01/2006, 16/01/2006, 17/01/2006, 18/01/2006, 19/01/2006,
       20/01/2006, 21/01/2006, 22/01/2006, 23/01/2006, 24/01/2006,
       25/01/2006, 26/01/2006, 27/01/2006, 28/01/2006, 29/01/2006,
       30/01/2006, 31/01/2006, 01/02/2006, 02/02/2006, 03/02/2006,
       04/02/2006, 05/02/2006, 06/02/2006, 07/02/2006, 08/02/2006,
       09/02/2006, 10/02/2006, 11/02/2006, 12/02/2006, 13/02/2006,
       14/02/2006, 15/02/2006, 16/02/2006, 17/02/2006, 18/02/2006,
       19/02/2006, 20/02/2006, 21/02/2006, 22/02/2006, 23/02/2006,
       24/02/2006, 25/02/2006, 26/02/2006, 27/02/2006, 28/02/2006,
       01/03/2006],
      dtype='|S10')

101>array([datetime.strftime(x, "%d/%m/%Y") for x in 
num2date(datestr2num(datas['Dates']) )])
Sortie[101]:
array([01/05/2006, 01/06/2006, 01/07/2006, 01/08/2006, 01/09/2006,
       01/10/2006, 01/11/2006, 01/12/2006, 13/01/2006, 14/01/2006,
       15/01/2006, 16/01/2006, 17/01/2006, 18/01/2006, 19/01/2006,
       20/01/2006, 21/01/2006, 22/01/2006, 23/01/2006, 24/01/2006,
       25/01/2006, 26/01/2006, 27/01/2006, 28/01/2006, 29/01/2006,
       30/01/2006, 31/01/2006, 02/01/2006, 02/02/2006, 02/03/2006,
       02/04/2006, 02/05/2006, 02/06/2006, 02/07/2006, 02/08/2006,
       02/09/2006, 02/10/2006, 02/11/2006, 02/12/2006, 13/02/2006,
       14/02/2006, 15/02/2006, 16/02/2006, 17/02/2006, 18/02/2006,
       19/02/2006, 20/02/2006, 21/02/2006, 22/02/2006, 23/02/2006,
       24/02/2006, 25/02/2006, 26/02/2006, 27/02/2006, 28/02/2006,
       03/01/2006],
      dtype='|S10')
-- 
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

I had a problem yesterday running a script in interactive mode in
ipython -pylab using the gtk backend.  In the script there was some
expensive computation, and I wanted to break out of it by hitting
CTRL-C.  Because of the nastiness of signal handling across threads,
this isn't possible, though Fernando tells me it is keeping him up
nights.  ipython prints a message that it got a CTRL-C, but can't
cause the code running in the gui thread to exit cleanly.

One hack/workaround solution is to check for the IPython.Shell state
variable at various points in your code

  # some script you want to run from ipython -pylab or ipython -gthread
  # with "run"
  import IPython.Shell
  from pylab import figure, show, setp

  def ipbreak():
      if IPython.Shell.KBINT:
          IPython.Shell.KBINT = False
          raise SystemExit

  for i in somerange:
      do_something_very_expensive()
      ipbreak()

  show()


Then if you want to interrupt the script, you can with CTRL-C.  Of
course the trick is to remember to sprinkle these throughout your code
ahead of time.  The script will run normally outside of ipython, and
as long as the function call and boolean check are cheap compared to
your expensive function, shouldn't hurt performance.

Thanks Fernando for the suggestion!


JDH