In matplotlib, if you want to alternate between two images while preserving the current view (location and zoom) as defined by user interaction (e.g., panning or zooming with the mouse), you can achieve this by:

1. Storing the current axes limits before switching images.
2. Restoring those limits after updating the image data.

Here’s how you can do it without hardcoding the view in your script, so it respects the user's manual adjustments:

Example Code

import numpy as np
import matplotlib.pyplot as plt

# Create two random images
image1 = np.random.rand(100, 100)
image2 = np.random.rand(100, 100)

fig, ax = plt.subplots()
ax.imshow(image1, cmap='viridis')
plt.title("Press '1' for Image 1, '2' for Image 2")

current_image = image1

def on_key(event):
    global current_image
    # Save current view
    xlim = ax.get_xlim()
    ylim = ax.get_ylim()

    if event.key == '1':
        current_image = image1
    elif event.key == '2':
        current_image = image2
    else:
        return

    # Update image data
    ax.clear()
    ax.imshow(current_image, cmap='viridis')
    ax.set_title(f"Image: {event.key}")

    # Restore view
    ax.set_xlim(xlim)
    ax.set_ylim(ylim)
    fig.canvas.draw()

fig.canvas.mpl_connect('key_press_event', on_key)
plt.show()

How It Works

• ax.get_xlim() and ax.get_ylim() capture the current view (zoom/pan) before switching.
• ax.set_xlim() and ax.set_ylim() restore the view after updating the image.
• The key press event ('1' or '2') triggers the switch.

Notes

• This approach preserves the user's manual zoom/pan between switches.
• Works for any 2D image (e.g., imshow, pcolormesh).
• For interactive backends (e.g., Qt5Agg, TkAgg), this will work as expected.