VBFHWW Lepton Optimization

Useful links:

Fakes

  • Currently present Wjets mc16a:
efakes:

-361100 /bkg/$(channel)/Wjets/Nom/ev/
-361103 /bkg/$(channel)/Wjets/Nom/ev/

mufakes:

-361101 /bkg/$(channel)/Wjets/Nom/mv/
-361104 /bkg/$(channel)/Wjets/Nom/mv/

Lepton Isolation:

Running instructions:

  • Follow the README to run 36.1/fb lepton isolation scan

Minutes from https://indico.cern.ch/event/794818/

  • Slide 4: What is the selection applied? We are looking at the last BDT bin.
  • In the last BDT bin, you select already VBF like events, maybe it is not optimal to selection.
  • Suggestion: Move to a more inclusive selection, without MVA regions
  • slides20-21-22: It seems that applying a tight cut we will remove a lot of backgrounds while keeping a higher efficiency on the signal. Is it due to the missing W+jets estimation? This could indeed the be case.
  • All scan: Do you apply isolation only on leading lepton? No, at the moment no isolation is applied on leading lepton. This should most probably be revisited to have a proper ID and Anti-ID definitions.
  • Slide 11: what is 40%? Do you apply 40% only on the background, right? Yes, it is 40% of the background. Why 40%? this was a rough estimation of the fake background uncertainty. In the prompt background, we were applying ~10%. Scan using different errors were also performed.
  • Slide 36: Do you plan to check the latest CP recommendation, including PFlow-based WPs? Yes, this will be done
  • Slide 36: Can you make data/MC plot for the input variables?

New Iso to try

HWWElectronsAuxDyn:
4.980 kb 0.948 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCTightTrackOnly_new #compare to FixedCutTrackCone40 #TRY NOW########
4.917 kb 0.954 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCLoose #looking for tighter cuts
5.005 kb 0.954 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCTightTrackOnly_FixedRad
4.961 kb 0.961 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCHighPtCaloOnly #optimizing for low pT
4.971 kb 0.961 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFixedCutPflowLoose #PFlow at a later date
4.961 kb 0.962 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCLoose_FixedRad #looking for tighter cuts
4.922 kb 0.963 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoGradient #it is used in pT>25GeV so should be used as a reference in pT< 25GeV #TRY NOW##########
4.917 kb 0.964 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCTight #Kathrin insisted not useful (confirm the argument later)
4.976 kb 0.966 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFixedCutTrackCone40 # R20.7 optimized WP #TRIED Already#
4.941 kb 0.967 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoGradient_exp #TRY NOW##############
4.971 kb 0.971 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFixedCutPflowTight #PFLOW at a later date#
4.961 kb 0.976 kb 0.003 kb 0.000 289 (B) HWWElectronsAuxDyn.passIsoFCTight_FixedRad #same as FCTight for pT<25GeV (ROI)#
HWWMuonsAuxDyn:
4.940 kb 0.949 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCLoose
4.994 kb 0.959 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutPflowLoose
4.984 kb 0.959 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCLoose_FixedRad
4.984 kb 0.962 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCTightTrackOnly #compare to FixedCutTrackCone40 #TRY NOW########
4.999 kb 0.963 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutHighMuLoose
4.940 kb 0.966 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCTight
5.019 kb 0.966 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutHighMuTrackOnly #TRY NOW########
5.028 kb 0.969 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCTightTrackOnly_FixedRad
4.984 kb 0.976 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFCTight_FixedRad
5.019 kb 0.979 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutHighPtTrackOnly
4.999 kb 0.980 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutHighMuTight #TRY NOW########
4.994 kb 0.995 kb 0.003 kb 0.000 289 (B) HWWMuonsAuxDyn.passIsoFixedCutPflowTight

Added in share/config/aliases/common/aliases-flat.cfg

aliases.lep0_ptvarcone20_TightTTVA_pt1000: [CPFD:Event$(cand)[0].part(0):ptvarcone20_TightTTVA_pt1000]
aliases.lep1_ptvarcone20_TightTTVA_pt1000: [CPFD:Event$(cand)[0].part(1):ptvarcone20_TightTTVA_pt1000]
aliases.lep0_ptvarcone30_TightTTVA_pt1000: [CPFD:Event$(cand)[0].part(0):ptvarcone20_TightTTVA_pt1000]
aliases.lep1_ptvarcone30_TightTTVA_pt1000: [CPFD:Event$(cand)[0].part(1):ptvarcone30_TightTTVA_pt1000]

[status: compiles for now]

>>> import ROOT

>>> f = ROOT.TFile("/eos/atlas/atlascerngroupdisk/phys-higgs/HSG3/R21/PAOD_V18/2LDF/mc/group.phys-higgs.mc16_13TeV.HWW_mc16a_HWWSignal.merge.PAOD_2L.V18.1_PAOD_2LDF/group.phys-higgs.341422.e4210_e5984_s3126_r9364_r9315_p3639.15672673.PAOD_2LDF._000005.pool.root", 'r')

>>> t = f.Get("CollectionTree")

>>> t.Print("HWWElectronsAuxDyn.*cone*")

*****************************************************************************
*Tree :CollectionTree: CollectionTree
*Entries : 7014 : Total = 79354710 bytes File Size = 24426391 *
: : Tree compression factor = 3.24 *
******************************************************************************
*Br 0 :HWWElectronsAuxDyn.ptvarcone20 : vector<float> *
*Entries : 7014 : Total Size= 128199 bytes File Size = 22470 *
*Baskets : 10 : Basket Size= 64512 bytes Compression= 5.68 *
............................................................................*
*Br 1 :HWWElectronsAuxDyn.ptvarcone40 : vector<float> *
*Entries : 7014 : Total Size= 128199 bytes File Size = 25023
*Baskets : 10 : Basket Size= 64512 bytes Compression= 5.10 *
............................................................................*
*Br 2 :HWWElectronsAuxDyn.topoetcone20 : vector<float> *
*Entries : 7014 : Total Size= 128213 bytes File Size = 53872
*Baskets : 10 : Basket Size= 64512 bytes Compression= 2.37 *
............................................................................*
*Br 3 :HWWElectronsAuxDyn.topoetcone40 : vector<float> *
*Entries : 7014 : Total Size= 128213 bytes File Size = 53809
*Baskets : 10 : Basket Size= 64512 bytes Compression= 2.37 *
............................................................................*

Update:

just backed up the loosened Iso for ID leptons as:

cp LeptonIDHelper.py LeptonIDHelper.py.loosenID

TODO: need to put a switch between this loosened config and gradIso in LeptonIDHelper.py

Some useful links

-- SouravSen - 2019-01-09

Edit | Attach | Watch | Print version | History: r17 < r16 < r15 < r14 < r13 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r17 - 2019-05-14 - SouravSen
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Sandbox All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback