QCD%20Meeting

1
QCD Meeting July 23, 2004
2
Asymmetry plots in beam-? and beam-K system
(ignores jets) Polarization allowed under parity
conservation is in the y-direction (second from
top).
Solid lines Data Dashed lines 20 hybrid
MC-events/real-event scaled by 1/20
3
Fits Without and With Mass Constraints
4
(No Transcript)
5
Work in Progress and Planned Summer-Fall, 2004

• ? Mass-constrained fits 2-?, 2-Ks, 1-ee.
• ? Sum of weights over solutions 1.0 per event.
• Mix MC events for events with both K and ?
  solutions
• as in the data, weighted by relative
  P(?2)P(Decay).
• Further study of sailor-cowboy problem.
• ?Add fits with mass constraints displaced by ?20
  MeV.
• Background subtraction.
• Decide whether analysis is viable or not.
• If so, do polarization in the other two
  coordinate systems.

Additional Tasks

1. ? Use ??2 ?2(MassConstraint) -
   ?2(NoMassConstraint)
2. Random 2-track crossings.

6
?24 Fit with mass-constraint 4 deg.
freedom ?23 Fit without mass-constraint 3
deg. freedom Irrelevant Fluctuations in ?2 e.g.
deviations from exact 3-D intersection at vertex
tend to cancel. Strong deviations from correct
mass DO NOT cancel in ??2 .
7
Repeat for Mass Constraint M? ? 20 MeV (Note
that p? threshold is only 38 MeV below M?)
8
Subtract sum of two sidebands from
SignalBackground (Sideband normalization used
here is from previous graphs.)
9
Problem The backgrounds from sidebands populate
the cos? plots at different values from real
backgrounds. Alternate Approach Generate hybrid
MC backgrounds

• For each real event, generate 20 MC events
• at same decay vertex as real event
• parent mass from background mass distribution
• random angles in CM system
• apply standard acceptance cuts
• apply tight cuts in ??2

10
Fit with no Mass Constraint Assume Daughters are
p? Subtract ? signal, Smooth Remainder (5-bins)
11
Normalize Probability Distribution P(M) to Unit
Area. Numerically Integrate. Find values of M(p?)
at intervals of 0.01 in the integral and form a
table of its integral. Choose a random number
R(01) Table Lookup and Interpolate to get M(p?)
12
Status and Plans
Three known contributions to dataset -- ?
Signal -- Ks Background -- Continuum 2-track
background MC code exists to generate first
two. Coding in progress for the third. Plan
Generate 20 MC of each type for each real event.
Mix three MC samples, in appropriate
proportions to produce 20 MC events for
each real event. Adjust proportions of each
and assumed polarization of MC ?s to fit
data.