Complete Guide to Saving Individual Subplots in Matplotlib

Introduction

In data visualization workflows, there is often a need to save individual subplots from multi-subplot figures as separate image files. Matplotlib, as Python's most popular plotting library, provides flexible figure saving capabilities, but saving individual subplots requires specific technical approaches.

Core Method: The bbox_inches Parameter

Matplotlib's savefig function accepts a bbox_inches parameter that can specify the region of the image to save. By obtaining the subplot's bounding box and transforming it to the correct coordinate system, individual subplot saving can be achieved.

Basic Implementation Steps

First, create a figure containing multiple subplots:

import matplotlib.pyplot as plt
import numpy as np

fig = plt.figure()
ax1 = fig.add_subplot(2, 1, 1)
ax1.plot(range(10), 'b-')

ax2 = fig.add_subplot(2, 1, 2)
ax2.plot(range(20), 'r^')

Obtaining Subplot Bounding Box

Use the get_window_extent() method to get the subplot's bounding box on the canvas, then transform it to inches using coordinate conversion:

# Get the second subplot's bounding box and transform coordinates
extent = ax2.get_window_extent().transformed(fig.dpi_scale_trans.inverted())

# Save the complete figure
fig.savefig('full_figure.png')

# Save only the second subplot
fig.savefig('ax2_figure.png', bbox_inches=extent)

Bounding Box Expansion

If padding is needed around the subplot, use the expanded method:

# Expand by 10% in x-direction and 20% in y-direction
fig.savefig('ax2_figure_expanded.png', bbox_inches=extent.expanded(1.1, 1.2))

Advanced Method: Complete Bounding Box Calculation

For cases requiring inclusion of axis labels and titles, use more precise bounding box calculation methods:

from matplotlib.transforms import Bbox

def full_extent(ax, pad=0.0):
    """Get complete bounding box including all elements"""
    ax.figure.canvas.draw()
    items = ax.get_xticklabels() + ax.get_yticklabels()
    items += [ax, ax.title]
    bbox = Bbox.union([item.get_window_extent() for item in items])
    return bbox.expanded(1.0 + pad, 1.0 + pad)

# Save using complete bounding box
extent = full_extent(ax2).transformed(fig.dpi_scale_trans.inverted())
fig.savefig('ax2_complete.png', bbox_inches=extent)

Understanding Coordinate Systems

Matplotlib uses multiple coordinate systems:

• Data coordinates: Original coordinates of the plotted data
• Axis coordinates: Coordinates relative to the axis (0-1 range)
• Figure coordinates: Coordinates relative to the entire figure
• Display coordinates: Screen coordinates in pixels

Proper coordinate transformation is key to successful subplot saving.

Practical Application Recommendations

In practical applications, it is recommended to:

1. Ensure the figure is fully rendered before calling savefig
2. Adjust bounding box expansion parameters based on specific needs
3. Test different DPI settings for optimal image quality
4. Consider using tight_layout or constrained_layout to optimize subplot arrangement

Conclusion

By properly utilizing the bbox_inches parameter and bounding box calculations, individual subplots from Matplotlib figures can be effectively saved as separate image files. This approach not only works for simple subplot extraction but can also accommodate complex saving requirements through custom bounding box calculations.