Complete the SuperBIT OBA module before the launch window

[sweverett]

As the title says - the launch window opens on March 21st. As this will be accompanied by a software-freeze on the telescope's computer, we should finalize the OBA before this date. To minimize the size of the repo on the onboard computer, we will likely create a tag and then fork the repo with all modules other than OBA and utility functionality (& the CookieCutter of course!) removed.

Issues that need to be solved before then:

• Use realistic rolls between bands in sims
• Compute PSF aberrations for all filters (Compute optical aberrations for all filters #103)
• Add "inactive pixels" to the calibration mask (Add "inactive pixels" to image simulation modules #108)
• On-the-fly masking configuration (Allow for flexible masking types in OBA #114)
• Astrometry failures & backups (OBA astrometry.net code fails #115)
• Add WCS to (copied) RAW science images (Add WCS solution to raw SCI OBA images if not present in astrometry module #116)
• Pass mask to La Cosmic (Pass masking information to La Cosmic #117)
• La Cosmic (or alternative) speedups (La Cosmic finds far too many cosmic rays & takes too long #119)
• Add relative flux weights for the detection image
• Handle NoOverlap's in the CookieCutter (Handle NoOverlapError in CookieCutter #118)
• Bright star masking (Bright star masking in OBA #120)
• Use estimated BKG noise instead of predicted during star masking
• Determining max bounding box for OBA coadds (Determine maximum bounding box for OBA coadds #121)
• Add cluster member galaxies to the OBA sims (Add cluster members to OBA sims #122)
• Finish CookieCutter mask implementation (Add segmentation into the CookieCutter mask #124)
• Write CookieCutter to Image method (and test on MEDS) (CookieCutter Image reconstruction #126)
• Save over any desired calibrated image file HDR fields to the copied RAW's for CookieCutter
• Add cluster center option for CookieCutter (Add capability to treat cluster center as a single cutout #125)
• Make final oba conda env (mostly done, but need to finalize minimal version)
• Either new script or incorporate sufficient image checker into run_oba.py
• Default the cleanup module to remove OBA_TEMP dir
• OBA base config + key:val replacement command
• OBA prepper / run command (target_name, band:req_num); det_bands in base config
• Update expected FITS header cols to match new QCC definitions
• SExtractor not running detection on full image (Source Extractor doesn't run detection on the full det image #127)
• Significantly speedup CookieCutter
• Figure out the CookieCutter WCS inconsistency CookieCutter cutouts are offset from original image #128
• Implement second "2D" CookieCutter output mode
• Speedup "2D" CookieCutter
• Validate stellar masking
• Finalize config values
• Write QCC OBA script to update a swarp/sextractor config
• Do IMG_QUAL parsing in preprocess (Add image quality parsing to OBA #130)
• Rework prep_oba.py to handle new IMG_QUAL parsing (Add image quality parsing to OBA #130)
• Incorporate Steve's GAIA cats into star masking (Bright star masking in OBA #120)
• Solve weird noise stripes when using COMBINE_TYPE: CLIPPED
• Fix coadd clipping for bands not in det image
• Add band-dependence on flat images in IOManager
• Add saturation bitmask flag
• Update QCC uniform flats for all bands
• Deeper GAIA cat to mag 15
• Update Starmasking spike scaling
• Brightest star not masked?
• Speedup 2D CC writing
• Change output format (no dirs)
• Reset qcc global OBA config after testing
• Check what is happening w/ COW command on bands, if time
• Remove final warnings

Right before flight:

• Clear out OBA dirs on QCC
• Final git pull's and tests
• Tag GitHub version

Will update & add as we go (though issues won't technically close until merged into master)