Closed stscieisenhamer closed 2 months ago
I have left this as draft because:
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 76.57%. Comparing base (
b61e70b
) to head (da44930
). Report is 18 commits behind head on main.:exclamation: Current head da44930 differs from pull request most recent head a6d72e5. Consider uploading reports for the commit a6d72e5 to get more accurate results
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Thanks Jonathan. Do you have an L3 file you can point me to that I can take a look at?
An example is found here
/grp/roman/eisenhamer/rcal-798-l3meta/roman_l3_metacheck.asdf
Note: This is saved from the test_l3_wcsinfo test because that test actually runs resample.
Okay, I see, the dictionary issue is from meta.ref_file.crds, which seems okay. If we wanted to do that better we would do some kind of dictionary flattening first. Presumably tests are failing because the RDM change isn't in yet?
Yes, the unknown dtype error is the error. Without the string-ing-fying, the dict is just an object which nothing likes. The ref_files.crds situation is the only one the currently exists, in which case flattening would be OK. But if more cases come up and flattening is attempted, one will need to ensure there is no key-name clashes.
Okay, let's hold this until @PaulHuwe can take a look at it and the corresponding RDM PR, and then merge after the RDM update goes in.
It would also be good to either run regtests or manually grab a couple cal files from artifactory and run the HLP on them with this PR, just in case the real exposure metadata is different enough from the filler metadata to cause issues. I don't think I expect that, though.
Latest relevant regtest. Differences found are expected.
Believe all is now clear and ready for final review.
Resolves RCAL-789 Requires roman_datamodels PR#348
This PR addresses implements the filling of the MosaicImage
meta.individual_image_meta
block.Checklist
CHANGES.rst
under the corresponding subsection