search

spacetelescope / stdatamodels

https://stdatamodels.readthedocs.io
Other
5 stars 24 forks source link

New dynamic_dq warnings appearing in CI #222

Closed braingram closed 8 months ago

braingram commented 11 months ago 
 WARNING  stdatamodels.dynamicdq:dynamicdq.py:50 Keyword CDP_PARTIAL_DATA does not correspond to an existing DQ mnemonic, so will be ignored
WARNING  stdatamodels.dynamicdq:dynamicdq.py:50 Keyword CDP_LOW_QUAL does not correspond to an existing DQ mnemonic, so will be ignored
WARNING  stdatamodels.dynamicdq:dynamicdq.py:50 Keyword CDP_UNRELIABLE_ERROR does not correspond to an existing DQ mnemonic, so will be ignored
WARNING  stdatamodels.dynamicdq:dynamicdq.py:50 Keyword DIFF_PATTERN does not correspond to an existing DQ mnemonic, so will be ignored

I'm not certain when these appeared by they did not exist in a CI run 2 days ago (prior to the new crds context going operational): https://github.com/spacetelescope/stdatamodels/actions/runs/6254501459/job/16982181097

braingram commented 11 months ago

The crds file: https://jwst-crds.stsci.edu/browse/jwst_miri_flat_0588.fits contains a dq_def table:

FITS_rec([(0,   1, 'DO_NOT_USE', 'Bad pixel, do not use'),
          (1,   2, 'NON_SCIENCE', 'Non science data'),
          (2,   4, 'UNRELIABLE_FLAT', 'Flat-field variance large'),
          (3,   8, 'CDP_PARTIAL_DATA', 'Data derived from incomplete input'),
          (4,  16, 'CDP_LOW_QUAL', 'Data of low quality'),
          (5,  32, 'CDP_UNRELIABLE_ERROR', 'Data without a reliable error estimate'),
          (6,  64, 'NO_FLAT_FIELD', 'Flatfield value not available'),
          (7, 128, 'DIFF_PATTERN', 'Diffraction & interference patterns')],
         dtype=(numpy.record, [('BIT', '<i4'), ('VALUE', '<u4'), ('NAME', 'S40'), ('DESCRIPTION', 'S80')]))

that are not defined in the pixel level dqflags: https://github.com/spacetelescope/stdatamodels/blob/cddf80cebd796f9f2e1510b0a277c5fef3ebbd7f/src/stdatamodels/jwst/datamodels/dqflags.py

braingram commented 11 months ago

From discussion with @tapastro and @jemorrison these aren't of immediate concern as the instrument teams are aware of the allowed flags and these are set for other reasons.

I'm going to leave this issue open for now to make it more visible for other PRs that might reference the warnings.

braingram commented 8 months ago

These appear to have gone away.