[#PIPE2D-875] Provide missing b-band dark, fiberprofiles and detectorMap SuNSS calibs

Comment by hassan [ 28/Sep/21 ]

Dark pfsDark-2021-01-15-041400-b1.fits, fiberProfiles pfsFiberProfiles-2021-06-11-063050-b1.fits and detectorMap pfsDetectorMap-063091-b1.fits added to /projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-06-16 CALIB.

All processing done using pfs_pipe2d w.2021.39.

Firstly, the dark was ingested:

ingestPfsCalibs.py /projects/HSC/PFS/Subaru/ --calib /projects/HSC/PFS/CALIB-SuNSS-2021-06-16 /projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-05-04/DARK/pfsDark-2021-01-15-041400-b1.fits --mode=copy --validity 1000 --config clobber=True

Then a DCB-derived detectormap pfsDetectorMap-045744-b1.fits was ingested as a basis:

ingestPfsCalibs.py /projects/HSC/PFS/Subaru/ --calib /projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-06-16

Then the fiberProfiles were constructed:

constructFiberProfiles.py /projects/HSC/PFS/Subaru --calib=/projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-06-16/ --rerun=hassans/pipe2d-875/fiberProfiles_b1 --doraise --cores=1 --id visit=63050..63061 arm=b -c isr.doFlat=False

Finally, a detectorMap was generated using SuNSS exposures of Subaru room lights and ingested.

 reduceArc.py /projects/HSC/PFS/Subaru/ --calib /projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-06-16 --rerun /projects/HSC/PFS/Subaru/rerun/hassans/pipe2d-875/reduceArc_b1_roomlightLinesOnly --id visit=63091..63100 arm=b --configfile=/scratch/hassans/calib/config/reducearc_pipe2d875_roomlights.config --longlog > /scratch/hassans/log.reducearc_b1_pipe2d875_roomlightsOnly

ingestPfsCalibs.py /projects/HSC/PFS/Subaru/ --calib /projects/HSC/PFS/Subaru/CALIB-SuNSS-2021-06-16/ /projects/HSC/PFS/Subaru/rerun/hassans/pipe2d-875/reduceArc_b1_roomlightLinesOnly/DETECTORMAP/pfsDetectorMap-063091-b1.fits --mode=copy --validity 1000 --config clobber=True

The line list was limited to 6 bright lines identified by eye (see attached subaruRoomLights.txt). This provided a good fit compared with using the full range of lines available in the 'standard' line lists provided in obs_pfs, which previous runs lead to a number of incorrect lines being identified, affecting the quality of the fit. This is expected to be addressed once the relative line intensities for each light source are updated the line list set. This is the subject of future work.

The fit quality looks good, with the x and y RMS errors being less than 0.03 px. Visual inspection of key lines (see screenshots attached) provide a sanity check of the fit.

INFO  2021-09-27T13:17:08.460-0400 reduceArc.fitDetectorMap ()(fitDistortedDetectorMap.py:556)- Final fit: chi2=34787.562814 dof=8540 xRMS=0.023609 yRMS=0.023602 (0.001609 nm) from 4300/9381 lines
INFO  2021-09-27T13:17:08.471-0400 reduceArc.fitDetectorMap ()(fitDistortedDetectorMap.py:562)- Fit quality from reserved lines: chi2=137824.630717 xRMS=0.044730 yRMS=0.043179 (0.002944 nm) from 1042 lines (10.0%)
INFO  2021-09-27T13:17:08.475-0400 reduceArc.fitDetectorMap ()(fitDistortedDetectorMap.py:709)- Softening errors by x=0.023144, y=0.023065 pixels (0.001573 nm) to yield chi^2/dof=1
INFO  2021-09-27T13:17:08.498-0400 reduceArc.fitDetectorMap ()(fitDistortedDetectorMap.py:715)- Softened fit: chi2=8538.190491 dof=8540 xRMS=0.023895 yRMS=0.023926 (0.001631 nm) from 4300 lines
INFO  2021-09-27T13:17:08.510-0400 reduceArc.fitDetectorMap ()(fitDistortedDetectorMap.py:721)- Softened fit quality from reserved lines: chi2=31172.169360 xRMS=0.044718 yRMS=0.043373 (0.002957 nm) from 1042 lines

Comment by price [ 28/Sep/21 ]

Did you iterate on the fiberProfiles and detectorMap? The two are coupled.

Comment by price [ 28/Sep/21 ]

Also, all the slit offsets are zero, so you need to run with --config fitDetectorMap.doSlitOffsets=True.

[PIPE2D-875] Provide missing b-band dark, fiberprofiles and detectorMap SuNSS calibs Created: 30/Jul/21 Updated: 28/Sep/21 Resolved: 28/Sep/21
Status:	Done
Project:	DRP 2-D Pipeline
Component/s:	None
Affects Version/s:	None
Fix Version/s:	None

[PIPE2D-875] Provide missing b-band dark, fiberprofiles and detectorMap SuNSS calibs Created: 30/Jul/21 Updated: 28/Sep/21 Resolved: 28/Sep/21