search

spacetelescope / jwst

Python library for science observations from the James Webb Space Telescope
https://jwst-pipeline.readthedocs.io/en/latest/
Other
571 stars 167 forks source link

calwebb_image2 resample step failing for NOps4+5 data #3628

Closed stscijgbot closed 5 years ago

stscijgbot commented 5 years ago

Issue JP-781 was created by Howard Bushouse:

JWSTDMS-23 reports that one or more MIRI images are failing in the resample step of calwebb_image2 with: {noformat} File "/dms/local/jwst/pipeline/pkgs/miniconda3/envs/jwstdp-cal-env/lib/python3.6/site-packages/jwst-0.13.3-py3.6.egg/jwst/resample/resample_utils.py", line 53, in make_output_wcs "{}".format(output_wcs.data_size)) ValueError: Invalid output frame shape: (0, 0){noformat} The ticket points to example file locations.

stscijgbot commented 5 years ago

Comment by Howard Bushouse: Data files have been copied to /grp/jwst/ssb/chartreuse/dms_bugs/jwstdms-23/

stscijgbot commented 5 years ago

Comment by Jane Morrison: Cdelt1 and cdelt2 = 1.0 for this data. resample_util is calling assign_wcs.util.wcs_from_footprints which uses these values to define the footprint of the resampled image. When I change the cdelt1 and cdelt2 to something reasonable for a MIRI Image resample runs with out errors. EXP_TYPE was set correctly for this data to MIR_IMAGE So now the question is why is CDELT1, CDELT2 = 1 in the rate files ? - continuing to research this.

jdavies-st commented 5 years ago

We should fix wcs_from_footprints() to compute CDELTn from the GWCS object, not rely on the FITS keywords. In the case of spectral data, these CDELTn keywords no longer exist anyway, though I don't believe we are using wcs_from_footprints() for spectral data at the moment.

jemorrison commented 5 years ago

The problem is with the data and not the pipeline. The EXP_TYPE for both files is MIR_IMAGE, but the APERNAME and PPS_APER is MIRIFU_CHANNEL1A. The incorrect APERNAME is causing the wrong data in the siaf file to be used resulting in CDELT1,2 being set to 1.0

stscijgbot commented 5 years ago

Comment by Jane Morrison: The two files (jw00623029001_03102_00002_mirimage_rate.fits jw00623030001_03102_00001_mirimage_rate.fits) have inconsistent EXP_TYPE and APERNAME

the EXP_TYPE for both files is MIR_IMAGE the APERNAME and PPS_APER for both files is MIRIFU_CHANNEL1A

the incorrect APERNAME likely the cause of CDELT1 and CDELT2 to be incorrectly set with values of 1.

stscijgbot commented 5 years ago

Comment by Jane Morrison: The problem is with the data not the pipeline