BUG: fix NaNs in off-center FRB slices from ds.r

[yut23]

PR Summary

FRBs of slices made with ds.r were selecting the wrong area when the center of the slice didn't match the center of the domain. The existing tests didn't catch this, as they only check regions that have the same center as the domain. I've added a test case that covers this.

Example code and resulting plots:

import yt
import matplotlib.pyplot as plt

ds = yt.load_sample("IsolatedGalaxy")
region = ds.r[0:0.5:100j, 0.5:1.0:100j, 0.5]

# display NaNs as red
cmap = plt.colormaps["viridis"]
cmap.set_bad(color="red")

plt.imshow(region["index", "x"], origin="lower", vmin=0, vmax=1, cmap=cmap)
plt.colorbar()

PR Checklist

• Adds a test for any bugs fixed. Adds tests for new features.

[chrishavlin]

nice catch, the fix looks good to me.