AP Pipeline Meeting, 2022-12-19

2pm Pacific Time

Zoom: https://washington.zoom.us/j/98846655020

attending:

• Eric Bellm 
• Ian Sullivan 
• John Parejko 
• Krzysztof Findeisen 
• Brianna Smart 
• Erin Howard 
• Nima Sedaghat 
• Kenneth Herner 

regrets

Topics for discussion:

• Project updates (Eric Bellm , Ian Sullivan ):
  • last Monday meeting of the year!  Next meeting Jan. 9 (during AAS; Ian will lead)
  • sprint planning likely Jan 17++
• Sprint closeout (all):
  • add slides here for Wednesday's 1pm Pacific Science Pipelines team meeting
    
    • John P demoing DIASource cutouts (creation and access)
    • Nima on getting RB scores from ap_verify
    • Ari on hand-classified cutouts
    • Erin on variable selection
    • Bri on simbad search, missed CRs
    • aim to have draft slides by Weds. pair coding
• Pipelines Infrastructure (Krzysztof Findeisen , John Parejko ): 
  • KF: not much to say about Prompt Processing; have mostly been working on ap_verify this week.  Almost ready with example dataset creation and update scripts soon.
  • JP: triaging things coming out of the sprint, some refcat stuff
• Precursor processing ( Kenneth Herner , Erin Howard ):
  • Erin: hoping to run a successful ap_pipe run on Saha Bulge–have calibs & templates ready
    • Ian: bps report should be working now, which should make it easier to manage
• Image differencing algorithms, DCR  ( Ian Sullivan , Bruno Sanchez ):
  • Ian: while Bruno was here, got diffim with preconvolution running.  Has been cleaning up some edge cases with unit tests since then, found a few bugs & issues.  Continuing to refactor to improve detect and measure over the quiet period.
• transinet (Nima Sedaghat, Eric Bellm):
  • worked with Ian on applying RB model.
  • would be useful to have some raw data transferred to GCP for work during the holidays–wants to try running nonstandard pipelines
    • DC2 data is already there (and should be enough for his tests for him), but Nima wasn't sure how to access it
      • Ian thinks ask in #ops-google-idf
• alert distribution (Brianna Smart , Eric Bellm ):
  • Bri: AD is still working but not quite as expected.  Stayed up for seven days, but number of messages doesn't seem to have decreased–deletion mechanism still not working as expected.  (The alert simulator isn't running right now so no new messages are being loaded.)
• Solar-system processing (Ari Heinze ):
  • out today
• Review CI (https://chronograf-demo.lsst.codes/):
  • no changes
  • DC2 catalog dataset shows HITS predicted SSObjects?  need to debug the dashboard
• Review outstanding action items
• QA meeting in February:
  • RB scoring?
  • Eric on detection thresholds/fakes forced photometry?
• AOB

Action Items