IPP Progress Report for the week 2010.05.17 - 2010.05.21

(Up to IPP Progress Reports)

Eugene Magnier

Heather Flewelling

• more addtool/minidvodb work. down to a finite number of outstanding tasks
• ifaps1 questions answered

Bill Giebink

Roy Henderson

• PSPS
  • Confusion over ra/dec ranges and corresponding Db slices. Now only slices for > -30 dec
  • IN batch issues: discussions as to whether to keep old for historical comparison
  • Wrote up a design (on IPP wiki) for a framework of a recovery system, based on conversations with Conrad
  • Schema changes:
    • survey name added to Survey table in IN batch
    • surveyID added to SkinnyObject
    • 64-bit (instead of 32-bit) ints for flags

• IPP
  • Updated 'photCode' parsing script to extract extinction values. Also updated with new zero points.
  • Now logging jobID and corresponding exposureID for auditing purposes. Ultimately, this will be stored in the Db.
  • Now populating detection flags from smf as well as pulling extra flags from DVO
  • surveyID stuff: pulling the right thing from IPP encoding in for PSPS

Paul Price

• Streaks in detector coordinates:
  • Added magic clusters input for ppCoord
  • Added 'exposure' mode for magictool to get astrometry file given magic_id
  • Modified magic_process.pl to run ppCoord on streaks and clusters
  • Sped up ppCoord using the 'WCS' file type (doesn't read sources in SMF file)
• Removed "FPA.OBS" --- was only used to generate warnings, was generating lots; no need for it
• Background restoration
  • Proof of concept with current tools: can restore pattern row at warp stage using ppVizPattern, ppImage, pswarp, ppArith
  • Planned pipeline design with Gene, developed draft workflow
• MOPS support
  • Updated ICD with new fields (some long promised, others from Gene's recent work to provide statistics for weeding out bad subtractions)
  • Updated ppMops to provide new fields
  • Designing interface for providing detection efficiencies
  • Decided with Larry to adopt an 'SMF-like' interface format, where IPP concatenates diff CMF files into a single FITS file, with duplicates (from overlapping skycells) removed; this removes complications in providing (and reading) detection efficiencies
• Magic recertification: ran w and y data through pipeline for verification
  • No streaks missed
  • w-band observations (from SweetSpot) have significantly more (false?) streaks identified
• Investigating centroid problem in diff photometry from Dave Young
  • Some SNe have centroids off by ~ 2 arcsec
  • Seems to be due to very flat tops on (S/N)² images so that peak can be on the side, and the peak is then set more or less set in stone as the position
  • PSF_MOMENTS_RADIUS wasn't getting into psphotReadoutMinimal; this also fixes the anti-dipole statistics
• Writing PSF-matching paper

Bill Sweeney

• Did some work insuring that the error codes returned by the postage stamp server for data

that is damaged or otherwise unavailable correctly reflect the situation.

• Updated pstamp_check_dependent to handle some error conditions encountered processing some requests.
• updated the pending update queries to wait until corresponding magicDSRun's have been cleaned up.
• Updated chip stage configuration and script to allow the raw images that have older burntool tables to be re-processed by 'update'.
• Built a dvo database of recent exposures for Eric Morganson of MPIA.
• Investigated query about incorrect stamp sizes from users. (The sizes were fine the coordinates were near a skycell boundary and we don't yet have overlapping MD skycells)
• Updated the postage stamp server interface control document (ICD). It is version 1 now.
• wrote a new postage stamp request file generator that has an interface that is easier to use.

Chris Waters

• mask fractions: Debugged all stages and merged branch. Change to behavior of camera stage (calculate stats for both 3.0 degree field and maximum unvignetted field) means this isn't implemented yet.
• exposure level JPEGs: Need to finish database interactions still, but jpegs are being created correctly.