130 likes | 231 Views
B J/ y p (K ) analysis on the full data sample. Francesco Fabozzi, Luca Lista Università and INFN - Napoli. Analysis status. September collaboration meeting PDFs parameters from data sidebands retuned with full data sample Still parameters from SP3 MC
E N D
B J/y p (K) analysis on the full data sample Francesco Fabozzi, Luca Lista Università and INFN - Napoli
Analysis status • September collaboration meeting • PDFs parameters from data sidebands retuned with full data sample • Still parameters from SP3 MC • Preliminary results on branching fraction and asymmetry • Fit is OK in describing the signal region • Not following data behaviour for DEp ~ –110 MeV • 1.8 s asymmetry effect in J/yK channelmostly coming from 2001 data • Now • another iteration performed • SP4 MC • Improved PDFs description F.Fabozzi
DE PDF for the signal • Same DE PDF is used in J/y p, J/y K hypotheses • Fit with sum of two Gaussians • Determination of m(core)-m(tail), sigma(core)/sigma(tail), Area(core)/Area(tail) m(c) m(t)(MeV) s(c)/s(t) A(c)/A(t) J/y mm -2.4 0.9 0.40 0.03 0.92 0.01 J/y ee -3.4 1.0 0.45 0.02 0.84 0.02 SP4 MC J/y mm events J/y ee events F.Fabozzi
D (=DEK -DEp) PDF for the signal • Good data-MC agreement SP4 MC • Different parameters for J/y ee, J/y mm events F.Fabozzi
Peaking background • Number of peaking background events are extracted from SP4 Inclusive JPsi MC (650K events) Npeaking (81.9fb-1) = 40.4 7.2 R = 0.014 0.002 mES F.Fabozzi
Improved fit • In our PDFs we have always neglected the effect of the cuts defining the data sample • |DEp|<-120 MeV && |DEK|<-120 MeV • OK if sample ~ 0 near the cut values • With more statistics we need to improve PDFs including this effect Data sideband B J/y p B J/y K DEK< 120 MeV DEp> -120 MeV F.Fabozzi
Improved signal PDFs • Example: PDF(DEK, D | K): • f (DEK| K) g(D | K)if DEp> -120 MeV • 0 if DEp< -120 MeV • f PDF is the signal (core + tail) Gaussians • To characterize g we study the projection of the PDF on the D axis: • Then we obtain g: In the old fit: g(D|K) = G(D|K) F.Fabozzi
Improved background PDFs (I) • Example: PDF(S, D | bkd) • Not easy as in the previous case, since both S and D PDFs are affected by the cuts • Simpler if we change variables: (S, D) (X, Y) • Cuts corresponds to oblique lines in the (S, D) plane • Cuts correspond to vertical lines in the (X,Y) plane B A C D A B C D F.Fabozzi
Improved background PDFs (II) • Characterization of X and Y PDFs • X distribution fitted with a line • Y distribution fitted with a multi-functional PDF 2nd order poly Gaussian 2nd order poly F.Fabozzi
DEp projection OLD FIT NEW FIT • Sample to fit: 7854 events (4053 ee, 3801 mm) • Now the fit describes also the behavior around –110 MeV J/y ee events OLD FIT NEW FIT J/y mm events F.Fabozzi
Branching ratio summary Np = 234 19 NK = 4521 70 Nbkd = 309859 ppee = 0.485 0.041 pKee = 0.4641 0.0077 pbkdee = 0.5941 0.0094 Br/Br (%) ee5.42 0.66 mm4.98 0.56 All 5.19 0.43 statistical error only Br. ratio in different sub-sample rpK= -0.015 F.Fabozzi
Asymmetry results rpK=-0.02 Ap AK Abkd All 0.116 0.087 0.028 0.015 0.019 0.019 Run1 0.01 0.17 0.011 0.029 0.040 0.037 Run2-2001 0.30 0.14 0.038 0.024 0.017 0.031 Run2-2002 0.01 0.14 0.031 0.027 0.003 0.034 F.Fabozzi
Conclusions • Complete iteration of the analysis done • All SP4 MC used • Improved PDFs both for signal and MC • Improved DEp projection • Study of the systematics is ongoing • Systematics from PDFs parametrization done • sPDF(Np)% = 2.0%, sPDF(NK)% = 0.1% sPDF( Brp/BrK )% = 2.0% • sPDF(Ap)% = 1.0%, sPDF(AK)% = 0.3% • A comparison of the fit on old and new run1 is needed • A new BAD must be released • Hopefully by the end 2002 / early 2003 F.Fabozzi