matplotlib-users

[Matplotlib-users] inverting y axis?

[Andrew S. <str...@as...>]

Simple question:  how can I plot something such that the y axis is 
"inverted"?  (Increasing downwards?)

(I thought I'd better ask before fubaring my copy of mpl's transAxes, 
transData, bbox code and going insane in the process.)

Cheers!
Andrew

Re: [Matplotlib-users] inverting y axis?

[Andrew S. <str...@as...>]

Andrew Straw wrote:

> Simple question:  how can I plot something such that the y axis is 
> "inverted"?  (Increasing downwards?)

Simple answer:

ymin,ymax= get(gca(),'ylim')
set(gca(),'ylim',[ymax,ymin])

Now, on to determine how to plot an image with origin='upper' with the 
extent going the other way...

Cheers!
Andrew

Re: [Matplotlib-users] inverting y axis?

[Andrew S. <str...@as...>]

Andrew Straw wrote:

>
> Now, on to determine how to plot an image with origin='upper' with the 
> extent going the other way...

Again, a no brainer:

(modifications to image_demo3.py)

w,h = lena.size
extent = 0,w,h,0
figure(figsize=figsize)

im = imshow(lena, origin='upper', aspect='preserve', extent=extent)

Re: [Matplotlib-users] inverting y axis?

[John H. <jdh...@ac...>]

>>>>> "Andrew" == Andrew Straw <str...@as...> writes:

    Andrew> Again, a no brainer:

Glad you found the answer :-)

See also Fernando's matshow command in CVS.  The data scaling with
aspect='preserve' in imshow is currently broken -- the axes
coordinates on the shorter dimension are wrong and because of this
zoom to rect doesn't do what it should.  Fernando's command works
around this problem by creating a figure and axes with the proper
aspect ratio .  Currently, it only works with arrays, but you could
easily patch it to work with pil images as you did for imshow.

Cheers,
JDH