matplotlib-users Mailing List for matplotlib

http://news.bbc.co.uk/1/hi/world/americas/4313978.stm

> The divide-by-zero behavior may have changed between Numeric versions.
> What are the version numbers on each of your machines?

In fact, we just figured that out.  The one that worked was with Numeric
24.0b2.  The one that blew up was 23.8.  For future reference, in case
anyone cares, you can make 23.8 work by compiling it with 

	-DHAVE_FINITE=1

E

Erik Curiel wrote:
> I've compiled MPL on a new Solaris 10 box, but now it won't initialize on
> this new box (though it still works fine on the old box).  Here's what
> happened:
> 
>>>>from pylab import *
[snip]
>   File
> "/usr/local/lib/python2.4/site-packages/matplotlib/numerix/_nc_imports.py",
> line 25, in ?
>     inf = infty = infinity = Infinity = (array([1])/0.0)[0]
> OverflowError: math range error

The divide-by-zero behavior may have changed between Numeric versions.
What are the version numbers on each of your machines?

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

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

I've compiled MPL on a new Solaris 10 box, but now it won't initialize on
this new box (though it still works fine on the old box).  Here's what
happened:

>>> from pylab import *
loaded rc file /home/erik/.matplotlib/matplotlibrc
matplotlib version 0.84
verbose.level debug
interactive is False
platform is sunos5
loaded modules: ['pylab', '_bisect', '__future__', 'copy_reg',
'sre_compile', 'distutils', '_sre', '__main__', 'site', '__builtin__',
'datetime', 'matplotlib.tempfile', 'encodings', 'pytz.datetime',
'sre_constants', 'distutils.string', 'dateutil', 'matplotlib.datetime',
'posixpath', '_random', 'tempfile', 'errno', 'matplotlib.warnings',
'binascii', 'encodings.codecs', 'matplotlib.sys', 're', 'os.path',
'pytz.sys', '_codecs', 'distutils.sysconfig', 'encodings.exceptions',
'pytz.sets', 'math', 'fcntl', 'stat', 'zipimport', 'string', 'warnings',
'encodings.types', 'UserDict', 'encodings.ascii', 'distutils.os',
'matplotlib', 'sys', 'pytz.tzinfo', 'pytz', 'matplotlib.__future__',
'codecs', 'distutils.re', 'readline', 'matplotlib.pytz', 'types', 'strop',
'matplotlib.dateutil', 'matplotlib.os', 'thread', 'sre', 'bisect',
'matplotlib.distutils', 'signal', 'distutils.errors', 'random',
'linecache', 'itertools', 'posix', 'encodings.aliases', 'sets',
'exceptions', 'sre_parse', 'pytz.bisect', 'distutils.sys', 'os']
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.4/site-packages/pylab.py", line 1, in ?
    from matplotlib.pylab import *
  File "/usr/local/lib/python2.4/site-packages/matplotlib/pylab.py", line
194, in ?
    import cm
  File "/usr/local/lib/python2.4/site-packages/matplotlib/cm.py", line 5,
in ?
    import colors
  File "/usr/local/lib/python2.4/site-packages/matplotlib/colors.py", line
33, in ?
    from numerix import array, arange, take, put, Float, Int, where, \
  File
"/usr/local/lib/python2.4/site-packages/matplotlib/numerix/__init__.py",
line 59, in ?
    from _nc_imports import nx, inf, infinity, Infinity
  File
"/usr/local/lib/python2.4/site-packages/matplotlib/numerix/_nc_imports.py",
line 25, in ?
    inf = infty = infinity = Infinity = (array([1])/0.0)[0]
OverflowError: math range error


Any help would be greatly appreciated.

Thanks!

E

>>>>> "Martin" == Martin Richter <law...@gm...> writes:


    Martin> Can you reproduce the segmentation fault?  Best,

Hi Martin,

Yes, I reproduced it yesterday <wink> The segfault arises when you
pass a path to draw_markers rather than a transformed path.  Obviously
this needs to be fixed to raise rather than segfault. I do test for
this in the extension code

why the check in _backend_agg.cpp

  agg::conv_transform<agg::path_storage> *ppath;
  swig_type_info * descr = SWIG_TypeQuery("agg.conv_transform<agg::path_storage> *");  
  assert(descr);
  if (SWIG_ConvertPtr(args[1].ptr(),(void **)(&ppath), descr, 0) == -1) 
    throw Py::TypeError("Could not convert agg.conv_transform<agg::path_storage>");

but clearly the check is not catching the bogus type and raising the
exception as it should.  I'll work on this.  In the meantime, you can
continue to work as long as you make sure you don't call draw_markers
with the wrong signature.  Eg, in Line2D.draw, something like

    if funcname in ('_draw_square', '_draw_triangle_up'):
        markerFunc = getattr(self, funcname)
        markerFunc(renderer, gc, xt, yt)

The segfault you are getting is happening on the tick drawing, which
are implemented as line markers (eg _draw_tickleft).  Since you
probably haven't ported the tick drawing to the new API, you are
getting the crash.

Guess I shouldn't have left this detail out yesterday!!

JDH

Hello everyone,
Hello John,

first of all: thanks for you mail, it was full of information!
I exchanged the lines in _backend_agg.cpp and tried to implement your
thoughts. While writing data.py wasn't a problem the trouble started when I
tried to run a little sample program with GTKAgg - I got a segmentation
fault. Unfortunately I'm not able to see where this one comes from. But when
using GTKAgg you're not able to draw markers at all; not even ticks ...
plotting a line is still possible when using
  from pylab import *
  axis('off')
  plot(arange(10))
so the error seems really due to the draw_markers() and not something else.

Below are the changed files.

Can you reproduce the segmentation fault?
Best,

Martin

P.S.: Once this particular example works, I will convert
the other markers (Arnd and I have worked out a scheme
for the size which seems consistent and simple to implement).

#### data.py 
"""
    Creates a dictionary containing information on how to draw markers.
    The values used within the paths are chosen so that the longest 
    dimension of a sourrounding rectangle is 1, e.g:
    triangle up ... has a height and a width of one
    square ........ has a height and a width of one
    ...
"""
from matplotlib import agg

markerd = {}

# triangle up
path = agg.path_storage()
path.move_to(   0,  0.5)
path.line_to(-0.5, -0.5)
path.line_to( 0.5, -0.5)
path.end_poly()
markerd['^'] = path

# square
path = agg.path_storage()
path.move_to(-0.5, -0.5)
path.line_to(-0.5,  0.5)
path.line_to( 0.5,  0.5)
path.line_to( 0.5, -0.5)
path.end_poly()
markerd['s'] = path

# tri_up                 (BTW this one change a bit)
offset = 2.0/3.0
offset1 = offset*0.8
offset2 = offset*0.5
path = agg.path_storage()
path.move_to(0, 0)
path.line_to(0, offset)
path.move_to(offset1, -offset2)
path.line_to(0, 0)
path.line_to(-offset1, -offset2)
path.end_poly()
markerd['2'] = path

####


#### changed routine for lines.py
def _draw_square(self, renderer, gc, xt, yt):
    side = renderer.points_to_pixels(self._markersize)
    rgbFace = self._get_rgb_face()
    path = d.markerd['s']
    if self._newstyle:
        scale = agg.trans_affine_scaling(side, side)
        transpath = agg.conv_transform_path(path, scale)
        renderer.draw_markers(gc, path, rgbFace, xt, yt,self._transform)
    else:
        offset = side*.5
        vertices = d.path_2_vertices(path)
        for (x,y) in zip(xt, yt):
            for verts in vertices:
                renderer.draw_polygon(gc, rgbFace, verts)
####


### Example call:
plot(arange(10),'bs')
###

-- 
Highspeed-Freiheit. Bei GMX supergünstig, z.B. GMX DSL_Cityflat,
DSL-Flatrate für nur 4,99 Euro/Monat*  http://www.gmx.net/de/go/dsl

>>>>> "Christopher" == Christopher Mutel <cm...@gm...> writes:

    Christopher> Hello again- Well, I think I have everything working,
    Christopher> thanks to your magicness. I had to manually set the
    Christopher> NUMERIX string to "Numeric", because even though I
    Christopher> had the latest python-numarray and
    Christopher> python-numarray-ext packages, setup.py still couldn't
    Christopher> find my numarray files. Observe:

Great!  I don't know why you had this numarray problem but it may be
because the dev headers aren't installed in the debian package.  If
you want it, it is trivial to compile and install from source.  Then
you would need to recompile matplotlib so it would pick up the
numarray extension code.

In any case, glad you have something working!

For future reference, I like to keep mpl discussions on list unless
they get bogged down into machine specific technical minutae because
others will benefit from having this information archived.

CC-ing to list.

Cheers,
JDH

    Christopher> cmutel@veter:~/src/matplotlib-0.84 $ sudo python
    Christopher> setup.py build TKAgg requires TkInter running build
    Christopher> running build_py running build_ext building
    Christopher> 'matplotlib._na_transforms' extension gcc -pthread
    Christopher> -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
    Christopher> -Wstrict-prototypes -fPIC -Isrc -I. -I
    Christopher> -I/usr/local/include -I/usr/include
    Christopher> -I. -I/usr/local/include/python2.4 -c
    Christopher> src/_na_transforms.cpp -o
    Christopher> build/temp.linux-i686-2.4/src/_na_transforms.o
    Christopher> -DNUMARRAY=1 In file included from
    Christopher> /usr/local/include/python2.4/Python.h:8, from
    Christopher> CXX/Objects.hxx:9, from CXX/Extensions.hxx:19, from
    Christopher> src/_transforms.h:12, from src/_na_transforms.cpp:2:
    Christopher> /usr/local/include/python2.4/pyconfig.h:835:1:
    Christopher> warning: "_POSIX_C_SOURCE" redefined In file included
    Christopher> from
    Christopher> /usr/include/c++/3.3/i486-linux/bits/os_defines.h:39,
    Christopher> from
    Christopher> /usr/include/c++/3.3/i486-linux/bits/c++config.h:35,
    Christopher> from /usr/include/c++/3.3/functional:53, from
    Christopher> src/_na_transforms.cpp:1:
    Christopher> /usr/include/features.h:131:1: warning: this is the
    Christopher> location of the previous definition
    Christopher> src/_na_transforms.cpp:6:35: numarray/arrayobject.h:
    Christopher> No such file or directory src/_na_transforms.cpp: In
    Christopher> member function `Py::Object
    Christopher> Bbox::update_numerix(const Py::Tuple&)':
    Christopher> src/_na_transforms.cpp:436: error: `PyArrayObject'
    Christopher> undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:436: error: (Each undeclared
    Christopher> identifier is reported only once for each function it
    Christopher> appears in.)  src/_na_transforms.cpp:436: error: `x'
    Christopher> undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:436: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp:442: error: `y'
    Christopher> undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:442: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp: In member function
    Christopher> `Py::Object Transformation::numerix_x_y(const
    Christopher> Py::Tuple&, const Py::Dict&)':
    Christopher> src/_na_transforms.cpp:810: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp:816: error: parse
    Christopher> error before `)' token src/_na_transforms.cpp:841:
    Christopher> error: `retx' undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:841: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp:848: error: `rety'
    Christopher> undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:848: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp: In member function
    Christopher> `Py::Object
    Christopher> Transformation::nonlinear_only_numerix(const
    Christopher> Py::Tuple&, const Py::Dict&)':
    Christopher> src/_na_transforms.cpp:895: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp:901: error: parse
    Christopher> error before `)' token src/_na_transforms.cpp:917:
    Christopher> error: parse error before `)' token
    Christopher> src/_na_transforms.cpp:924: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp:932: error:
    Christopher> `retmask' undeclared (first use this function)
    Christopher> src/_na_transforms.cpp:935: error: parse error before
    Christopher> `)' token src/_na_transforms.cpp: In function `void
    Christopher> init_na_transforms()': src/_na_transforms.cpp:1978:
    Christopher> error: `import_array' undeclared (first use this
    Christopher> function) error: command 'gcc' failed with exit
    Christopher> status 1

    Christopher> ...

    Christopher> cmutel@veter:/usr/local/lib/python2.4/site-packages $
    Christopher> ls /usr/include/python2.4/Numeric/ arrayobject.h
    Christopher> f2c.h ranlib.h ufuncobject.h
    Christopher> cmutel@veter:/usr/local/lib/python2.4/site-packages $
    Christopher> ls /usr/include/python2.4/numarray/ arraybase.h
    Christopher> ieeespecial.h libteacup.h numconfig.h arrayobject.h
    Christopher> libnumarray.h numarray.h nummacro.h cfunc.h
    Christopher> libnumeric.h numcomplex.h tc.h

    Christopher> And of course Numeric built fine.

    Christopher> Ultimately, of course, this doesn't matter, because I
    Christopher> can import matplotlib and pyalb as * without a
    Christopher> problem. I had a presentation yesterday, and I had to
    Christopher> export my data, and then import it on a clean build
    Christopher> of matplotlib on my windows machine, which wasn't
    Christopher> much fun.

    Christopher> Kepp on rockin'-

    Christopher> Chris Mutel

    Christopher> On 10/6/05, Christopher Mutel <cm...@gm...>
    Christopher> wrote:
    >> John-
    >> 
    >> I have been following the matplotlib mailing list for a while,
    >> and you have been unfailing kind and generous with your time in
    >> every message I have seen. I think what you're doing is damn
    >> great.
    >> 
    >> I actually already did to install your python package - I got
    >> errors with PyUnicodeUCS4... after seraching around on the net,
    >> this seems to be an error that is not resolvable without
    >> recompiliing either python or... wait for
    >> it... matplotlib. Here is the actual error code I get:
    >> 
    >> >>> from pylab import *
    >> /usr/lib/python2.4/site-packages/matplotlib/__init__.py:792:
    >> UserWarning: Could not find matplotlibrc; using defaults
    >> warnings.warn('Could not find matplotlibrc; using defaults')
    >> /usr/lib/python2.4/site-packages/matplotlib/__init__.py:813:
    >> UserWarning: could not find rc file; returning defaults
    >> warnings.warn(message) Traceback (most recent call last): File
    >> "<stdin>", line 1, in ?  File
    >> "/usr/local/lib/python2.4/site-packages/pylab.py", line 1, in ?
    >> from matplotlib.pylab import * File
    >> "/usr/local/lib/python2.4/site-packages/pylab.py", line 198, in
    >> ?
    >> 
    >> File
    >> "/home/jdhunter/debs/matplotlib/usr/lib/python2.4/site-packages/matplotlib/axes.py",
    >> line 15, in ?  File
    >> "/home/jdhunter/debs/matplotlib/usr/lib/python2.4/site-packages/matplotlib/axis.py",
    >> line 25, in ?  File
    >> "/home/jdhunter/debs/matplotlib/usr/lib/python2.4/site-packages/matplotlib/font_manager.py",
    >> line 39, in ?  ImportError:
    >> /usr/lib/python2.4/site-packages/matplotlib/ft2font.so:
    >> undefined symbol: PyUnicodeUCS4_AsUnicode
    >> 
    >> Is there any reason that this should happen? I don't know. I
    >> also don't know why apt doesn't install a matplotlibrc file.
    >> 
    >> I am pretty sure that my problems are the result of my having a
    >> machine that is out of whack in some respects. I installed
    >> python 2.4 form the standard Ubuntu package, and there is no
    >> reason that my package should break matplotlib while yours
    >> didn't. I don't know enough about linux to fit everything - on
    >> the other hand, the day after I installed Ubuntu the CD-Rom
    >> failed on my Linux box, so I don't have any way of
    >> re-installing everything.
    >> 
    >> I will try what you told me.
    >> 
    >> -Chris
    >> 
    >> On 10/6/05, John Hunter <jdh...@ac...> wrote:
    >> > >>>>> "Christopher" == Christopher Mutel <cm...@gm...>
    >> writes:
    >> >
    >> > Christopher> I'm having trouble compiling 0.84 on Python 2.4
    >> / > Christopher> Ubuntu Hoary (5.04).  It seems like when GCC
    >> tries to > Christopher> build ft2font it can't find the two
    >> files referenced > Christopher> in CXX/extensions.hxx and
    >> CXX/objects.hxx.
    >> >
    >> > Do you mean CXX/Extensions.hxx and CXX/Objects.hxx?  I just
    >> untarred > the 0.84 distribution and they are certainly there
    >> >
    >> > peds-pc311:/var/tmp/matplotlib-0.84> ls CXX/ > Config.hxx
    >> cxxsupport.cxx IndirectPythonInterface.cxx Version.hxx >
    >> cxxextensions.c Exception.hxx IndirectPythonInterface.hxx >
    >> cxx_extensions.cxx Extensions.hxx Objects.hxx
    >> >
    >> > Are you running setup.py from the dir in which it resides, eg
    >> >
    >> > > python setup.py build
    >> >
    >> > If so CXX/Extensions.hxx and CXX/Objects.hxx will be in your
    >> '.' (ie > current dir) include path.  Perhaps there is a
    >> problem with include > flags in setupext.py.  Try replacing the
    >> add_base_flags function in > that file with this one:
    >> >
    >> > def add_base_flags(module): > incdirs = [os.path.join(p,
    >> 'include') for p in basedir[sys.platform] > if
    >> os.path.exists(p)] > libdirs = [os.path.join(p, 'lib') for p in
    >> basedir[sys.platform] > if os.path.exists(p)] >
    >> module.include_dirs.extend(incdirs) >
    >> module.include_dirs.append('.')  >
    >> module.library_dirs.extend(libdirs)
    >> >
    >> >
    >> >
    >> > Christopher> I installed (through apt) python-cxx, but this
    >> didn't > Christopher> seem to change anything.
    >> >
    >> > No, we use our own CXX
    >> >
    >> > Christopher> This seems like a completely trivial problem >
    >> Christopher> (especially compared to the other frutrations I
    >> have > Christopher> had with matplotlib in the past), but I
    >> don't know > Christopher> how to fix it...
    >> >
    >> > Perhaps the approach above will fix it.  If not, you may want
    >> to > enable universe and multiverse in your
    >> /etc/apt/sources.list and then > add
    >> >
    >> > deb http://peds-pc311.bsd.uchicago.edu binary/
    >> >
    >> > and do
    >> >
    >> > > sudo apt-get update > > sudo apt-get python-matplotlib-jdh
    >> >
    >> > These are hoary debs that I maintain for our local system.
    >> I'm > certainly an amateur debian package maker, but they seem
    >> to work and > usually track matplotlib CVS fairly closely.
    >> >
    >> > JDH
    >> >
    >> >
    >> 

>>>>> "Mark" == Mark Bakker <ma...@gm...> writes:

    Mark> I just installed mpl 0.84 for Python23 on win32 and it works
    Mark> great, interactively on TkAgg, running from IDLE (sorry, I
    Mark> don't use gtk) Thanks a million, John. I will try to make
    Mark> the fixes to the axis('scaled') next week.

    Mark> Can you update us on the new Callback for resizing a window.
    Mark> Is that operational?

Only in gtk*.  Perhaps you could take a look at the tk backend and add
the functionality for it?  It should be fairly simple.  See this
thread for more info:

  http://sourceforge.net/mailarchive/message.php?msg_id=12947502

Thanks,
JDH

How did you find out how to use a patch? Trial and Error?
I couldn't find any documentation other than http:// 
matplotlib.sourceforge.net/matplotlib.patches.html
and that doesn't show how to use it.


On 06 Oct, 2005, at 20:26, Mark Bakker wrote:

> Thanks for pointing out this drawing arrow patch.
> I wrote a quick function to use this.
> Figuring out the correct width to use requires some experimentation.
> It would be nice if we could eventually have a line with a regular
> line width, and a nicely sized arrow head. (Any ideas?).
> But this gives nices results.
> Mark
>
> from pylab import *
> def draw_arrow( x, y, dx, dy, width=1, color = 'k' ):
>     ax = gca()
>     a = Arrow( x, y, dx, dy, width )
>     a.set_edgecolor(color)
>     a.set_facecolor(color)
>     ax.add_patch( a )
>     draw_if_interactive()
>     return a
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Power Architecture Resource Center: Free content, downloads,  
> discussions,
> and more. http://solutions.newsforge.com/ibmarch.tmpl
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************

Thanks for pointing out this drawing arrow patch.
I wrote a quick function to use this.
Figuring out the correct width to use requires some experimentation.
It would be nice if we could eventually have a line with a regular
line width, and a nicely sized arrow head. (Any ideas?).
But this gives nices results.
Mark

from pylab import *
def draw_arrow( x, y, dx, dy, width=3D1, color =3D 'k' ):
    ax =3D gca()
    a =3D Arrow( x, y, dx, dy, width )
    a.set_edgecolor(color)
    a.set_facecolor(color)
    ax.add_patch( a )
    draw_if_interactive()
    return a

I just installed mpl 0.84 for Python23 on win32 and it works great,
interactively on TkAgg, running from IDLE (sorry, I don't use gtk)
Thanks a million, John. I will try to make the fixes to the axis('scaled')
next week.

Can you update us on the new Callback for resizing a window.
Is that operational?

Thanks again, Mark

On 10/6/05, John Hunter <jdh...@ac...> wrote:
> >>>>> "Mark" =3D=3D Mark Bakker <ma...@gm...> writes:
>
>     Mark> Hey John - Any chance you can put a win32 install file of
>     Mark> mpl 0.84 for Python 2.3 on the download page?  I am teaching
>     Mark> a class on computational methods, and I really would like
>     Mark> the students to use the updated axis('equal') functionality.
>     Mark> Would be much appreciated.
>
> Hey Mark -- Coincidentally, I've been working on this this morning.
> It's been a pain mainly because I have a new build machine and getting
> gtk to compile on win32 is no fun.  I just uploaded the file so take
> it for a test drive and good luck with your class!
>
> JDH
>