matplotlib-users Mailing List for matplotlib (Page 3)

If I run the script below (taken straight from 
the docs except for the imports) the script fails 
with a message that matshow does not return a tuple, which 
appears true. Dropping the extra LHS variables allows the 
script to run, but produces 4 separate figures.
Thanks, 
Alan Isaac 
============================================ 
from scipy import * 
from pylab import * 
def samplemat(dims): 
 aa = zeros(dims) 
 for i in range(min(dims)): 
 aa[i,i] = i 
 return aa 
dimlist = [(12,12),(128,64),(64,512),(2048,256)] 
for d in dimlist: 
 fig, ax, im = matshow(samplemat(d)) 
show() 

Hi Danny,
Danny Shevitz wrote:
> howdy,
> 
> I have some beginner type questions with matshow.
> 
> I want the origin of the axes to be the lower left of the figure. I tried:
> 	fig=matshow(data, cmap, origin='lower')
> 
> but that didn't seem to do anything. Actually, I want to flip the figure 
> over upper left->lower right diagonal and I want to
> do this without explicitly transposing the matrix if possible.
> 
> I also am having trouble with the figure size. I am working on a weird dpi 
> (93.65) screen. I tried
> 	fig.set_figsize_inches((5.,5.))
> 	fig.set_dpi(93.6585)
> 
> But the screen dimensions still come out off. They are in fact 5 5/16" x 5 
> 5/16". Any ideas?
Even though I wrote matshow, I have to admit that it was very much a 'stumble 
in the dark' kind of code. I basically just mucked around with imshow enough 
to get a matrix to display in the 'traditional' way (top-down, with properly 
preserved aspect ratio) and that was about it. I'm sure the code can use 
improvements, but at the time I can't commit any effort into it, I'm afraid.
If you can extend it (the code is extremely simple) in a useful way, just send 
a patch over and I'm sure we can apply it.
Regards,
f

>>>>> "zori" == zori <zo...@vi...> writes:
 zori> Hi, Exploring the agg.py module of the matplotlib library, I
 zori> have encauntered that I cannot represent smooth arcs. For
 zori> example, circle appears as octagon. Could you give me an
 zori> idea how to manage this and what is the reason for such
 zori> behaviour? Here is the code of the agg_test.py changed to
 zori> draw arc, but not giving the desired result:
Hi Zori,
I updated the agg_test.py example to show how to do a few more things
-- drawing curves, filling and stroking, and making affine
transformations of paths and curves. Down the road I would like to
replace refactor backend_agg.py and _backend_agg.cpp to use this swig
wrapper of agg, so that agg users would have access to the low level
agg library for customized drawing over a mpl figure.
# this example uses the agg python module directly there is no
# documentation -- you have to know how to use the agg c++ API to use
# it
import matplotlib.agg as agg
from math import pi
## Define some colors
red = agg.rgba8(255,0,0,255)
blue = agg.rgba8(0,0,255,255)
green = agg.rgba8(0,255,0,255)
black = agg.rgba8(0,0,0,255)
white = agg.rgba8(255,255,255,255)
yellow = agg.rgba8(192,192,255,255)
## Create the rendering buffer, rasterizer, etc
width, height = 600,400
stride = width*4
buffer = agg.buffer(width, height, stride)
rbuf = agg.rendering_buffer()
rbuf.attachb(buffer)
pf = agg.pixel_format_rgba(rbuf)
rbase = agg.renderer_base_rgba(pf)
rbase.clear_rgba8(blue) 
renderer = agg.renderer_scanline_aa_solid_rgba(rbase);
renderer.color_rgba8( red )
rasterizer = agg.rasterizer_scanline_aa()
scanline = agg.scanline_p8()
## A polygon path
path = agg.path_storage()
path.move_to(10,10)
path.line_to(100,100)
path.line_to(200,200)
path.line_to(100,200)
path.close_polygon()
# stroke it
stroke = agg.conv_stroke_path(path)
stroke.width(3.0)
rasterizer.add_path(stroke)
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
## A curved path
path = agg.path_storage()
path.move_to(200,10)
path.line_to(350,50)
path.curve3(150,200)
path.curve3(100,70)
path.close_polygon()
curve = agg.conv_curve_path(path)
# fill it
rasterizer.add_path(curve)
renderer.color_rgba8( green )
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
# and stroke it
stroke = agg.conv_stroke_curve(curve)
stroke.width(5.0)
rasterizer.add_path(stroke)
renderer.color_rgba8( yellow )
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
## Transforming a path
path = agg.path_storage()
path.move_to(0,0)
path.line_to(1,0)
path.line_to(1,1)
path.line_to(0,1)
path.close_polygon()
rotation = agg.trans_affine_rotation(pi/4)
scaling = agg.trans_affine_scaling(30,30)
translation = agg.trans_affine_translation(300,300)
trans = rotation*scaling*translation
transpath = agg.conv_transform_path(path, trans)
stroke = agg.conv_stroke_transpath(transpath)
stroke.width(2.0)
rasterizer.add_path(stroke)
renderer.color_rgba8( black )
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
## Converting a transformed path to a curve
path = agg.path_storage()
path.move_to(0,0)
path.curve3(1,0)
path.curve3(1,1)
path.curve3(0,1)
path.close_polygon()
rotation = agg.trans_affine_rotation(pi/4)
scaling = agg.trans_affine_scaling(30,30)
translation = agg.trans_affine_translation(300,250)
trans = rotation*scaling*translation
transpath = agg.conv_transform_path(path, trans)
curvetrans = agg.conv_curve_trans(transpath)
stroke = agg.conv_stroke_curvetrans(curvetrans)
stroke.width(2.0)
rasterizer.add_path(stroke)
renderer.color_rgba8( white )
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
## Display it with PIL
s = buffer.to_string()
print len(s)
import Image
im = Image.fromstring( "RGBA", (width, height), s)
im.show()

>>>>> "Jeff" == Jeff Peery <jef...@se...> writes:
 Jeff> Hello,
 Jeff> I'm plotting a simple xy plot and I'd like to highlight
 Jeff> regions that are above a threshold, say for example all
 Jeff> points y>=1. The method I am thinking of applying is
 Jeff> getting the x,y position of the point that is above the
 Jeff> threshold and filling the region between two vertical lines
 Jeff> that are right and left of the point with a bright color.
 Jeff> I was having some difficulty filling between two vertical
 Jeff> lines, I don't think I can use axvline() with fill()?
It sounds like axvspan is what you want
 def axvspan(self, xmin, xmax, ymin=0, ymax=1, **kwargs):
 """
 AXVSPAN(xmin, xmax, ymin=0, ymax=1, **kwargs)
 axvspan : Axis Vertical Span. xcoords are in data units and y coords
 are in axes (relative 0-1) units
 Draw a vertical span (regtangle) from xmin to xmax. With the default
 values of ymin=0 and ymax=1, this always span the yrange, regardless
 of the ylim settings, even if you change them, eg with the ylim
 command. That is, the vertical extent is in axes coords: 0=bottom,
 0.5=middle, 1.0=top but the y location is in data coordinates.
 kwargs are the kwargs to Patch, eg
 antialiased, aa
 linewidth, lw
 edgecolor, ec
 facecolor, fc
 the terms on the right are aliases
 return value is the patches.Polygon instance.
 # draw a vertical green translucent rectangle from x=1.25 to 1.55 that
 # spans the yrange of the axes
 axvspan(1.25, 1.55, facecolor='g', alpha=0.5)
For another approach, see examples/fill*.py, in particular
examples/fill_between.py.
But if all you want is shading across the vertical extent over an
xrange, axvspan is the function for you,\.
JDH

>>>>> "zori" == zori <zo...@vi...> writes:
 zori> Hi, Exploring the agg.py module of the matplotlib library, I
 zori> have encauntered that I cannot represent smooth arcs. For
 zori> example, circle appears as octagon. Could you give me an
 zori> idea how to manage this and what is the reason for such
 zori> behaviour? Here is the code of the agg_test.py changed to
 zori> draw arc, but not giving the desired result:
You have to call conv_curve on the path, and then stroke the result of
that
 curve = agg.conv_curve_path(path)
 stroke = agg.conv_stroke_curve(curve)
agg is a heavily templated library, and one or the challenges of
wrapping a template library is deciding which template combinations
need to be instantiated. Since matplotlib's agg wrapper is
undocumented, I'll explain some of the combinations and the naming
convention now. In the agg C++ library, the equivalent calls are
 
 typedef agg::conv_stroke<path_t> stroke_t;
 typedef agg::conv_curve<path_t> curve_t;
 agg::conv_curve<path_t> curve(path);
 agg::conv_stroke<curve_t> stroke(curve)
matplotlib basically takes the template argument and appends it to the
class name, eg
Agg C++ : matplotlib wrapper :
agg::conv_curve<path_t> agg.conv_curve_path
agg::conv_stroke<curve_t> agg.conv_stroke_curve
The typedefs (path_t, curve_t, etc) are defined in swig/agg_typedefs.h
in the swig directory of the matplotlib src distribution. The various
curves are defined in swig/agg_conv_curve.i and the strokes in
swig/agg_conv_stroke.i
The stroke and converter classes thus defined are
 conv_curve_path - convert path to curve, maps "path"->"curve"
 conv_curve_trans - convert a transformed path, maps "transpath"->"curvetrans"
 conv_stroke_path - stroke a path
 conv_stroke_transpath - stroke a transpath
 conv_stroke_curve - stroke a curve
 conv_stroke_transcurve - stroke a transform of a curve
 conv_stroke_curvetrans - stroke a curve of a transformed path
Hope this helps,
JDH

Dear matplotlib-users,
I have made an application for tsunami wave travel time prediction 
(slowmo.sf.net). The application uses the basemap toolkit and is developed on 
Linux. I would like to offer potential Windows users an easier way to install 
and test it than is currently available.
For this I would need a binary windows package of the basemap toolkit in a 
newer version than 0.21 which is currently available. Unfortunately I do not 
have access to the windows compilers that are necessary to make this binary 
package. I would therefore be very grateful if anyone from this list has the 
binary or could easily produce it.
Kind regards,
Jesper

Hi,
Exploring the agg.py module of the matplotlib library, I have 
encauntered that I cannot represent smooth arcs. For example, circle 
appears as octagon. Could you give me an idea how to manage this and 
what is the reason for such behaviour? Here is the code of the 
agg_test.py changed to draw arc, but not giving the desired result:
##########################
import matplotlib.agg as agg
width, height = 600,400
stride = width*4
buffer = agg.buffer(width, height, stride)
rbuf = agg.rendering_buffer()
rbuf.attachb(buffer)
red = agg.rgba8(255,0,0,255)
blue = agg.rgba8(0,0,255,255)
green = agg.rgba8(0,255,0,255)
black = agg.rgba8(0,0,0,255)
white = agg.rgba8(255,255,255,255)
yellow = agg.rgba8(192,192,255,255)
path = agg.path_storage()
path.move_to(10,10)
path.line_to(100,100)
path.line_to(200,200)
path.arc_to(100,100,0,0,1,100,200)
path.close_polygon()
stroke = agg.conv_stroke_path(path)
stroke.width(3.0)
pf = agg.pixel_format_rgba(rbuf)
rbase = agg.renderer_base_rgba(pf)
rbase.clear_rgba8(blue)
renderer = agg.renderer_scanline_aa_solid_rgba(rbase);
renderer.color_rgba8( red )
rasterizer = agg.rasterizer_scanline_aa()
rasterizer.add_path(stroke)
scanline = agg.scanline_p8()
agg.render_scanlines_rgba(rasterizer, scanline, renderer);
s = buffer.to_string()
print len(s)
import Image
im = Image.fromstring( "RGBA", (width, height), s)
im.show()
Thank you!
Regards,
Z.Danevska

Steve Schmerler wrote:
> Hi
> 
> I'm using MPL 0.81 on a Debian box. When I call ipython with the pylab 
> option I get a GTK message which bothers me.
> 
> 
> ===============================================================================
> elcorto@ramrod:~$ ipython -pylab
> /usr/lib/python2.3/site-packages/IPython/Shell.py:628: 
> GtkDeprecationWarning: gtk.timeout_add is deprecated, use 
> gobject.timeout_add instead
> self.gtk.timeout_add(self.TIMEOUT, self.on_timer)
> Python 2.3.5 (#2, May 4 2005, 08:51:39)
> Type "copyright", "credits" or "license" for more information.
> 
> IPython 0.6.13 -- An enhanced Interactive Python.
> ? -> Introduction to IPython's features.
> %magic -> Information about IPython's 'magic' % functions.
> help -> Python's own help system.
> object? -> Details about 'object'. ?object also works, ?? prints more.
> 
> Welcome to pylab, a matplotlib-based Python environment.
> For more information, type 'help(pylab)'.
> ===============================================================================
> 
> So far I discovered no problems while using MPL but I'm wondering what 
> this message is.
This is harmless, and will go away when you upgrade to ipython 0.6.15.
Best,
f

howdy,
I have some beginner type questions with matshow.
I want the origin of the axes to be the lower left of the figure. I tried:
	fig=matshow(data, cmap, origin='lower')
but that didn't seem to do anything. Actually, I want to flip the figure 
over upper left->lower right diagonal and I want to
do this without explicitly transposing the matrix if possible.
I also am having trouble with the figure size. I am working on a weird dpi 
(93.65) screen. I tried
	fig.set_figsize_inches((5.,5.))
	fig.set_dpi(93.6585)
But the screen dimensions still come out off. They are in fact 5 5/16" x 5 
5/16". Any ideas?
thanks,
danny
Here's a full file:
#####################################
from pylab import *
from matplotlib import cm
data=zeros((10,10))
for i in range(10):
	for j in range(10):
		data[i,j]=i+j
		
fig=matshow(data, cm.jet)#, origin='lower'
fig.set_figsize_inches((5.,5.))
fig.set_dpi(72.) #93.6585
savefig('tempPic.png')

Hi
I'm using MPL 0.81 on a Debian box. When I call ipython with the pylab 
option I get a GTK message which bothers me.
===============================================================================
elcorto@ramrod:~$ ipython -pylab
/usr/lib/python2.3/site-packages/IPython/Shell.py:628: 
GtkDeprecationWarning: gtk.timeout_add is deprecated, use 
gobject.timeout_add instead
 self.gtk.timeout_add(self.TIMEOUT, self.on_timer)
Python 2.3.5 (#2, May 4 2005, 08:51:39)
Type "copyright", "credits" or "license" for more information.
IPython 0.6.13 -- An enhanced Interactive Python.
? -> Introduction to IPython's features.
%magic -> Information about IPython's 'magic' % functions.
help -> Python's own help system.
object? -> Details about 'object'. ?object also works, ?? prints more.
 Welcome to pylab, a matplotlib-based Python environment.
 For more information, type 'help(pylab)'.
===============================================================================
So far I discovered no problems while using MPL but I'm wondering what 
this message is.
cheers,
steve
	

The problem is that nearly all of the video codecs are proprietary, and 
encoding, in particular, usually incurs money changing hands somewhere. 
 Some distributions (Debian) therefore are philosophically unable to 
include much in the way of movie tools, and others go through hoops to 
avoid incurring the wrath of the lawyers. I've recently heard that 
ffmpeg works pretty well. After spending many frustrating hours trying 
to get something to work, I've decided for now that using QuickTime Pro 
is what I'll use (30ドル for Windows or Mac version). It can generate AVIs 
(limited number of codecs, at least on Macs), MOVs, and MPEG-4s from 
image sequences, and It Just Works. Good luck, and let us know if you 
find anything that works well in linux.
Cheers!
Andrew
On Jun 21, 2005, at 2:39 PM, Danny Shevitz wrote:
> This is not strictly speaking a matplotlib question, but I have been 
> trying to get the movie_demo.py file working on
> a windows box using the os.system command with mencoder.
>
> I'm curious if anyone has gotten this to work with vcodec=mpeg4 so 
> that the subsequent video can be opened with Window Media player.
> I have spent all day searching the web and playing with stuff. I have 
> gotten readable output with wmv2 and mpeg1video. That's it.
>
> If you have had some success can you please contact me and I'll ask 
> you some questions off the list.
>
> thanks,
> Danny
>
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hello, 
I'm plotting a simple xy plot and I'd like to highlight regions that are
above a threshold, say for example all points y>=1. The method I am
thinking of applying is getting the x,y position of the point that is
above the threshold and filling the region between two vertical lines
that are right and left of the point with a bright color. 
I was having some difficulty filling between two vertical lines, I don't
think I can use axvline() with fill()?
Is there a more efficient way to do this... maybe creating a filled
shape behind the highlight points? Any help would be much appreciated!
Thanks.
Jeff

>>>>> "Eric" == Eric Emsellem <ems...@ob...> writes:
 Eric> followed what you suggested, everything was fine until the
 Eric> 'grep' : for some obscure reason every new install produced
 Eric> an empty (0 byte) backend_ps.py file (I think this may have
 Eric> come from a previously interrupted 'build'). I therefore
 Eric> erased the original matplotlib source dir (as well as the
 Eric> installed ones), and redetared the archive starting from
 Eric> scratch.. It works...
Oh -- both Darren and I forgot to suggest rm -rf "build" in addition
to site-packages/matplotlib. This is, however, suggested in
http://matplotlib.sourceforge.net/faq.html#FAILURE
 The first thing to try is to remove site-packages/matplotlib and
 reinstall. If you are running the windows installer or using some
 other package, that is all you need to do. If you are compiling
 matplotlib yourself, you also need to remove the build subdirectory
 of the matplotlib src tree you are compiling from. It is not enough
 to do python setup.py clean. Try rebuilding and reinstalling
 matplotlib after removing these two directories and see if that
 helps. If 
Glad you found and fixed the problem!
JDH

followed what you suggested, everything was fine until the 'grep' : for 
some obscure reason every new install produced an empty (0 byte) 
backend_ps.py file (I think this may have come from a previously 
interrupted 'build'). I therefore erased the original matplotlib source 
dir (as well as the installed ones), and redetared the archive starting 
from scratch..
It works...
thanks a lot for the support!!!!
Eric
John Hunter wrote:
>>>>>>"Eric" == Eric Emsellem <ems...@ob...> writes:
>>>>>> 
>>>>>>
>
> Eric> Thanks for the tip, but as you and John suggested I tried a
> Eric> new clean installation (both for matplotlib - removing the
> Eric> site-packages and /usr/share matplotlib dir - and ipython)
> Eric> but the small test I sent you still react the same. The
> Eric> specific import command:
>
> Eric> from matplotlib.backends.backend_ps import FigureCanvasPS
>
> Eric> does indeed give an ImportError:
>
> Eric> ImportError: cannot import name FigureCanvasPS
>
> Eric> not sure what to do now. Thanks for any input there!
>
>Strange indeed. Issue these commands and let me know what you find
>
> >>> import matplotlib.backends.backend_ps
> >>> matplotlib.__file__
> '/usr/lib/python2.4/site-packages/matplotlib/__init__.pyc'
> >>> matplotlib.backends.backend_ps.__file__
> '/usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.pyc'
>
> >>> import matplotlib
> >>> matplotlib.__version__
> '0.82'
> >>> matplotlib.__file__
> '/usr/lib/python2.4/site-packages/matplotlib/__init__.pyc'
>
>
>Also, what do these commands from the unix shell return (substituting
>in your path to backend_ps.py)
>
> peds-pc311:~> md5sum /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
> 2bb2161d70c11f56e803d7749b4fe05a /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
> peds-pc311:~> grep '^class FigureCanvas' /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
> class FigureCanvasPS(FigureCanvasBase):
>
>The only things I can think of are 
>
> 1) you have an errant copy of matplotlib.backends.backend_ps
> somewhere in your path
>
> 2) you have a corrupted backend_ps (hence the md5sum check)
>
> 3) You did not successfully remove site-packages/matplotlib and have
> a corrupt install.
>
>JDH
>
> 
>
-- 
===============================================================
Observatoire de Lyon ems...@ob...
9 av. Charles-Andre tel: +33 4 78 86 83 84
69561 Saint-Genis Laval Cedex fax: +33 4 78 86 83 86
France http://www-obs.univ-lyon1.fr/eric.emsellem
===============================================================

Hello, I'm having some difficulty getting a simple timer working. I
attached a quick example. I setup the timer, gave it a event handler and
start the timer with .Start(). It seems to start, but the timer event
doesn't happen... that is in my example the event function that I
created never is called. Any help would be much appreciated! Thanks.
Jeff

>>>>> "Eric" == Eric Emsellem <ems...@ob...> writes:
 Eric> Thanks for the tip, but as you and John suggested I tried a
 Eric> new clean installation (both for matplotlib - removing the
 Eric> site-packages and /usr/share matplotlib dir - and ipython)
 Eric> but the small test I sent you still react the same. The
 Eric> specific import command:
 Eric> from matplotlib.backends.backend_ps import FigureCanvasPS
 Eric> does indeed give an ImportError:
 Eric> ImportError: cannot import name FigureCanvasPS
 Eric> not sure what to do now. Thanks for any input there!
Strange indeed. Issue these commands and let me know what you find
 >>> import matplotlib.backends.backend_ps
 >>> matplotlib.__file__
 '/usr/lib/python2.4/site-packages/matplotlib/__init__.pyc'
 >>> matplotlib.backends.backend_ps.__file__
 '/usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.pyc'
 >>> import matplotlib
 >>> matplotlib.__version__
 '0.82'
 >>> matplotlib.__file__
 '/usr/lib/python2.4/site-packages/matplotlib/__init__.pyc'
Also, what do these commands from the unix shell return (substituting
in your path to backend_ps.py)
 peds-pc311:~> md5sum /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
 2bb2161d70c11f56e803d7749b4fe05a /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
 peds-pc311:~> grep '^class FigureCanvas' /usr/lib/python2.4/site-packages/matplotlib/backends/backend_ps.py
 class FigureCanvasPS(FigureCanvasBase):
The only things I can think of are 
 1) you have an errant copy of matplotlib.backends.backend_ps
 somewhere in your path
 2) you have a corrupted backend_ps (hence the md5sum check)
 3) You did not successfully remove site-packages/matplotlib and have
 a corrupt install.
JDH

>>>>> "Jo=E3o" =3D=3D Jo=E3o Lu=EDs Silva <rom...@ya...> writes:
 Jo=E3o> Hello all, I'm making a small visualization application
 Jo=E3o> based on matplotlib which is in turn embedded in gtk2 using
 Jo=E3o> pygtk. The code goes like this:
 Jo=E3o> fig =3D matplotlib.figure() ax =3D fig.add_subplot(111) im =3D
 Jo=E3o> ax.imshow(A) canvas =3D FigureCanvas(fig) [...]
 Jo=E3o> and it works fine. However, after an user action I need to
 Jo=E3o> update the image, and so I do a
 Jo=E3o> im.set_data(NewData)
 Jo=E3o> but the image remains unchanged, and will only update if I
 Jo=E3o> somehow move it using the toolbar. I googled a bit and found
 Jo=E3o> that perhaps fig.draw(renderer) would work, but I have no
 Jo=E3o> idea as how to get a reference to the renderer. Any ideas?
call canvas.draw() which creates the renderer and calls
fig.draw(renderer). The segregation between the canvas and the figure
was done precisely to decouple the backend (eg renderer) from the
frontend (eg figure).
See examples/dynamic_image_gtkagg.py for an example of how to
dynamically update the gtkagg canvas.
JDH

>>>>> "Nils" == Nils Wagner <nw...@me...> writes:
 Nils> Hi all, Why do I get a math domain error if I import scipy
 Nils> before pylab in cdiff.py (see attachment) ?
Hi Nils,
It helps to try and reduce your error to a minimal example that
exposes the bug. That will save my time and help you find the real
cause of the bug. If you had done this, you would have seen that the
problem is in the Numeric call to sqrt of a negative number, and does
not involve matplotlib or scipy
 from Numeric import sqrt, sin, cos
 z = 2.5
 print 'arg is negative', sin(z)**3+cos(z)**3
 print 'this will fail', sqrt(sin(z)**3+cos(z)**3)
numarray would return nan.
JDH

Hi all,
Why do I get a math domain error if I import scipy before pylab in 
cdiff.py (see attachment) ?
numerix Numeric 24.0b2
Traceback (most recent call last):
 File "cdiff.py", line 28, in ?
 e1=abs((f(z+h)-f(z))/h-fd(z))/abs(fd(z))
 File "cdiff.py", line 6, in f
 return exp(z)/sqrt(sin(z)**3+cos(z)**3)
ValueError: math domain error
/home/nwagner> vi cdiff.py
/home/nwagner> /usr/bin/python cdiff.py
numerix Numeric 24.0b2
Traceback (most recent call last):
 File "cdiff.py", line 30, in ?
 e1=abs((f(z+h)-f(z))/h-fd(z))/abs(fd(z))
 File "cdiff.py", line 6, in f
 return exp(z)/sqrt(sin(z)**3+cos(z)**3)
ValueError: math domain error
Nils

Hello all,
I'm making a small visualization application based on matplotlib which 
is in turn embedded in gtk2 using pygtk. The code goes like this:
fig = matplotlib.figure()
ax = fig.add_subplot(111)
im = ax.imshow(A)
canvas = FigureCanvas(fig)
[...]
and it works fine. However, after an user action I need to update the 
image, and so I do a
im.set_data(NewData)
but the image remains unchanged, and will only update if I somehow move 
it using the toolbar. I googled a bit and found that perhaps 
fig.draw(renderer) would work, but I have no idea as how to get a 
reference to the renderer. Any ideas?
Thanks,
João Silva

Thanks for the tip, but as you and John suggested I tried a new clean 
installation (both for matplotlib - removing the site-packages and 
/usr/share matplotlib dir - and ipython) but the small test I sent you 
still react the same. 
The specific import command:
from matplotlib.backends.backend_ps import FigureCanvasPS
does indeed give an ImportError:
ImportError: cannot import name FigureCanvasPS
not sure what to do now. Thanks for any input there!
Eric
Darren Dale wrote:
>Hi Eric,
>
>On Tuesday 21 June 2005 11:16 am, Eric Emsellem wrote:
> 
>
>>for some reason I have now (with the new matplotlib) problems with ps
>>files saving figures with savefig.
>>(it works with gif, jpg, etc).
>>
>>[I am running Python 2.3.4, IPython 0.6.12, matplotlib 0.82
>>and verbose.level helpful, interactive is False, platform is linux2,
>>numerix numarray 1.3.0]
>>
>>So here is what I get with a simple example:
>>
>>scatter(arange(0,10,1), arange(10,20,1))
>>savefig('toto.ps')
>> 
>>
>
>I didnt have any trouble with this example. Your traceback shows backend_agg 
>cant import FigureCanvasPS. Does this work in ipython?
>
>from matplotlib.backends.backend_ps import FigureCanvasPS
>
>Did you try to removing the matplotlib directory in site-packages, and 
>reinstalling?
>
> 
>
-- 
===============================================================
Observatoire de Lyon ems...@ob...
9 av. Charles-Andre tel: +33 4 78 86 83 84
69561 Saint-Genis Laval Cedex fax: +33 4 78 86 83 86
France http://www-obs.univ-lyon1.fr/eric.emsellem
===============================================================

This is not strictly speaking a matplotlib question, but I have been trying 
to get the movie_demo.py file working on
a windows box using the os.system command with mencoder.
I'm curious if anyone has gotten this to work with vcodec=mpeg4 so that the 
subsequent video can be opened with Window Media player.
I have spent all day searching the web and playing with stuff. I have 
gotten readable output with wmv2 and mpeg1video. That's it.
If you have had some success can you please contact me and I'll ask you 
some questions off the list.
thanks,
Danny

Jeff, 
On 2005年6月21日, Jeff Peery wrote:
> Hello, is there a way to check if a plot is currently open? I'm using
> wxAgg to show plots in a frame. I want the user to be able to show
> different plots in the same frame, although only one at a time. So I
> need to know if a plot is already there so I can close it and create a
> new one. Thanks!
 
I think this may need to be backend dependent. For wxAgg, I put
the matplotlib figure canvas on a wx.Frame, and then check
whether that wx.Frame exists. That is, with 'self.plotframe'
being my toplevel frame that has a FigureCanvasWxAgg, and is
built with PlotFrame(), I run this prior to each plot:
 def ShowPlotFrame(self, do_raise=True):
 "make sure plot frame is enabled, and, optionally visible"
 # if plotframe does not exist, build it
 if self.plotframe == None: 
 self.plotframe = PlotFrame(self)
 try:
 self.plotframe.Show()
 except wx.PyDeadObjectError:
 self.plotframe = PlotFrame(self) 
 self.plotframe.Show()
	# make sure the plot frame is raised.
 if do_raise: 
 self.plotframe.Raise()
The 'except wx.PyDeadObjectError:' checks if the User has killed
the Frame from the Window Manager. This could probably be made
more robust, but it works well for me.
Hope that helps,
--Matt

Nadia, John,
Thanks for the help/pointers. Unfortunately nothing worked so far.
I started with gcc 3.4.3 and even compiled latest available 3.X gcc (3.4.4) but that did not help.
It sounds like older versions of Solaris work. In my case downgrading the OS is not an option.
Miro
> The only way I can compile matplotlib on Solaris is using gcc
> (by setting CC to gcc).
> I am on Solaris 8 and use gcc 3.2.2.
> Python was compiled with Sun's compilers, but I think as long
> as python version is 2.3 or later, distutils will pick up the
> compiler from the environment variable CC.
> 
> Nadia
> 
> John Hunter wrote:
> >>>>>>"mmadecki" == mmadecki <mma...@co...> writes:
> > 
> > 
> > mmadecki> Hi, I'm trying to compile Matplotlib 0.82 on Solaris 10
> > mmadecki> (SunOS 5.10_x86). I've seen similar posts but no
> > mmadecki> solutions to this problem (see errors below). Any help
> > mmadecki> is greatly appreciated.
> > 
> > If you search for "error" from the beginning of your gcc output, the
> > first match is 
> > 
> > error: `::btowc' has not been declared
> > 
> > If you paste that into google, the first result you'll see is a post
> > to matplotlib from Mike Rightmire, who offlist and with the help of
> > Thomas Wessell, worked around this problem with a brute force
> > approach. I believe Erik Curiel had a similar problem on a solaris
> > box recently.
> > 
> > The most informative google result I saw was when I took you error
> > message and added some quotes to reduce false matches
> > 
> > error: `::btowc' "has not been declared"
> > 
> > [Aside: putting double quotes around the entire string did not match
> > this result; apparently google does not handled nested quotes very
> > well.] Fourth from the top and was
> > http://lists.schmorp.de/pipermail/rxvt-unicode/2005q2/000092.html .
> > This thread is not related to matplotlib, and first response there
> > isn't too helpful, but if you follow the rest of the thread you'll get
> > lots of good information. In particular, this response looks helpful
> > http://lists.schmorp.de/pipermail/rxvt-unicode/2005q2/000104.html
> > 
> > Nadia is the only person I know personally who actively compiles mpl
> > on a solaris box; I hope she has some suggestions for you.
> > 
> > Cheers,
> > JDH
> > 
> > 
> > -------------------------------------------------------
> > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> > from IBM. Find simple to follow Roadmaps, straightforward articles,
> > informative Webcasts and more! Get everything you need to get up to
> > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users

The only way I can compile matplotlib on Solaris is using gcc
(by setting CC to gcc).
I am on Solaris 8 and use gcc 3.2.2.
Python was compiled with Sun's compilers, but I think as long
as python version is 2.3 or later, distutils will pick up the
compiler from the environment variable CC.
Nadia
John Hunter wrote:
>>>>>>"mmadecki" == mmadecki <mma...@co...> writes:
> 
> 
> mmadecki> Hi, I'm trying to compile Matplotlib 0.82 on Solaris 10
> mmadecki> (SunOS 5.10_x86). I've seen similar posts but no
> mmadecki> solutions to this problem (see errors below). Any help
> mmadecki> is greatly appreciated.
> 
> If you search for "error" from the beginning of your gcc output, the
> first match is 
> 
> error: `::btowc' has not been declared
> 
> If you paste that into google, the first result you'll see is a post
> to matplotlib from Mike Rightmire, who offlist and with the help of
> Thomas Wessell, worked around this problem with a brute force
> approach. I believe Erik Curiel had a similar problem on a solaris
> box recently.
> 
> The most informative google result I saw was when I took you error
> message and added some quotes to reduce false matches
> 
> error: `::btowc' "has not been declared"
> 
> [Aside: putting double quotes around the entire string did not match
> this result; apparently google does not handled nested quotes very
> well.] Fourth from the top and was
> http://lists.schmorp.de/pipermail/rxvt-unicode/2005q2/000092.html .
> This thread is not related to matplotlib, and first response there
> isn't too helpful, but if you follow the rest of the thread you'll get
> lots of good information. In particular, this response looks helpful
> http://lists.schmorp.de/pipermail/rxvt-unicode/2005q2/000104.html
> 
> Nadia is the only person I know personally who actively compiles mpl
> on a solaris box; I hope she has some suggestions for you.
> 
> Cheers,
> JDH
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users