Fix Version/s: None
Team:Data Release Production
DM-33279, we can write a task that re-estimates the PSF model using uniquely/consistently reserved stars.
Discussion with Jim Bosch has led to the following plan:
- This will be run after fgcmcal/jointcal (but at the moment it will not use these outputs as inputs)
- This will be run before skycorr
- This is run per-ccd at the moment.
The inputs will be:
- calexp (to measure the psf)
- src catalog which a subset will be input to psf estimator task
- tract-by-tract isolated stars
In the task it will:
- Read src, apply a schema mapper to cut down to relevant columns.
- Run psf estimator (excluding reserved stars)
- On the psf stars (+ reserve stars) run each algorithm that requires an aperture correction
- Compute aperture correction for each algorithm
The outputs will be:
- The psf model (in its own file)
- The apcorr model (in its own file)
- The psf catalog (source catalog with measurements + flags including psf_used and psf_reserved).
The coadd make warp code will have to be updated to know to use these psfs, aperture corrections, and psf catalog (for flags).