No weekly run this week because of LAM datamodel evolution

First feedback on the analysis of the weekly run with input spectra from the scientific catalogs and taking into account the LSF :

• The LSF (even when simply described as 3 gaussian) improves the results (see attached plot)
• The results are quite good for a first run : 88% success rate at the 0.001 error 
• It seems that the estimated variance is quite overestimated, the +/- 1 sigma area is displayed in pink on the spectrum, and seem to be larger that the actual dispersed values.

It seem that (at least) one sky emission line has been forgotten  in the list (or in the process of subtraction), as in some failure on the weekly run an emission line is identified at this location (see  Forgotten_skyline.png )

In the catalog0 file, there is no correspondance between the IDs and the ones in the data processed, therefore we cannot check the results. The problem does not exist with the two other files  (catalog5 and catalog6)

catId=0 is for flux standard stars and sky fibers.

I don't know the mapping of objId, but the input spectra can be found at https://github.com/Subaru-PFS/drp_instdata/tree/master/data/objects/fluxCal .

OK but it does not explain why we don't have any match. Aren't there any of those fibers in the weekly run ?

And it means that we are missing one catalog then ...

Missing catalog spectra are listed in catalog0_list.txt