A_LL Binned in Run Regions

 This page presents a study of A_LL vs. pT, plotted over 21 divisions with 50 runs each for the runs considered for analysis in run13. 

The summary is that there is a region of runs where A_LL for 1<=pT<=5 GeV is inconsistent with the other runs; this region correpsonds directly to 
the problematic region in the relative luminosity analysis (run indices 500-600) that still failed some diagnostic tests. This region is the same one 
where we saw the "after-pulsing" shaped effect of rising scaler counts coming out of abort gaps, but was concluded to not be after-pulsing. 

The attached pdf shows A_LL vs. pT where each page is a set of 50 runs.

The run numbers for each page are:

page 1: runs 14096079-14100014
page 2: runs 14100018-14105037
page 3: runs 14105038-14108079
page 4: runs 14108080-14110059
page 5: runs 14110060-14117020
page 6: runs 14117021-14118057
page 7: runs 14118059-14122012
page 8: runs 14122013-14124026
page 9: runs 14124027-14132024
page 10: runs 14132025-14136054
page 11: runs 14136055-14138061
page 12: runs 14138062-14140030
page 13: runs 14140043-14142005
page 14: runs 14142007-14145023
page 15: runs 14145024-14146024
page 16: runs 14146025-14147057
page 17: runs 14147061-14149021
page 18: runs 14149022-14151089
page 19: runs 14151090-14153013
page 20: runs 14153014-14154042
page 21: runs 14154043-14157017

For each run period, I fit the region of 1<=pT<=5 GeV to a constant constrained at A_LL=0, to extract the chi2/ndf for the fit. This image shows
the chi2/ndf for each run region, where the horizontal axis corresponds to the run region via the page number from the table above; 
The image on the right is a zoom in of the image on the left, showing the variation of chi2/ndf around 1. 

For the zoomed in image, cutting out points with chi2/ndf >= 10; the pol0 fit returns chi2/ndf = 0.987966 +/- 0.15

The spike at run regions 10-11 corresponds directly to a "problematic region" in the relative luminosity that I addressed earlier. It is a region
that had worse agreement between E,W,X scaler bit combos as well as worse agreement between ZDC and VPD rellum means. 

Here is the link to the relative luminosity analysis: