matplotlib-users Mailing List for matplotlib (Page 7)

The example assumes that the file is a dump of uin16 512x512 array.
So, no doubt that it won't work with a png file.
See
http://matplotlib.sourceforge.net/users/image_tutorial.html
-JJ
On Tue, Mar 23, 2010 at 9:41 AM, yogesh karpate <yog...@gm...> wrote:
> The 1st code snippet of image_demo2.py is as follows
>
> #!/usr/bin/env python
> from pylab import *
>
> w, h = 512, 512
> s = file('../data/ct.raw', 'rb').read()
> A = fromstring(s, uint16).astype(float)
>
> A *= 1.0/max(A)
> A.shape = w, h
>
> I replaced s = file('../data/ct.raw', 'rb').read() with s =
> file('/home/jaguar/Developemnt/image_demo2.png', 'rb').read()
>
> The image is same as in given demo example(its given in .png format) and
> rest of code is also same.....
>
> when I run it it gives the error as "
> A = fromstring(s, uint16).astype(float)
> ValueError: string size must be a multiple of element size"
>
> WHat can be the error in program? I may asking stupid question Thanks in
> advance !
> Regards
> Yogesh
>
>
>
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

How was the name Axes chosen for the Axes component? :)
It did confuse me for at least two days while I was first learning 
mpl. It's in my thoughts again as I'm writing some wrapper classes for 
it; what were the alternatives considered, even after the fact? :)

You need to define your own path (or you may combine a wedge and rectangles).
This may be helpful.
http://matplotlib.sourceforge.net/users/path_tutorial.html
Regards,
-JJ
On Sat, Mar 20, 2010 at 2:33 PM, Gary Jaffe <gf...@gm...> wrote:
> Hi all --
>
> I'm new to Matplotlib, and it looks like a great project, but I'm having
> trouble figuring out one thing.
>
> I would like to draw a pie chart, but have each wedge appear to have
> thickness. I see from the examples and the docs that it is possible to draw
> a wedge in relief so that it appears to cast a shadow, but I don't see how
> to make it look something like this. Pardon my bad ascii art. :)
>       ___________________
>       | \        /
>       \ \       /
>        \ \      /
>        \ \     /
>         \ \    /
>         \ \   /
>          \ \  /
>          \ \ /
>           \ \/
>           \|
>
> Is there a way to do this?
>
> Thanks,
> Gary
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

In the current implementation, sharing the axis does not mean sharing its scale.
This is not a subplots-specific issue, but applies to all kind of axes sharing.
So you need to change the scale of all the axes even though they have
shared axis.
What seems to be a better approach to me is to initialize "subplots"
with proper scale.
f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, sharey=True,
 subplot_kw=dict(yscale="log"))
Regards,
-JJ
On Mon, Mar 22, 2010 at 1:18 PM, Gökhan Sever <gok...@gm...> wrote:
> Hello,
>
> I am testing the newly added subplots function in ipython -pylab with the
> following code:
>
> f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, sharey=True)
> ax1.plot(np.random.random(20))
> ax2.plot(np.random.random(20))
> ax3.plot(np.random.random(20))
> ax4.plot(np.random.random(20))
>
> For some reason scaling the y-axes logaritmically works only on the focused
> figure canvas, the rest of the subplots are scaled in a distorted fashion.
> Axes labels change to proper notation but the scaling stays as if linear
> along with the data. See for better description:
> http://img408.imageshack.us/img408/7149/logscale.png
>
> Any ideas?
>
> --
> Gökhan
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

Hello,
I am having some issues generating pie charts, when some of the slices
become very small, their labels
will draw on top of each other, making it impossible to distinguish between
them. And I am trying to avoid using a legend.
Does anyone know if there is a way to properly position labels of pie charts
to avoid overlapping.
(By for example distributing them vertically with lines pointing to their
respective slices)
Similar to what is done here:
http://chart.apis.google.com/chart?cht=p&chd=s:Uf9a&chs=250x100&chl=January|February|March|April
--
Regards
Rune V. Sjoen

The 1st code snippet of image_demo2.py is as follows
#!/usr/bin/env python
from pylab import *
w, h = 512, 512
s = file('../data/ct.raw', 'rb').read()
A = fromstring(s, uint16).astype(float)
A *= 1.0/max(A)
A.shape = w, h
I replaced s = file('../data/ct.raw', 'rb').read() with s =
file('/home/jaguar/Developemnt/image_demo2.png', 'rb').read()
The image is same as in given demo example(its given in .png format)
and rest of code is also same.....
when I run it it gives the error as *" *
*A = fromstring(s, uint16).astype(float)*
*ValueError: string size must be a multiple of element size"*
WHat can be the error in program? I may asking stupid question Thanks
in advance !
Regards
Yogesh

I managed to resolve it, by following Mike's hint thanks Mike.
It was simply to change the setup.py script to
	import cx_Freeze
	import sys
	import matplotlib
	base = None
	if sys.platform == "win32":
	base = "Win32GUI"
	executables = [
		cx_Freeze.Executable("script.py", base = base)
	]
	cx_Freeze.setup(
		name = "script",
 options = {"build_exe": {"include_files":[( matplotlib.get_data_path(),"mpl-data")],}},
		version = "0.1",
		description = "Sample matplotlib script",
		executables = executables)
Peter
On Friday 19 March 2010, Michael Droettboom wrote:
> I have no experience with cx_freeze, but the page on packaging
> matplotlib with py2exe may be relevant. You do need to find a way to
> convince cx_freeze to include the data files and then a way for
> matplotlib to find them at run time.
> 
> Mike
> 
> Peter Bloomfield wrote:
> > Hi,
> >
> > I want to build executables from python scripts that call matplotlib
> > under linux. To this end I have installed cxfreeze on my SuSE 11.2
> > machine
> >
> > I have tried two methods
> > 	1. Execute the command 'cxfreeze script.py'
> > and
> > 	2. Creating a setup.py script
> > 		import cx_Freeze
> > 		import sys
> > 		base = None
> > 		if sys.platform == "win32":
> > 		 base = "Win32GUI"
> >
> > 		executables = [
> > 		 cx_Freeze.Executable("script.py", base = base)
> > 		]
> > 		cx_Freeze.setup(
> > 		 name = "script",
> > 		 version = "0.1",
> > 		 description = "Sample matplotlib script",
> > 		 executables = executables)
> > 	and then execute 'python setup.py build'
> >
> > In both cases I get an executable, but when executed I get the following
> > error RuntimeError: Could not find the matplotlib data files
> >
> > The version of matplotlib I am running is 0.99.1.1 and Python 2.6.2
> >
> > Does anyone have any thoughts/suggestions to resolve this, thanks
> >
> >
> > Peter

Thank you!
I just thought there must be a solution.
I saw someone posted he can modified contour.py to fix this, but I failed.
:(
Marius 't Hart-3 wrote:
> 
> Actually, it does not draw the polygon edges, but leaves small gaps 
> between them. Through those gaps you can see the background. (This also 
> happens with polar plots and other polygons by the way.) I consider this 
> a bug, though there are ways around it. For contour plots one can plot 
> two contourplots over each other, one with different levels (or a 
> different number of levels) than the other. Messy, but it works.
> 
> Marius.
> 
> lmkli wrote:
>> When I use contourf to plot a filled contour map, I get some white-like
>> lines
>> between each two color patches, or you can call them gaps.
>> This is not like the contourf doc string says: "it does not draw the
>> polygon edges.", actually, it does.
>> http://old.nabble.com/file/p27982822/contourf1.png contourf1.png 
>> The figure above is what I get by contourf, but this is not exactlly what
>> I
>> want.
>>
>> The filled contour map I want should be like this one:
>> http://old.nabble.com/file/p27982822/contourf2.png contourf2.png 
>>
>> Any advice?
>>
>> Thank you!
>> 
> 
> 
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> 
-- 
View this message in context: http://old.nabble.com/contourf-creats-white-like-lines-%28or-gaps%29-between-each-two-color-patches-tp27982822p27996271.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

> I had this problem when I installed python from source code and did 
> not use the --enable-framework flag at the ./configure stage 
> (learned I should have it from some document - but can't remember if 
> it was in matplotlib doc, on the internet, in a python doc etc).
> I am using the enthought distribution, I would hope that it was done 
> right but maybe not.
I am not sure, I am unfamiliar with the enthought distribution.
As a possible temporary "fix", when I had this problem, I was able to 
copy and paste the name of a file into the OSX save box from another 
window/terminal etc, and so could still make new names for the files.
Andre
>
> 		
> Vincent Davis
> 720-301-3003
> vi...@vi...
>
> my blog | LinkedIn
>
>
> On Mon, Mar 22, 2010 at 9:37 PM, Andre Walker-Loud <wal...@gm... 
> > wrote:
> Hi Vincent,
>
> I had this problem when I installed python from source code and did 
> not use the --enable-framework flag at the ./configure stage 
> (learned I should have it from some document - but can't remember if 
> it was in matplotlib doc, on the internet, in a python doc etc).
>
> I then went back and recompiled python with this flag (at 
> configure), but still had some warnings.
>
> So then I built python from the a dmg file from the python website, 
> and everything works great (I don't think I even needed to rebuild 
> matplot lib).
>
> The enable-framework option is what you need for proper GUI backend 
> support (here I don't really know what I am talking about, but I 
> think I remember the key jargon) and the OSX save box is part of 
> this GUI interface.
>
>
> Maybe this is what is causing your problem also.
>
>
> Andre
>
>
>
>
>
>
>
>
>
> On Mar 22, 2010, at 11:10 PM, Vincent Davis wrote:
>
>> I am not able to name plots when saving them from the plot menu in 
>> OSX. When I click the save menu item I get the nice OSX save dialog 
>> box but I am not able to type in the text field. I can select an 
>> existing file which then fills the text field with that name but I 
>> am still not able to change the name. At this point I can save the 
>> file and it does not overwrite the file as long as it is not a png 
>> which is the file type the plot gets saved as.
>> And the magic fix is ?????? :)
>> 		
>> Vincent Davis
>> 720-301-3003
>> vi...@vi...
>>
>> my blog | LinkedIn
>>
>> ------------------------------------------------------------------------------
>> Download Intel&#174; Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev_______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev_______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

>
> I had this problem when I installed python from source code and did not use
> the --enable-framework flag at the ./configure stage (learned I should have
> it from some document - but can't remember if it was in matplotlib doc, on
> the internet, in a python doc etc).
I am using the enthought distribution, I would hope that it was done right
but maybe not.
 *Vincent Davis
720-301-3003 *
vi...@vi...
 my blog <http://vincentdavis.net> |
LinkedIn<http://www.linkedin.com/in/vincentdavis>
On Mon, Mar 22, 2010 at 9:37 PM, Andre Walker-Loud <wal...@gm...>wrote:
> Hi Vincent,
>
> I had this problem when I installed python from source code and did not use
> the --enable-framework flag at the ./configure stage (learned I should have
> it from some document - but can't remember if it was in matplotlib doc, on
> the internet, in a python doc etc).
>
> I then went back and recompiled python with this flag (at configure), but
> still had some warnings.
>
> So then I built python from the a dmg file from the python website, and
> everything works great (I don't think I even needed to rebuild matplot lib).
>
> The enable-framework option is what you need for proper GUI backend support
> (here I don't really know what I am talking about, but I think I remember
> the key jargon) and the OSX save box is part of this GUI interface.
>
>
> Maybe this is what is causing your problem also.
>
>
> Andre
>
>
>
>
>
>
>
>
>
> On Mar 22, 2010, at 11:10 PM, Vincent Davis wrote:
>
> I am not able to name plots when saving them from the plot menu in OSX.
> When I click the save menu item I get the nice OSX save dialog box but I am
> not able to type in the text field. I can select an existing file which then
> fills the text field with that name but I am still not able to change the
> name. At this point I can save the file and it does not overwrite the file
> as long as it is not a png which is the file type the plot gets saved as.
> And the magic fix is ?????? :)
>
> *Vincent Davis
> 720-301-3003 *
> vi...@vi...
> my blog <http://vincentdavis.net> | LinkedIn<http://www.linkedin.com/in/vincentdavis>
>
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
>
> http://p.sf.net/sfu/intel-sw-dev_______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
>

Hi Vincent,
I had this problem when I installed python from source code and did 
not use the --enable-framework flag at the ./configure stage (learned 
I should have it from some document - but can't remember if it was in 
matplotlib doc, on the internet, in a python doc etc).
I then went back and recompiled python with this flag (at configure), 
but still had some warnings.
So then I built python from the a dmg file from the python website, 
and everything works great (I don't think I even needed to rebuild 
matplot lib).
The enable-framework option is what you need for proper GUI backend 
support (here I don't really know what I am talking about, but I think 
I remember the key jargon) and the OSX save box is part of this GUI 
interface.
Maybe this is what is causing your problem also.
Andre
On Mar 22, 2010, at 11:10 PM, Vincent Davis wrote:
> I am not able to name plots when saving them from the plot menu in 
> OSX. When I click the save menu item I get the nice OSX save dialog 
> box but I am not able to type in the text field. I can select an 
> existing file which then fills the text field with that name but I 
> am still not able to change the name. At this point I can save the 
> file and it does not overwrite the file as long as it is not a png 
> which is the file type the plot gets saved as.
> And the magic fix is ?????? :)
> 		
> Vincent Davis
> 720-301-3003
> vi...@vi...
>
> my blog | LinkedIn
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev_______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

I am not able to name plots when saving them from the plot menu in OSX. When
I click the save menu item I get the nice OSX save dialog box but I am not
able to type in the text field. I can select an existing file which then
fills the text field with that name but I am still not able to change the
name. At this point I can save the file and it does not overwrite the file
as long as it is not a png which is the file type the plot gets saved as.
And the magic fix is ?????? :)
*Vincent Davis
720-301-3003 *
vi...@vi...
 my blog <http://vincentdavis.net> |
LinkedIn<http://www.linkedin.com/in/vincentdavis>

I identify the problem.
There is no basemap data directory and the app search those datas in
[MY_DIR]\dist\library.zip\mpl_toolkits\basemap\data
But I don't know how to add it at the compilation....
2010年3月22日 Friedrich Romstedt <fri...@gm...>:
> I'm not shure whether the following suggestion solves your problem,
> but it would simplify your script anyway.
>
> import matplotlib
> ...
> setup(..., data_files = matplotlib.get_py2exe_datafiles())
>
> And maybe don't forget to exclude 'libgdk_pixbuf-2.0-0.dll' (on my
> system) in 'dll_excludes'. But I actually don't remember for what
> reason I had to exclude it.
>
> hth,
> Friedrich
>

Mathes, Gary C wrote:
> Installation question:
> 
> 
> 
> The *import pylab* command reports an error. See bold below for 
> details. What installation step(s) am I missing?
it looks like you've got GTK set as your default back-end, but are not 
running and X-server on this box.
If you want to use MPL without a GUI, you need to set your default 
backend to "AGG", but you probably shouldn't use pylab for that anyway, 
but rather use the OO API -- some googling will tell you about that.
-Chris
> 
> 
> Thanks!
> 
> 
> 
> *# uname -a*
> 
> Linux SESB01-RE01 2.6.18-92.el5 #1 SMP Tue Jun 10 18:51:06 EDT 2008 
> x86_64 x86_64 x86_64 GNU/Linux
> 
> 
> 
> #!/usr/bin/python
> 
> 
> 
> # File: *mp_versions.py*
> 
> 
> 
> import matplotlib
> 
> import numpy
> 
> import matplotlib as mpl
> 
> 
> 
> print "matplotlib version = " + matplotlib.__version__
> 
> print "matplotlib file = " + matplotlib.__file__
> 
> print "numpy version = " + numpy.__version__
> 
> print "mpl.get_configdir = " + mpl.get_configdir()
> 
> 
> 
> *import pylab*
> 
> 
> 
> *# mp_versions.py*
> 
> matplotlib version = 0.99.1.1
> 
> matplotlib file = /usr/lib64/python2.4/site-packages/matplotlib/__init__.pyc
> 
> numpy version = 1.2.1
> 
> mpl.get_configdir = /root/.matplotlib
> 
> *Traceback (most recent call last):*
> 
> * File "./mp_versions.py", line 12, in ?*
> 
> * import pylab*
> 
> * File "/usr/lib64/python2.4/site-packages/pylab.py", line 1, in ?*
> 
> * from matplotlib.pylab import **
> 
> * File "/usr/lib64/python2.4/site-packages/matplotlib/pylab.py", line 
> 247, in ?*
> 
> * from matplotlib.pyplot import **
> 
> * File "/usr/lib64/python2.4/site-packages/matplotlib/pyplot.py", line 
> 78, in ?*
> 
> * new_figure_manager, draw_if_interactive, show = pylab_setup()*
> 
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/__init__.py", 
> line 25, in pylab_setup*
> 
> * globals(),locals(),[backend_name])*
> 
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtkagg.py", 
> line 10, in ?*
> 
> * from matplotlib.backends.backend_gtk import gtk, FigureManagerGTK, 
> FigureCanvasGTK,\*
> 
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtk.py", 
> line 8, in ?*
> 
> * import gtk; gdk = gtk.gdk*
> 
> * File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", 
> line 76, in ?*
> 
> * _init()*
> 
> * File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", 
> line 64, in _init*
> 
> * _gtk.init_check()*
> 
> *RuntimeError: could not open display*
> 
> 
> 
> # *ipython*
> 
> Python 2.4.3 (#1, May 24 2008, 13:57:05)
> 
> Type "copyright", "credits" or "license" for more information.
> 
> 
> 
> IPython 0.8.4 -- An enhanced Interactive Python.
> 
> ? -> Introduction and overview of IPython's features.
> 
> %quickref -> Quick reference.
> 
> help -> Python's own help system.
> 
> object? -> Details about 'object'. ?object also works, ?? prints more.
> 
> 
> 
> *Installation steps:*
> 
> # rpm -ivh refblas3-3.0-11.2.x86_64.rpm lapack3-3.0-19.2.x86_64.rpm
> 
> # rpm –ivh python-dateutil-1.2-1.el5.noarch.rpm pytz-2006p-1.el5.noarch.rpm
> 
> # rpm -ivh python-setuptools-0.6c5-2.el5.noarch.rpm
> 
> # rpm -ivh python-nose-0.10.4-1.el5.noarch.rpm
> 
> # rpm -ivh atlas-3.8.3-1.el5.x86_64.rpm
> 
> # rpm -ivh numpy-1.2.1-2.el5.x86_64.rpm
> 
> # rpm -ivh python-matplotlib-0.99.1.2-1.el5.x86_64.rpm
> 
> # rpm -ivh ipython-0.8.4-1.el5.noarch.rpm
> 
> 
> 
> *Location of RPMs:*
> 
> http://download.opensuse.org/repositories/home:/ashigabou/CentOS_5/x86_64/
> 
> http://www.rpmfind.net/linux/rpm2html/search.php?query=python-dateutil
> 
> http://www.rpmfind.net/linux/RPM/epel/5/ppc/pytz-2006p-1.el5.noarch.html
> 
> http://fr2.rpmfind.net/linux/RPM/epel/testing/5/ppc/python-nose-0.10.4-1.el5.noarch.html
> 
> http://rpm.pbone.net/index.php3/stat/4/idpl/12391980/com/atlas-3.8.3-1.el5.x86_64.rpm.html
> 
> http://www.rpmfind.net/linux/rpm2html/search.php?query=numpy
> 
> http://www.rpmfind.net/linux/rpm2html/search.php?query=python-matplotlib
> 
> http://rpm.pbone.net/index.php3/stat/4/idpl/9548551/com/ipython-0.8.4-1.el5.noarch.rpm.html
> 
> 
> 
> 
> 
> **/Gary/**
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
-- 
Christopher Barker, Ph.D.
Oceanographer
Emergency Response Division
NOAA/NOS/OR&R (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
Chr...@no...

It looks like you're using the gtk backend but you're running on a 
remote terminal without an X11 connection. That's what "could not open 
display" probably means. Are you logged in remotely by any chance? Does 
'python -c "import gtk" ' work?
If you want to run headless, you can use the "Agg" backend. See here:
http://matplotlib.sourceforge.net/faq/installing_faq.html#what-is-a-backend
Mike
Mathes, Gary C wrote:
>
> Installation question:
>
> The *import pylab* command reports an error. See bold below for 
> details. What installation step(s) am I missing?
>
> Thanks!
>
> *# uname -a*
>
> Linux SESB01-RE01 2.6.18-92.el5 #1 SMP Tue Jun 10 18:51:06 EDT 2008 
> x86_64 x86_64 x86_64 GNU/Linux
>
> #!/usr/bin/python
>
> # File: *mp_versions.py*
>
> import matplotlib
>
> import numpy
>
> import matplotlib as mpl
>
> print "matplotlib version = " + matplotlib.__version__
>
> print "matplotlib file = " + matplotlib.__file__
>
> print "numpy version = " + numpy.__version__
>
> print "mpl.get_configdir = " + mpl.get_configdir()
>
> *import pylab*
>
> *# mp_versions.py*
>
> matplotlib version = 0.99.1.1
>
> matplotlib file = 
> /usr/lib64/python2.4/site-packages/matplotlib/__init__.pyc
>
> numpy version = 1.2.1
>
> mpl.get_configdir = /root/.matplotlib
>
> *Traceback (most recent call last):*
>
> * File "./mp_versions.py", line 12, in ?*
>
> * import pylab*
>
> * File "/usr/lib64/python2.4/site-packages/pylab.py", line 1, in ?*
>
> * from matplotlib.pylab import **
>
> * File "/usr/lib64/python2.4/site-packages/matplotlib/pylab.py", line 
> 247, in ?*
>
> * from matplotlib.pyplot import **
>
> * File "/usr/lib64/python2.4/site-packages/matplotlib/pyplot.py", line 
> 78, in ?*
>
> * new_figure_manager, draw_if_interactive, show = pylab_setup()*
>
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/__init__.py", 
> line 25, in pylab_setup*
>
> * globals(),locals(),[backend_name])*
>
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtkagg.py", 
> line 10, in ?*
>
> * from matplotlib.backends.backend_gtk import gtk, FigureManagerGTK, 
> FigureCanvasGTK,\*
>
> * File 
> "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtk.py", 
> line 8, in ?*
>
> * import gtk; gdk = gtk.gdk*
>
> * File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", 
> line 76, in ?*
>
> * _init()*
>
> * File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", 
> line 64, in _init*
>
> * _gtk.init_check()*
>
> *RuntimeError: could not open display*
>
> # *ipython*
>
> Python 2.4.3 (#1, May 24 2008, 13:57:05)
>
> Type "copyright", "credits" or "license" for more information.
>
> IPython 0.8.4 -- An enhanced Interactive Python.
>
> ? -> Introduction and overview of IPython's features.
>
> %quickref -> Quick reference.
>
> help -> Python's own help system.
>
> object? -> Details about 'object'. ?object also works, ?? prints more.
>
> *Installation steps:*
>
> # rpm -ivh refblas3-3.0-11.2.x86_64.rpm lapack3-3.0-19.2.x86_64.rpm
>
> # rpm –ivh python-dateutil-1.2-1.el5.noarch.rpm 
> pytz-2006p-1.el5.noarch.rpm
>
> # rpm -ivh python-setuptools-0.6c5-2.el5.noarch.rpm
>
> # rpm -ivh python-nose-0.10.4-1.el5.noarch.rpm
>
> # rpm -ivh atlas-3.8.3-1.el5.x86_64.rpm
>
> # rpm -ivh numpy-1.2.1-2.el5.x86_64.rpm
>
> # rpm -ivh python-matplotlib-0.99.1.2-1.el5.x86_64.rpm
>
> # rpm -ivh ipython-0.8.4-1.el5.noarch.rpm
>
> *Location of RPMs:*
>
> http://download.opensuse.org/repositories/home:/ashigabou/CentOS_5/x86_64/
>
> http://www.rpmfind.net/linux/rpm2html/search.php?query=python-dateutil
>
> http://www.rpmfind.net/linux/RPM/epel/5/ppc/pytz-2006p-1.el5.noarch.html
>
> http://fr2.rpmfind.net/linux/RPM/epel/testing/5/ppc/python-nose-0.10.4-1.el5.noarch.html
>
> http://rpm.pbone.net/index.php3/stat/4/idpl/12391980/com/atlas-3.8.3-1.el5.x86_64.rpm.html
>
> http://www.rpmfind.net/linux/rpm2html/search.php?query=numpy
>
> http://www.rpmfind.net/linux/rpm2html/search.php?query=python-matplotlib
>
> http://rpm.pbone.net/index.php3/stat/4/idpl/9548551/com/ipython-0.8.4-1.el5.noarch.rpm.html
>
> **/Gary/**
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> ------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

Installation question:
The import pylab command reports an error. See bold below for details. What installation step(s) am I missing?
Thanks!
# uname -a
Linux SESB01-RE01 2.6.18-92.el5 #1 SMP Tue Jun 10 18:51:06 EDT 2008 x86_64 x86_64 x86_64 GNU/Linux
#!/usr/bin/python
# File: mp_versions.py
import matplotlib
import numpy
import matplotlib as mpl
print "matplotlib version = " + matplotlib.__version__
print "matplotlib file = " + matplotlib.__file__
print "numpy version = " + numpy.__version__
print "mpl.get_configdir = " + mpl.get_configdir()
import pylab
# mp_versions.py
matplotlib version = 0.99.1.1
matplotlib file = /usr/lib64/python2.4/site-packages/matplotlib/__init__.pyc
numpy version = 1.2.1
mpl.get_configdir = /root/.matplotlib
Traceback (most recent call last):
 File "./mp_versions.py", line 12, in ?
 import pylab
 File "/usr/lib64/python2.4/site-packages/pylab.py", line 1, in ?
 from matplotlib.pylab import *
 File "/usr/lib64/python2.4/site-packages/matplotlib/pylab.py", line 247, in ?
 from matplotlib.pyplot import *
 File "/usr/lib64/python2.4/site-packages/matplotlib/pyplot.py", line 78, in ?
 new_figure_manager, draw_if_interactive, show = pylab_setup()
 File "/usr/lib64/python2.4/site-packages/matplotlib/backends/__init__.py", line 25, in pylab_setup
 globals(),locals(),[backend_name])
 File "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtkagg.py", line 10, in ?
 from matplotlib.backends.backend_gtk import gtk, FigureManagerGTK, FigureCanvasGTK,\
 File "/usr/lib64/python2.4/site-packages/matplotlib/backends/backend_gtk.py", line 8, in ?
 import gtk; gdk = gtk.gdk
 File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", line 76, in ?
 _init()
 File "/usr/lib64/python2.4/site-packages/gtk-2.0/gtk/__init__.py", line 64, in _init
 _gtk.init_check()
RuntimeError: could not open display
# ipython
Python 2.4.3 (#1, May 24 2008, 13:57:05)
Type "copyright", "credits" or "license" for more information.
IPython 0.8.4 -- An enhanced Interactive Python.
? -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help -> Python's own help system.
object? -> Details about 'object'. ?object also works, ?? prints more.
Installation steps:
# rpm -ivh refblas3-3.0-11.2.x86_64.rpm lapack3-3.0-19.2.x86_64.rpm
# rpm -ivh python-dateutil-1.2-1.el5.noarch.rpm pytz-2006p-1.el5.noarch.rpm
# rpm -ivh python-setuptools-0.6c5-2.el5.noarch.rpm
# rpm -ivh python-nose-0.10.4-1.el5.noarch.rpm
# rpm -ivh atlas-3.8.3-1.el5.x86_64.rpm
# rpm -ivh numpy-1.2.1-2.el5.x86_64.rpm
# rpm -ivh python-matplotlib-0.99.1.2-1.el5.x86_64.rpm
# rpm -ivh ipython-0.8.4-1.el5.noarch.rpm
Location of RPMs:
http://download.opensuse.org/repositories/home:/ashigabou/CentOS_5/x86_64/
http://www.rpmfind.net/linux/rpm2html/search.php?query=python-dateutil
http://www.rpmfind.net/linux/RPM/epel/5/ppc/pytz-2006p-1.el5.noarch.html
http://fr2.rpmfind.net/linux/RPM/epel/testing/5/ppc/python-nose-0.10.4-1.el5.noarch.html
http://rpm.pbone.net/index.php3/stat/4/idpl/12391980/com/atlas-3.8.3-1.el5.x86_64.rpm.html
http://www.rpmfind.net/linux/rpm2html/search.php?query=numpy
http://www.rpmfind.net/linux/rpm2html/search.php?query=python-matplotlib
http://rpm.pbone.net/index.php3/stat/4/idpl/9548551/com/ipython-0.8.4-1.el5.noarch.rpm.html
Gary

I'm not shure whether the following suggestion solves your problem,
but it would simplify your script anyway.
import matplotlib
...
setup(..., data_files = matplotlib.get_py2exe_datafiles())
And maybe don't forget to exclude 'libgdk_pixbuf-2.0-0.dll' (on my
system) in 'dll_excludes'. But I actually don't remember for what
reason I had to exclude it.
hth,
Friedrich

On Sun, Mar 21, 2010 at 05:52:55PM -0500, Ryan May wrote:
> On Sun, Mar 21, 2010 at 3:35 PM, David Carmean <dl...@ha...> wrote:
 [snip]
> > One of the things I'm trying to figure out is whether I can build
> > a graphic in "reverse order". The standard M.O. seems to be to
 [snip]
> Certainly things like lines, collections can be added to existing axes
> (and don't need one around to be created). This is actually what's
> used under the hood. You *should* also be able to create an axes
> object and then set its figure, but I've never personally done it.
> 
> What you really want to look at are a lot of Axes methods:
> set_figure()
> add_artists()
> add_collection()
> add_line()
I did spend a couple of hours looking at the code, and there are a few 
places where the child components do depend on data/methods in the parent 
container, even though the "child" objects can be created without the 
parent; I'm about 80% sure that there are no guards against calling those 
child methods without a parent, so it seems safer to rework my model to 
use the top-down approach.
Perhaps I'll find some time soon to learn enough to create some unit tests that 
confirm/refute the above, and if true I'll file some bugs. 

Hello World,
I'm using py2exe to create an executable on windows .
There is a setup.py file that allows me to create this executable under eclipse.
I have a problem with basemap.
Probably, the app needs data basemap directory but I don't know where to put it.
Thanks for your help.
Yagua
The contents of the log file is as follows:
Traceback (most recent call last)
 File "traitement_points.py", line 40, in <module>
 File "mpl_toolkits\basemap\__init__.pyc", line 774, in __init__
 File "mpl_toolkits\basemap\__init__.pyc", line 848, in _readboundarydata
IOError: Unable to open boundary dataset file. Only the 'crude', 'low',
'intermediate' and 'high' resolution datasets are installed by default.
If you are requesting a 'full' resolution dataset, you may need to
download and install those files separately
(see the basemap README for details).
The contents of the setup file is :
from distutils.core import setup
import py2exe
import glob
opts = {
 'py2exe': { "includes" : ["sip", "PyQt4", "matplotlib.backends",
"matplotlib.backends.backend_qt4agg",
 "matplotlib.figure","pylab", "numpy",
"matplotlib.numerix.fft",
 "matplotlib.numerix.linear_algebra",
"matplotlib.numerix.random_array",
 "matplotlib.backends.backend_tkagg"],
 'excludes': ['_gtkagg', '_tkagg', '_agg2', '_cairo',
'_cocoaagg',
 '_fltkagg', '_gtk', '_gtkcairo', ],
 'dll_excludes': ['libgdk-win32-2.0-0.dll',
 'libgobject-2.0-0.dll']
 }
 }
data_files = [(r'mpl-data',
glob.glob(r'C:\Python25\Lib\site-packages\matplotlib\mpl-data\*.*')),
 (r'mpl-data',
[r'C:\Python25\Lib\site-packages\matplotlib\mpl-data\matplotlibrc']),
(r'mpl-data\images',glob.glob(r'C:\Python25\Lib\site-packages\matplotlib\mpl-data\images\*.*')),
(r'mpl-data\fonts',glob.glob(r'C:\Python25\Lib\site-packages\matplotlib\mpl-data\fonts\*.*'))]
setup(windows=[{"script" : "traitement_points.py", "icon_resources":
[(1, "lambda.ico")]}],
 description="test py2exe.",
 version = "1.0",
 options=opts, data_files=data_files)

Hello,
I am testing the newly added subplots function in ipython -pylab with the
following code:
f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, sharey=True)
ax1.plot(np.random.random(20))
ax2.plot(np.random.random(20))
ax3.plot(np.random.random(20))
ax4.plot(np.random.random(20))
For some reason scaling the y-axes logaritmically works only on the focused
figure canvas, the rest of the subplots are scaled in a distorted fashion.
Axes labels change to proper notation but the scaling stays as if linear
along with the data. See for better description:
http://img408.imageshack.us/img408/7149/logscale.png
Any ideas?
-- 
Gökhan

Actually, it does not draw the polygon edges, but leaves small gaps 
between them. Through those gaps you can see the background. (This also 
happens with polar plots and other polygons by the way.) I consider this 
a bug, though there are ways around it. For contour plots one can plot 
two contourplots over each other, one with different levels (or a 
different number of levels) than the other. Messy, but it works.
Marius.
lmkli wrote:
> When I use contourf to plot a filled contour map, I get some white-like lines
> between each two color patches, or you can call them gaps.
> This is not like the contourf doc string says: "it does not draw the
> polygon edges.", actually, it does.
> http://old.nabble.com/file/p27982822/contourf1.png contourf1.png 
> The figure above is what I get by contourf, but this is not exactlly what I
> want.
>
> The filled contour map I want should be like this one:
> http://old.nabble.com/file/p27982822/contourf2.png contourf2.png 
>
> Any advice?
>
> Thank you!
> 

Well, to be clear.
How to get tick labels at the very end of the axes?
If the x-axis goes from is 0, 1 the axis ticks will be at 0.2,0.4,0.6 and
0.8 as you can here: see
http://matplotlib.sourceforge.net/examples/mplot3d/2dcollections3d_demo.html
How do I additionally get labels at 0 and 1?
Hope this triggers a few responses. :D Thanks in advance.
Martin
On Sun, Mar 21, 2010 at 12:43 PM, Martin Bothe
<mar...@go...>wrote:
> Hello everyone,
>
> got a question concerning the right tick format in mplot3d.
> Using the example from
> http://matplotlib.sourceforge.net/plot_directive/mpl_examples/mplot3d/lines3d_demo.py with
> python 2.6.4 and mpl Version: 0.99.0-1ubuntu1, I get a nice 3dplot, but the
> ticks are not reaching the corners of the z-axis.
> But if I ask him for the ticklabels and locations, on the terminal he gives
> out what I would like to have: 9 labels from -2 to 2.
>
> In [3]: ax.w_zaxis.get_ticklabels
> ------> ax.w_zaxis.get_ticklabels()
> Out[3]: <a list of 9 Text major ticklabel objects>
>
> In [4]: ax.w_zaxis.get_ticklocs
> ------> ax.w_zaxis.get_ticklocs()
> Out[4]: array([-2. , -1.5, -1. , -0.5, 0. , 0.5, 1. , 1.5, 2. ])
>
> However, in the figure (attached as png) the labels are only reaching from
> -1.5 to 1.5, meaning 7 tick labels.
> How to fix this? What is wrong here?
> Thanks for your help.
>
> Greetings from Berlin
> Martin
>

When I use contourf to plot a filled contour map, I get some white-like lines
between each color patches.
This is not like the contourf doc string says: "it does not draw the
polygon edges.", actually, it does.
http://old.nabble.com/file/p27982822/contourf1.png contourf1.png 
The figure above is what I get by contourf, but this is not exactlly what I
want.
The filled contour map I want should be like this one:
http://old.nabble.com/file/p27982822/contourf2.png contourf2.png 
Any advices?
Thank you!
-- 
View this message in context: http://old.nabble.com/contourf-creats-white-like-lines-%28or-gaps%29-between-each-color-patches-tp27982822p27982822.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

On Sun, Mar 21, 2010 at 3:35 PM, David Carmean <dl...@ha...> wrote:
>
> Has anyone seen/created any UML models/diagrams for Matplotlib?
>
> I'm currently doing so to help me understand how to use it in an
> application, but wondered if somebody has already built the wheel
> that I'm trying to invent.
>
> One of the things I'm trying to figure out is whether I can build
> a graphic in "reverse order". The standard M.O. seems to be to
> traverse a self-building tree by invoking convenience methods on
> each new child, Figure-->Subplot-->plot_date, etc.
>
> What I'm looking to do is build from the bottom up, e.g. create a line
> or scatter plot instance *first*, then append that to the children of
> an existing Axes instance, which is in-turn appended to the childrend of
> a Figure instance.
Certainly things like lines, collections can be added to existing axes
(and don't need one around to be created). This is actually what's
used under the hood. You *should* also be able to create an axes
object and then set its figure, but I've never personally done it.
What you really want to look at are a lot of Axes methods:
set_figure()
add_artists()
add_collection()
add_line()
Ryan
-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma

Has anyone seen/created any UML models/diagrams for Matplotlib? 
I'm currently doing so to help me understand how to use it in an 
application, but wondered if somebody has already built the wheel 
that I'm trying to invent.
One of the things I'm trying to figure out is whether I can build 
a graphic in "reverse order". The standard M.O. seems to be to 
traverse a self-building tree by invoking convenience methods on 
each new child, Figure-->Subplot-->plot_date, etc.
What I'm looking to do is build from the bottom up, e.g. create a line 
or scatter plot instance *first*, then append that to the children of 
an existing Axes instance, which is in-turn appended to the childrend of 
a Figure instance. 
Thanks.