Re: [Matplotlib-users] Large figure sizes get squashed or clipped?

Scrollbars seem like the only reasonable solution I can think of.  Of 
course, this will require adding support in all the backends -- not a 
small undertaking.

I've filed a feature request for this here:

https://sourceforge.net/tracker/?func=detail&aid=2906107&group_id=80706&atid=560723

Maybe if we can select a developer for each backend it won't be too much 
work.  Any takers?

Mike

On 11/28/2009 12:03 AM, Michiel de Hoon wrote:
> I think the best solution to this problem is to have scroll bars in the figure window if needed. Depending on the size of the figure (in physical units), the size of the window containing the figure, and the zoom factor the scroll bars would appear or disappear. This would require some modifications to all GUI backends.
>
> The same problem occurs in the Mac OS X backend, where the figure window size can be smaller than the requested size if the requested size is larger than the monitor size. See bug 2891502:
>
> http://sourceforge.net/tracker/?func=detail&atid=560720&aid=2891502&group_id=80706
>
> Any comments, opinions?
>
>
> --Michiel.
>
> --- On Wed, 11/25/09, do...@us...<do...@us...>  wrote:
>
>    
>> From: do...@us...<do...@us...>
>> Subject: Re: [Matplotlib-users] Large figure sizes get squashed or clipped?
>> To: "Michael Droettboom"<md...@st...>
>> Cc: mat...@li...
>> Date: Wednesday, November 25, 2009, 7:43 PM
>> Yes, I should have mentioned that;
>> saving an image works fine.  But
>> then if I want to display it 1:1 with figimage() or such, I
>> can't.  :(
>>   Silly GUIs, not wanting to display a window larger than my
>> screen....
>>
>>
>>
>> On Wed, Nov 25, 2009 at 9:43 AM, Michael Droettboom<md...@st...>
>> wrote:
>>      
>>> There may be a limitation on window sizes in the
>>>        
>> various GUI backends.  Have
>>      
>>> you tried using the non-GUI backend (agg), and now
>>>        
>> "show"ing it, but just
>>      
>>> using "savefig"?
>>>
>>> Mike
>>>
>>> do...@us...
>>>        
>> wrote:
>>      
>>>> I'm trying to make a 10 inch wide by 30 inch high,
>>>>          
>> 72 dpi figure and
>>      
>>>> display it interactively.  Matplotlib seems to
>>>>          
>> squash the height for
>>      
>>>> anything over a certain size, depending on the
>>>>          
>> backend:
>>      
>>>>
>>>>          
>> ---------------------------------------------------
>>      
>>>> #!/usr/bin/env python
>>>>
>>>> import sys, os, matplotlib
>>>> matplotlib.use('TkAgg')
>>>> import matplotlib.pyplot as plt
>>>>
>>>> print os.uname()
>>>> print sys.version
>>>> print matplotlib.__version__
>>>> print
>>>>
>>>> f = plt.figure(figsize=(10,30), dpi=72)
>>>> print "figheight before show(): %f" %
>>>>          
>> f.get_figheight()
>>      
>>>> plt.show()
>>>> print "figheight after show():  %f" %
>>>>          
>> f.get_figheight()
>>      
>>>>          
>> --------------------------------------------------
>>      
>>>> Prints this:
>>>>
>>>> ==========================
>>>> ('Linux', 'prime', '2.6.31-14-generic',
>>>>          
>> '#48-Ubuntu SMP Fri Oct 16
>>      
>>>> 14:05:01 UTC 2009', 'x86_64')
>>>> 2.6.4 (r264:75706, Nov  2 2009, 14:44:17)
>>>> [GCC 4.4.1]
>>>> 0.99.0
>>>>
>>>> figheight before show(): 30.000000
>>>> figheight after show():  22.027778
>>>> ==========================
>>>>
>>>> Tk squashes heights over 12 inches (the heights do
>>>>          
>> get larger as you
>>      
>>>> request larger figures, but not as large as what
>>>>          
>> you request); GTK
>>      
>>>> goes up to 11; and Qt4Agg only goes up to 7.3.
>>>>          
>>   WX seems to be the
>>      
>>>> only backend that will give me a 30 inch figure,
>>>>          
>> but even then I have
>>      
>>>> to manually resize the window to make it fit.
>>>>          
>>   This happens in scripts
>>      
>>>> with pyplot, in ipython with or without -pylab,
>>>>          
>> and via the matplotlib
>>      
>>>> API.
>>>>
>>>> Is there some limitation on figure sizes?
>>>>
>>>>
>>>>
>>>>          
>> ------------------------------------------------------------------------------
>>      
>>>> Let Crystal Reports handle the reporting - Free
>>>>          
>> Crystal Reports 2008
>>      
>>>> 30-Day trial. Simplify your report design,
>>>>          
>> integration and deployment - and
>>      
>>>> focus on what you do best, core application
>>>>          
>> coding. Discover what's new with
>>      
>>>> Crystal Reports now.  http://p.sf.net/sfu/bobj-july
>>>> _______________________________________________
>>>> 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
>>>
>>>
>>>        
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal
>> Reports 2008 30-Day
>> trial. Simplify your report design, integration and
>> deployment - and focus on
>> what you do best, core application coding. Discover what's
>> new with
>> Crystal Reports now.  http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>>      
>
>
>