On the philosophy on lepton isolation

[José]

Dear Delphes authors,

I have a few questions regarding lepton isolation. Reading previous tickets, the source code and the manual gave me some insight, but I would like to confirm the following statements, which I assume to be true:

a) Lepton isolation uses only information from tracks (not from the calorimeter).
b) The sum over tracks is restricted to hadronic tracks (e.g: this would allow for two leptons with similar momentum to be close to each other).
c) The non-isolated leptons are moved into the jet-clustering.

Just to be on the safe side, I copy below the MuonIsolation module from the input card, the electron is the same. My input card is essentially the FCC one, but my "muons" in the output come from the MuonMomentumSmearing module instead from the UniqueObjectFinder (I understand this is not a problem, please correct me if I am wrong).

Best Regards,

José

module Isolation MuonIsolation {
set CandidateInputArray MuonEfficiency/muons
set IsolationInputArray EFlowMerger/eflow
set OutputArray muons
set DeltaRMax 0.3
set PTMin 0.5
set PTRatioMax 0.1
}

[Michele Selvaggi]

Hi,

a) no, lepton isolation uses eflow objects, which are a combination of tracks and towers.
if you want to use only tracks, you should provide as IsolationInputArray the collection TrackMerger/tracks, instead of the default EFlowMerger/eflow.

b) see a), the sum is done over all eflow objects. but two high pt leptons close to each will result in neither of them being isolated.

c) every eflow object, which include isolated leptons are given as input to jet clustering. Only after the jets are found clustered a check is done by the UniqueObjectFinder to remove overlaps and possible double countings according to a priority that goes like photon>electron>muon>jet. In other words if an isolated muon is also forming a jet, that particular jet is removed from the jet collection.

If your muons come from MuonMomentumSmearing, they won't be isolated, which might be an issue depending on what you are trying to do...

Michele