Adjust saturation threshold depending on the read pattern

[schlafly]

Resolves RCAL-618

This PR in concert with spacetelescope/stcal#188 updates the saturation threshold to depend on the read pattern.

The idea is that the saturation reference file gives the number of counts at which a given pixel saturates. Roman downlinks resultants, which are usually averages of many reads. The current saturation flagging checks if a resultant is above the saturation threshold, but we want to know if any read in the resultant is above the threshold. The resultants in a sense dilute the saturated signal---a saturated last read is averaged with unsaturated earlier reads, bringing the resultant value beneath the saturation threshold. This PR corrects for that affect by passing stcal an additional argument, read_pattern, that specifies how reads are allocated to resultants, and using that information to adjust the saturation threshold resultant-by-resultant.

Checklist

• added entry in CHANGES.rst under the corresponding subsection
• updated relevant tests
• updated relevant documentation
• updated relevant milestone(s)
• added relevant label(s)

[codecov]

Codecov Report

Patch has no changes to coverable lines.

Files Changed	Coverage
romancal/saturation/saturation.py	ø

📢 Thoughts on this report? Let us know!.

[schlafly]

Regression tests are failing for unrelated reasons (presumably the forthcoming release?), though more broadly I think we expect the change to the saturation algorithm to introduce some differences in the saturation flagging.

[schlafly]

I think this is good to go, though I doubt we actually want to merge it until the corresponding stcal PR is merged.

[ddavis-stsci]

LGTM, but of course the real work is in stcal.