BUG: ensure YTSlice.coord has units

[chrishavlin]

PR Summary

Found that for sph data, calling ds.slice(axis, center).to_frb(...)[field] would error if the center argument was a plain float. For other dataset types (and other data objects), a plain float is assumed to be in units of code_length. Easiest fix is to ensure that YTSlice.coord attribute is a unyt object, following the other data objects.

Here's the stack trace from the error on main:

import yt
ds = yt.load_sample("snapshot_033")
frb = ds.slice(0, 0.25).to_frb(ds.domain_width[0], (64, 64))
print(frb["gas", "density"])
# Traceback (most recent call last):
#   File "/Users/chavlin/.pyenv/versions/yt_dev/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3548, in run_code
#     exec(code_obj, self.user_global_ns, self.user_ns)
#   File "<ipython-input-3-76bca84cf85c>", line 6, in <module>
#     print(frb["gas", "density"])
#   File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 209, in __getitem__
#     return self.get_image(item)
#   File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 213, in get_image
#     self._generate_image_and_mask(key)
#   File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 179, in _generate_image_and_mask
#     buff, mask = self.ds.coordinates.pixelize(
#   File "/Users/chavlin/src/yt_/yt_dev/yt/yt/geometry/coordinates/cartesian_coordinates.py", line 238, in pixelize
#     buff, mask = self._ortho_pixelize(
#   File "/Users/chavlin/src/yt_/yt_dev/yt/yt/geometry/coordinates/cartesian_coordinates.py", line 554, in _ortho_pixelize
#     data_source.coord.to("code_length").v,
# AttributeError: 'float' object has no attribute 'to'

PR Checklist

• [ ] New features are documented, with docstrings and narrative docs N/A
• Adds a test for any bugs fixed. Adds tests for new features.

[neutrinoceros]

random github markdown tip of the day:
python-traceback is a supported syntax coloring scheme:

Traceback (most recent call last):
  File "/Users/chavlin/.pyenv/versions/yt_dev/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3548, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "<ipython-input-3-76bca84cf85c>", line 6, in <module>
    print(frb["gas", "density"])
  File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 209, in __getitem__
    return self.get_image(item)
  File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 213, in get_image
    self._generate_image_and_mask(key)
  File "/Users/chavlin/src/yt_/yt_dev/yt/yt/visualization/fixed_resolution.py", line 179, in _generate_image_and_mask
    buff, mask = self.ds.coordinates.pixelize(
  File "/Users/chavlin/src/yt_/yt_dev/yt/yt/geometry/coordinates/cartesian_coordinates.py", line 238, in pixelize
    buff, mask = self._ortho_pixelize(
  File "/Users/chavlin/src/yt_/yt_dev/yt/yt/geometry/coordinates/cartesian_coordinates.py", line 554, in _ortho_pixelize
    data_source.coord.to("code_length").v,
AttributeError: 'float' object has no attribute 'to'

[neutrinoceros]

This looks fine to me unfortunately it seems I cannot compile yt these days (I assume because I'm being too aggressive with upgrading my system dependencies), so I'm not able to verify that your test captures the problem. Nonetheless it seems trivial enough that I can approve to give you the option of self-merging here.

[neutrinoceros]

nice catch

[chrishavlin]

Nonetheless it seems trivial enough that I can approve to give you the option of self-merging here.

It's not a critical fix, so I'll let it hang out for now and self merge at the end of the day if no one else jumps in.