[Hps-analysis] per-file beam charge

[Sho Uemura]

Sebouh found a problem causing 20% undercounting of integrated charge for 
runs where every file is unblinded ("calibration" runs). I fixed this and 
updated the spreadsheet. Regular physics runs are not affected.

I also found two more golden runs, and identified files with SVT DAQ 
errors. The golden runs table has been updated to exclude those files.

On Fri, 16 Oct 2015, Sho Uemura wrote:

> Another new sheet ("golden runs") has a list of 18 production runs. I'm still 
> looking at these runs, but they comprise 80% of the lat147 data and more than 
> half of the 0.5 mm data.
>
> The unblinded recon LCIO and DST files for these runs have been copied to 
> SLAC (/nfs/slac/g/hps3/data/engrun2015/pass3/golden_recon, golden_dst).
>
> The table includes an estimated luminosity. This uses the as-built target 
> thickness given in Clive's HPS note 
> (https://misportal.jlab.org/mis/physics/hps_notes/viewFile.cfm/2015-008.pdf?documentId=11). 
> Rafo pointed this out to me, and it's important since it is 8% thinner than 
> the design thickness of 0.125% X_0.
>
> Run 5725 is known to have an SVT DAQ error, which is not accounted for in the 
> integrated charge. It's possible other runs are affected. Pelle is checking.
>
> On Tue, 13 Oct 2015, Sho Uemura wrote:
>
>> The run spreadsheet has a new sheet ("filesummary") with per-file totals of 
>> integrated charge (gated and ungated, with bias on, with bias on and SVT at 
>> nominal position).
>> 
>> These numbers allow people to normalize their blind analyses with something 
>> better than 0.1*(total run luminosity).
>> 
>> The file start and end times are taken from TI timestamps; bias-on and 
>> position-good periods, Faraday cup current and livetime are from MYA. I'm 
>> using current*livetime in lieu of the actual gated Faraday cup scaler since 
>> that's not available in MYA; ultimately we should use the gated Faraday cup 
>> scaler.
>> 
>