AP Pipeline Meeting, 2020-06-01

2pm Pacific Time

BlueJeans: https://bluejeans.com/426716450

attending:

• Eric Bellm
• Unknown User (cmorrison)
• John Swinbank
• Meredith Rawls
• Unknown User (emorganson)
• Unknown User (gkovacs)
• Siegfried Eggl

regrets:

Topics for discussion:

• Discussion of monthly reprocessing results (Meredith Rawls and all, starting next week (June 8)):
  • Overview of goals of this session (Eric Bellm )–give us a chance to engage in more detail in looking at our monthly processing, bring commissioning team up to speed
  • Meredith Rawls to walk through the "standard" notebook–good for first meetings
• Gen 3 conversion (all):
  • Krzysztof Findeisen is out today

• Precursor processing (Meredith Rawls ):
  • copying "new" January templates with updated pixel scales into our repos
• report from NCSA (Unknown User (emorganson)  + Unknown User (manninga) et al):
  • Unknown User (emorganson) says Unknown User (manninga) is working to make Kubernetes accessible more broadly, which should enable a more end-to-end processing→alerts workflow
• Alert Distribution (Spencer Nelson ):
  • Spencer Nelson is out
• ap_association/SDM/fakes/alerts (Unknown User (cmorrison)):
  • got reviews back from John S
  • starting to look at astropy CCDData code for stamp serializations.  thinking some about local WCS
• Crowded fields (Ian Sullivan ):
  • on vacation
• Image differencing (Unknown User (gkovacs) ):
  • working on ZOGY code improvements to have ready for review when Ian returns
  • starting a draft technical note to capture technical details
• Review CI (https://chronograf-demo.lsst.codes/):
  • nothing compelling–maaaaybe a small runtime jump in diffim?
• Review outstanding action items
• AOB
  • Siegfried Eggl preparing some sample products for the Solar System science collab, wants to generate some false detections that are more realistic than previous
    
    • asking about psf total flux ~0 objects; consensus is that these are likely to be dipoles
    • this is highlighting again that we need to record the detection SNR in addition to the measured SNR
    • showing fairly strong trend in number of DIASources with seeing–a plot we should start making regularly
      • Meredith Rawlsto add to reprocessing notebook
    • Eric Bellm cautioning that we don't represent this false detection rate to the SSC as characteristic of final performance

Action Items