matplotlib, place a transparent subplot on top of another subplot

Question

My goal is to have a plot with a transparent background exactly on top of another plot, but with the y-axis on the other side. However, I do not know how to get the size right.

This code

plt.figure(1, figsize=(9, 3))
plt.subplot(1,1,1)  
plt.plot([1, 2], [3, 2], color='black')
ax2 = plt.axes([0, 0, 1, 1], facecolor='none')
ax2.yaxis.set_label_position("right")
ax2.yaxis.tick_right()
ax2.plot([1, 2], [3.1,2.1])

produces

So apparently, the size of the box needs to be set differently.

In case you are wondering why, here is the backdrop. My goal is to have a plot with 2 y-axes, one on the left and one on the right, where the left y-axis is interrupted and the right one is not. With interrupted I mean like this:

The figure is created with 2 subplots like this:

top_ratio = 1
bot_ratio = 4
gsdict={'height_ratios':[top_ratio, bot_ratio]}
f, (ax1, ax2) = plt.subplots(2, 1, sharex=True, gridspec_kw=gsdict)

However, if I now use twinx to get another y-axis, that new y-axis is only for one of the subplots, but I want it to go all the way from top to bottom. My idea is to create an additional axis with transparent background and add it on top.

Answer 1

You would want to create a new subplot at the same position as the existing ones. Because, by default, all subplots use the same subplot specifications, this is as easy ax3 = fig.add_subplot(111).

import matplotlib.pyplot as plt

gsdict={'height_ratios':[1, 4]}
fig, (ax1, ax2) = plt.subplots(2, 1, sharex=True, gridspec_kw=gsdict)

ax3 = fig.add_subplot(111, label="new subplot", facecolor="none")
ax3.yaxis.set_label_position("right")

ax3.tick_params(left=False, right=True, labelleft=False, labelright=True,
                bottom=False, labelbottom=False)
ax1.get_shared_x_axes().join(ax1,ax3)

# just to see the effect, make spines green
plt.setp(ax3.spines.values(), color="limegreen", linestyle=":")
plt.show()