matplotlib-users Mailing List for matplotlib

Hi
Thanks for the replies. I replaced all the fonts of my script to
Helvetica and it works fine. Yes, I'm using ps.fonttype=3. The files
are pretty small.
But I still couldn't find out why matplotlib was getting that ttf Vera...
Cheers,
Marco
On Tue, Aug 26, 2008 at 4:48 AM, Michael Droettboom <md...@st...> wrote:
> The "useafm" option is designed to work with the 14 standard Postscript
> fonts. I suppose it's possible to use a third-party afm, but such a thing
> would be extremely rare. The closest thing to Arial in the standard PS
> fonts is Helvetica.
>
> But yes, this is a bug, only in that the error message could be better, not
> that using Arial as an AFM would ever work.
>
> As for reducing file sizes, are you already using ps.fonttype set to 3?
> That will subset a TTF font, and should reduce file sizes over type 42.
>
> Cheers,
> Mike
>
> Marco A. S. Netto wrote:
>>
>> Hi everyone
>>
>> I'm a matplotlib's beginner user and I had a problem that seems to be
>> a bug. If I enable the option ps.useafm and I ask for a font that is
>> not available in my OS, the matplotlib's font manager tries to get a
>> default font, which is Vera (I don't know how the default font is set
>> to Vera, since I removed it from the font.serif and font.sans-serif
>> lists). The problem is that the font manager is getting a ttf file, in
>> my OS it is located at
>> /usr/share/fonts/truetype/tf-bitstream-vera/Vera.ttf. Therefore I'm
>> getting the error:
>>
>> Could not match Arial, normal, 400. Returning
>> /usr/share/fonts/truetype/tf-bitstream-vera/Vera.ttf (this line is
>> showed when I use the option --verbose-helpful)
>> ...
>> ...
>> ...
>> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 295,
>> in __init__
>> parse_afm(fh)
>> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 281,
>> in parse_afm
>> _sanity_check(fh)
>> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 76,
>> in _sanity_check
>> raise RuntimeError('Not an AFM file')
>> RuntimeError: Not an AFM file
>>
>> Is this a bug? Shouldn't matplotlib try to get an AFM default font,
>> rather than a ttf font, in case it cannot find what I'm looking for
>> (with ps.useafm=True)? In my case I put the font Arial but matplotlib
>> was not able to find it. Some details: I'm using ubuntu hardy and
>> matplotlib version 0.98.3. backend PS to generated eps files. My
>> decision to use AFM is to reduce the file sizes.
>>
>> Kind regards,
>> Marco
>>
>> -------------------------------------------------------------------------
>> This SF.Net email is sponsored by the Moblin Your Move Developer's
>> challenge
>> Build the coolest Linux based applications with Moblin SDK & win great
>> prizes
>> Grand prize is a trip for two to an Open Source event anywhere in the
>> world
>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>
>

Thanks for spotting this. I don't think a lot of us ever use 
non-English locales, so that's fallen through the cracks.
backend_svg.py was using '%s' to convert floats, to limit the number of 
digits written to the file. This works fine with regular floats (since 
they do not follow the locale unless explicitly asked to do so), but 
numpy arrays seem to always follow the locale. (Not certain whether 
that inconsistency could be considered a bug in Numpy).
In any case, I've committed a workaround to matplotlib, which 
unfortunately results in slightly larger SVG files. This is now fixed 
in SVN r6049. Look there for a patch if you just want to fix your local 
backend_svg.py.
Cheers,
Mike
Mathieu Leplatre wrote:
> Hi,
>
> If I change the locale using cairo backend, the result is fine.
> But with SVG backend, every objects collapse on top-left corner.
>
> As you can see here :
> http://mathieu-leplatre.info/media/matplotlib-svg/localechange-svg.svg
>
> I exported it to png with inkscape to reveal objects outside of frame :
> http://mathieu-leplatre.info/media/matplotlib-svg/localechange-svg.png
>
> Cairo output is fine :
> http://mathieu-leplatre.info/media/matplotlib-svg/localechange-cairo.svg
>
> I guess it's not a unicode problem, since the strings have no
> non-ascii characters.
>
> The code is quite straightforward, it allows me to choose the date
> language for the formatters.
> Let me know if I should fill a bug report.
> Thank you all !
>
> ----
> import locale, matplotlib
> from datetime import datetime
> matplotlib.use('svg')
> import pylab, numpy
>
> locale.setlocale(locale.LC_ALL, 'fr_FR.UTF-8')
> today = datetime.today().strftime("%A")
>
> pylab.title( today )
> pylab.plot( range(10), numpy.random.randn( 10 ))
> pylab.savefig('localechange-%s' % matplotlib.get_backend())
> ----
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> 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

Hi,
If I change the locale using cairo backend, the result is fine.
But with SVG backend, every objects collapse on top-left corner.
As you can see here :
http://mathieu-leplatre.info/media/matplotlib-svg/localechange-svg.svg
I exported it to png with inkscape to reveal objects outside of frame :
http://mathieu-leplatre.info/media/matplotlib-svg/localechange-svg.png
Cairo output is fine :
http://mathieu-leplatre.info/media/matplotlib-svg/localechange-cairo.svg
I guess it's not a unicode problem, since the strings have no
non-ascii characters.
The code is quite straightforward, it allows me to choose the date
language for the formatters.
Let me know if I should fill a bug report.
Thank you all !
----
import locale, matplotlib
from datetime import datetime
matplotlib.use('svg')
import pylab, numpy
locale.setlocale(locale.LC_ALL, 'fr_FR.UTF-8')
today = datetime.today().strftime("%A")
pylab.title( today )
pylab.plot( range(10), numpy.random.randn( 10 ))
pylab.savefig('localechange-%s' % matplotlib.get_backend())
----

On 2008年8月25日, Michael Droettboom apparently wrote:
> The closest thing to Arial in the standard PS fonts is 
> Helvetica. 
OT:
Just a reminder that this is because Arial was designed as 
a Helvetica clone for Windows users.
Cheers,
Alan Isaac

The "useafm" option is designed to work with the 14 standard Postscript 
fonts. I suppose it's possible to use a third-party afm, but such a 
thing would be extremely rare. The closest thing to Arial in the 
standard PS fonts is Helvetica.
But yes, this is a bug, only in that the error message could be better, 
not that using Arial as an AFM would ever work.
As for reducing file sizes, are you already using ps.fonttype set to 3? 
That will subset a TTF font, and should reduce file sizes over type 42.
Cheers,
Mike
Marco A. S. Netto wrote:
> Hi everyone
>
> I'm a matplotlib's beginner user and I had a problem that seems to be
> a bug. If I enable the option ps.useafm and I ask for a font that is
> not available in my OS, the matplotlib's font manager tries to get a
> default font, which is Vera (I don't know how the default font is set
> to Vera, since I removed it from the font.serif and font.sans-serif
> lists). The problem is that the font manager is getting a ttf file, in
> my OS it is located at
> /usr/share/fonts/truetype/tf-bitstream-vera/Vera.ttf. Therefore I'm
> getting the error:
>
> Could not match Arial, normal, 400. Returning
> /usr/share/fonts/truetype/tf-bitstream-vera/Vera.ttf (this line is
> showed when I use the option --verbose-helpful)
> ...
> ...
> ...
> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 295,
> in __init__
> parse_afm(fh)
> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 281,
> in parse_afm
> _sanity_check(fh)
> File "/usr/lib/python2.5/site-packages/matplotlib/afm.py", line 76,
> in _sanity_check
> raise RuntimeError('Not an AFM file')
> RuntimeError: Not an AFM file
>
> Is this a bug? Shouldn't matplotlib try to get an AFM default font,
> rather than a ttf font, in case it cannot find what I'm looking for
> (with ps.useafm=True)? In my case I put the font Arial but matplotlib
> was not able to find it. Some details: I'm using ubuntu hardy and
> matplotlib version 0.98.3. backend PS to generated eps files. My
> decision to use AFM is to reduce the file sizes.
>
> Kind regards,
> Marco
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 

On Aug 22, 2008, at 5:08 AM, Jeff Whitaker wrote:
> Zane Selvans wrote:
>> I'm drawing several hundred lines at a time, each consisting of 
>> 10-100 points, and it takes a couple of minutes for them all to 
>> display, which makes me think I must be doing something stupid.
>>
> Zane: You can set up a LineCollection like this
>
> lcoll = LineCollection(segments)
>
> then add it to the current axes
>
> ax = pylab.gca()
> ax.add_collection(lcoll)
>
> (instead of using the Basemap plot method).
It turned out that just adding:
interactive(False)
before the drawing loop, and
show()
after it, sped things up tremendously. Instead of 5 minutes, it now 
takes about 2 seconds for the map to render. Which, ironically, is 
much more "interactive" :)
Thanks for the pointer to LineCollection though. I'd never seen that.
Zane
--
Zane Selvans
Amateur Earthling
http://zaneselvans.org
za...@id...
303/815-6866
PGP Key: 55E0815F

Hi,
 I have a problem with the loglog function with ls='steps' (matplotlib 
version: 0.98.3). When I try to plot arrays with loglog(x,y,ls='steps'), 
nothing is drawn whereas loglog(x,y) and plot(x,y,ls='steps'):
x=[10.,20.,30.,40.,50.]
y=[0.09,0.02,0.02,0.03,0.01]
pylab.loglog(x,y) #it works
pylab.loglog(x,y,ls='steps') #nothing is drawn...
pylab.plot(x,y,ls='steps') #it works
 Is there something I can do to use loglog with ls='steps'?
thanks in advance,
Nicolas

On Fri, Aug 15, 2008 at 10:39 AM, CliftonH <tf...@we...> wrote:
> I'm plotting data that I get out of a database, which works fine. The only
> thing that doesn't work, although it works with any other simple plot, is
> the xlim()- and ylim()-function.
> ylim(-9,13)
> xlim(-2,25)
> axhline(color='k')
Here is your problem -- all matplotlib plotting functions trigger the
autoscaling functionality, so your previous xlim/ylim setting is being
overridden. To prevent this, either call xlim/ylim *after* the
plotting function, or turn autoscaling off
 gca().set_autoscale_on(False)
JDH