Replace cudf column.full with cudf.as_column

[mroeschke]

Description

I have a PR in cudf removing full in favor of as_column which has equivalent functionality (rapidsai/cudf#14698), so replacing the single usage I found here

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.

[vyasr]

@mroeschke looks like there is one real error caused by RangeIndexes looking different

[harrism]

#1349 is now failing because of the failed import of full. I would like to get this merged. @mroeschke what is the blocker?

[bdice]

@mroeschke FYI, this PR is also blocking cuxfilter CI.

[bdice]

One legitimate failure is copied below (to separate from the noise of unrelated CI failures):
https://github.com/rapidsai/cuspatial/actions/runs/8086937898/job/22098408527?pr=1350#step:7:17280

=========================== short test summary info ============================
FAILED tests/spatial/distance/test_distance.py::test_geoseries_distance_indices_different - AssertionError: ColumnBase are different

values are different (100.0 %)
[left]:  {'[0.0, <NA>]'}
[right]: {'[0.0, nan]'}

[galipremsagar]

One legitimate failure is copied below (to separate from the noise of unrelated CI failures): https://github.com/rapidsai/cuspatial/actions/runs/8086937898/job/22098408527?pr=1350#step:7:17280

=========================== short test summary info ============================
FAILED tests/spatial/distance/test_distance.py::test_geoseries_distance_indices_different - AssertionError: ColumnBase are different

values are different (100.0 %)
[left]:  {'[0.0, <NA>]'}
[right]: {'[0.0, nan]'}

Following change: #1350 (comment) will fix the CI failure.

[bdice]

I'll pre-emptively trigger a merge on this since we expect things to pass and it's blocking other work.

[bdice]

/merge

[mroeschke]

Looks like this PR is green except for the timed-out falures (I don't have permission to re-run those jobs)

[bdice]

I reran the failed jobs.