matplotlib-users Mailing List for matplotlib

On 2/26/07, Mark Bakker <ma...@gm...> wrote:
> ax = gca()
> if len(args)==0:
> aspect = ax.get_aspect()
> elif len(args)==1:
> aspect = args[0]
> adjustable = popd(kwargs,'adjustable','box')
> anchor = popd(kwargs,'anchor','C')
> fixlim = popd(kwargs,'fixlim',True)
Whenever you are mutating a kwargs dictionary, eg with popd, you
should first copy it. The user may be creating, saving, and resuing a
kwarg dict
 myaspectprops = dict(adjustable='box', anchor='NW')
 # some code
 aspect(**myaspectprops)
 # some more code
 aspect(**myaspectprops)
This will fail if you pop off of the kwargs in the aspect function w/o
first copying it. So always do
 kwargs = kwargs.copy()
before calling popd and friends.
This is covered in the "CODING_GUIDE" document in the svn repository.
JDH

I installed gsview and gs and tried it. AFAIK, gs didn't seen to complain=
. I also tried GIMP. While it didn't complain, the picture displayed is=
 horribly bad.=0A=0A> =0A> An "invalidrestore" error signals an improper re=
store.=0A> E.g., a string, dictionary, or procedure is left on the =0A> sta=
ck that needs to be discarded before restore.=0A> Probably a dictionary wit=
h no 'end'.=0A> =0A> Two obvious possibilities:=0A> =0A> - there is an inva=
lid restore. This should be fixed.=0A> - something is wrong with the Word =
interpreter (quite =0A> possible given its history, but perhaps surprisin=
g=0A> for this particular problem)=0A> =0A> Does GhostScript report the s=
ame error?=0A> =0A> fwiw,=0A> Alan Isaac=0A=0A =0A--=0AJohn Henry=0A=0A

Switching to PDF works (eventually - see below) but bring with it yet anoth=
er set of problems.=0A=0AFirst of all, I had to fix backend_pdf.py for MPL =
to work (w Python 2.3, anyway).=0A=0ATwo problems:=0A=0Aa) 2.3 doesn't have=
 the sorted function - it uses a .sort() function. So, I had to change lin=
e 487 from:=0A=0Achars=3Dsorted(charmap.keys())=0A=0Ato:=0A=0Achars =3D cha=
rmap.keys()=0Achars.sort()=0A=0Ab) No update() function (line 396)=0A=0ASo,=
 instead of:=0A xobjects.update([(name, value[0]) for (name, value) =
in self.markers.items()])=0A=0AHad to do:=0A=0A for (name, value) in=
 self.markers.items():=0A xobjects[name]=3Dvalue[0]=0A=0AAfter t=
hat, I got my pdf file. However, if I print the PDF directly, works fine b=
ut when I create an object link to the PDF and print from inside Word, the =
printout is degraded (kind of fuzzy with texts).=0A=0ASo, for a temporary w=
ork around, I can use PDF instead of EPS and hopefully that problem will ge=
t fixed someday.=0A=0A> -----Original Message-----=0A> From: matplotlib-use=
rs-...@li... =0A> [mailto:matplotlib-users-bounces@list=
s.sourceforge.net] On =0A> Behalf Of Jouni K. Sepp=E4nen=0A> Sent: Monday, =
February 26, 2007 9:56 AM=0A> To: mat...@li...=0A=
> Subject: Re: [Matplotlib-users] invalidrestore=0A> =0A> =0A> kc106_2005-m=
atp...@ya... writes:=0A> =0A> > Back to the subject at hand. Using P=
NG files works but =0A> brings with it =0A> > other problems and so I would=
 really appreciate a =0A> resolution to this =0A> > "invalidrestore" issue.=
 I can't use PDF because I have to =0A> embedd the =0A> > plots in a Word d=
ocument. Not sure what SVG is (I'll look it up).=0A> =0A> In my (admittedly=
 limited) experience Word handles pdf files =0A> much better than eps files=
. I just tried (in MS Word 2004 for =0A> Mac) Insert / Picture / From File =
and selected a file =0A> produced by Matplotlib's pdf backend, and Word see=
ms to embed =0A> it just fine. (I can't test printing right now, though.)=
=0A> =0A> -- =0A> Jouni K. Sepp=E4nen=0A> http://www.iki.fi/jks=0A=0A =0A--=
=0AJohn Henry=0A=0A

Eric, list -
Here is the new aspect function for possible inclusion in pylab.
It works great and with the default values for the kwargs, it works exactly
the way I think it is useful for a combination of contouring and plotting.
What do you think, should we include this?
Mark
def aspect(*args, **kwargs):
 """
 Set/Get the aspect ratio (ylength/xlength) of the current axis
 Aspect ratio is defined as unit-length-along-y-axis /
unit-length-along-x-axis
 If no arguments are passed, the current aspect ratio is returned
 Aspect ratio may be met by changing size of axes while keeping data
limits fixed (called 'box'),
 or by changing data limits while keeping the lengths of the axes fixed
(called 'datalim')
 One point remains fixed, which is called the anchor, for example the
center (called 'C')
 Autoscaling may be turned on (limits may change upon next plotting
command)
 or off (limits will remain fixed)
 Keyword arguments:
 adjustable: 'box' (default), 'datalim'
 anchor: 'C' (default), 'SW', 'S', 'SE', 'E', 'NE', 'N', 'NW', 'W'
 fixlim: True (default), False
 """
 ax = gca()
 if len(args)==0:
 aspect = ax.get_aspect()
 elif len(args)==1:
 aspect = args[0]
 adjustable = popd(kwargs,'adjustable','box')
 anchor = popd(kwargs,'anchor','C')
 fixlim = popd(kwargs,'fixlim',True)
 ax.set_aspect(aspect,adjustable,anchor)
 if fixlim:
 ax.set_autoscale_on(False)
 else:
 ax.set_autoscale_on(True)
 ax.autoscale_view()
 draw_if_interactive()
 return aspect

Lionel Roubeyrie <lro...@li...> writes:
> Fonts are BitStream VeraSans-Roman, TrueType.
Then this is not the usual font problem where fonts get converted
either into Type 3 or into raw drawing commands. Can you put example
files (both eps and pdf) on some website so we can try to understand
what is wrong with them?
-- 
Jouni K. Seppänen
http://www.iki.fi/jks
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An "invalidrestore" error signals an improper restore.
E.g., a string, dictionary, or procedure is left on the 
stack that needs to be discarded before restore.
Probably a dictionary with no 'end'.
Two obvious possibilities:
- there is an invalid restore. This should be fixed.
- something is wrong with the Word interpreter (quite 
 possible given its history, but perhaps surprising
 for this particular problem)
Does GhostScript report the same error?
fwiw,
Alan Isaac

kc1...@ya... writes:
> Back to the subject at hand. Using PNG files works but brings with it
> other problems and so I would really appreciate a resolution to this
> "invalidrestore" issue. I can't use PDF because I have to embedd the
> plots in a Word document. Not sure what SVG is (I'll look it up).
In my (admittedly limited) experience Word handles pdf files much
better than eps files. I just tried (in MS Word 2004 for Mac) Insert /
Picture / From File and selected a file produced by Matplotlib's pdf
backend, and Word seems to embed it just fine. (I can't test printing
right now, though.)
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Thanks for the answer, John.=0A=0AUp to this point, the other software pack=
age we use produces EPS files only and so I've been using it this way for a=
 while. May not be the best way but I've tried other EPS viewers and so fa=
r Word actually is the best way since I have to embedd the plots into Word =
reports.=0A=0ABack to the subject at hand. Using PNG files works but bring=
s with it other problems and so I would really appreciate a resolution to t=
his "invalidrestore" issue. I can't use PDF because I have to embedd the p=
lots in a Word document. Not sure what SVG is (I'll look it up).=0A=0AReg=
ards,=0A=0A=0A> =0A> Using PS/EPS in Word or any office product is a pretty=
 =0A> unusual combination. My experience is that they are not =0A> support=
ed s well as other image formats. Will PNG work for =0A> you? That is wha=
t I usually use for Office. You may also =0A> consider PDF or SVG if you n=
eed vector graphics.=0A> =0A> JDH=0A>=0A =0A--=0AJohn Henry=0A=0A

Hi Nicolas,
I just want to set xlabels with :
MOIS=3D['janvier',=20
u'f=E9vrier', 'mars', 'avril', 'mai', 'juin', 'juillet', 'aout',
'septembre', 'octobre', 'novembre', u'd=E9cembre']
resulting in 'f?vrier' and 'd?cembre' in the pdf file.
Le Lundi 26 F=E9vrier 2007 17:01, Nicolas Grilly a =E9crit=A0:
> Hello Lionel,
>
> On 2/26/07, Lionel Roubeyrie <lro...@li...> wrote:
> > sorry to resend this problem, I didn't have any response and the problem
> > always exists, all accentued characters are set to "?" when I save
> > figures in pdf format. Does someone have a solution to that?
> > MPL 0.87.7 on Edgy Eft
>
> Can you send me a little test script showing what you are trying to
> do? I'll run it with the patched PDF backend. I use this backend daily
> to produce PDF files containing accentuated characters.
>
> Thanks,
>
> Nicolas
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
> your opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=3Djoin.php&p=3Dsourceforge&CID=3D=
DEVDEV
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
=2D-=20
Lionel Roubeyrie - lro...@li...
Chag=E9 d'=E9tudes et de maintenance
LIMAIR - la Surveillance de l'Air en Limousin
http://www.limair.asso.fr

Hi,
=46onts are BitStream VeraSans-Roman, TrueType.
Le Lundi 26 F=E9vrier 2007 16:43, Jouni K. Sepp=E4nen a =E9crit=A0:
> What exactly is the problem with the converted pdf file? Try opening
> it in Adobe Reader and viewing Document Properties / Fonts. Do you see
> any "Type 3" fonts, and what are their names?
=2D-=20
Lionel Roubeyrie - lro...@li...
Chag=E9 d'=E9tudes et de maintenance
LIMAIR - la Surveillance de l'Air en Limousin
http://www.limair.asso.fr

Hello Lionel,
On 2/26/07, Lionel Roubeyrie <lro...@li...> wrote:
> sorry to resend this problem, I didn't have any response and the problem
> always exists, all accentued characters are set to "?" when I save figures in
> pdf format. Does someone have a solution to that?
> MPL 0.87.7 on Edgy Eft
Can you send me a little test script showing what you are trying to
do? I'll run it with the patched PDF backend. I use this backend daily
to produce PDF files containing accentuated characters.
Thanks,
Nicolas

Lionel Roubeyrie <lro...@li...> writes:
> Le Lundi 26 Février 2007 14:20, Darren Dale a écrit :
>> What are you using to do the conversion? epstopdf should maintain the
>> integrity of the fonts.
> epstopdf should maintain... but not, it doesn't do that! 
What exactly is the problem with the converted pdf file? Try opening
it in Adobe Reader and viewing Document Properties / Fonts. Do you see
any "Type 3" fonts, and what are their names?
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

On 2/25/07, kc1...@ya...
<kc1...@ya...> wrote:
> After doing some investigation, I discovered that this problem is reproducible using sample programs such as axes_demo.py. Here are the steps I took:
>
> a) Run axes_demo.py
> b) Save the plot as eps file
> c) Start Word
> d) Insert the eps file
> e) Print
Using PS/EPS in Word or any office product is a pretty unusual
combination. My experience is that they are not supported s well as
other image formats. Will PNG work for you? That is what I usually
use for Office. You may also consider PDF or SVG if you need vector
graphics.
JDH

Le Lundi 26 F=E9vrier 2007 14:20, Darren Dale a =E9crit=A0:
> What are you using to do the conversion? epstopdf should maintain the
> integrity of the fonts.
epstopdf should maintain... but not, it doesn't do that!=20
=2D-=20
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

On Sunday 25 February 2007 12:41:58 pm kc1...@ya... wrote:
> After doing some investigation, I discovered that this problem is
> reproducible using sample programs such as axes_demo.py. Here are the
> steps I took:
>
> a) Run axes_demo.py
> b) Save the plot as eps file
> c) Start Word
> d) Insert the eps file
> e) Print
>
> I got a 2 page print: 1 for the plot, and 1 for the error message
> invalidrestore.
>
> Could somebody please try and see if it fails the same on their system?
I don't have Microsoft Word, so I can't test this. Considering how matplotlib 
creates eps files that are compatible with latex, ghostscript, etc., I wonder 
if this is a problem with Word.
Darren

On Monday 26 February 2007 06:53:22 am Lionel Roubeyrie wrote:
> Le Lundi 26 F=E9vrier 2007 12:35, Jouni K. Sepp=E4nen a =E9crit=A0:
> > Does saving the figure as eps and converting to pdf work for you? The
> > pdf backend doesn't support unicode, since I've never been able to
> > really understand how encoding strings in pdf works. (Nicolas Grilly's
> > recent patch fixed something related to encodings, so perhaps he
> > understands them better?)
> >
> > Patches are of course welcome. I'm afraid I'm currently too busy with
> > my day job to do any major hacking on the pdf backend.
>
> Hi Jouni,
> converting eps to pdf works, but fonts are not well rendered, and I need
> very high quality results for our publisher.
What are you using to do the conversion? epstopdf should maintain the=20
integrity of the fonts.

Le Lundi 26 F=E9vrier 2007 12:35, Jouni K. Sepp=E4nen a =E9crit=A0:
> Does saving the figure as eps and converting to pdf work for you? The
> pdf backend doesn't support unicode, since I've never been able to
> really understand how encoding strings in pdf works. (Nicolas Grilly's
> recent patch fixed something related to encodings, so perhaps he
> understands them better?)
>
> Patches are of course welcome. I'm afraid I'm currently too busy with
> my day job to do any major hacking on the pdf backend.
Hi Jouni,
converting eps to pdf works, but fonts are not well rendered, and I need ve=
ry=20
high quality results for our publisher.
=2D-=20
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

Lionel Roubeyrie <lro...@li...> writes:
> sorry to resend this problem, I didn't have any response and the
> problem always exists, all accentued characters are set to "?" when
> I save figures in pdf format. Does someone have a solution to that?
Does saving the figure as eps and converting to pdf work for you? The
pdf backend doesn't support unicode, since I've never been able to
really understand how encoding strings in pdf works. (Nicolas Grilly's
recent patch fixed something related to encodings, so perhaps he
understands them better?)
Patches are of course welcome. I'm afraid I'm currently too busy with
my day job to do any major hacking on the pdf backend.
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Hi,
sorry to resend this problem, I didn't have any response and the problem 
always exists, all accentued characters are set to "?" when I save figures in 
pdf format. Does someone have a solution to that?
MPL 0.87.7 on Edgy Eft
thanks
-- 
Lionel Roubeyrie - lro...@li...
LIMAIR
http://www.limair.asso.fr

Hi all,
Sorry about the steady stream of help requests. I haven't been able to 
figure out a couple of things:
- How can I make my figures and axes transparent by default?
- When I inserted some of my old pdf plots into a latex presentation, to 
my surprise their foreground color had changed from black to the color 
of the text in the presentation. Is there a way to signal to Matplotlib 
that I would like this to happen? Can I make this behavior default?
Many thanks,
Anand Patil

It works! What beeyootiful plots. Man, that is SO MUCH nicer than 
Matlab's stanky old X11 b@#?*&^%. And contrary to something I think I 
remember you saying on an earlier thread, I could swear it's faster than 
TkAgg by a long shot.
Two wrinkles: It pops up a blank plot before showing the first plot the 
user requests, and I can't get it to work in interactive mode. On the 
(good) chance you're aware of those and working on it, sorry.
Thanks,
Anand
Charlie Moad wrote:
> On 2/21/07, Anand Patil <an...@so...> wrote:
>
>>
>> From the matplotlib-devel list:
>>
>> >On 2/14/07, Michiel Jan Laurens de Hoon <mdehoon@c2...> wrote:
>> >> Dear Charles,
>> >>
>> >> I was trying to use your cocoa-agg backend for matplotlib, but it 
>> seems
>> >> to have a problem to read Matplotlib.nib. Opening this nib with
>> >> Interface Builder also gives an error. It appears that the 
>> problem is
>> >> caused by the file keyedobjects.nib in the Matplotlib.nib folder. 
>> If I
>> >> run plutil on keyedobjects.nib, it crashes. The other two files in
>> >> Matplotlib.nib look fine. So I was wondering if it is possible 
>> that the
>> >> keyedobjects.nib file included in matplotlib is damaged. If so, 
>> do you
>> >> have a valid copy of this file?
>> >>
>> >> Many thanks in advance,
>> >>
>> >> --Michiel de Hoon.
>> >
>> >We recently moved the data files, and I think the nib files got
>> >interpreted as text instead of binary files. I grabbed an old copy
>> >from a previous source release and committed them as binary. They
>> >should work now.
>> >
>> >- Charlie
>>
>> I just downloaded
>> /trunk/matplotlib/lib/matplotlib/backends/Matplotlib.nib, rev 3022 on
>> svn. When I tried to use it, this happened:
>>
>> In [1]: figure()
>> 2007年02月21日 15:00:51.611 Python[813] *** NSThread: ignoring exception
>> '*** -[NSKeyedUnarchiver decodeObjectForKey:]: missing class information
>> for object' that raised during delayed perform of target 0x3b308c0 and
>> selector 'startWithBundle:'
>>
>> I wasn't able to open the nib package with the interface builder, and
>> plutil crashed on keyedobjects.nib (I don't know what plutil is supposed
>> to do). I'm using a PPC G4 running OS X 10.4.
>>
>> Thanks much,
>> Anand
>
>
> Sorry it took some time, but I tried committing fresh non-binary nibs.
> Please try again now.

On 2/21/07, Anand Patil <an...@so...> wrote:
>
> From the matplotlib-devel list:
>
> >On 2/14/07, Michiel Jan Laurens de Hoon <mdehoon@c2...> wrote:
> >> Dear Charles,
> >>
> >> I was trying to use your cocoa-agg backend for matplotlib, but it seems
> >> to have a problem to read Matplotlib.nib. Opening this nib with
> >> Interface Builder also gives an error. It appears that the problem is
> >> caused by the file keyedobjects.nib in the Matplotlib.nib folder. If I
> >> run plutil on keyedobjects.nib, it crashes. The other two files in
> >> Matplotlib.nib look fine. So I was wondering if it is possible that the
> >> keyedobjects.nib file included in matplotlib is damaged. If so, do you
> >> have a valid copy of this file?
> >>
> >> Many thanks in advance,
> >>
> >> --Michiel de Hoon.
> >
> >We recently moved the data files, and I think the nib files got
> >interpreted as text instead of binary files. I grabbed an old copy
> >from a previous source release and committed them as binary. They
> >should work now.
> >
> >- Charlie
>
> I just downloaded
> /trunk/matplotlib/lib/matplotlib/backends/Matplotlib.nib, rev 3022 on
> svn. When I tried to use it, this happened:
>
> In [1]: figure()
> 2007年02月21日 15:00:51.611 Python[813] *** NSThread: ignoring exception
> '*** -[NSKeyedUnarchiver decodeObjectForKey:]: missing class information
> for object' that raised during delayed perform of target 0x3b308c0 and
> selector 'startWithBundle:'
>
> I wasn't able to open the nib package with the interface builder, and
> plutil crashed on keyedobjects.nib (I don't know what plutil is supposed
> to do). I'm using a PPC G4 running OS X 10.4.
>
> Thanks much,
> Anand
Sorry it took some time, but I tried committing fresh non-binary nibs.
 Please try again now.