Bug summary

I'm trying to produce a density plot of logarithmically spaced data. Using NonUniformImage used to work before matplotlib version 3.10, where I know exhibit a weirdly scaled axis.

Code for reproduction

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.image import NonUniformImage

x = np.linspace(0, 1, 2)
y = np.geomspace(1e1, 1e5, 3)
z = np.random.uniform(size=(len(x), len(y)))

_, (ax1, ax2) = plt.subplots(ncols=2)

# create the non-uniform image
im = NonUniformImage(ax1, extent=[0, 1, 1e0, 1e6], origin="lower")
im.set_data(x, y, z.T)
ax1.add_image(im)
ax1.set_yscale("log")
ax1.set_ylim(1e0, 1e6)  # omitting this leads to very weird behavior
ax1.set_title("NonUniformImage")

# use pcolormesh to get reference image
ax2.pcolormesh([0, 0.5, 1], [1e0, 1e2, 1e4, 1e6], z.T)
ax2.set_yscale("log")
ax2.set_title("pcolormesh")

Actual outcome

The code above produces the following image

Expected outcome

I expect the left panel to look like the right panel, which was produced using pcolormesh. The reason I don't want to use pcolormesh for these plots is that the quality in PDFs is often much worse (with lines showing up between patches) and the file size can easily be 10 times larger.

Additional information

I recently asked a related question on Stackoverflow, which has some additional details.

Operating system

OS/X

Matplotlib Version

3.10.0

Matplotlib Backend

module://matplotlib_inline.backend_inline

Python version

3.12.8

Jupyter version

7.3.1

Installation

conda