Run 9 200GeV Dijet Cross Section Scaler Investigation

First look at the scaler files to get an estimate of the accidental and multiples corrections for the VPD ...

The VPDMB trigger cross section can be found by looking at the VPDMB trigger rate from the MuDsts during a vernier scan. For the VPDMB cross section to be useful, we need an estimation of the size of the accidental and multiples corrections. To do this, we need to look at the scaler files. Zilong provided me with code to read the scaler files and I've taken a first look at the size of the corrections using the runs from fill 101746. All data is from Scaler board 6.

Figure 1: The total number of the seven interesting combinations of the east, west, and coincidence bits for the BBC, ZDC, and VPD summed over all bunch crossings

 

Figure 2:  The bunch crossing specra for each of the seven interesting combinations of the east, west, and coincidence bits for the BBC. The number of events in each bin have been scaled by the total number of events seen from the board for that bunch crossing.

 

Figure 3: Same as figure 2 but now for the ZDC

 

Figure 4: Same as figure 2 but now for the VPD

 

I calculate the Accidentals and Multiples corrections for the BBC, ZDC, and VPD using two different combinations of scaler bits, the TAC Difference scheme and the Global scheme as defined in the Run 9 relative luminosity analysis note.

TAC Diff: East = (0,1,0) + (0,1,1) + (1,1,1) {X,E,W}
              West = (0,0,1) + (0,1,1) + (1,1,1)
              Coin = (0,1,1) + (1,1,1)

Global: East = (0,1,0) + (0,1,1) + (1,0,0) + (1,0,1) + (1,1,0) + (1,1,1)
           West = (0,0,1) + (0,1,1) + (1,0,0) + (1,0,1) + (1,1,0) + (1,1,1)
           Coin = (1,0,0) + (1,0,1) + (1,1,0) + (1,1,1)

Modified TAC Diff: East = (0,1,0) + (0,1,1) + (1,1,0) + (1,1,1)
                           West = (0,0,1) + (0,1,1,) + (1,0,1) + (1,1,1)
                           Coin = (0,1,1) + (1,1,1)

The accidental corrections were calculated according to the formulas:

P_E = (N_E - N_X)/(sum - P_W)

P_W = (N_W - N_X)/(sum - P_E)

P_X = (N_X - (N_E * N_W)/sum)/(sum + N_X - N_E - N_W)

Where the P are the corrected probabilities and the N are the number of scaler counts for the east, west, or coincidences defined above. Sum is the scaler counts for all bit combinations.

The final multiples correction was calculated using:

Mult = -sum * Ln(1 - P_X)

In the figure below, I look at the 11 runs in my analysis which came from Fill 101746. The runs in this fill are: 10134021, 10134024, 10134025, 10134026, 10134027, 10134028, 10134030, 10134035, 10134036, 10134037, 10134044.
 

 
 
Figure 5: Summary of the accidental and multiple corrections for the runs in fill 101746. The values show the ratio: (uncorrected-corrected)/corrected coincidence numbers where the coincidence is defined in the TAC difference scheme (circles), the global scheme (squares), and the Modified TAC difference scheme (triangles). The ratio is shown for the BBC (red), ZDC (blue), and VPD (green). The bottom figure is just a zoom in of the top.

Figure 6: Summary of the accidental corrections only. The values are again the ratio: (uncorrected-corrected)/corrected coincidence numbers. Here the corrected values are the P_X (which are probabilities) multiplied by the total scaler combination sum. The color/symbol scheme is the same as in figure 5.