Corrections to ptmiss
Summary
This MR includes soft jets in the computation of the type 1 correction to ptmiss (#32 (closed)) and enables systematic variations in the unclustered component of the ptmiss. The first change has required a substantial rework of parts related to jets and ptmiss. The computation now starts from raw ptmiss instead of the default type 1 corrected one. JECs need to be evaluated, and files with their definitions are added to the repository and specified in the master configuration. Access to JECs, their uncertainties, and JER smearing have been moved to a dedicated class JetCorrector, which is used from JetBuilder.
Validation
I checked the final mT histograms produced by the main analysis, comparing them against results with b9c144a7. I did it for data (full DoubleMuon sample) and simulation (2M events from the qq → ZZ sample). The difference between the histograms in reasonably populated bins is within 7% (typically much less) in both cases.
I checked the variation of the unclustered ptmiss in simulation. It seems to be consistent with fluctuations, i.e. there is almost no real effect.