DRP Metrics Monitoring 2023-05-29

Use this to fill in our review of the w_2023_19 week of May 22nd:

Agenda:

• Announcements
• Review Action items from last month

  • Eli solved the mysteries of the stellar locus jumps (please read full comment on: DM-38733 - Getting issue details... STATUS )

    w_2023_09: 7.4 mmag
    w_2023_10: 9.6 mmag	Jump caused by DM-37955 - Getting issue details... STATUS (ApCorr changed star selection). being adresssed on DM-39203 - Getting issue details... STATUS (new single frame extendedness definition)
    w_2023_11: 10.6 mmag	Jump caused by DM-37141 - Getting issue details... STATUS (turning on gbdes without pm fitting... I think that without pm fitting there was excess noise/band-to-band offsets.).
    w_2023_15: 7.81 mmag	Drop caused by DM-23781 - Getting issue details... STATUS (sky object change) Fix done on DM-39141 - Getting issue details... STATUS (use detect_isPrimary on source selectors)
    w_2023_16: 7.85 mmag
    w_2023_17: 7.85 mmag
    w_2023_18: 5.1 mmag	Decrease caused by DM-37943 - Getting issue details... STATUS (pm ref cat bugfix and pm fitting).
    w_2023_19: 5.1 mmag
    w_2023_20: 10.4 mmag	Jump caused DM-38246 - Getting issue details... STATUS (excluding EDGE pixels from SourceDetectionTask).  Fix on  DM-39306 - Getting issue details... STATUS (We expect this to be back to 5.2 mmag on w_2023_21)
    w_2023_21: 5.2 mmag	After DM-39306 - Getting issue details... STATUS but before DM-39141 - Getting issue details... STATUS

• Processing Status
• Review the w_2023_19 rerun:
  • Orion Eiger New errors? File tickets with the how-to-reproduce! 
  • Pipelines team: Clare Saunders Lauren MacArthur Arun Kannawadi Lee Kelvin  Do your plots for w19 look OK? https://usdf-rsp.slac.stanford.edu/plot-navigator/dashboard_gen3
    • (Clare) The astrometric residuals-with-the-reference-catalog plots look way better, as expected from the bug-fix in meas_algorithms. Some bands looked fine before, and there isn't much change there. Whether a band improved should depend on the epochs that went into the coadd, and I haven't checked, but assume that the bands that looked bad before had visits that were farther from the Gaia reference epoch. Some of the improvement should also be credited to turning on proper motion and parallax fitting, though the proper motions are not actually applied to the coadd objects. The plots below show the i-band results in w15 (on the left) and w19 (on the right).
    • 
  • Pipelines team: Do the metrics on Sasquatch look OK?  Sasquatch + analysis_tools chronograph 
  • Any changes in the resource usages? (Yusra AlSayyad will post plot from notebook)
• What do we expect for w_2023_23?? 

Comments:

• Lauren MacArthur : updated stellar locus plots in analysis_drp  look good.  I see only subtle differences between the w_2023_19 processing run and the one I ran against w_2023_11  on DM-35336 - Getting issue details... STATUS , e.g.

• • w_2023_11 + DM-35336

• w_2023_19 

Action Items