P07ib : CuCu200-MB-RFF
Instead of using the FitSlices command from root to evaluate the pull, I took the distribution of the pullXY and pullZ for different bins in P : &delta P =0.2 GeV/cThen for each bin, a double gaussian was better than a simple or a simple gaussian+pol2
pullXY for Nfitted = 3
 |
 |
 |
| Figure 1 : 0.1 <= P < 0.3 | Figure 2 : 0.9 <= P < 1.1 | Figure 3 : 1.7 <= P < 01.9 |
figure 2 : &sigma1 = 0.388 ; &sigma2 = 0.6477
figure 3 : &sigma1 = 0.4118 ; &sigma2 = 0.6924