masterCode - comments

[amartelli]

Dear all, some comments having a look at the code currently in the master. I write here for the record...

ElectronMerger

*Concerning the check for BDT value for the lowpttracks (and the MVA_ID), it is already done in the PATLowPtElectronSeedingEmbedder.cc
https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/ElectronMerger.cc#L159-L163
And the embedding of the variables is also replicated
https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/ElectronMerger.cc#L200-L201
Or it means that PATLowPtElectronSeedingEmbedder is no more needed?
Maybe now that the mva_id is available, we want to cut on this?

*Sorting in pT, as introduced by Mauro originally (I think) was a good idea, because when building pairs one can loop on j = i+1 (i = leading and j = subleading)
https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/ElectronMerger.cc#L208-L212
No?

MuonTriggerSelector

• It seems that SelectedMuons are only probe, am I wrong? And why in that case duplicating the trigger muons?
  https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/MuonTriggerSelector.cc#L85

*https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/MuonTriggerSelector.cc#L207
should be off to save B on the tag side for muon final states

• Are we sure that these should be required for all muons? and not only for the triggering muons? (in any case it is off in the config python right now...)
  https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/MuonTriggerSelector.cc#L202

*concerning the use of innerTrack or bestTrack:
https://github.com/cms-sw/cmssw/blob/bab3e4b74749fb0b3c0b31bfb49daf3b18fc30f8/DataFormats/PatCandidates/interface/Muon.h#L82-L83

TrackMerger

*https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/TrackMerger.cc#L121
We want also muons, maybe can ask 211 or 13 for packedPfcandidates and only 211 for lost tracks, not to accept subleading muon candidates from lost tracks and increase cleanliness

*https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/TrackMerger.cc#L134
Should be off to save the candidate on tag side for muons

*should be called only if useful (currently value set to -10000 in the cfg...)
https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/TrackMerger.cc#L154-L159

*the method DCA might be useful in a separate helper if used somewhere else
https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/plugins/TrackMerger.cc#L189

*Agreed on DCA wrt beam spot. The change to use the triggerMuon vtx was to deal with a same reference when splitting probe from tag in the collections...

** python/electronsBPark**

*these selections do not correspond to what agreed (pT > 0.5 dZ < 1, dR > 0.4)
Plus decide on the use of PATLowPTGsfEmbedder and clean code https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/python/electronsBPark_cff.py#L46-L51

** python/muonsBPark**

• pT > 0.5, no?
  https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/python/muonsBPark_cff.py#L21

python/tracksBPark
*values need update dz < 1 for example... https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/python/tracksBPark_cff.py#L11-L13

• why removing vz? it is not available by default and it can be needed for dZ wrt trigger muon in case...
  https://github.com/CMSBParking/BParkingNANO/blob/master/BParkingNano/python/tracksBPark_cff.py#L29-L31

test
if useful, the is also a ready to use script https://github.com/cms-sw/cmssw/blob/master/PhysicsTools/NanoAOD/test/inspectNanoFile.py
that produces a log like this https://amartell.web.cern.ch/amartell/Analysis/BParking/infoDATA.html