matplotlib-users

[Matplotlib-users] Getting contour positions without interacting with figure/axes

[Thomas R. <tho...@gm...>]

Hello,

I am using matplotlib to show an image using:

fig = figure()
ax  = fig.add_subplot(111)
ax.imshow(image)

After doing this, I want to find the contours for a different image 
(with different dimensions), but I do not want to interact with the 
current figure or axes, I just want to retrieve the set of 
LineCollection objects for the contours. The issue is that if I do

c = contour(image2)

the contours are stored inside c, but at the same time they are plotted 
in the current figure. If I use

ax.contour(image2)

then the contours are not plotted immediately, but the view interval has 
already been changed inside ax.

So essentially, I am wondering if it is possible to retrieve a set of 
LineCollection objects without acting in any way on the current figure/axes.

Thanks for any help,

Thomas

Re: [Matplotlib-users] Getting contour positions without interacting with figure/axes

[Eric F. <ef...@ha...>]

Thomas Robitaille wrote:
> Hello,
> 
> I am using matplotlib to show an image using:
> 
> fig = figure()
> ax  = fig.add_subplot(111)
> ax.imshow(image)
> 
> After doing this, I want to find the contours for a different image 
> (with different dimensions), but I do not want to interact with the 
> current figure or axes, I just want to retrieve the set of 
> LineCollection objects for the contours. The issue is that if I do
> 
> c = contour(image2)
> 
> the contours are stored inside c, but at the same time they are plotted 
> in the current figure. If I use
> 
> ax.contour(image2)
> 
> then the contours are not plotted immediately, but the view interval has 
> already been changed inside ax.

The workaround for now may be to call ax.set_autoscale_on(False) before 
your call to ax.contour.   You could also save the datalim before and 
restore them afterwards.

This sort of request has come up before, though, and the longer-term 
solution might be some refactoring in contour.py.  As it is, everything 
is done when the ContourSet is instantiated; one does not have the 
option of simply calculating the contours, for example.

Eric

> 
> So essentially, I am wondering if it is possible to retrieve a set of 
> LineCollection objects without acting in any way on the current figure/axes.

> 
> Thanks for any help,
> 
> Thomas
> 
> ------------------------------------------------------------------------------
> Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
> -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
> -Strategies to boost innovation and cut costs with open source participation
> -Receive a $600 discount off the registration fee with the source code: SFAD
> http://p.sf.net/sfu/XcvMzF8H
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Re: [Matplotlib-users] Getting contour positions without interacting with figure/axes

[Thomas R. <tho...@gm...>]

Hi Eric,

Thanks for your help! Unfortunately, ax.set_autoscale_on(False) does not 
work - I had already tried this before unsuccessfully. Consider the 
following:

from pylab import *
import numpy as np

def test():
	fig = figure()
	ax = fig.add_subplot(111)
	a = np.zeros((100,100))
	b = np.zeros((10,10))
	ax.imshow(a)
	ax.set_autoscale_on(False)
	ax.contour(b)
	fig.canvas.draw()

Running test() should result in a subplot with the view interval 0 to 
99, but instead goes from 0 to 9.

Is this normal?

Thanks,

Thomas

Eric Firing wrote:
> Thomas Robitaille wrote:
>> Hello,
>>
>> I am using matplotlib to show an image using:
>>
>> fig = figure()
>> ax = fig.add_subplot(111)
>> ax.imshow(image)
>>
>> After doing this, I want to find the contours for a different image
>> (with different dimensions), but I do not want to interact with the
>> current figure or axes, I just want to retrieve the set of
>> LineCollection objects for the contours. The issue is that if I do
>>
>> c = contour(image2)
>>
>> the contours are stored inside c, but at the same time they are
>> plotted in the current figure. If I use
>>
>> ax.contour(image2)
>>
>> then the contours are not plotted immediately, but the view interval
>> has already been changed inside ax.
>
> The workaround for now may be to call ax.set_autoscale_on(False) before
> your call to ax.contour. You could also save the datalim before and
> restore them afterwards.
>
> This sort of request has come up before, though, and the longer-term
> solution might be some refactoring in contour.py. As it is, everything
> is done when the ContourSet is instantiated; one does not have the
> option of simply calculating the contours, for example.
>
> Eric
>
>>
>> So essentially, I am wondering if it is possible to retrieve a set of
>> LineCollection objects without acting in any way on the current
>> figure/axes.
>
>>
>> Thanks for any help,
>>
>> Thomas
>>
>> ------------------------------------------------------------------------------
>>
>> Open Source Business Conference (OSBC), March 24-25, 2009, San
>> Francisco, CA
>> -OSBC tackles the biggest issue in open source: Open Sourcing the
>> Enterprise
>> -Strategies to boost innovation and cut costs with open source
>> participation
>> -Receive a $600 discount off the registration fee with the source
>> code: SFAD
>> http://p.sf.net/sfu/XcvMzF8H
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Re: [Matplotlib-users] Getting contour positions without interacting with figure/axes

[Eric F. <ef...@ha...>]

Thomas Robitaille wrote:
> Hi Eric,
> 
> Thanks for your help! Unfortunately, ax.set_autoscale_on(False) does not 
> work - I had already tried this before unsuccessfully. Consider the 
> following:
> 
> from pylab import *
> import numpy as np
> 
> def test():
>     fig = figure()
>     ax = fig.add_subplot(111)
>     a = np.zeros((100,100))
>     b = np.zeros((10,10))
>     ax.imshow(a)
>     ax.set_autoscale_on(False)
>     ax.contour(b)
>     fig.canvas.draw()
> 
> Running test() should result in a subplot with the view interval 0 to 
> 99, but instead goes from 0 to 9.
> 
> Is this normal?

On my system, it does go from 0-99.  What mpl version are you using?  I 
made the change to using autoscale_view in the ContourSet initializer, 
which obeys the setting by set_autoscale_on, on November 17.  Prior to 
that, xlim and ylim were being set directly, so the workaround if you 
have an earlier version would be to save those values (using 
ax.get_xlim(), ax.get_ylim()) and then restore them with ax.set_xlim etc.

Eric

> 
> Thanks,
> 
> Thomas
> 
> Eric Firing wrote:
>> Thomas Robitaille wrote:
>>> Hello,
>>>
>>> I am using matplotlib to show an image using:
>>>
>>> fig = figure()
>>> ax = fig.add_subplot(111)
>>> ax.imshow(image)
>>>
>>> After doing this, I want to find the contours for a different image
>>> (with different dimensions), but I do not want to interact with the
>>> current figure or axes, I just want to retrieve the set of
>>> LineCollection objects for the contours. The issue is that if I do
>>>
>>> c = contour(image2)
>>>
>>> the contours are stored inside c, but at the same time they are
>>> plotted in the current figure. If I use
>>>
>>> ax.contour(image2)
>>>
>>> then the contours are not plotted immediately, but the view interval
>>> has already been changed inside ax.
>>
>> The workaround for now may be to call ax.set_autoscale_on(False) before
>> your call to ax.contour. You could also save the datalim before and
>> restore them afterwards.
>>
>> This sort of request has come up before, though, and the longer-term
>> solution might be some refactoring in contour.py. As it is, everything
>> is done when the ContourSet is instantiated; one does not have the
>> option of simply calculating the contours, for example.
>>
>> Eric
>>
>>>
>>> So essentially, I am wondering if it is possible to retrieve a set of
>>> LineCollection objects without acting in any way on the current
>>> figure/axes.
>>
>>>
>>> Thanks for any help,
>>>
>>> Thomas
>>>
>>> ------------------------------------------------------------------------------ 
>>>
>>>
>>> Open Source Business Conference (OSBC), March 24-25, 2009, San
>>> Francisco, CA
>>> -OSBC tackles the biggest issue in open source: Open Sourcing the
>>> Enterprise
>>> -Strategies to boost innovation and cut costs with open source
>>> participation
>>> -Receive a $600 discount off the registration fee with the source
>>> code: SFAD
>>> http://p.sf.net/sfu/XcvMzF8H
>>> _______________________________________________
>>> Matplotlib-users mailing list
>>> Mat...@li...
>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>

Re: [Matplotlib-users] Getting contour positions without interacting with figure/axes

[Thomas R. <tho...@gm...>]

Hi Eric,

I'm using matplotlib 0.98 (which comes with the Enthought distribution) 
so that would explain why there was an issue.

Thanks for your help!

Thomas

Eric Firing wrote:
> Thomas Robitaille wrote:
>> Hi Eric,
>>
>> Thanks for your help! Unfortunately, ax.set_autoscale_on(False) does
>> not work - I had already tried this before unsuccessfully. Consider
>> the following:
>>
>> from pylab import *
>> import numpy as np
>>
>> def test():
>> fig = figure()
>> ax = fig.add_subplot(111)
>> a = np.zeros((100,100))
>> b = np.zeros((10,10))
>> ax.imshow(a)
>> ax.set_autoscale_on(False)
>> ax.contour(b)
>> fig.canvas.draw()
>>
>> Running test() should result in a subplot with the view interval 0 to
>> 99, but instead goes from 0 to 9.
>>
>> Is this normal?
>
> On my system, it does go from 0-99. What mpl version are you using? I
> made the change to using autoscale_view in the ContourSet initializer,
> which obeys the setting by set_autoscale_on, on November 17. Prior to
> that, xlim and ylim were being set directly, so the workaround if you
> have an earlier version would be to save those values (using
> ax.get_xlim(), ax.get_ylim()) and then restore them with ax.set_xlim etc.
>
> Eric
>
>>
>> Thanks,
>>
>> Thomas
>>
>> Eric Firing wrote:
>>> Thomas Robitaille wrote:
>>>> Hello,
>>>>
>>>> I am using matplotlib to show an image using:
>>>>
>>>> fig = figure()
>>>> ax = fig.add_subplot(111)
>>>> ax.imshow(image)
>>>>
>>>> After doing this, I want to find the contours for a different image
>>>> (with different dimensions), but I do not want to interact with the
>>>> current figure or axes, I just want to retrieve the set of
>>>> LineCollection objects for the contours. The issue is that if I do
>>>>
>>>> c = contour(image2)
>>>>
>>>> the contours are stored inside c, but at the same time they are
>>>> plotted in the current figure. If I use
>>>>
>>>> ax.contour(image2)
>>>>
>>>> then the contours are not plotted immediately, but the view interval
>>>> has already been changed inside ax.
>>>
>>> The workaround for now may be to call ax.set_autoscale_on(False) before
>>> your call to ax.contour. You could also save the datalim before and
>>> restore them afterwards.
>>>
>>> This sort of request has come up before, though, and the longer-term
>>> solution might be some refactoring in contour.py. As it is, everything
>>> is done when the ContourSet is instantiated; one does not have the
>>> option of simply calculating the contours, for example.
>>>
>>> Eric
>>>
>>>>
>>>> So essentially, I am wondering if it is possible to retrieve a set of
>>>> LineCollection objects without acting in any way on the current
>>>> figure/axes.
>>>
>>>>
>>>> Thanks for any help,
>>>>
>>>> Thomas
>>>>
>>>> ------------------------------------------------------------------------------
>>>>
>>>>
>>>> Open Source Business Conference (OSBC), March 24-25, 2009, San
>>>> Francisco, CA
>>>> -OSBC tackles the biggest issue in open source: Open Sourcing the
>>>> Enterprise
>>>> -Strategies to boost innovation and cut costs with open source
>>>> participation
>>>> -Receive a $600 discount off the registration fee with the source
>>>> code: SFAD
>>>> http://p.sf.net/sfu/XcvMzF8H
>>>> _______________________________________________
>>>> Matplotlib-users mailing list
>>>> Mat...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>>
>