The actual valid visits are 76389..76396^76398, as visit 76397 is not an SPS exposure.

Note that for those valid visits, a 100W halogen lamp was used, replacing a 50W halogen lamp. The latter lamp was used for earlier data taken during Sept 2021 for a similar experiment (visits 68424..68427). See also PIPE2D-902.

June 2022 data processed by DRP using the following command line:

reduceExposure.py /projects/HSC/PFS/Subaru --calib /projects/HSC/PFS/Subaru/CALIB-20220630 --rerun hassans/dev/pipe2d-1069-20220812a --id visit=76389..76397^76398 arm=b^r -c isr.doFlat=False isr.doWrite=True repair.doCosmicRay=False

The Sep 2021 data were re-processed following the same prescription.

The integrated flux for each fiber across the r band were computed from the pfsArm data and scaled according to the following formula: (flux-refFlux)/np.median(refFlux) where flux is the original integrated flux array (one element per fiber), refFlux and the corresponding integrated flux array for a 'reference' visit, 76398 with instrument rotator InR of 90deg, from the same June 2022 group of visits.

For each visit, the scaled flux for each fiber is plotted in PFI_mm coordinates using the cobra design position taken from the Grand Fiber Map, as shown in PFI_scaledflux_June2022.png.

The data from the Sept 2021 run were plotted similarly, using the reference visit 68426, again with InR=90deg. See PFI_scaledflux_Sep2021.png.

The structure that was noticed originally in the Sept 2021 data is unfortunately still apparent in the June 2022 data. Look in particular at visit 76391 (InR=-90deg) in the June 2022 data, and compare that with that for visit 68424 (same InR) in the Sep 2021 data.

Jupyter notebook used committed to github: https://github.com/Subaru-PFS/dev_2ddrp/blob/master/notebooks/PIPE2D-1069%20(June2022%20PFI%20non-uniformity).ipynb