matplotlib-users Mailing List for matplotlib

Smith, Frank wrote:
> I'm really just starting to use it. I've picked wxFloatCanvas mainly
> because it's a link between Matplotlib and PythonCard. PythonCard
> recently added a FloatCanvas widget as a "component" that can be added
> to the main background window of a PythonCard application. Since
> FloatCanvas inherits from wxFloatCanvas, and in turn from wxPanel (which
> is the basis of the Figure Canvas of MatPlotLib) In answer to your last
> question: I was hoping to build a simple plot editor application -
> something that can add or remove points, change titles, ticks, fonts,
> etc. Although limited, PythonCard has a nice clean and simple structure
> that I was hoping to utilize.
It sounds like you need MPL, not FloatCanvas. Trying to put a MPL figure 
Canvas on top of a FloatCanvas makes no Sense.
The other option is that you could use FloatCanvas all by itself. If 
your needs for the axis, etc are simple, it could be pretty easy to 
build what you want, and the interactive part is pretty simple. I've 
enclosed a simple sample of a moving line plot.
If you decide to go with MPL, I'd start by looking at the code for the 
PythonCard FloatCanvas widget, and see how it is done. Making a similar 
wrapper for an MPL Figure Canvas should be straight forward, and I'm 
sure you'll get help from the PythonCard folks.
However, if you want to edit the plot by point & clicking, you've got a 
job ahead of you, and it will probably take learning some about wxPython 
internals.
Good luck,
-Chris
PS Have you check out Chaco? I haven't been following its development 
lately, but a while back they had some very nice GUI plot-editing 
features, using wxPython.
-- 
Christopher Barker, Ph.D.
Oceanographer
 		
NOAA/OR&R/HAZMAT (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
Chr...@no...

Thanks for your advice with installing matplotlib on cygwin. I =
downloaded and installed the windows binaries and it worked.
Anyway, the reason that I didn't want to use binaries in the first place =
was because I wanted to modify the matplotilb source code. But it seems =
like even with the binaries, if I change the source code then it will =
still affect the operation of the program when I run it, which is what I =
want.
In particular, I am looking to speed up the pcolor() function because it =
runs exceedingly slow with large mesh sizes. I believe the reason it is =
running slow is because of a memory leak. When I do the following:
from pylab import *
n=3D200
[x,y]=3Dmeshgrid(arange(n+1)*1./n,arange(n+1)*1./n)
z=3Dsin(x**2 + y**2)
and then do
pcolor(x,y,z)
repeatedly, the memory usage increases by about 15 MB each time, and it =
runs progressively slower.each time. By using the profiler I can see =
that almost all the time is spent in the pcolor function itself, rather =
than any functions it calls. When I take out the line =
"self.add_collection(collection)", there is no memory leak and it runs =
much faster (of course, since I'm not actually adding the collection on, =
when I do that i can no longer actually see the plot). It seems to me =
like what is happening is that it is repeatedly appending the same =
PolyCollection to the collections list, using up more and more memory. =
What I want to know is what the "collections" are and how they are used =
so I can figure out a way of getting rid of extraneous collections.
-Alexander Mont

On Monday 31 October 2005 3:13 pm, Robert Hetland wrote:
> When I make a plot, and export it in postscript (or eps), often I
> like to modify it in Adobe Illustrator. However, the fonts that
> maplotlib produces seem to be postscript representations of the
> actual fonts. That is, Illustrator does not recognize the text as
> text, but as some object (that is, of course, shaped like text).
>
> How do I make matplotlib export images with text that is recognizable
> and editable in Illustrator?
Postscript output, by default, does export text as fonts. So I am guessing 
that you are using mpl's latex support, which unfortunately does convert 
fonts to outlines. It is a deficiency with ghostscript, not mpl. There is a 
hack to get around this, which is published at 
http://www.scipy.org/wikis/topical_software/UsingTex. The only other 
alternative at the moment is to set usetex=False in your RC file.

When I make a plot, and export it in postscript (or eps), often I 
like to modify it in Adobe Illustrator. However, the fonts that 
maplotlib produces seem to be postscript representations of the 
actual fonts. That is, Illustrator does not recognize the text as 
text, but as some object (that is, of course, shaped like text).
How do I make matplotlib export images with text that is recognizable 
and editable in Illustrator?
-Rob.
-----
Rob Hetland, Assistant Professor
Dept of Oceanography, Texas A&M University
p: 979-458-0096, f: 979-845-6331
e: he...@ta..., w: http://pong.tamu.edu

Smith, Frank wrote:
> Unfortunately I'm more familiar with PythonCard which hides just
> about everything about wxWindows/wxPython.
I really need to check it out again, it's been a while.
> Unfortunately wxPyPlot doesn't
> have the functionality I need. PythonCard does support wxFloatCanvas.
> The application I'm developing has a main startup GUI comoposed of
> several components, one of which is a FloatCanvas. What I want to do
> is display the matplotlib plot (or plots) in this already existing
> component. I'm not sue this is even possible: I guess I want to
> assign the figure container to the PythonCard component.
I don't know how to "wrap" an particular wx.window() for use with 
PythonCard, but MPL simple uses a custom wx.Panel, just like 
FloatCanvas, so it certainly can be done. I'd ask on the PythonCard 
lists, I'm sure someone can help you out, it shouldn't be a big deal.
By the way, what are you using FloatCanvas for? I'm always curious how 
it's being used.
-Chris
-- 
Christopher Barker, Ph.D.
Oceanographer
 		
NOAA/OR&R/HAZMAT (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
Chr...@no...

John Hunter <jdh...@ac...> writes:
> >>>>> "Alexander" == Alexander Mont <ale...@co...> writes:
> Alexander> line 100, in add_base_flage incdirs = [os.path.join(p,
> Alexander> 'include') for p in basedir[sys, platform] KeyError:
> Alexander> 'cygwin'
> 
> This is almost certainly a typo. I guess you mean 'sys.platform'
> 
> Alexander> How do I fix this? I installed numarray and I changed the NUMERIX value in the matplotlibrc file from Numeric to numarray but I still have the same problem.
> 
> This has nothing to do with the numerix setting. Check out
> setupext.py and add an entry to the basedir dictionary to reflect your
> local path information. If you get something that works, please post
> it so we can include it in the standard distro.
I don't have my cygwin machine with me at the moment, but I recall
needing to add a line to setupext.py something like:
 'linux' : ['/usr/local', '/usr',], ### This line already present
 'cygwin' : ['/usr/local', 'usr',],
Once I'd solved this, I had problems with the path for Tk - but I
can't remember exactly what they were
> 
> Is there a reason you don't want to use the pre-built win32 binaries?
> Most windows users are reasonably happy with that.
I can't speak for Alexander, but I went down the same path (and have
managed to get matplotlib working under cygwin [1]). I did so mainly
because of familiarity with the unix style of doing things. I use
cygwin to get a unix style way of working, but still have word etc
available without dual booting.
Specifically, I have a python/c module that I wanted to use to read in
the data I wanted to plot with matplotlib - and while this module was
easy to compile under cygwin, I couldn't work out how to compile it
for use with win32 (using gcc as I don't have MSVC).
Chris
[1] If sending you this source would be useful, then I can do so - I'd
hoped to write a recipe, but have been too busy. The main problem was
getting all the other things that are necessary for matplotlib, but
not included with cygwin to build, rather than matplotlib itself.

Thank you, Ken, Ryan, and Charlie, for your help.
I'm sorry for the late reply, but I wasn't online over the weekend and
have got no time before.
Anyway, I managed to install matplotlib with your help and got no
serious complaint during compilation (simply neglecting a few warnings).
Importing matplotlib and pylab works. I didn't do any testing so far,
but at least I got to this point and you deserved some feedback before
I'll get to the testing, since this may take until tomorrow.
Cheers
Christian

On 2005年10月30日, Stephen Walton wrote:
> Arnd Baecker wrote:
>
> >To be clear (hopfully):
> >one clicks with button1, keeps it pressed
> >and moves the mouse. While moving one releases button1
> >but continues moving.
> >
> >
> Confirmed here with matplotlib 0.84, GTKAgg backend, Fedora Core 4.
> Maybe the solution is simple: stop the mouse before releasing the
> button :-)
>
> "Doctor, it hurts when I do this." "Well don't do that."
For me that directly translates into "zooming hurts" ;-).
Actually, it took us quite a while to find out in
which situation the problem occurs.
How often it happens does depend on the way the user
interacts with the mouse, but:
Any tiny movement of the mouse while releasing the left mouse
button will have the same effect.
Best, Arnd

>>>>> "Gene" == Gene <ce...@nm...> writes:
 Gene> version 8.4 FPS: 34.4115620704 Fatal Python error:
 Gene> PyEval_RestoreThread: NULL tstate
 Gene> This application has requested the Runtime to terminate it
 Gene> in an unusual way. Please contact the application's support
 Gene> team for more information.
Search the archives for this message -- it has occurred in a number of
guises. We've never found a way to make tkagg work in every supported
mode on winxp w/o this message. It's mostly considered harmless since
it only happens at program termination but it is annoying.
Suggested solution. Don't use "pylab animation" (ie repeated calls to
draw in interactive mode). Instead, use tk's idle handler or timer to
trigger animation calls with a proper tk event loop, as in
examples/anim_tk.py.
Hope this helps,
JDH

Dirk Junghanns wrote:
>
> Hi,
>
> I have a countourf() plot, overlayed by countour().
> Now I want to plot additional points (here: errors in data) like 
> scatter over the existing figure.
Maybe I don't understand the question, but I just ran the contourf_demo 
example, then did
hold(True)
plot([2.0],[1.0],'+')
and saw the plotted point.

Rick Muller wrote:
> I'm finding some really weird behavior in matplotlib that I just can't 
> figure out.
> 
> When I plot out the following short script:
> 
> >>> from pylab import *
> >>> y = [-48.0694,-48.0754,-48.0772,-48.0781,-48.0782,-48.0781,
> -48.0772,-48.0720]
> >>> plot(y)
> >>> savefig('bs.png')
> 
> I get a y-axis that goes from -10 to 0, instead of -48.0782 - -48.0694. 
> And in the upper left hand corner of the plot I see the string 
> "x1e-3-4.8069e1".
> 
> I'm running python 2.3.5 on Mac OS X 10.4.2. Matplotlib 0.82.
> 
> Any hints?
> 
> Rick Muller
> rm...@sa...
> 
I get the same on WinXP, MPL 0.82.
cheers,
Steve
-- 
grep -ir fuck /usr/src/linux

I'm finding some really weird behavior in matplotlib that I just 
can't figure out.
When I plot out the following short script:
 >>> from pylab import *
 >>> y = [-48.0694,-48.0754,-48.0772,-48.0781,-48.0782,-48.0781,
 -48.0772,-48.0720]
 >>> plot(y)
 >>> savefig('bs.png')
I get a y-axis that goes from -10 to 0, instead of -48.0782 - 
-48.0694. And in the upper left hand corner of the plot I see the 
string "x1e-3-4.8069e1".
I'm running python 2.3.5 on Mac OS X 10.4.2. Matplotlib 0.82.
Any hints?
Rick Muller
rm...@sa...

Arnd Baecker wrote:
>To be clear (hopfully):
>one clicks with button1, keeps it pressed
>and moves the mouse. While moving one releases button1
>but continues moving.
> 
>
Confirmed here with matplotlib 0.84, GTKAgg backend, Fedora Core 4. 
Maybe the solution is simple: stop the mouse before releasing the 
button :-)
"Doctor, it hurts when I do this." "Well don't do that."

Darren Dale wrote:
>For those who are not on the scipy-dev list, Travis just informed me that 
>upgrading to Numeric 24 will solve the problem I reported. I was working with 
>23.8, the new version seems to have solved my problem.
> 
>
And for those running Fedora Core 4, which has a dependency of pygtk2 on 
Numeric and therefore installs the package python-numeric-23.7, I 
sort-of recommend building Numeric 24.0 from source ("python setup.py 
bdist_rpm") and then installing the resulting RPM with 
"--replacefiles". I haven't thoroughly tested everything, but so far 
this setup seems to work for me.

Hi,
there is something weird going on when zooming
(matplotlib.__version__: '0.85.cvs').
If I do
from pylab import *
x=arange(0.0,10.0,0.01)
plot(x,sin(x**2))
show()
and then choose the "zoom to rect mode"
it is possible to create, one after another,
several zoom boxes which are all visible and the interiour is not
updated, while the axes ranges are.
(Moving the mouse outside of the window with the figure
will update the plot.)
This seems to happen reproducibly,
when one zooms in such a way that
one release button1 while still dragging.
To be clear (hopfully):
one clicks with button1, keeps it pressed
and moves the mouse. While moving one releases button1
but continues moving.
Somehow the redraw for the inside is not issued
when the mouse is still moving but the button already released.
(Note that this is not an artificial problem, but
does happen a lot).
Is this behaviour reproducible on other systems/platforms
(I am running this on debian sarge with the default backend)?
Best, Arnd

So far, when I run a script having the command draw() including two I 
copied from http://matplotlib.sourceforge.net/tutorial.html a fatal 
python error occurs when the script completes. I've tried different 
backends and numeric libraries with out success. Win XP. The scripts 
are run in a cmd box. 
Any ideas?
Thanks,
Gene
 from pylab import *
 import time
 # turn interactive mode on for dynamic updates. If you aren't in
 # interactive mode, you'll need to use a GUI event handler/timer.
 ion() 
 tstart = time.time() # for profiling
 x = arange(0,2*pi,0.01) # x-array
 line, = plot(x,sin(x))
 for i in arange(1,50):
 line.set_ydata(sin(x+i/10.0)) # update the data
 draw() # redraw the canvas
 print 'FPS:' , 200/(time.time()-tstart)
e:\Data\NRAO\DG DTE>m1.py --verbose-helpful
matplotlib data path E:\Programs\Python24\share\matplotlib
$HOME=E:\Documents and Settings\Gene
CONFIGDIR=E:\Documents and Settings\Gene\.matplotlib
loaded rc file E:\Documents and Settings\Gene\.matplotlib\matplotlibrc
matplotlib version 0.84
verbose.level helpful
interactive is False
platform is win32
numerix numarray 1.4.1
font search path ['E:\\Programs\\Python24\\share\\matplotlib']
loaded ttfcache file E:\Documents and 
Settings\Gene\.matplotlib\ttffont.cache
backend TkAgg version 8.4
FPS: 34.4115620704
Fatal Python error: PyEval_RestoreThread: NULL tstate
This application has requested the Runtime to terminate it in an unusual 
way.
Please contact the application's support team for more information.
e:\Data\NRAO\DG DTE>

Thanks Ken. I will try this tomorrow after I make sure I have the
right wxPython.h.
Ryan
On 10/28/05, Ken McIvor <mc...@ii...> wrote:
> Ryan,
>
> I'm afraid there isn't a Debian package for Python2.4+wxPython2.6, so I c=
an't
> test that combination. Our best bet is to go for the GDB stack trace, li=
ke so:
>
> 1. Write an iPython script that reproduces the segfault:
> $ ipython <test-script>
> <snip>
> Segmentation fault
> 2. Find out where Python lives:
> $ which python
> /usr/bin/python
> 2. Find out where iPython lives:
> $ which ipython
> /usr/bin/ipython
> 3. Launch gdb:
> $ gdb <path-to-python>
> <snip>
> (gdb) run <path-to-ipython> <test-script>
> 4. Get the backtrack:
> Program received signal SIGSEGV, Segmentation fault.
> <snip>
> (gdb) bt
> #0 ...
> #1 ...
> <snip>
> #N ...
>
> Please email me everything that is printed by the "bt" command. You'll
> probably have to copy and paste several screens of backtrace. Thanks!
>
> Ken
>
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VGhhbmtzIGZvciByZXNwb25kaW5nIHRvIG15IHBsZWEsIEtlbiEgSSd2ZSBhbHJlYWR5IGxvb2tl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>>>>> "Alexander" == Alexander Mont <ale...@co...> writes:
 Alexander> line 100, in add_base_flage incdirs = [os.path.join(p,
 Alexander> 'include') for p in basedir[sys, platform] KeyError:
 Alexander> 'cygwin'
This is almost certainly a typo. I guess you mean 'sys.platform'
 Alexander> How do I fix this? I installed numarray and I changed the NUMERIX value in the matplotlibrc file from Numeric to numarray but I still have the same problem.
This has nothing to do with the numerix setting. Check out
setupext.py and add an entry to the basedir dictionary to reflect your
local path information. If you get something that works, please post
it so we can include it in the standard distro.
Is there a reason you don't want to use the pre-built win32 binaries?
Most windows users are reasonably happy with that.
JDH

Ryan,
I'm afraid there isn't a Debian package for Python2.4+wxPython2.6, so I can't 
test that combination. Our best bet is to go for the GDB stack trace, like so:
	1. Write an iPython script that reproduces the segfault:
		$ ipython <test-script>
		<snip>
		Segmentation fault
	2. Find out where Python lives:
		$ which python
		/usr/bin/python
	2. Find out where iPython lives:
		$ which ipython
		/usr/bin/ipython
	3. Launch gdb:
		$ gdb <path-to-python>
		<snip>
		(gdb) run <path-to-ipython> <test-script>
	4. Get the backtrack:
		Program received signal SIGSEGV, Segmentation fault.
		<snip>
		(gdb) bt
		#0 ...
		#1 ...
		<snip>
		#N ...
Please email me everything that is printed by the "bt" command. You'll 
probably have to copy and paste several screens of backtrace. Thanks!
Ken

>>>>> "James" == James Boyle <bo...@ll...> writes:
 James> I am trying to build matplotlib 0.84 on my Linux box:
 James> python 2.2, Redhat 3.2.3-53, gcc version 3.2.3 20030502
 James> (Red Hat Linux 3.2.3-53)
 James> The error I got is listed below - I am at a loss as to what
 James> is up. I have built 0.84 with no problem on my OS X
 James> machine. The point in the code indicated by the error is
 James> just after a #if / else checking whether Numeric of
 James> Numarray is present. I have only Numeric on this machine
 James> and the error occurs just after the #else indicating
 James> Numeric is present. My OS X box has numarray, so I guess
 James> this branch of code is not hit when I compile there. I did
 James> comment out the py-modules line as suggested in the
 James> setup.py since I am using python 2.2.
My guess is that this is because cntr.c is using a python2.3 only
macro, and we haven't sufficiently tested with python2.2. I've been
meaning to drop python2.2 support, but haven't had the time <wink>
PyMODINIT_FUNC
Try replacing all occurrences of that with 
extern "C"
DL_EXPORT(void)
and let us know. Just a guess....
JDH

Smith, Frank wrote:
> Greetings:
> 
> Is it possible to plot onto an existing application window?
I'm not entirely sure what you mean by this.
> In
> particular, I'm developing a simple application under PythonCard
> whose main application window that already has a component that
> inherits from wxFloatCanvas
Do you want to plot with MPL on top of the FloatCanvas? If so, that will 
be tricky. However, if you want to create another Panel, and plot to 
that, you want to look at the examples and docs for "embedding in wx" 
and similar titles.
By the way, I'm the author of FloatCanvas, so feel free to send me any 
questions, and let me know if you want the latest and greatest version.
-Chris
> I've looked at the WX, WXAgg backends but at first glance they
> are a little unyielding!
yup. If all goes well, you should be able to just use those without 
having to understand them!
-Chris
-- 
Christopher Barker, Ph.D.
Oceanographer
 		
NOAA/OR&R/HAZMAT (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
Chr...@no...

Smith, Frank wrote:
> Is it possible to plot onto an existing application window? In particular,
> I'm developing a simple application under PythonCard whose main application
> window that already has a component that inherits from wxFloatCanvas (which in
> turn inherits, appropriately, from wxPanel). In broad strokes can anyone
> provide some hints, direction, main steps, ..... I feel a little clueless to
> tell the truth. I've looked at the WX, WXAgg backends but at first glance they
> are a little unyielding! I would appreciate any help .... anything at all!
There's a page on this topic in the matplotlib cookbook which ought to help 
you get started:
	http://www.scipy.org/wikis/topical_software/EmbeddingInWx
Ken
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