All
- Task review
- SPIE faro paper
- faro failure due to jointcal bug has been fixed (
).
- Jeff has a PR for using ForcedSource table in faro.
- Was going to enable switching between afw and parquet Source tables, but appears it will take more work and not worth including at the moment.
- Do we need to keep the ability to load all the (parquet or afw) source tables and match them? Yes.
added in backlog to remind us to shift catalog-matching over to parquet tables at some point.
- Should we implement reading the parquet source tables, or just keep the current code which reads the afw tables? Keith: high priority is computing metrics on forced photometry.
Bugs, issues, topics, etc of the week
Data processing campaigns
- AuxTel
- AP DiffIm
- First step in closing the loop from observing to real-time feedback on the mountain.
- Have 1 sq deg of imaging (SDSS g band) with auxtel in hand. Once Merlin has everything sorted (still finalizing the gen3 processing), then Erik will process the data. Then hand over to this meeting for analysis and metric computation.
- After this runs at NCSA, then will try to get it running at the base (antu commissioning cluster).
- Should start thinking about what metrics we want to compute for auxtel in real-time, but not urgent yet. Aim for the auxtel run in December?
- Keith: S/N as a function of magnitude, could help in running the observing process? Erik: can't process fast enough yet. But could develop on HSC data and then have ready.
- Observing run next week, still refining the data-taking.
- Quantities automatically generated by SFM: PSF model, photometric zeropoint and uncertainty, astrometric zeropoint and uncertainty, some sky brightness quantification
- SIT-Com should help to define what is needed for commissioning needs
- Compare with what Operations wants to define survey quality visits / detector images (Lauren)
- There are raws, processCCD metdadata, visit summary files, logs, headers, and metadata ... we should distinguish between numbers that are generated before and after processCCD. Headers and metadata generally discouraged as the primary place to get data of long-term value.
- RHL: Question of running Faro on calibs? We should come back to this topic next week.
Reprocessing status and metrics review
- Nightly CI with new rc2_subset
- RC2/DC2 reprocessing epic :
- w_2021_38 RC2:
- w_2021_36 DC2 :
- No updates on reprocessing.
- Pulled up nightly metrics dashboard, one day where the metrics changed and then went back to "normal". Ellipticity residuals and the stellar locus width. Repeatability stayed the same. October 23rd.
Faro for DP0.2
- Fall 2021 epic
- Backlog epic:
- Erik Dennihy Jeffrey Carlin Status of conversion to using parquet files as input to faro
- Talked about this at the beginning of the meeting. Erik and Keith's tickets on this topic are merged.
- TEx metric: turning out to be difficult to do direct comparison between the parquet version and the src table version. Dependent on analysis contexts. Jeff and Erik will connect on this for cowork.
Potential co-working session ideas here
