Version 125 (modified by 6 years ago) ( diff ) | ,
---|
Code Changes
add references to gen particles (added TRef to Particle for Track, Electron and Muon)fix isolationchange default CMS config file: produce PF objects (electrons, muons, charged hadrons ..)sort leptons in pt (all objects written to ROOT tree are now sorted)set by default PF met calculation (instead of of CaloMET) in CMS cardprotect input file from overwrite if wrong command was giveninefficient tracks should deposit energy in the CALOsadd frog visualization (replaced frog by ROOT based event display)include Ht in variablesremove muons from calorimeter (be aware of met)treat equally taus and bs (collection or flag?)include jet constituents in the outputsadd to candidates a variable (double) for isolation + flag.nevents to process specified by the userdefine treatement for charged particle orginating from pile-up for eta>2.5add information about selected events in the outputadd neutralinos in calorimeter (should not deposit)change input array in pileup card for jet pileup subtraction (btagging)write jet energy scale module (in pt,eta)add genmet calculation by default.correct bug in LHEF readers (look for "<events>" string rather)apply logN smearing everywhere?add python code from Christophecommit tuned cardscomment code about b and tauspossibility to compute rho and subtract in different rapidity regions (done bad in my own release)add muons isolated in the UniqueObjecFinder and change muons in tree writerthink about where apply CHS in presence of PUmake Calorimeter module compatible with what is in the papermake study of jesthink about b-jet matching (cf. SWGuideBTagMCTools)add on Delphes website a MinBias.pileup file.think about muons in PFalgorithm? why are they out?in EFlowMerger add TrackPileUpSubtractor/muons instead of MuonMomentumSmearing/muonsadd flag for pile up distribution (gauss,poisson,flat)add flag for isolation (relative/absolute)correct for efficiency/isolation at high pt (for e/mu/gamma)particle-flow optimisation (start from Calorimeter.cc.bk.4, think about dynamical Particle Flow for high pt tracks)look for discrepancies in taus.add charged particles from pu in the eflow met calculation.compute jet energy scale for ATLASremove pile-up subtraction from the missing ET calculationsunderstand what is the correct ECAL resolution for CMSsqrt(energy^2*0.007^2 + energy*0.07^2 + 0.35^2)
orsqrt(energy^2*0.005^2 + energy*0.05^2 + 0.25^2)
independent calos ECAL HCALaddset RandomSeed 0
to all the example cards and close ticket #229develop FCC card based on May CERN WSadd gridMedianEstimatoradd AngularSmearing module from /storage/Delphes_dev/Delphes-3.1.2/modulesinclude Pythia8 fix ticket 288think about how to properly define genets (remove neutrinos, see what CMS does)Cms PileUp card, remove pileup subtraction from MET calculationwrite library example for fast jet/rivetadd a zero bias suppression for the calos and check that everything works fine.write list of hepmc particles post detector simulation before fast jet clustering…develop and finalise FCC detector card.remove muon from jet constituent loop in Example3.Cwrite pf for independent calls with separate photons and Neutral Hadronscalculate dynamically cutoff sigmatrk ~sigmacalo for pfinclude HepTopTaggercheck also resolution e/mu/gamma at high energycheck if basic things can be changed in default CMS/ATLAS conf. (checkMate numbers/MadAnalysis numbers)include vertexing in delphes: https://indico.cern.ch/event/342233/session/2/contribution/1/material/slides/2.pdfcheck whether all formats get saved with same status convention in Delphes output (probably not, but should we do it ?)update the isolation by only using charged hadrons, photons, neutral hadrons.photon conversionsproduce ILC cardwrite mass plot production macro for testing new releasesfix btagging efficiencies (too pessimistic at low pt, too optimistic at high pt)fix TrackCountingBTagging module, seems to be broken in ILD cardfix particle flow at high pthow to treat unknown particles (meaning they are not in root pdg_table)fix tau-tagging by counting tracks (à la checkmate)fix Nsubjettiness to most recent usageimplement CHS based on comparison between zIP smeared and vertex resolution, for more realistic high eta performance.think about how to implement track merging (more complicated than just smearing angles) in high density, as a function of track pt, eta- write real tau-tagging module like btagging (based on id prongs)
- check efficiencies e/mu/gamma at high energy
- fake-track producer (can be done easily like pile-up producer)
add tracking efficiency loss inside jet module- instead of rescaling jets why not rescaling directly pf objects (maybe to match to total gen HT)
- correctly process long-lived particles
- adapt example Python analysis code (DelphesAnalysis) to ROOT 6
- fill code and config file with comments
- change tau tagging module (add bits like b-tagging)
- understand how to treat light fake rate (seems to be too high, due to secondary splittings) - btagging
- PF a` la papas -- treat complex case of non exclusive deposit in ECAL/HCAL
- Simplify Jet substructure , remove TLorentzVector arrays in favor or Jet constituents
- Multithreading?
- recover daughters and mothers when filtering genparticles (important !!)
- Allow for batch jobs in validation
- add jet substructure in validation
- why di-photon mass resolution degrades at high higgs pT?
- flat tree converter? some users find hard to read Root Tree with structure? also need interface then
Note:
See TracWiki
for help on using the wiki.