Update BJT and JetUncertainties to fix the bug for the consolidated W/top tagger recommendation
For jets that not pass the W/top tagging requirement, we apply the SF(ineff): SF(ineff) = (1 - SF*efficiency) / (1 - efficiency).
We have a NP varying the efficiency value. But in case of SF=1.0, the efficiency variation is not propagated to SF(ineff). We agreed to use SF for signal jet in that case.
What we want to add in this MR:
- add sigeffSF variable, decorated to the jet in BJT.
- add the accessor to sigeffSF in JetUncertainties so that we can use it only when we calculate the efficiency variation NP for jets with SF=1.0.
Added a function getTruthLabelStr(jet), which was originally a part of getSF(jet) function. Updated the argument of getSF(jet) to getSF(jet, truthLabel). Then, for example,
SF = getSF(jet, truthLabel);
sigeffSF = getSF(jet, "t_qqb");
Edited by Takuya Nobe