Matplotlib能否创建带有寄生轴的子图？

Absolutely, Matplotlib does support creating subplots with parasite axes—you just need to make sure each subplot has its own independent host/parasite axis setup. The blank subplots you're seeing are likely because you're reusing the same axis grid structure across both subplots, or not properly initializing the parasite axes for each individual subplot.

Let's walk through a corrected example that creates two subplots, each with their own parasite axis:

from mpl_toolkits.axes_grid1 import host_subplot
import mpl_toolkits.axisartist as AA
import matplotlib.pyplot as plt

# Create a figure with 2 side-by-side subplots
fig = plt.figure(figsize=(12, 5))

# --- Set up first subplot with parasite axis ---
host1 = host_subplot(121, axes_class=AA.Axes, figure=fig)
plt.subplots_adjust(right=0.75)

# Add parasite axis to the first subplot
parasite1 = host1.twinx()
parasite1.axis["right"].toggle(all=True)

# Plot sample data on host and parasite axes
host1.plot([1, 2, 3], [10, 20, 30], label="Host Dataset 1")
parasite1.plot([1, 2, 3], [0.1, 0.2, 0.3], label="Parasite Dataset 1", color="crimson")

# Configure labels and legend for first subplot
host1.set_xlabel("X Values")
host1.set_ylabel("Host Y Axis")
parasite1.set_ylabel("Parasite Y Axis")
host1.legend()

# --- Set up second subplot with parasite axis ---
host2 = host_subplot(122, axes_class=AA.Axes, figure=fig)

# Add parasite axis to the second subplot
parasite2 = host2.twinx()
parasite2.axis["right"].toggle(all=True)

# Plot different sample data on the second subplot
host2.plot([1, 2, 3], [5, 15, 25], label="Host Dataset 2")
parasite2.plot([1, 2, 3], [0.05, 0.15, 0.25], label="Parasite Dataset 2", color="darkgreen")

# Configure labels and legend for second subplot
host2.set_xlabel("X Values")
host2.set_ylabel("Host Y Axis 2")
parasite2.set_ylabel("Parasite Y Axis 2")
host2.legend()

# Adjust layout to prevent label overlap
plt.tight_layout()
plt.show()

Key Fixes & Notes: #

• Each subplot gets its own host_subplot instance, explicitly tied to the figure and its position (using 121 and 122 for 1 row, 2 columns, positions 1 and 2).
• We explicitly toggle the parasite axis visibility with parasite.axis["right"].toggle(all=True) to ensure it renders properly.
• Avoid reusing a single AxesGrid or host axis across multiple subplots—each needs its own independent setup.
• Using plt.subplots_adjust or plt.tight_layout() helps prevent overlapping axis labels and improves readability.

内容的提问来源于stack exchange，提问作者Alex S