- Some of our metrics have been dropped and missing from the CI jobs since ~ 1m month.
- TE1/2 now operate on coadds and the CI jobs only calculate matched visit metrics. No coadds in the CI dataset.
- Robert Lupton to as Yusra about the best course of action
- PA1 is producing NaNs, missing for about a month. 27 days ago the algorithm was updated. The algorithm requires >= 3 visits (to calculate an RMS).
- Number of visits that faro used was set to test parity with validate_drp. This can be relaxed now.
- When we move to parquet?
- w_2021_14 (RC2)
- w_2021_16 (DC2 gen3) (thanks Dan)
- Draft of Characterization Metric Report for v22: https://dmtr-311.lsst.io/v/DM-29601/index.html (results of w_2021_14 9RC2)
- Large changes in teh TE1/2 values.
- Monika has run faro on w_2021_18 (RC2) and using BPS with no errors.
- Robert Lupton/ Jeffrey Carlin Check if there is a flag that allows us to run faro only on isolated point sources (i.e not deblended by scarlet) This may change the TE1/2 metric values
- Jeffrey CarlinRerun faro with the deblending flag and compare the TE1/2 outputs
Review progress on the current development epic
Backlog epic:
