Details
-
Type:
Story
-
Status: Done
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: meas_algorithms
-
Labels:None
-
Story Points:1
-
Epic Link:
-
Sprint:AP F21-6 (November)
-
Team:Alert Production
-
Urgent?:No
Description
Often I have wanted to be certain that I'm loading the refcat I think I am: there are a few different ways to configure it, and I've gotten them tangled. The refcat loader has an INFO level "Loading reference objects from region..." message that should also include the name of the refcat it's loading to help with such debugging. Lee Kelvin's jointcal debugging this morning was the final impetus to do this, but I think it will be generally useful.
Attachments
Activity
This works very nicely during my testing with some Merian DECam data, for example:
old:
====== Now processing astrometry...
|
Loading reference objects from region bounded by [128.82970545, 132.26135235], [-4.04593700, -0.61713353] RA Dec
|
...
|
====== Now processing photometry...
|
Loading reference objects from region bounded by [128.82970545, 132.26135235], [-4.04593700, -0.61713353] RA Dec
|
new:
====== Now processing astrometry...
|
Loading reference objects from gaia_dr2_20200414 in region bounded by [128.82970545, 132.26135235], [-4.04593700, -0.61713353] RA Dec
|
...
|
====== Now processing photometry...
|
Loading reference objects from ps1_pv3_3pi_20170110 in region bounded by [128.82970545, 132.26135235], [-4.04593700, -0.61713353] RA Dec
|
For reference, the above log outputs were generated from runs instigated using the following command:
pipetask --long-log run --register-dataset-types -j 12 \
|
-b $REPO --instrument lsst.obs.decam.DarkEnergyCamera \
|
-i DECam/defaults/merian,u/lskelvin/drp-pipe/testing \
|
--output-run u/lskelvin/drp-pipe/jointcal-noticket \
|
-p $DRP_PIPE_DIR/recipes/DECam/Merian.yaml#jointcal \
|
-d "instrument='DECam' AND skymap='hsc_rings_v1' AND tract=9071 AND patch=0"
|
Only a minor comment on the PR. During testing running the jointcal task, the new code ran 7s slower (~2%). I put this down to variable load on my test machine, and likely not something to worry about. Otherwise, this looks great, thanks! As Jenkins hasn't complained, I think this looks good to merge to me.
Lee Kelvin: this ticket should help you with your "what refcat did I load?" debugging. See the two separate commits: I cleaned up some of the gen3 config stuff along the way, so it was slightly more than completely trivial.