Define aspect ratio when using twinx

Question

How do I set the aspect ratio for a plot that uses twinx?

---

Below, I provide three examples:

1. without twinx to highlight how I set the aspect ratio
2. with only twinx to illustrate how the previously defined aspect ratio is reset
3. my attempt to define the aspect ratio for both y-axis while using twinx (which does not work)

I am using matplotlib 1.3.0
The following script illustrates how I set the aspect ratio of a simple plot:

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(0,1.6,50) + 50.0

fig = plt.figure()

ax = fig.add_subplot(111)

XLIM = [50.0, 51.6]
YLIM = [0.0, 1.1, 0.0, 11.0] 

ax.plot(x,np.sin(x-50.0),'b')

ax.set_xlim([XLIM[0], XLIM[1]])
ax.set_ylim([YLIM[0], YLIM[1]])

ax.set_xticks(np.arange(XLIM[0], XLIM[1], 0.2))
ax.set_yticks(np.arange(YLIM[0], YLIM[1]+0.1, 0.1)[:-1])

ax.grid(True,which='major',linestyle='solid')

ax.set_aspect((XLIM[1]-XLIM[0])/(YLIM[1]-YLIM[0]))

plt.show()

The figure this produces has a defined aspect ratio:

However, if I want to add an additional y-axis using the twinx() function, the aspect ratio is reset:

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(0,1.6,50) + 50.0

fig = plt.figure()

ax = fig.add_subplot(111)
ax2 = ax.twinx()

XLIM = [50.0, 51.6]
YLIM = [0.0, 1.1, 0.0, 11.0] 

ax.plot(x,np.sin(x-50.0),'b')

ax.set_xlim([XLIM[0], XLIM[1]])
ax.set_ylim([YLIM[0], YLIM[1]])

ax.set_xticks(np.arange(XLIM[0], XLIM[1], 0.2))
ax.set_yticks(np.arange(YLIM[0], YLIM[1]+0.1, 0.1)[:-1])

ax.grid(True,which='major',linestyle='solid')

ax.set_aspect((XLIM[1]-XLIM[0])/(YLIM[1]-YLIM[0]))

plt.show()

Resulting in this figure (in which the previously defined y-limits on the left-hand-side are reset):

Now, trying to specify the aspect ratio again does not yield the expected result:

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(0,1.6,50) + 50.0

fig = plt.figure()

ax = fig.add_subplot(111)
ax2 = ax.twinx()

XLIM = [50.0, 51.6]
YLIM = [0.0, 1.1, 0.0, 11.0] 

ax.plot(x,np.sin(x-50.0),'b')
ax2.plot(x,np.cos(x-50.0)*10.,'r')

ax.set_xlim([XLIM[0], XLIM[1]])
ax.set_ylim([YLIM[0], YLIM[1]])
ax2.set_ylim([YLIM[2], YLIM[3]])

ax.set_xticks(np.arange(XLIM[0], XLIM[1], 0.2))
ax.set_yticks(np.arange(YLIM[0], YLIM[1]+0.1, 0.1)[:-1])
ax2.set_yticks(np.arange(YLIM[2], YLIM[3]+1.0, 1.0))

ax.grid(True,which='major',linestyle='solid')

ax.set_aspect((XLIM[1]-XLIM[0])/(YLIM[1]-YLIM[0]))
ax2.set_aspect((XLIM[1]-XLIM[0])/(YLIM[3]-YLIM[2]))

plt.show()

This produces:

What am I missing in order to generate a plot like the first one, but with two y-axis of different scales? How do I define the aspect ratio when using twinx?

Peele

Answer 1

Have a look at the adjustable parameter of the axes. This controls whether the data limits or the bounding rectangle's shape are changed when the aspect/limits are changed.

Normally, the options are "box" and "datalim", but there's a special case for shared axes. In your case, you want adjustable='box-forced'.

As a quick example (I took the liberty of simplifying things slightly, feel free to use the separate set_foo methods if they're clearer to you. If you take that route, the method is ax.set_adjustable.):

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(0,1.6,50) + 50.0

fig, ax = plt.subplots()
ax2 = ax.twinx()

XLIM = [50.0, 51.6]
YLIM = [0.0, 1.1, 0.0, 11.0]

ax.plot(x,np.sin(x-50.0),'b')
ax2.plot(x,np.cos(x-50.0)*10.,'r')

ax.set(adjustable='box-forced',
       xlim=XLIM, ylim=YLIM[:2],
       xticks=np.arange(XLIM[0], XLIM[1], 0.2),
       yticks=np.arange(YLIM[0], YLIM[1]+0.1, 0.1)[:-1],
       aspect=(XLIM[1]-XLIM[0])/(YLIM[1]-YLIM[0]))

ax2.set(adjustable='box-forced',
        ylim=YLIM[2:],
        yticks=np.arange(YLIM[2], YLIM[3]+1.0, 1.0),
        aspect=(XLIM[1]-XLIM[0])/(YLIM[3]-YLIM[2]))

ax.grid(True, which='major',linestyle='solid')

plt.show()

Nari

Unfortunately, box-forced is not supported in new versions of matplotlib, it seems there is not any direct solution for that. There is an indirect one in another post of SO. A direct one will be much better if it could. – Cinelli

Answer 2

Matplotlib 3.1 removes the option 'box-forced'. Joe's solution now yields a ValueError: 'box-forced' is not a valid value for adjustable; supported values are 'box', 'datalim'. Workarounds can be found at Define aspect ratio when using twinx in new version of matplotlib.

Raven