New ROLL_REF, V3I_YANG, VPARITY keywords for primary extension

[stscijgbot-jp]

Issue JP-3836 was created on JIRA by Rachel Cooper:

Add ROLL_REF, V3I_YANG, and VPARITY keywords. These keywords usually exist in SCI extension; we want to add them to the "NIRISS AMI information" section of the dictionary with a modified definition so they can be in the PRIMARY fits extension. Changes are reflected in stdatamodels updates here: spacetelescope/stdatamodels#357

All fields except those capturing the changes (fits_hdu, section, level, si, mode) should be the same as existing WCS Info keywords for consistency; fields on the "additional data required for new keyword" tab have been copied from the existing entries (except V3I_YANG, which doesn't have a unit currently but should be degrees as far as I know). The data type for VPARITY should also be integer instead of float but I could only pick one on the dropdown.

[stscijgbot-jp]

Comment by Linda Thompson on JIRA:

These keywords already exist in the SCI extension for all instruments.
core.wcs.schema.json: "fits_keyword" : "VPARITY",
core.wcs.schema.json: "fits_keyword" : "V3I_YANG",
science.wcs.schema.json: "fits_keyword" : "ROLL_REF",

This is fine for everything else, but the NIRISS oi_fits standards call for them in the primary HDU.
All of them are archived, so moving them into PRIMARY only for NIRISS AMI level-3 is odd. The NIRISS AMI FITS files do not include the SCI extension and the standard does not call for it; if we add it, it would contain only header keywords.

The requested implementation will cause MAST search issues. The board recommends leaving them in the SCI extension and adding it for NIRISS AMI files, even though the standard doesn't call for it. Present to NIRISS team. If then need to be added to the PRIMARY header, do not archive it.

We should investigate how these keywords are searched in MAST. If you search for these keywords, level-3 files will be excluded, but when you retrieve the data, will you also get the level-3 files?

We could tweak the names for NIRISS AMI.

[stscijgbot-jp]

Comment by Rachel Cooper on JIRA:

Hi Linda,

We tried using the existing keywords from the wcsinfo schema, but as you point out, this results in output files with an empty 'SCI' extension and a header containing only these keywords, which is odd and potentially confusing for people who might expect to find science data there. While adding new extensions is not forbidden by the oifits standard, having the first extension be an empty 'SCI' HDU is a significant difference from what users familiar with OIFITS files produced by other facilities might be expecting.
We thought duplicating the keywords to be in the primary header only for AMI level 3 products would be less confusing than having keywords with slightly different names but the same definition and usage being created only for AMI. The full discussion leading to this request is on this PR: spacetelescope/stdatamodels#357

 

[stscijgbot-jp]

Comment by Linda Thompson on JIRA:

Ticket JWSTKD-580 was discussed at today's meeting. The JWST Keyword Dictionary board agrees that Cal can add these keywords to the primary header for NIRISS AMI exposures and the KD JSON files will remain unchanged. This will be a permanent difference that CAL needs to keep in mind when running the JSON/YAML comparison tool.
The JWSTKD ticket was moved to JP.

[stscijgbot-jp]

Comment by Tyler Pauly on JIRA:

Moved to JP project to track any necessary updates to AMI datamodels.