matplotlib-users Mailing List for matplotlib

On 06/04/2010 02:25 PM, Daniele Nicolodi wrote:
> On 04/06/10 20:08, Michael Droettboom wrote:
>    
>> Set rc.Params['path.simplify'] to False, or upgrade to 0.99.3.
>>      
> Setting path.simplify = False solved my issue. Has been the issue solved
> in another way on 0.99.3 or path.simplify = False is simply the new default?
>
> Thanks. Cheers,
>    
The path simplification algorithm has a bug in 0.99.1 that has been 
fixed in 0.99.3.

Mike

-- 
Michael Droettboom
Science Software Branch
Space Telescope Science Institute
Baltimore, Maryland, USA

On 04/06/10 20:08, Michael Droettboom wrote:
> Set rc.Params['path.simplify'] to False, or upgrade to 0.99.3.

Setting path.simplify = False solved my issue. Has been the issue solved
in another way on 0.99.3 or path.simplify = False is simply the new default?

Thanks. Cheers,
-- 
Daniele

Le jeudi 03 juin 2010 à 11:41 -0400, Jae-Joon Lee a écrit :
> On Wed, Jun 2, 2010 at 2:46 PM, Denis Laxalde <dla...@gm...> wrote:
> > That would indeed be a better approach. Can somebody points me to the
> > particular methods/attributes to look at ?
> >
> 
> As far as I can see, there is no public methods/attributes.
> Can you file a bug so that Reinier (or others) can pick this up later?

Done: https://sourceforge.net/tracker/?func=detail&atid=560723&aid=3011616&group_id=80706

Cheers,

    Denis

Set rc.Params['path.simplify'] to False, or upgrade to 0.99.3.

Mike

On 06/04/2010 02:04 PM, Daniele Nicolodi wrote:
> Hello.
>
> I'm observing a quite annoying behavior of matplotlib generated plots.
>
> I plot signal time series with continuous lines. When in those time
> series I have single points laying far from the median, those are not
> represented on the plot. I think this must be due to the anti aliasing
> algorithm, because if I add + markers to the plot, those are correctly
> placed.
>
> There is a way to avoid this behavior?
>
> Thanks. Cheers,
>    


-- 
Michael Droettboom
Science Software Branch
Space Telescope Science Institute
Baltimore, Maryland, USA

Hello.

I'm observing a quite annoying behavior of matplotlib generated plots.

I plot signal time series with continuous lines. When in those time
series I have single points laying far from the median, those are not
represented on the plot. I think this must be due to the anti aliasing
algorithm, because if I add + markers to the plot, those are correctly
placed.

There is a way to avoid this behavior?

Thanks. Cheers,
-- 
Daniele

By default, when the xy coordinate of the annotate is given in the
data coordinate, it draws the arrow only if the xy point is inside the
axes bbox. And, when the xy point is at the boundary of the axes bbox,
the inside-test results seems to depend on the backend. So, some
backends draws the arrow but others do not.

I'll try to improve the situation but I'm not quite sure how.

Meanwhile, you can turn off the inside-test by setting annotate_clip=False.

ax.annotate("", xy = (0.1, -0.5), xytext = (0.1, -0.2),
            arrowprops = dict(arrowstyle = "->"), annotation_clip=False)

IHTH,

Regards,

-JJ



On Fri, Jun 4, 2010 at 10:59 AM, Matthias Michler
<Mat...@gm...> wrote:
> Hello list,
>
> I'm encountering a problem with annotate using the latest svn and GTKAgg-
> backend.
>
> I plot several annotate-arrows in my script, but some of them aren't saved to
> the eps/pdf - file. More precisely the lower left arrow in the attached
> example, who touches the lower (negative) boundary of the plot arrea shows up
> in the matplotlib figure and the png-picture, but not in the pdf / eps output.
>
> It seems to be important that in my case the lower limit of the y-axis is
> negative.
>
> I found two solutions to this problems, which are the middle and right of the
> lower arrows, but I don't understand why this is needed. The middle arrow does
> not touch the lower boundary and the right arrow has an inverted direction and
> inverted arrowstyle.
>
> Is this a bug or did I miss something?
>
> Kind regards and thanks in advance for any hints.
> Matthias
>
> ------------------------------------------------------------------------------
> ThinkGeek and WIRED's GeekDad team up for the Ultimate
> GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the
> lucky parental unit.  See the prize list and enter to win:
> http://p.sf.net/sfu/thinkgeek-promo
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

Hello list; 

I'm new to python/matplotlib, migrating from IDL. I need to do some
interactive point selection with mouse, and the pyplot.ginput() routine
seemed to be just the right thing here. I do however need to be able to
make a not previously specified number of clicks, so ginput(n=0) is a
requirement. 

However, when I press enter on my laptop (with no middle mouse button),
I get the error message:

TypeError: stop_event_loop() takes exactly 1 argument (2 given)

I've tried two different pieces of example code from the matplotlib SF
page.
I'm using Matplotlib 0.99.1 on Ubuntu 10.04.
Is there any solution or workaround? If I cannot have this
functionality, it is, unfortunately, kind of a deal breaker for at least
some of my work, and I'd like to keep it purely Python.

Best regards; 

Emil

On Tue, May 25, 2010 at 4:44 PM, Kaushik Ghose
<Kau...@hm...> wrote:
> Hi Gang,
>
> I don't know if it is a problem from nabble, but the 'archives' link from the
> main matplotlib pages goes to a decidedly non-matplotlib page.
>
> The link is http://www.nabble.com/matplotlib---users-f2906.html

Thanks -- I've updated the website.

JDH

On Fri, Jun 4, 2010 at 3:21 AM, Eric Firing <ef...@ha...> wrote:
> On 05/25/2010 11:44 AM, Kaushik Ghose wrote:
>>
>> Hi Gang,
>>
>> I don't know if it is a problem from nabble, but the 'archives' link from
>> the
>> main matplotlib pages goes to a decidedly non-matplotlib page.
>>
>> The link is http://www.nabble.com/matplotlib---users-f2906.html
>
> Evidently nabble has been reconfigured, and everything that was on
> www.nabble.com is now on old.nabble.com, unless someone requested a
> migration.  What I don't understand is how our old url got mapped to that
> Russian site.  In any case, we need to migrate and/or change the url in the
> docs.  Presumably migration is the thing to do.
>
> http://nabble-support.1.n2.nabble.com/Migrating-forums-from-Nabble1-to-Nabble2-td3791390.html#a3791390
>
> John, if you set up the nabble archiving in the first place, I suspect you
> need to make the request.

I did not, and don't know who did.

JDH

Fixed in r8373.

Mike

On 06/03/2010 09:01 PM, Eric Firing wrote:
> On 06/03/2010 02:29 PM, Christoph Gohlke wrote:
>    
>>
>> On 6/3/2010 5:20 PM, PH...@Ge... wrote:
>>      
>>>> -----Original Message-----
>>>> From: Christoph Gohlke [mailto:cg...@uc...]
>>>> Sent: Sunday, May 30, 2010 3:41 PM
>>>> To: mat...@li...
>>>> Subject: [Matplotlib-users] leading whitespace in text
>>>>
>>>> Hello,
>>>>
>>>> consider the following simple code:
>>>>
>>>> import matplotlib
>>>> from matplotlib import pyplot
>>>> pyplot.text(0, 0, "<--", family='monospace')
>>>> pyplot.show()
>>>>
>>>> Using matplotlib-0.99.3.win32-py2.6, the text '<--' is drawn with an
>>>> x-offset from the y-axis, as expected from the presence of leading
>>>> whitespace.
>>>>
>>>> Using matplotlib-1.0.svn.win32-py2.6 rev 8347, the leading whitespace is
>>>> apparently removed from the text and '<--' is drawn right at the y-axis
>>>> with no x-offset.
>>>>
>>>> In both cases the font is monospace.
>>>>
>>>> Is this change intended or a regression in 1.0svn?
>>>>          
>>> Christoph,
>>>
>>> I haven't had a chance to try this out with my machine that has MPL 1.0svn on it, but is there any chance that you have mpl.rcParams['text.usetex'] = True when you get the error?
>>>
>>>        
>>
>> Hi Paul,
>>
>> no, usetex is not set, I made sure of that. I have traced this to a
>> recent commit and opened a ticket on SF.
>>
>> <http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7635&r2=7838>
>> <http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7838&r2=7839>
>>
>> <http://sourceforge.net/tracker/?func=detail&aid=3010045&group_id=80706&atid=560720>
>>
>>      
> Thanks for doing that.  I have assigned it to Mike D.  I hope he has a
> little time to look at it.  I suspect he will instantly see what the
> problem is and how to fix it.
>
> Eric
>
>
> ------------------------------------------------------------------------------
> ThinkGeek and WIRED's GeekDad team up for the Ultimate
> GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the
> lucky parental unit.  See the prize list and enter to win:
> http://p.sf.net/sfu/thinkgeek-promo
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>    


-- 
Michael Droettboom
Science Software Branch
Space Telescope Science Institute
Baltimore, Maryland, USA

On 05/25/2010 11:44 AM, Kaushik Ghose wrote:
> Hi Gang,
>
> I don't know if it is a problem from nabble, but the 'archives' link from the
> main matplotlib pages goes to a decidedly non-matplotlib page.
>
> The link is http://www.nabble.com/matplotlib---users-f2906.html

Evidently nabble has been reconfigured, and everything that was on 
www.nabble.com is now on old.nabble.com, unless someone requested a 
migration.  What I don't understand is how our old url got mapped to 
that Russian site.  In any case, we need to migrate and/or change the 
url in the docs.  Presumably migration is the thing to do.

http://nabble-support.1.n2.nabble.com/Migrating-forums-from-Nabble1-to-Nabble2-td3791390.html#a3791390

John, if you set up the nabble archiving in the first place, I suspect 
you need to make the request.

Eric


>
> Best
> -Kaushik
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

On Thu, Jun 3, 2010 at 12:22 PM, Jae-Joon Lee <lee...@gm...> wrote:

> On Thu, Jun 3, 2010 at 12:52 PM, Benjamin Root <ben...@ou...> wrote:
> > Well, the link is still not back, so I will probe you a bit further.  You
> > say that axes_grid is provided for backward compatibility, does that mean
> > that I should be using axes_grid1 for new code?  I have noticed
> differences
> > in behavior if I import axes_grid versus axes_grid1.  For example, my
> > colorbar labels wouldn't appear unless I used axes_grid1.
> >
> > Ben Root
> >
>
> Yes, axes_grid1 is recommended for new code.
> The "axes_grid" in the svn actually imports the "axes_grid1" and
> "axisaritst" modules.
>
> With axes_grid1, I tried to fix some compatibility issue that
> axes_grid had with the original matplolib. And I guess the colorbar
> label issue was one of them. With axes_grid, the colorbar labels are
> invisible by default, but they are visible in axes_grid1.
>
> Regards,
>
> -JJ
>

Thanks, that helps clear up some things for me.  I am curious if axes_grid1
is intended to be a final name.  When looking at the docs, it almost seems
like axes_grid1 is some supporting module while axes_grid is the main
module, just from a naming convention point-of-view.  But, that's just my
first impression of seeing those modules in that folder.

Ben Root

On 06/03/2010 02:29 PM, Christoph Gohlke wrote:
>
>
> On 6/3/2010 5:20 PM, PH...@Ge... wrote:
>>> -----Original Message-----
>>> From: Christoph Gohlke [mailto:cg...@uc...]
>>> Sent: Sunday, May 30, 2010 3:41 PM
>>> To: mat...@li...
>>> Subject: [Matplotlib-users] leading whitespace in text
>>>
>>> Hello,
>>>
>>> consider the following simple code:
>>>
>>> import matplotlib
>>> from matplotlib import pyplot
>>> pyplot.text(0, 0, "<--", family='monospace')
>>> pyplot.show()
>>>
>>> Using matplotlib-0.99.3.win32-py2.6, the text '<--' is drawn with an
>>> x-offset from the y-axis, as expected from the presence of leading
>>> whitespace.
>>>
>>> Using matplotlib-1.0.svn.win32-py2.6 rev 8347, the leading whitespace is
>>> apparently removed from the text and '<--' is drawn right at the y-axis
>>> with no x-offset.
>>>
>>> In both cases the font is monospace.
>>>
>>> Is this change intended or a regression in 1.0svn?
>>
>> Christoph,
>>
>> I haven't had a chance to try this out with my machine that has MPL 1.0svn on it, but is there any chance that you have mpl.rcParams['text.usetex'] = True when you get the error?
>>
>
>
> Hi Paul,
>
> no, usetex is not set, I made sure of that. I have traced this to a
> recent commit and opened a ticket on SF.
>
> <http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7635&r2=7838>
> <http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7838&r2=7839>
>
> <http://sourceforge.net/tracker/?func=detail&aid=3010045&group_id=80706&atid=560720>
>

Thanks for doing that.  I have assigned it to Mike D.  I hope he has a 
little time to look at it.  I suspect he will instantly see what the 
problem is and how to fix it.

Eric

On 6/3/2010 5:20 PM, PH...@Ge... wrote:
>> -----Original Message-----
>> From: Christoph Gohlke [mailto:cg...@uc...]
>> Sent: Sunday, May 30, 2010 3:41 PM
>> To: mat...@li...
>> Subject: [Matplotlib-users] leading whitespace in text
>>
>> Hello,
>>
>> consider the following simple code:
>>
>> import matplotlib
>> from matplotlib import pyplot
>> pyplot.text(0, 0, "<--", family='monospace')
>> pyplot.show()
>>
>> Using matplotlib-0.99.3.win32-py2.6, the text '<--' is drawn with an
>> x-offset from the y-axis, as expected from the presence of leading
>> whitespace.
>>
>> Using matplotlib-1.0.svn.win32-py2.6 rev 8347, the leading whitespace is
>> apparently removed from the text and '<--' is drawn right at the y-axis
>> with no x-offset.
>>
>> In both cases the font is monospace.
>>
>> Is this change intended or a regression in 1.0svn?
>
> Christoph,
>
> I haven't had a chance to try this out with my machine that has MPL 1.0svn on it, but is there any chance that you have mpl.rcParams['text.usetex'] = True when you get the error?
>


Hi Paul,

no, usetex is not set, I made sure of that. I have traced this to a 
recent commit and opened a ticket on SF.

<http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7635&r2=7838>
<http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/src/ft2font.cpp?r1=7838&r2=7839>

<http://sourceforge.net/tracker/?func=detail&aid=3010045&group_id=80706&atid=560720>

-- 
Christoph

> -----Original Message-----
> From: Christoph Gohlke [mailto:cg...@uc...]
> Sent: Sunday, May 30, 2010 3:41 PM
> To: mat...@li...
> Subject: [Matplotlib-users] leading whitespace in text
> 
> Hello,
> 
> consider the following simple code:
> 
> import matplotlib
> from matplotlib import pyplot
> pyplot.text(0, 0, "    <--", family='monospace')
> pyplot.show()
> 
> Using matplotlib-0.99.3.win32-py2.6, the text '<--' is drawn with an
> x-offset from the y-axis, as expected from the presence of leading
> whitespace.
> 
> Using matplotlib-1.0.svn.win32-py2.6 rev 8347, the leading whitespace is
> apparently removed from the text and '<--' is drawn right at the y-axis
> with no x-offset.
> 
> In both cases the font is monospace.
> 
> Is this change intended or a regression in 1.0svn?
 
Christoph,

I haven't had a chance to try this out with my machine that has MPL 1.0svn on it, but is there any chance that you have mpl.rcParams['text.usetex'] = True when you get the error?

-paul