Run 9 Simu: Primordial kT Investigation
Updated on Tue, 2014-09-30 12:24. Originally created by pagebs on 2014-09-29 01:32.
Look at how changing the primordial kT distribution in the simulation affects my results ...
To test the effect that the primordial kT has on the simulation, I reweight the default which has a gaussian kT distribution with sigma = 2 to a distribution with sigma = 1. This is done by looking at the x and y components of momentum for the parton in each proton which initiates the hard scattering (3 and 4 in the pythia record). The px and py distributions are gaussian and relatively independent of particle type, pt bin, or which proton they come from. Thus, the weight factor used to correct the simulation is just the ratio of two gaussians, one having a sigma = sqrt(2) and the other having sigma = 0.5*sqrt(2). So the weight factor (over 4 dimensions for px and py for two particlse) is:
W_kT = 16*Exp[-0.75*((pt_3)^2 + (pt_4)^2)]
This factor multiplies the already existing weight factor from the partonic pt bin and z-vertex reweighting.
Figure 1: Comparison between data (blue) and simulation with kT = 2.0 (red) and the simulation with kT = 1.0 (green) for the L2JetHigh (left) and JP1 (right) triggers. The ratio plots show data/simulation for the kT=2.0 simu (blue) and the kT=1.0 simu (red). Comparisons for jet pT, jet eta, and dijet mass are shown.
Figure 2: Pythia / NLO dijet cross section ratio for the pythia sample with kT = 2.0 and the sample with kT = 1.0.
To test the effect that the primordial kT has on the simulation, I reweight the default which has a gaussian kT distribution with sigma = 2 to a distribution with sigma = 1. This is done by looking at the x and y components of momentum for the parton in each proton which initiates the hard scattering (3 and 4 in the pythia record). The px and py distributions are gaussian and relatively independent of particle type, pt bin, or which proton they come from. Thus, the weight factor used to correct the simulation is just the ratio of two gaussians, one having a sigma = sqrt(2) and the other having sigma = 0.5*sqrt(2). So the weight factor (over 4 dimensions for px and py for two particlse) is:
W_kT = 16*Exp[-0.75*((pt_3)^2 + (pt_4)^2)]
This factor multiplies the already existing weight factor from the partonic pt bin and z-vertex reweighting.
Figure 1: Comparison between data (blue) and simulation with kT = 2.0 (red) and the simulation with kT = 1.0 (green) for the L2JetHigh (left) and JP1 (right) triggers. The ratio plots show data/simulation for the kT=2.0 simu (blue) and the kT=1.0 simu (red). Comparisons for jet pT, jet eta, and dijet mass are shown.
Figure 2: Pythia / NLO dijet cross section ratio for the pythia sample with kT = 2.0 and the sample with kT = 1.0.
»
- pagebs's blog
- Login or register to post comments