Mondays 12pm - 12:50pm EST
Yusra's Zoom: https://princeton.zoom.us/my/yusra
Attendees:
Yusra, Andrés, Arun, Chris (Waters), Clare, Colin, Dan, Eli, Hsin-Fang, Ian, Jim, Josh, Lauren, Leanne, Lee, Merlin, Morgan, Nate, Robert, Simon, Sophie, Tim (Morton)
Agenda:
- Review Action items from last week
- Review w_2021_14
- What was noteworthy in this rerun?
- Gen2: Done and converted to Gen3 (last few gen2 reruns are in /repos/main
- Gen3: Lots of little issues revealed, most fixes in by 16. 9813-only of w16 to see if we can do it in 2 QGs. Not knowing how much memory takes is the big question and just testing that it all works after w16.
- Likewise, Dan is running the DC2 dataset through the Gen3 pipeline and is working through configs.
- LM: Have parity up to the end of single processing (calexps), and am moving on to jointcal/FGCM.
- Still need to run and compare faro on both the gen2 and gen3 reruns.
- DM-29729 → pair-coding
- Ensure that isPrimary column is set/consistent in fields where we don't have full band coverage and scarlet doesn't run (see
for some plots and details). Fred Moolekamp
- What changed? Annotate
- JM: There’s a new algorithm for measuring moments available to remove a bias that scales with signal to noise. Basically by adding noise to the high-SNR to look like low SNR.
- Joshua MeyersTurn on debiased HSM moments in obs_subaru, obs_decam, and obs_lsst.
- Arun Kannawadi Modify rho stats in pipe_analysis to use debiased moments (see
).
- What do we expect from w_2021_18?
- New features:
- Debiased HSM moments.
- High-level pipeline files to new packages. See draft RFC out later today.
- analysis_drp? Probably not.
- Changes toward Gen2-Gen3 parity:
- What new metric can we expect next time
- Review DC2
- w_2021_12:
- Jim will convert DC2 gen2 rerun w_2021_12 → /repo/dc2. Someone from the faro team will run faro.
- Gen3 Pipeline for DP0.2
- AOB
Action Items
