1 / 13

tth, (h → bb): Cuts-Based analysis status/progress

tth, (h → bb): Cuts-Based analysis status/progress. Chris Collins-Tooth 06-Feb-2007. Samples. Using Common Analysis, v00-01-05 (see TWiki) https://uimon.cern.ch/twiki/bin/view/Atlas/TthhbbAnalysis - ttH_rootples_Example_Analysis Signal tth(h → bb): 5340 (forced l + )

jeslyn
Télécharger la présentation

tth, (h → bb): Cuts-Based analysis status/progress

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. tth, (h→bb): Cuts-Based analysis status/progress Chris Collins-Tooth 06-Feb-2007

  2. Samples • Using Common Analysis, v00-01-05 (see TWiki) • https://uimon.cern.ch/twiki/bin/view/Atlas/TthhbbAnalysis - ttH_rootples_Example_Analysis • Signal tth(h→bb): • 5340 (forced l + ) • 5341 (forced l - ) • Backgrounds used: • 5855 AcerMC gg->ttbb (QCD) • 5214 AcerMC gg->ttbb (EW) • 5212 MC@NLO/Herwig ttjj (jj=bb removed).

  3. Cuts Used (see v00-01-05) • Lepton: • e: Pt>25 GeV; EtCone<15 GeV;isEM 0x4FF; auth=1||3. • .mStaco: Pt>20 GeV; 2/dof<50 < ; EtCone40<50 GeV. • Jets: 6j; Pt>20 GeV; ||<5. • bJets: 4bj; bjetWeight 5. • Neutrino Solution: use colinear appx if no solution. • Sample 5212 (ttjj) only: remove if 4 unique bquarks • MWlep/had: 80.4  25 GeV. • Mt: 175  20 GeV. • Background + Combinatorics: Pt(J12+J22)>75 GeV (see later)......

  4. Signal and BG Efficiencies. • Latest “full sim” is Compared to Cammin [ATL-PHYS-2003-024] • Applied same cuts as Cammin and saved v00-01-05 • On the whole encouraging. • Problem with ttjj • many events had >=4 unique b-quarks • Had to remove them, as ttjj(jj=bb)  ttbb

  5. Signal and BG Efficiencies. higher jet mult. in MC@NLO? • Latest “full sim” is Compared to Cammin [ATL-PHYS-2003-024] • Applied same cuts as Cammin and saved v00-01-05 • On the whole encouraging. • Problem with ttjj • many events had >=4 unique b-quarks • Had to remove them, as ttjj(jj=bb)  ttbb

  6. Signal and BG Efficiencies. higher jet mult. in MC@NLO? cut on 2 tops too harsh? • Latest “full sim” is Compared to Cammin [ATL-PHYS-2003-024] • Applied same cuts as Cammin and saved v00-01-05 • On the whole encouraging. • Problem with ttjj • many events had >=4 unique b-quarks • Had to remove them, as ttjj(jj=bb)  ttbb

  7. Signal and BG Efficiencies. higher jet mult. in MC@NLO? 25% 47% cut on 2 tops too harsh? 15% 30% 17% 47% • Latest “full sim” is Compared to Cammin [ATL-PHYS-2003-024] • Applied same cuts as Cammin and saved v00-01-05 • On the whole encouraging. • Problem with ttjj • many events had >=4 unique b-quarks • Had to remove them, as ttjj(jj=bb)  ttbb

  8. Moving on..... • Look at signal and BG normalised to 30 fb-1 • Problem: BG samples not big enough! Use AF? • No sensible distributions yet! – Can’t do shape analysis!

  9. Optimisations/Cuts.. • Lots of cut optimisations to consider; also new cuts like Pt(J12+J22)>75 GeV on bJets forming Higgs • I will add this cut to the CVS-based analysis

  10. A Cut: Pt(J12+J22)>70 GeV on bJets forming H • Red circles: signal (+ black dot = “correct” bjets). • Others: backgrounds #5855 ttbb[QCD] #5340 poslep #5341 neglep Pt J1 Pt J1 Pt J1 Pt J2 Pt J2 Pt J2 #5214 ttbb[EW] #5212 ttjj

  11. “Higgs bJets”: order by bjetWeight • Currently, analysis in CVS produces Higgs from highest Et bJets if >2 remain after t/tbar. • Tried using bjetWeight to order the jets if >2 “left over”. • No difference really, as most (~99%) of the selected events have only 2 bJets! • we have to improve how we reconstruct the tops to beat combinatorics!

  12. Common Analysis: things I don’t like... • Removal of ttjj where jj=bb • Truly awful to devise algorithm to do this in tthhbbClass.cxx (!) • Settled on removing events with 4 or more UNIQUE TRUTH b-quarks in ttjj. • Typical Herwig b->bMeson->b: check mother, grandmother etc. to see if b is unique • We also see eg: b-g-g-g-g-g-bbar where ‘b’ is mother of ‘bbar’, and ‘b’ has no mother. • Output ntuple ‘improved-data’ has a new attribute to allow one to remove these events for ttjj: “my_truthUniqueBQuarks” (remove ttjj event if >=4). • doplots.C is a root macro in CVS and cuts out these events • runtth.sh makes 10 ntuples for samples 5212 • the list of input files is too large for my shell (zsh) • these have to be merged manually into “5212All.root” • merge: TChain ch("improved-data");ch.Add("5212*.root");ch.Merge("5212All.root"); • top-mass window seems too harsh. • jet calibration.. is it even done??

  13. Other things to implement • Add soft muons(+n??) back to Jets. • Jet energy corrections based on v12. • See if there’s any way of improving combinatorics. • D=0, instead of colinear solution. • Improve Whad light jets (use 3+ jets?- Simon?). • S/BG discrimination with Neural Net (Helen). • Bill’s fitting code. • Others people would like to see / I have forgotten?? • Volunteers????? • Not enough events for a BG shape analysis.

More Related