BlueJeans: https://bluejeans.com/426716450

attending:

Topics for discussion:

  • Gen 3 conversion (

    T Key Summary Assignee Reporter P Status Resolution Created Updated Due
    Refresh

    ):
  • HiTS processing (Meredith Rawls ):
    • plotting tools updates ongoing
    • then HSC datasets, January rerun
  • report from NCSA (Unknown User (emorganson) et al.):
    • had meeting about AP pipelines; lots of late cancellation, so focused mainly on technical challenging of keeping templates available
    • will become a regular meeting
    • starting to think about standing up dedicated computing for testing–will check documentation/with Meredith; Meredith will update docs if there are deficiencies
  • ap_association/SDM (Unknown User (cmorrison)):
    • met with Yusra, generalizing some unit tests to test values running through the functors, and doesn't have to use fragile fake data
    • two tickets spawned: use local CP products & make them available to multi-band driver; update obs_subaru to output fully-calibrated data products; then hook up on AP good.  Yusra has a ticket to generate parquet outputs for single-frame
  • DCR (Ian Sullivan ):
    • just Gen 3; everything is in review
    • rerunning HiTS with good seeing selection done correct, no major conclusions–but it does look better
  • Image differencing (Unknown User (gkovacs) ):
    • some discussion of the weirdness in the new December template batch:  Gabor left some comments on DM-22729 - Getting issue details... STATUS .  Eric points out that this is the first time we've made templates in bulk when we have the ScaleVariance enabled, wonders if that's at issue
    • the December templates had the wrong seeing range; so January should have correct bestSeeing range, and see if that fixes the issue
  • Algorithms Workshop prep (Eric Bellm ):
    • nothing to report
  • Review CI (https://chronograf-demo.lsst.codes/):
    • Should we drop fraction of DIAObjects updated?  Fragile to execution order, hard to understand in practice.  Slack discussion: https://lsstc.slack.com/archives/C2B6X08LS/p1578956211031700 
      • Replace with N_associated/N_DIAsources ?
      • Eric will write a low-priority ticket and see when it crosses the bar
    • Mysterious change in Ratio of DIASources to Direct Image Sources on 1-16.  Chris says these are both outputs from diffim and SFM directly, so no connection to ap association. John did not find any obvious candidate merges on the 15th or 16th–will poke a bit more.
  • AOB
    • nothing