Re: [Matplotlib-users] pylab axis query and possible bug
Beautiful!
Many thanks John.
Gary R.
John Hunter wrote:
> You can manually turn off autoscaling on the axes instance with the
> following, and both scatter and plot should then work as you want.
>
> ax1 = subplot(121)
> axis('off')
> ax1.imshow(rand(20,20))
> ax2 = subplot(122)
> axis('off')
> ax2.imshow(rand(20,20))
>
> ax1.set_autoscale_on(False)
> #ax1.scatter([5,10],[5,10])# note 2
> ax1.plot([5,10],[5,10], 'o') # note 3
> show()
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] pylab axis query and possible bug
On Dec 21, 2007 11:50 AM, John Hunter <[EMAIL PROTECTED]> wrote:
> On Dec 20, 2007 9:22 PM, Gary Ruben <[EMAIL PROTECTED]> wrote:
> > Hi listees,
> >
> > I often generate plots using the pylab interface plot() function to
> > overlay an imshow() image. The minimal script below demonstrates a
> > problem, which may be a bug, or may be a deliberate change introduced
> > into mpl 0.91.1. It works fine with mpl 0.90.1 but gives a traceback
> > with 0.91.1 - it seems not to be happy with the subplot limits.
> > Commenting out the "note 1" line lets it run and demonstrates my real
> > question. With scatter(), the first subplot doesn't rescale, but if line
> > "note 2" is commented out and "note 3" is uncommented, it rescales. How
> > do I prevent the rescaling? I prefer plot() instead of scatter() in this
> > case because of the plot origin.
You can manually turn off autoscaling on the axes instance with the
following, and both scatter and plot should then work as you want.
ax1 = subplot(121)
axis('off')
ax1.imshow(rand(20,20))
ax2 = subplot(122)
axis('off')
ax2.imshow(rand(20,20))
ax1.set_autoscale_on(False)
#ax1.scatter([5,10],[5,10])# note 2
ax1.plot([5,10],[5,10], 'o') # note 3
show()
> >
>
> Hey Gary, thanks for the resend.
>
> On your first question vis-a-vis subplots_adjust, it looks like the
> validate limits are wrong for these params. I'm CC-ing Darren because
> he has done the most work recently for rc param validation.
>
> They currently read:
>
> 'figure.subplot.left' : [0.125, ValidateInterval(0, 1,
> closedmin=False, closedmax=False)],
> 'figure.subplot.right' : [0.9, ValidateInterval(0, 1,
> closedmin=False, closedmax=False)],
> 'figure.subplot.bottom' : [0.1, ValidateInterval(0, 1,
> closedmin=False, closedmax=False)],
> 'figure.subplot.top': [0.9, ValidateInterval(0, 1,
> closedmin=False, closedmax=False)],
> 'figure.subplot.wspace' : [0.2, ValidateInterval(0, 1,
> closedmin=False, closedmax=True)],
> 'figure.subplot.hspace' : [0.2, ValidateInterval(0, 1,
> closedmin=False, closedmax=True)],
>
> and I think they should read:
>
>
> 'figure.subplot.left' : [0.125, ValidateInterval(0, 1,
> closedmin=True, closedmax=True)],
> 'figure.subplot.right' : [0.9, ValidateInterval(0, 1,
> closedmin=True, closedmax=True)],
> 'figure.subplot.bottom' : [0.1, ValidateInterval(0, 1,
> closedmin=True, closedmax=True)],
> 'figure.subplot.top': [0.9, ValidateInterval(0, 1,
> closedmin=True, closedmax=True)],
> 'figure.subplot.wspace' : [0.2, ValidateInterval(0, 1,
> closedmin=True, closedmax=False)],
> 'figure.subplot.hspace' : [0.2, ValidateInterval(0, 1,
> closedmin=True, closedmax=False)],
>
> That is, 0 and 1 should be legal for all values except the max of the
> wspace and hspace. Darren, unless I am missing something, I'll commit
> these changes.
>
> I'll look at your second question in a followup
>
> JDH
>
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] pylab axis query and possible bug
On Dec 20, 2007 9:22 PM, Gary Ruben <[EMAIL PROTECTED]> wrote: > Hi listees, > > I often generate plots using the pylab interface plot() function to > overlay an imshow() image. The minimal script below demonstrates a > problem, which may be a bug, or may be a deliberate change introduced > into mpl 0.91.1. It works fine with mpl 0.90.1 but gives a traceback > with 0.91.1 - it seems not to be happy with the subplot limits. > Commenting out the "note 1" line lets it run and demonstrates my real > question. With scatter(), the first subplot doesn't rescale, but if line > "note 2" is commented out and "note 3" is uncommented, it rescales. How > do I prevent the rescaling? I prefer plot() instead of scatter() in this > case because of the plot origin. > Hey Gary, thanks for the resend. On your first question vis-a-vis subplots_adjust, it looks like the validate limits are wrong for these params. I'm CC-ing Darren because he has done the most work recently for rc param validation. They currently read: 'figure.subplot.left' : [0.125, ValidateInterval(0, 1, closedmin=False, closedmax=False)], 'figure.subplot.right' : [0.9, ValidateInterval(0, 1, closedmin=False, closedmax=False)], 'figure.subplot.bottom' : [0.1, ValidateInterval(0, 1, closedmin=False, closedmax=False)], 'figure.subplot.top': [0.9, ValidateInterval(0, 1, closedmin=False, closedmax=False)], 'figure.subplot.wspace' : [0.2, ValidateInterval(0, 1, closedmin=False, closedmax=True)], 'figure.subplot.hspace' : [0.2, ValidateInterval(0, 1, closedmin=False, closedmax=True)], and I think they should read: 'figure.subplot.left' : [0.125, ValidateInterval(0, 1, closedmin=True, closedmax=True)], 'figure.subplot.right' : [0.9, ValidateInterval(0, 1, closedmin=True, closedmax=True)], 'figure.subplot.bottom' : [0.1, ValidateInterval(0, 1, closedmin=True, closedmax=True)], 'figure.subplot.top': [0.9, ValidateInterval(0, 1, closedmin=True, closedmax=True)], 'figure.subplot.wspace' : [0.2, ValidateInterval(0, 1, closedmin=True, closedmax=False)], 'figure.subplot.hspace' : [0.2, ValidateInterval(0, 1, closedmin=True, closedmax=False)], That is, 0 and 1 should be legal for all values except the max of the wspace and hspace. Darren, unless I am missing something, I'll commit these changes. I'll look at your second question in a followup JDH - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users
