Update 09.20.2018 -- Run 9 pp: Default vs. Efficiency +- 5% With No Errors on Efficiency (R = 0.2 Charged Jets)
Yesterday I checked what happens when I fix the response matrix, but change the efficiency:
So the next step is to simultaneously adjust the efficiency (+- 5%) and the response matrix accordingly. The priors remain fixed across all 3 variants as the particle- and detector-level R = 0.2 charged recoil jet spectra from the Run9 dijet embedding sample (charged hadron triggers with 'eTtrg = 9 - 20' GeV). In all 3 cases, the triggers used for the training samples had 'eTtrg = 9 - 20' GeV.
Note that the statistical errors on the efficiency curves were set to 0 in order to make the errors on the unfolded distributions more transparent.
As you can see, the errors on all 3 distributions are roughly equivalent. And to get a better idea of the comparisons, below are the same plots but with the default Pythia distribution (no adjustment to the tracking efficiency) included.
- dmawxc's blog
- Login or register to post comments