Details
-
Type:
Story
-
Status: Done
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: jointcal
-
Labels:
-
Story Points:2
-
Epic Link:
-
Sprint:AP S22-1 (December)
-
Team:Alert Production
-
Urgent?:No
Description
gen3 jointcal does source selection on the visit-level Parquet output, so various field names are different. This also required some adjustment to how the sourceFluxType config field was used. For now, this is dealt with via gen3-specific configs (see e.g. tests/config/config-gen3.py), but as part of the gen2 deprecation process, we should move those configs into the default JointcalConfig fields and `setDefaults` (where the sourceSelector is configured).
Jointcal currently defaults to the astrometry source selector, but unless that is converted to be gen3 and parquet compatible, we'll need to switch the sourceSelector to science in addition to the setDefaults changes.
Attachments
Issue Links
- blocks
-
DM-20572 Remove "Generation 2" Middleware
- Done
- is blocked by
-
DM-29535 Create gen3 cfht testdata_jointcal repos
- Done
-
DM-29542 Move jointcal cfht tests to gen3
- Done
-
DM-32601 Move remaining jointcal tests to gen3
- Done
-
DM-32584 Create gen3 decam testdata_jointcal repo
- Won't Fix
- is cloned by
-
DM-29012 Create obs_subaru gen3 jointcal pipeline file
- Invalid
- is duplicated by
-
DM-31440 Jointcal can't be run on the RSP
- Done
-
DM-31861 HSC DRP.yaml should not import from jointcal tests
- Done
- is triggering
-
DM-33298 Identify correct set of flags to use for jointcal source selection
- To Do
- relates to
-
DM-29615 Fix connections problems in running Gen3 jointcal on HSC RC2
- Done
-
DM-29821 Compare the jointcal SkyWcs's of the gen2 vs. gen3 w_2021_22 RC2 runs
- Done
-
DM-30891 Create drp_pipe and move all DRP pipeline definitions there
- Done
-
DM-31889 Update testdata_jointcal sourceTable_visit catalogs with new field names
- Done
-
DM-33760 Compare the jointcal SkyWcs's of the gen2 vs. gen3 w_2022_04 RC2 runs
- Done
-
DM-29659 convert astrometrySourceSelector to gen3 or remove it
- Won't Fix
-
DM-29885 Disable jointcal photometry in HSC via config
- Done
-
DM-31145 Update jointcal check_logged_chi2 to reflect new chi2 logging
- Won't Fix
Ok, as a test, I ran the following on this branch:
Gen2 + configJointcalGen2_lam.py:
Gen3:
The diff in the configs for the above is:
These look pretty synced to me.
To "look" at the results, I ran compareVisitAnalysis from pipe_analysis and here is where we stand now for Gen2 vs. Gen3 for jointcal:
First have a look at where things stood with the most recent weekly run w_2021_46:
And now with this branch + Gen2 config file:
Looking even better now! My suspicion is the remaining difference boils down to order-of-input differences.