-
Type:
Story
-
Status: Done (View Workflow)
-
Priority:
Normal
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: None
-
Component/s: None
-
Labels:
-
Sprint:PreEngRun15Mar, Engrun15Mar, postRun15Mar_or_preRun16May, EngRun16May
We have observed quartzes with every fourth fiber exposed (mod4, a.k.a. "shaka"), with slit dithers to measure the profile at high resolution. To deal with fibers that cannot move, we have also observed the quartzes with all fibers hidden, using the same slit dithers. These "dot-roach darks" need to be subtracted from the mod4 exposures.
- relates to
-
PIPE2D-1405 Generate detectorMaps for Mar. 2024 engineering run data
-
- Done
-
Kiyoto Yabe, could you please confirm that you understand how to create profiles using the new code (from the posted examples), and that the new profiles are of reasonable quality for inclusion in the calibs (from the attached QA plots)? I don't think it's necessary for you to review the code changes.
Sorry for being slow. At least, I confirmed that I could make the same profiles with those commands and the branch. I understand what those try to do but probably need to look into the codes to deeply understand what those actually do. I don't have any expertise to do the code review. By the way, is there any reason why the first 2 visits in allHidden for each group (e.g. visit=106946..106947 for group1) are not used?
I believe the first two visits have the same slit offset, so they don't add value to the process. Perhaps we could try coadding them in the future (so long as the visits are consecutive and the slit hasn't moved).
yeah, the dither sequence has always 7 visits with those positions : [0, -0.4, -0.2, 0, 0.2, 0.4, 0]
and we always add couple of visit in 0, so that we get 5 duplicate of the 0 position.
Those were meant to be coadded, unless we show they are different enough, I don't think so but worth checking.
Thank you. I made profiles by adding additional visits but the results don't change at all as far as I see the some statistics plots.
Reducing nominally-identical data would provide useful QA; if they are really consistent we'd gain S/N and robustness to CRs by using all the data.
I modified how profiles are constructed: we measure the profiles for the different groups (different fiber settings) separately and then merge them. Below are the command-lines I used to construct the profiles from Run 14 (December 2023) and Run 15 (March 2024). I also added some QA metrics and plots for evaluating the profiles, which I'll attach here.