BlueJeans: https://bluejeans.com/426716450
attending:
Topics for discussion:
- Gen 3 conversion (
):
- diffim soon to be merged
- unit tests soon to be merged
- end to end run blocked on decam crosstalk–Eric suggested simply disabling it for now, but we need to figure out whose job it is to fix crosstalk
- Precursor processing (Meredith Rawls ):
- reviews, butler bugs, figures for Eric
- ran HSC but it blew up a lot of sources–hasn't had a chance to dig in to it yet
- report from NCSA (Unknown User (emorganson) et al):
- nothing today, kid duty
ap_association/SDM/fakes (Unknown User (cmorrison)):- got fakes injected in all of HiTS
- completeness for calexp-injected sources looks quite good
- coadd or both-injections look less good–turns out some of the coadd PSFs return NaNs when computing aperture fluxes.
- rough exclusion of the missing region gets us back in the 90% completion range
- DCR (Ian Sullivan ):
- got three pages of comments back from Friday's practice talk, working through those
- thinking about what to say about measurements of quasars
- crowded fields probably going to be left on the floor; first pass didn't look great (even in sims with constant seeing), which may not be surprising given the pixel representation but is a concern.
- Image differencing (Unknown User (gkovacs) ):
- working on merging Gen 3
- met with RHL & discussed decorrelation afterburner; confirmed expectations, made a few tickets
- Algorithms Workshop prep (Eric Bellm ):
- starts tomorrow
- Algorithms Workshop: Diffim Plenary Outline
- major tasks:
- diffim plenary talk (Eric):
- walked through Thursday, responded to comments
- came together pretty well ultimately–both completeness and raw purity are pretty good out of the box
- will point people to slides once final figures from Meredith are in
- HSC Cosmos
- skipping showing this because the detections were so bad
- update HiTS census
- selecting DIASources at random and classifying proved pretty fruitful. Some obvious low-hanging fruit to clean up
- fakes
- thanks to help from Chris we got this run (see above for discussion)
- Eric thanks everyone for their hard work getting everything together the last few weeks, especially given all the upheaval in Seattle. We've got some nice results to show and have put together a first pass at some valuable infrastructure. After the workshop we can get some of this stuff cleaned up and off ticket branches (especially Eric).
- diffim plenary talk (Eric):
- Review CI (https://chronograf-demo.lsst.codes/):
- Chronograph error has occurred, we could
- AOB
- none