JP-3232: Fix NRC coron duplicate product names #7826
Conversation
| # Check and continue initialization. | ||
| super(Asn_Lv3NRCCoronImage, self).__init__(*args, **kwargs) | ||
|
|
||
| @property |
There was a problem hiding this comment.
For reviewers, this just adds an override for the regular "dms_product_name" function, to be used by only this particular ASN class.
| return product_name.lower() | ||
|
|
||
|
|
||
| def dms_product_name_coronimage(asn): |
There was a problem hiding this comment.
This is a line-for-line copy of the regular "dms_product_name" function, but with the addition of the "-image3" suffix to the end of the optical element field(s).
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## master #7826 +/- ##
==========================================
- Coverage 76.54% 76.52% -0.03%
==========================================
Files 456 456
Lines 36947 36964 +17
==========================================
+ Hits 28282 28285 +3
- Misses 8665 8679 +14
*This pull request uses carry forward flags. Click here to find out more.
☔ View full report in Codecov by Sentry. |
|
For the record, when the generator is run on pool "jw00016_20230331t130733_pool.csv" it creates the level-3 ASN's "jw00016-c1002_coron3_00001_asn.json" and "jw00016-c1002_image3_00001_asn.json". Before this rule change, the product name in both of those ASN's was "jw00016-c1002_t004_nircam_f210m-maskbar". With this change, the coron3 ASN still has that original product name, but the image3 ASN now uses the product name "jw00016-c1002_t004_nircam_f210m-maskbar-image3". All other aspects of the ASN's are unchanged. |
|
Regtest run started at https://plwishmaster.stsci.edu:8081/job/RT/job/JWST-Developers-Pull-Requests/868 |
|
@stscieisenhamer We already have the jw00016 pool included in our association regtests, which will check which asn's are created. But the only thing that changes as a result of this PR is the product name listed inside of one asn file, so I'm guessing that won't show up as a difference, right? The tests only check which asn files were created and not their content, right? |
|
If there are pools in the artifactory area, the tests themselves should see the change and report failure, along the lines of different/extra/missing products. If the changes are the expected ones, okifying will rectify the changes. |
The jw00016 is in the associations/sdp/pools/ directory on artifactory, which I believe is the input area for all the tests based on sdp-generated pools, but I also see a directory /truth/sdp/pools/ that contains a lot of pool files. 1) Why would there be pool files in a truth area, and 2) if they're necessary, do I need to put a copy of the jw00016 pool in there too? |
|
That is unfortunate. The tests should be fully self-contained under For your purposes, it is only the top level |
|
Regtests are clean. |
3 participants
Resolves JP-3232
Closes #7730
This PR adds an override for the construction of the level-3 product name for NIRCam coronagraphic associations that are to be processed as regular imaging, i.e. through the image3 pipeline. The data are also processed as a coron3 ASN, which previously resulted in identical "i2d" level-3 product names being created: one from coron3 and one from image3. This update breaks the file name degeneracy by simply adding the additional suffix "-image3" to the optical element field of the product name for the data processed as regular imaging.
Checklist for maintainers
CHANGES.rstwithin the relevant release sectionHow to run regression tests on a PR