Section A)-----------------------------

Section B)-------------------- Weighting of ZB events for W embedding

Fig 3a,b,c shows different characteristic of runs & fills used for W-analysis. 

Fig 3a shows ZDC rate (kHz) for all used runs from all used fills (black histo), yellow marks runs from which
ZB events will be used in embedding. Blue marks 3 luminosity ranges for which different weight of ZB events will be assigned.

Fig 3b shows # of ZB events w/ ADC values per fill. The same runs as in fig 3a. The total is 12.5k events useful for W-embedding.
To simplify the process 2 fills in the middle were rejected. Only 'yellow' fills will be used.

Fig 3c shows # of W-candidates passing all W-cuts. The total is ~800. Those are the same events as show in 'preliminary' Jacobian peak.
I used any 2x2 ET to retain more stats.

 Fig 4a,b,c shows how final ZB event weight was assigned to different runs 
There will be 12.2K ZB events from 427 runs used in W-embedding.

Fig 4a shows # of ZB events vs. ZDC rate.

Fig 4b shows # of W-events vs. ZDC rate.

Fig 4c shows the ratio of W-events/ZB-events vs. ZDC rate, scaled with arbitrary constant to set it to close to 1. for the most probable ZDC rate.
I decided to split ZDC rate at 3 ranges (marked by the blue lines) where the ratio is approximately constant (red lines).

Conclusion for W-embedding.

1. The pool of 12.2k ZB events [3 LISTs of ZB daq files] should be sampled non-uniformly,
   depending on the ZDC rate interval of <80 kHz, <140 kHz, or >140 kHz. nRuns= 427  nZBeve L/M/H= 2634 8547 991,  sum= 12172
2. The relative probability of ZB sampling should be of 0.68, 1.0, or 1.1 respectively.
3. the total # of usable ZB events is 12.2k. For the embedding of  (40K) Pythia W-events we will reuse ZB events ~4 times.
   For the future embedding of the filtered QCD events (~500K) every ZB event will be used ~50 times

Section C)-------------------- Distribution of 40K thrown W-events vs. RHIC fill

Table 1.

Section D ------------- pilot embedding of single electrons

Embed 500 e+ and 500 e- in the ZB events from 4 RHIC fills.   

Use y2009a geometry, beam line offset as stored in DB for given fill. Kumac+ ZB daq file list is in attachment C.

 Generating     1  POSITRON events in interval:

             20.00 < pT  <    50.00

             -1.00 < eta <     1.50

             0.000 < phi <    6.283

               0.0 <  Z  <      0.0