matplotlib-users Mailing List for matplotlib

Thanks Darren. The win32 binaries are updated now. TkAgg is the
default on all of them, and the OSX binaries already had TkAgg as the
default.
- Charlie
On Mon, Jun 2, 2008 at 6:13 PM, Darren Dale <dar...@co...> wrote:
> On Monday 02 June 2008 05:33:52 pm John Hunter wrote:
>> On Mon, Jun 2, 2008 at 4:32 PM, Charlie Moad <cw...@gm...> wrote:
>> > I am pretty sure setupext.py used to have these defaults coded in, but
>> > something must have changed. I haven't had to check default backends
>> > in a long time. I'll get new builds up asap.
>>
>> Darren has added a setup.cfg to configure the builds external to
>> setup.py. You'll probably want to use it to configure the default
>> backend.
>
> When I did all that stuff, I tried to make setup.py choose the backend for the
> default matplotlibrc intelligently. It defaults to agg, but then if a gui
> toolkit is available it will use that. BUT, maybe you want to tell matplotlib
> what the default backend should be, so you can set that in setup.cfg and THAT
> setting has precedence, see the template. That feature was added specifically
> for package managers, and I hoped it would make your life easier, Charlie.
> You should be able to edit setup.cfg, set the backend there, and it will be
> propagated to the default matplotlibrc.
>

KURT PETERS wrote:
> Oh, Yes, I must have got the "nines" mixed up. Can you talk me through 
> installing 0.99 without the binary installer in Win XP?
> And, you're right; I don't have a "working copy"; my old code keeps 
> crashing, especially when I moved the basemaps to mpl_toolkits.
>
> Kurt
> 
Kurt: I managed to break into my colleague's office and use his PC to 
make windows installers for basemap-0.99. They're up on the sourceforge 
site now, please give them a try and let me know how it goes.
-Jeff
> ----Original Message Follows----
> From: Jeff Whitaker <js...@fa...>
> To: KURT PETERS <pet...@ms...>
> CC: jd...@gm..., cw...@gm..., 
> mat...@li...
> Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy 
> producing	error
> Date: 2008年6月02日 12:23:36 -0600
>
> KURT PETERS wrote:
> 
>> Unfortunately, I think someone else was on the maillist with a similar 
>> problem, and perhaps our two mails got confused.
>> I have "always been" using 0.99 as far as I know.
>> 
>
> Kurt: I only released 0.99 yesterday. You were actually using 0.9.9, which 
> you installed from a binary installer. There is no binary installed for 
> 0.99 yet - which is the crux of your problem.
> 
>> It got "nuked" when I deleted the old matplotlib. I wonder why it 
>> couldn't find geos the second time around?
>>
>> 
> You installed 0.9.9 from a binary installer, so it didn't need to find 
> libgeos. It only looks for it when building from source.
> 
>> Once again, though, I'll let you know how things go. It's a good idea 
>> to pull the toolkit out of the matplotlib directory to avoid this very 
>> problem. I'll give that a try, by direct copying and changing my code. 
>> Probably a lot of example code might need changing as well.
>> Regards,
>> Kurt
>>
>>
>> 
> All the examples in 0.99 have been updated to use 'import 
> mpl_toolkits.basemap'. Until you actually have 0.99 installed, you don't 
> want to change your code.
>
> Hope this is clear - I'm getting the feeling that we're talking past each 
> other ...
>
> -Jeff
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

On Monday 02 June 2008 05:33:52 pm John Hunter wrote:
> On Mon, Jun 2, 2008 at 4:32 PM, Charlie Moad <cw...@gm...> wrote:
> > I am pretty sure setupext.py used to have these defaults coded in, but
> > something must have changed. I haven't had to check default backends
> > in a long time. I'll get new builds up asap.
>
> Darren has added a setup.cfg to configure the builds external to
> setup.py. You'll probably want to use it to configure the default
> backend.
When I did all that stuff, I tried to make setup.py choose the backend for the 
default matplotlibrc intelligently. It defaults to agg, but then if a gui 
toolkit is available it will use that. BUT, maybe you want to tell matplotlib 
what the default backend should be, so you can set that in setup.cfg and THAT 
setting has precedence, see the template. That feature was added specifically 
for package managers, and I hoped it would make your life easier, Charlie. 
You should be able to edit setup.cfg, set the backend there, and it will be 
propagated to the default matplotlibrc.

Oh, Yes, I must have got the "nines" mixed up. Can you talk me through 
installing 0.99 without the binary installer in Win XP?
And, you're right; I don't have a "working copy"; my old code keeps 
crashing, especially when I moved the basemaps to mpl_toolkits.
Kurt
----Original Message Follows----
From: Jeff Whitaker <js...@fa...>
To: KURT PETERS <pet...@ms...>
CC: jd...@gm..., cw...@gm..., 
mat...@li...
Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy 
producing	error
Date: 2008年6月02日 12:23:36 -0600
KURT PETERS wrote:
>Unfortunately, I think someone else was on the maillist with a similar 
>problem, and perhaps our two mails got confused.
> I have "always been" using 0.99 as far as I know.
Kurt: I only released 0.99 yesterday. You were actually using 0.9.9, which 
you installed from a binary installer. There is no binary installed for 
0.99 yet - which is the crux of your problem.
> It got "nuked" when I deleted the old matplotlib. I wonder why it 
>couldn't find geos the second time around?
>
You installed 0.9.9 from a binary installer, so it didn't need to find 
libgeos. It only looks for it when building from source.
> Once again, though, I'll let you know how things go. It's a good idea 
>to pull the toolkit out of the matplotlib directory to avoid this very 
>problem. I'll give that a try, by direct copying and changing my code. 
>Probably a lot of example code might need changing as well.
>Regards,
>Kurt
>
>
All the examples in 0.99 have been updated to use 'import 
mpl_toolkits.basemap'. Until you actually have 0.99 installed, you don't 
want to change your code.
Hope this is clear - I'm getting the feeling that we're talking past each 
other ...
-Jeff

On Mon, Jun 2, 2008 at 4:32 PM, Charlie Moad <cw...@gm...> wrote:
> I am pretty sure setupext.py used to have these defaults coded in, but
> something must have changed. I haven't had to check default backends
> in a long time. I'll get new builds up asap.
Darren has added a setup.cfg to configure the builds external to
setup.py. You'll probably want to use it to configure the default
backend.
JDH

I am pretty sure setupext.py used to have these defaults coded in, but
something must have changed. I haven't had to check default backends
in a long time. I'll get new builds up asap.
- Charlie
On Mon, Jun 2, 2008 at 12:01 PM, John Hunter <jd...@gm...> wrote:
> On Mon, Jun 2, 2008 at 10:41 AM, KURT PETERS <pet...@ms...> wrote:
>
>> File "C:\Python25\Lib\site-packages\matplotlib\backends\backend_gtk.py",
>> line 6, in <module>
>> import gobject
>> ImportError: No module named gobject
>
> This is a problem with our installer since the default backend we have
> set is GTKAgg (we usually set it to GTKAgg). Fortunately, it is
> relatively easy for you to fix:. Edit
> site-packages/matplotlib/mpl-data/matplotlib and change the 'backend :
> GTKAgg' line to
>
> backend : TkAgg
>
> Charlie -- we will need to roll out a new binary builds which fix this
> ASAP because this will bite all the win32 users who don't have a
> custom rc, which is most of them I suspect. We'll also need to check
> the settings in the 91.3 release. My apologies for not testing this
> as you requested (I did test on linux though). You probably will also
> want to check the default backend for the OS X eggs too.....
>
> JDH
>

Vincent Noel wrote:
> If you rename matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg to
> matplotlib-0.98.0-py2.5.egg,
> easy_install will install it from the disk.
It installed just fine for me with the long name....
>> easy_install ./matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
>>
>> Processing matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
>> removing '/Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5-
This is a bit odd -- a Framework Build of Python should be in 
/Library/Frameworks... (or System/Library/Frameworks). Maybe that egg 
doesn't work right with a non-framework build.
I think it should work with either the 2.5 Apple supplies with OS-X 
10.5, or the python.org 2.5 build.
-Chris
-- 
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...

Ah, that makes much more sense for the labels... what about the
commands for assigning weights and fonts to the default numbers along
an axis? I know I saw a post a while back where someone was able to
make the numbers boldface ... is that not rendered via TeX? And if so,
why do none of the options I changed seem to have any effect on those
numbers?
Thanks!
On Mon, Jun 2, 2008 at 5:53 AM, Michael Droettboom <md...@st...> wrote:
> First to correct some confusion (and a good candidate to add to the new docs
> that everyone is working so hard on...): mathtext and usetex are completely
> independent systems, and you can only use either one of the other.
> Therefore, if usetex is True, none of the mathtext settings will have any
> effect.
>
> Erik Tollerud wrote:
>>
>> I am having trouble figuring out how to control the weight of labels
>> and text with the usetex option set to True. I would like to force
>> all mathmode labels and text to be at least bold. I can imagine doing
>> this either by substituting a font (i.e. changing the mathtext.fontset
>> to 'custom' and all of the different matext.?? to some boldface font),
>> or some the font.weight rc parameter, but neither of these has any
>> effect - whatever I do, the mathmode and other TeX formatted text
>> stays the same weight. How can I make it just turn up the weights to
>> bold without changing anything else?
>>
>
> This is not something that is directly supported. When usetex is True, all
> of the rendering happens with (La)TeX, so you're limited by what (La)TeX can
> do. You would probably need to find some LaTeX package that does this, and
> then \include it in the LaTeX preamble using the "text.latex.preamble"
> setting. My quick Googling didn't reveal anything that can do that,
> however.
>
> You can get bold in a math expression by enclosing content in \mathbf{}, but
> that is by definition "upright" bold. The reason, I believe, is that (core)
> LaTeX simply doesn't include the fonts for Greek characters and other
> symbols in bold.
>
> Cheers,
> Mike
>
> --
> Michael Droettboom
> Science Software Branch
> Operations and Engineering Division
> Space Telescope Science Institute
> Operated by AURA for NASA
>
>

I can confirm this in at least version 0.91.3. The problem seems to be
caused by the midpoint of the stem being included in the path twice in a
row. The following patch removes the midpoint altogether and fixes the
rendering problem at least on my version of xpdf. I'm kind of swamped
right now, so I hope someone else can check whether this breaks any
other uses of arrows, or some other backend. (If this doesn't seem to
happen, please file a bug report in the tracker so it isn't forgotten.)
The patch is against the maintenance branch, but the trunk has
similar-looking code.
Index: lib/matplotlib/patches.py
===================================================================
--- lib/matplotlib/patches.py	(revision 5366)
+++ lib/matplotlib/patches.py	(working copy)
@@ -634,7 +634,13 @@
 if shape == 'right':
 coords = right_half_arrow
 elif shape == 'full':
- coords=npy.concatenate([left_half_arrow,right_half_arrow[::-1]])
+ # Concatenating the full paths caused the midpoint
+ # of the stem to be included twice, which was
+ # rendered badly by xpdf. Since the point is right
+ # between the corners of the stem, we can drop it
+ # from both halves.
+ coords=npy.concatenate([left_half_arrow[:-1],
+ right_half_arrow[-2::-1]])
 else:
 raise ValueError, "Got unknown shape: %s" % shape
 cx = float(dx)/distance
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

On Mon, Jun 2, 2008 at 1:04 PM, Les Schaffer <sch...@op...> wrote:
>> yep, its right there. very weird. does matplotlib fiddle with sys.path
>> or some such?????
No. Have you nuked all references to pylab.* and matplotlib* (esp
including any egg files?) Search your system for matplotlib and kill
all the old stuff-- egg files can do evil things. Also, what
directory are you running from?
Oh, I think I suddenly get it. If you are running from the examples
directory itself, eg
> python pylab/simple_plot.py
matplotlib i seeing "pylab" and trying to import it. Try
> cd examples/pylab
> python simple_plot.py --verbose-helpful
I think we need to rename that examples subdir. Is this the cause of
your problem?
JDH

KURT PETERS wrote:
> Unfortunately, I think someone else was on the maillist with a similar 
> problem, and perhaps our two mails got confused.
> I have "always been" using 0.99 as far as I know.
Kurt: I only released 0.99 yesterday. You were actually using 0.9.9, 
which you installed from a binary installer. There is no binary 
installed for 0.99 yet - which is the crux of your problem.
> It got "nuked" when I 
> deleted the old matplotlib. I wonder why it couldn't find geos the second 
> time around?
> 
You installed 0.9.9 from a binary installer, so it didn't need to find 
libgeos. It only looks for it when building from source.
> Once again, though, I'll let you know how things go. It's a good idea to 
> pull the toolkit out of the matplotlib directory to avoid this very problem. 
> I'll give that a try, by direct copying and changing my code. Probably a 
> lot of example code might need changing as well.
> Regards,
> Kurt
>
> 
All the examples in 0.99 have been updated to use 'import 
mpl_toolkits.basemap'. Until you actually have 0.99 installed, you 
don't want to change your code.
Hope this is clear - I'm getting the feeling that we're talking past 
each other ...
-Jeff
> ----Original Message Follows----
> From: Jeff Whitaker <js...@fa...>
> To: KURT PETERS <pet...@ms...>
> CC: jd...@gm..., cw...@gm..., 
> mat...@li...
> Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy 
> producing	error
> Date: 2008年6月02日 11:51:43 -0600
>
> KURT PETERS wrote:
> 
>> I am using 0.99.
>> 
> Kurt: Not really. You tried to install it but the install failed.
> 
>> I ended up copying the toolkits directory from my "old" installation of 
>> matplotlib to the new matplotlib directory and at least things seem to be 
>> working (although the HD seems to churn a lot more).
>>
>> 
> Hmm. I wouldn't expect this to work - there were changes in the matplotlib 
> API in version 0.98.0 that should break the old version of basemap.
> 
>> Let me test this a little more and I'll let you know how that is working 
>> for me (Windows could have started some other process in the background).
>> Kurt
>>
>> 
> OK.
>
> -Jeff
> 
>> ----Original Message Follows----
>> From: "John Hunter" <jd...@gm...>
>> To: "Jeff Whitaker" <js...@fa...>
>> CC: "KURT PETERS" <pet...@ms...>, cw...@gm..., 
>> mat...@li...
>> Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy producing 
>> error
>> Date: Mon, 2 Jun 2008 12:38:15 -0500
>>
>> On Mon, Jun 2, 2008 at 12:28 PM, Jeff Whitaker <js...@fa...> wrote:
>>
>> > Kurt: As you discovered, you'll need basemap 0.99 to use with 
>> matplotlib
>> > 0.98.0. I only have a source tarball on the sf site now, and you 
>> apparently
>> > are looking for the windows binary installer. I hope to have that up
>> > sometime next week - problem is that the person whose PC I use to 
>> build it
>> > is on vacation now. Perhaps someone will step up and build a basemap
>> > windows installer before then, or I can talk you through how to build 
>> one
>> > yourself with MingW32.
>>
>> I've updated the newsbox on the website with a link to a brief text
>> document including this info for basemap users.
>>
>> JDHr
>>
>>
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>> 
>
>
> --
> Jeffrey S. Whitaker Phone : (303)497-6313
> Meteorologist FAX : (303)497-6449
> NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
> 325 Broadway Office : Skaggs Research Cntr 1D-113
> Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

Unfortunately, I think someone else was on the maillist with a similar 
problem, and perhaps our two mails got confused.
 I have "always been" using 0.99 as far as I know. It got "nuked" when I 
deleted the old matplotlib. I wonder why it couldn't find geos the second 
time around?
 Once again, though, I'll let you know how things go. It's a good idea to 
pull the toolkit out of the matplotlib directory to avoid this very problem. 
 I'll give that a try, by direct copying and changing my code. Probably a 
lot of example code might need changing as well.
Regards,
Kurt
----Original Message Follows----
From: Jeff Whitaker <js...@fa...>
To: KURT PETERS <pet...@ms...>
CC: jd...@gm..., cw...@gm..., 
mat...@li...
Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy 
producing	error
Date: 2008年6月02日 11:51:43 -0600
KURT PETERS wrote:
>I am using 0.99.
Kurt: Not really. You tried to install it but the install failed.
> I ended up copying the toolkits directory from my "old" installation of 
>matplotlib to the new matplotlib directory and at least things seem to be 
>working (although the HD seems to churn a lot more).
>
Hmm. I wouldn't expect this to work - there were changes in the matplotlib 
API in version 0.98.0 that should break the old version of basemap.
> Let me test this a little more and I'll let you know how that is working 
>for me (Windows could have started some other process in the background).
>Kurt
>
OK.
-Jeff
>----Original Message Follows----
>From: "John Hunter" <jd...@gm...>
>To: "Jeff Whitaker" <js...@fa...>
>CC: "KURT PETERS" <pet...@ms...>, cw...@gm..., 
>mat...@li...
>Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy producing 
>error
>Date: Mon, 2 Jun 2008 12:38:15 -0500
>
>On Mon, Jun 2, 2008 at 12:28 PM, Jeff Whitaker <js...@fa...> wrote:
>
> > Kurt: As you discovered, you'll need basemap 0.99 to use with 
>matplotlib
> > 0.98.0. I only have a source tarball on the sf site now, and you 
>apparently
> > are looking for the windows binary installer. I hope to have that up
> > sometime next week - problem is that the person whose PC I use to 
>build it
> > is on vacation now. Perhaps someone will step up and build a basemap
> > windows installer before then, or I can talk you through how to build 
>one
> > yourself with MingW32.
>
>I've updated the newsbox on the website with a link to a brief text
>document including this info for basemap users.
>
>JDHr
>
>
>
>-------------------------------------------------------------------------
>This SF.net email is sponsored by: Microsoft
>Defy all challenges. Microsoft(R) Visual Studio 2008.
>http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>_______________________________________________
>Matplotlib-users mailing list
>Mat...@li...
>https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
--
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

Les Schaffer wrote:
> John Hunter wrote:
> 
>
>> After running the 0.98 installer do you have
>> pylab.py in site-packages? 
>> 
>
> yep, its right there. very weird. does matplotlib fiddle with sys.path 
> or some such?????
actually, it was right there in site-packages/matplotlib, which is what 
i assume you meant. i accidentally installed matplotlib/examples in my 
site-packages folder and when i cleaned that up, i inadvertently removed 
the pylab you put up there.
so i re-installed matplotlib once more, and now pylab is up in 
site-packages, and the pylab examples run fine, after i realized i 
needed to change the backend again in matplotlibrc -- i had deleted the 
one on Docs and Settings\XXXX per your advice, and got bit ;-)
Les

If you rename matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg to
matplotlib-0.98.0-py2.5.egg,
easy_install will install it from the disk.
I'm not sure why this happens. I've noticed lots of Mac OS X eggs with
cruft at the end of the
filename, which prevents their installation through easy_install.
Renaming is all it takes
to install them; still, it's annoying.
Cheers
V
On Mon, Jun 2, 2008 at 7:43 PM, Tommy Grav <tg...@ma...> wrote:
> I downloaded the egg for 0.98 from the matplotlib webpages,
> and I am trying to install it with
>
> easy_install ./matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
>
> Processing matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
> removing '/Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5-
> macosx-10.3-fat.egg' (and everything under it)
> creating /Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5-
> macosx-10.3-fat.egg
> Extracting matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg to /Library/
> Python/2.5/site-packages
> Removing matplotlib 0.91.2 from easy-install.pth file
> Adding matplotlib 0.98.0 to easy-install.pth file
>
> Installed /Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5-
> macosx-10.3-fat.egg
> Processing dependencies for matplotlib==0.98.0
> Searching for matplotlib==0.98.0
> Reading http://cheeseshop.python.org/pypi/matplotlib/
> Reading http://cheeseshop.python.org/pypi/matplotlib/0.98.0
> Reading http://matplotlib.sourceforge.net
> Reading http://cheeseshop.python.org/pypi/matplotlib/0.91.3
> No local packages or download links found for matplotlib==0.98.0
> error: Could not find suitable distribution for
> Requirement.parse('matplotlib==0.98.0')
> [skathi:~/Downloads] tgrav%
>
> Why is it trying to download the egg, when it is right there in the
> directory where
> I am running easy_install from. I do now see the point some people
> made on the
> python list about easy_install. This is an annoying behavior by the
> software.
>
> Cheers
> Tommy
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

KURT PETERS wrote:
> I am using 0.99. 
Kurt: Not really. You tried to install it but the install failed.
> I ended up copying the toolkits directory from my "old" 
> installation of matplotlib to the new matplotlib directory and at least 
> things seem to be working (although the HD seems to churn a lot more).
> 
Hmm. I wouldn't expect this to work - there were changes in the 
matplotlib API in version 0.98.0 that should break the old version of 
basemap.
> Let me test this a little more and I'll let you know how that is working 
> for me (Windows could have started some other process in the background).
> Kurt
> 
OK.
-Jeff
> ----Original Message Follows----
> From: "John Hunter" <jd...@gm...>
> To: "Jeff Whitaker" <js...@fa...>
> CC: "KURT PETERS" <pet...@ms...>, cw...@gm..., 
> mat...@li...
> Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy producing 
> error
> Date: Mon, 2 Jun 2008 12:38:15 -0500
>
> On Mon, Jun 2, 2008 at 12:28 PM, Jeff Whitaker <js...@fa...> wrote:
>
> > Kurt: As you discovered, you'll need basemap 0.99 to use with matplotlib
> > 0.98.0. I only have a source tarball on the sf site now, and you 
> apparently
> > are looking for the windows binary installer. I hope to have that up
> > sometime next week - problem is that the person whose PC I use to build 
> it
> > is on vacation now. Perhaps someone will step up and build a basemap
> > windows installer before then, or I can talk you through how to build one
> > yourself with MingW32.
>
> I've updated the newsbox on the website with a link to a brief text
> document including this info for basemap users.
>
> JDHr
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

On Mon, Jun 2, 2008 at 12:45 PM, John Hunter <jd...@gm...> wrote:
> for line in ax.get_xticklines() + ax.get_yticklines():
> line.set_markeredgewidth(10)
Correction ( I should have read the link myself first). The property
you are probably interested is the markersize (marker size in points)
though markeredgewidth may be useful too....
JDH

John Hunter wrote:
> By the way, on windows you can edit the rc file in place (ie leave it
> in mpl-data). I need to update the instructions in that file -- they
> are out of date.
> 
please do, they indicate movement is required.
>
> This is weird: it is working on my end once I update my "backend"
> setting to "TkAgg". 
i tried TkAgg as well but still no go.
> After running the 0.98 installer do you have
> pylab.py in site-packages? 
yep, its right there. very weird. does matplotlib fiddle with sys.path 
or some such?????
> Make sure you remove all matplotlib traces
> from site-packages before installing.
i did that and re-installed, same problem.
Les

I am using 0.99. I ended up copying the toolkits directory from my "old" 
installation of matplotlib to the new matplotlib directory and at least 
things seem to be working (although the HD seems to churn a lot more).
 Let me test this a little more and I'll let you know how that is working 
for me (Windows could have started some other process in the background).
Kurt
----Original Message Follows----
From: "John Hunter" <jd...@gm...>
To: "Jeff Whitaker" <js...@fa...>
CC: "KURT PETERS" <pet...@ms...>, cw...@gm..., 
mat...@li...
Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy producing 
error
Date: Mon, 2 Jun 2008 12:38:15 -0500
On Mon, Jun 2, 2008 at 12:28 PM, Jeff Whitaker <js...@fa...> wrote:
 > Kurt: As you discovered, you'll need basemap 0.99 to use with matplotlib
 > 0.98.0. I only have a source tarball on the sf site now, and you 
apparently
 > are looking for the windows binary installer. I hope to have that up
 > sometime next week - problem is that the person whose PC I use to build 
it
 > is on vacation now. Perhaps someone will step up and build a basemap
 > windows installer before then, or I can talk you through how to build one
 > yourself with MingW32.
I've updated the newsbox on the website with a link to a brief text
document including this info for basemap users.
JDHr

On Mon, Jun 2, 2008 at 12:32 PM, Berit Hinnemann <be...@to...> wrote:
> There is one thing, which I cannot get to work, namely to change the
> linewidth of the tick lines.
>
> I have tried
>
> ticklines = ax.get_xticklines()
> ticklines.extend(ax.get_yticklines())
>
> for line in ticklines:
> line.set_linewidth(10)
>
> but nothing changes. I can change the color and other properties, but no
> matter what I set for the linewidth, it does not change. I have also tried
> to go in and
This is the correct approach but it is a tricky one. In matplotlib,
the ticks are *markers*. All Line2D objects support a line (solid,
dashed, etc) and a marker (circle, square, tick). The tick linewidth
is controlled by the "markeredgewidth" property.
from pylab import figure, show
fig = figure()
ax = fig.add_subplot(111)
ax.plot(range(10))
for line in ax.get_xticklines() + ax.get_yticklines():
 line.set_markeredgewidth(10)
show()
BTW -- this is covered in the somewhat new artist api tutorial at
http://matplotlib.sf.net/pycon/artist_api_tut.pdf

I downloaded the egg for 0.98 from the matplotlib webpages,
and I am trying to install it with
easy_install ./matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
Processing matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg
removing '/Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5- 
macosx-10.3-fat.egg' (and everything under it)
creating /Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5- 
macosx-10.3-fat.egg
Extracting matplotlib-0.98.0-py2.5-macosx-10.3-fat.egg to /Library/ 
Python/2.5/site-packages
Removing matplotlib 0.91.2 from easy-install.pth file
Adding matplotlib 0.98.0 to easy-install.pth file
Installed /Library/Python/2.5/site-packages/matplotlib-0.98.0-py2.5- 
macosx-10.3-fat.egg
Processing dependencies for matplotlib==0.98.0
Searching for matplotlib==0.98.0
Reading http://cheeseshop.python.org/pypi/matplotlib/
Reading http://cheeseshop.python.org/pypi/matplotlib/0.98.0
Reading http://matplotlib.sourceforge.net
Reading http://cheeseshop.python.org/pypi/matplotlib/0.91.3
No local packages or download links found for matplotlib==0.98.0
error: Could not find suitable distribution for 
Requirement.parse('matplotlib==0.98.0')
[skathi:~/Downloads] tgrav%
Why is it trying to download the egg, when it is right there in the 
directory where
I am running easy_install from. I do now see the point some people 
made on the
python list about easy_install. This is an annoying behavior by the 
software.
Cheers
 Tommy

On Mon, Jun 2, 2008 at 12:28 PM, Jeff Whitaker <js...@fa...> wrote:
> Kurt: As you discovered, you'll need basemap 0.99 to use with matplotlib
> 0.98.0. I only have a source tarball on the sf site now, and you apparently
> are looking for the windows binary installer. I hope to have that up
> sometime next week - problem is that the person whose PC I use to build it
> is on vacation now. Perhaps someone will step up and build a basemap
> windows installer before then, or I can talk you through how to build one
> yourself with MingW32.
I've updated the newsbox on the website with a link to a brief text
document including this info for basemap users.
JDHr

Hi,
First of all, thanks to the developers for such a great Python plotting 
package. I use matplotlib more and more for my work.
There is one thing, which I cannot get to work, namely to change the 
linewidth of the tick lines.
I have tried
ticklines = ax.get_xticklines()
ticklines.extend(ax.get_yticklines())
for line in ticklines:
 line.set_linewidth(10)
but nothing changes. I can change the color and other properties, but no 
matter what I set for the linewidth, it does not change. I have also tried 
to go in and
ticklines[0]._linewidth = 10
but again no change. If I do a similar thing for the color
ticklines[0]._color = 'b'
the color is changed.
I use matplotlib 0.91 on a RedHat EL 4 machine.
Any ideas on how to solve this would be much appreciated. I have searched 
the mailing lists and the site, but could not find a solution (apologies 
if I should have missed it)
Best regards,
Berit
Berit Hinnemann
Research Scientist | Characterization Department | Research & Development 
Haldor Topsøe A/S
Nymøllevej 55
DK-2800 Kgs. Lyngby
Phone: +45 4527 2130 (direct)
Haldor Topsøe is market leading within heterogeneous catalysis and 
supplies catalysts and process design for oil refineries, chemical plants 
and the energy sector. The environment plays an important role in Topsøe's 
research and production, and through the introduction of new catalysts and 
technologies the company supports a more sustainable use of the world's 
resources. Annual revenues total DKK 3.5 billion generated by our 1,700 
dedicated employees. For additional information please see our web site 
www.topsoe.com
This e-mail message (including attachments, if any) is confidential and 
may be privileged. It is intended only for the addressee.
Any unauthorised distribution or disclosure is prohibited. Disclosure to 
anyone other than the intended recipient does not constitute waiver of 
privilege.
If you have received this e-mail in error, please notify the sender by 
e-mail and delete it and any attachments from your computer system and 
records.
HALDOR TOPSOE (www.topsoe.com)

On Mon, Jun 2, 2008 at 12:10 PM, KURT PETERS <pet...@ms...> wrote:
> Thanks,
> I made that change after an unfruitful attempt at installing gtk, which
> should have worked in my estimation. Now, at least I get by that original
> point.
>
> NOW basemaps won't work any more! When I run, I get:
>
> Traceback (most recent call last):
> File "C:\Documents and
> Settings\kpeters\smworkspace\parsesmExcelFile\src\readParse\ExcelClass.py",
> line 13, in <module>
> from matplotlib.toolkits.basemap import Basemap as Basemap
> ImportError: No module named toolkits.basemap
One of the api changes is that the import line for basemap (and all
matplotlib toolkits) is
 import mpl_toolkits.basemap
rather than
 import matplotlib.toolkits.basemap
> I'm supposing that I NOW have to reinstall basemaps. The basemaps installer
> won't work for me any more though :-(
> This is that message I get when trying to reinstall basemaps:
>Traceback (most recent call last):
> File "setup.py", line 80, in <module>
> geos_include_dirs=[os.path.join(GEOS_dir,'include'),numpy.get_include()]
> File "C:\Python25\lib\ntpath.py", line 90, in join
> assert len(path) > 0
> TypeError: object of type 'NoneType' has no len()
This is an error that you will get if you are trying to install
basemap yourself from source and the GEOS_DIR is set to None (Jeff, a
more helpful error would be good here). I think we are still waiting
on a binary installer for basemap.
JDH

KURT PETERS wrote:
> Thanks,
> I made that change after an unfruitful attempt at installing gtk, which 
> should have worked in my estimation. Now, at least I get by that original 
> point.
>
> NOW basemaps won't work any more! When I run, I get:
> 
Kurt: As you discovered, you'll need basemap 0.99 to use with 
matplotlib 0.98.0. I only have a source tarball on the sf site now, and 
you apparently are looking for the windows binary installer. I hope to 
have that up sometime next week - problem is that the person whose PC I 
use to build it is on vacation now. Perhaps someone will step up and 
build a basemap windows installer before then, or I can talk you through 
how to build one yourself with MingW32.
BTW; with the new basemap you'll have to use
from mpl_toolkits.basemap import Basemap
instead of the old
from matplotlib.toolkits.basemap import Basemap
-Jeff
> Traceback (most recent call last):
> File "C:\Documents and 
> Settings\kpeters\smworkspace\parsesmExcelFile\src\readParse\ExcelClass.py", 
> line 13, in <module>
> from matplotlib.toolkits.basemap import Basemap as Basemap
> ImportError: No module named toolkits.basemap
>
> I'm supposing that I NOW have to reinstall basemaps. The basemaps installer 
> won't work for me any more though :-(
> This is that message I get when trying to reinstall basemaps:
>
> C:\Python25\Lib\basemap-0.99>c:\Python25\python.exe setup.py install
> checking for GEOS lib in /usr/local ....
> checking for GEOS lib in /sw ....
> checking for GEOS lib in /opt ....
> checking for GEOS lib in /opt/local ....
> checking for GEOS lib in C:\Documents and Settings\kpeters ....
> Traceback (most recent call last):
> File "setup.py", line 80, in <module>
> geos_include_dirs=[os.path.join(GEOS_dir,'include'),numpy.get_include()]
> File "C:\Python25\lib\ntpath.py", line 90, in join
> assert len(path) > 0
> TypeError: object of type 'NoneType' has no len()
>
> I hope no one else has this pain when upgrading.
> Regards,
> Kurt
> ----Original Message Follows----
> From: "John Hunter" <jd...@gm...>
> To: "KURT PETERS" <pet...@ms...>, "Charlie Moad" <cw...@gm...>
> CC: mat...@li...
> Subject: Re: [Matplotlib-users] Update to Matplotlib and Numpy producing 
> error
> Date: Mon, 2 Jun 2008 11:01:48 -0500
>
> On Mon, Jun 2, 2008 at 10:41 AM, KURT PETERS <pet...@ms...> wrote:
>
> > File "C:\Python25\Lib\site-packages\matplotlib\backends\backend_gtk.py",
> > line 6, in <module>
> > import gobject
> > ImportError: No module named gobject
>
> This is a problem with our installer since the default backend we have
> set is GTKAgg (we usually set it to GTKAgg). Fortunately, it is
> relatively easy for you to fix:. Edit
> site-packages/matplotlib/mpl-data/matplotlib and change the 'backend :
> GTKAgg' line to
>
> backend : TkAgg
>
> Charlie -- we will need to roll out a new binary builds which fix this
> ASAP because this will bite all the win32 users who don't have a
> custom rc, which is most of them I suspect. We'll also need to check
> the settings in the 91.3 release. My apologies for not testing this
> as you requested (I did test on linux though). You probably will also
> want to check the default backend for the OS X eggs too.....
>
> JDH
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

On Mon, Jun 2, 2008 at 12:09 PM, Les Schaffer <sch...@op...> wrote:
> there are two files there: matplotlibrc and matplotlib.conf. is there a
> selection principle for choosing one or the other?
matplotlib.conf is used at build time, so you can ignore it with the
binary installer.
> i moved matplotlibrc to the indicated folder, set backed to WXAgg and some
By the way, on windows you can edit the rc file in place (ie leave it
in mpl-data). I need to update the instructions in that file -- they
are out of date.
> of the examples run. but i am getting failures finding pylab when i run the
> pylab examples:
> $ python geo_demo.py
> Traceback (most recent call last):
> File "geo_demo.py", line 4, in <module>
> from pylab import *
> ImportError: No module named pylab
>
> am not sure why the interpreter isnt finding pylab module even tho, for
> example, examples/{api, animation} runs, with a few exceptions.
This is weird: it is working on my end once I update my "backend"
setting to "TkAgg". After running the 0.98 installer do you have
pylab.py in site-packages? Make sure you remove all matplotlib traces
from site-packages before installing.
JDH