Repository navigation
pcolormesh in log polar coordinates #12803
Description
Bug report
Bug summary
radius range setting does not work for pcolormesh() in log polar coodinates.
Code for reproduction
%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt
r = np.geomspace(1, 100, num=10)
th = np.linspace(0, 360, num=50)
th = np.radians(th)
v = np.random.rand(len(r), len(th))
ax = plt.subplot(1, 1, 1, projection='polar')
ax.set_rlim(1, 100)
ax.set_rscale('log')
plt.pcolormesh(th, r, v)Actual outcome
Expected outcome
I think the pixels should spread to the outer edge of the circle, and no warnings should be shown.
Matplotlib version
- Operating system: Windows 8.1
- Matplotlib version: 2.2.3
- Matplotlib backend (print(matplotlib.get_backend())): module://ipykernel.pylab.backend_inline
- Python version: 3.6.6
- Jupyter version (if applicable): Jupyterlab 0.34.1
- Other libraries:
Python, jupyterlab numpy and matplotlib were installed with conda.
Activity
timhoffm commented on Nov 18, 2018
There's more going wrong: The radial tick labels are missing:
%matplotlib inline import numpy as np import matplotlib.pyplot as plt r = np.geomspace(1, 100, num=10) th = np.linspace(0, 360, num=50) th = np.radians(th) v = np.random.rand(len(r), len(th)) fig, (ax1, ax2) = plt.subplots(1, 2, subplot_kw={'projection' :'polar'}, figsize=(10, 5)) ax1.set_rlim(1, 100) ax1.pcolormesh(th, r, v) ax2.set_rlim(1, 100) ax2.set_rscale('log') ax2.pcolormesh(th, r, v)timhoffm commented on Nov 20, 2018
Here's a diff of transData of the two versions.
I suspect that the issue is the x1=5.0 instead of x1=10.0 in the very last Bbox. Can somebody with more transform experience please check this.
dstansby commented on Dec 20, 2022
I suspect this is the same bug as #24383 - I think it will be easier to fix #24383 first (there's an even simpler example of radial log scale that doesn't work there), then come back to this issue and see if it's fixed as a side effect.