Note: This is from a stackoverflow question.

Consider

import numpy as np
import matplotlib.pyplot as plt

data = np.zeros((3, 3))

fig = plt.figure()

ax1 = plt.subplot(131)
ax2 = plt.subplot(132)
ax3 = plt.subplot(133)
axes = [ax1, ax2, ax3]

for ax in axes:
    im = ax.imshow(data)

The resulting figure fig has 3 AxesSubplot children, one for each subplot. Their geometry

for child in fig.get_children():
    try:
        print(child.get_geometry())
    except AttributeError:
        pass

(1, 3, 1)
(1, 3, 2)
(1, 3, 3)

makes sense.

Now let's add some color bars:

for ax in axes:
    fig.colorbar(im, ax=ax, orientation='horizontal')

Now the geometries are reported as

(2, 1, 1)
(2, 1, 1)
(2, 1, 1)
(1, 3, 2)
(1, 3, 2)
(1, 3, 2)

All information about how the plots are arranged seems lost.

I guess I would have expected a nested structure: The three plots are children of fig with the ordering
(1, 3, 1), (1, 3, 2), (1, 3, 3), and each of those has two subplots with the ordering (2, 1, 1), (2, 1, 2).