Review of TD cuts

Agenda for 16 July 2003 meeting:

1. EV5 (Dima)
2. Everything else (Tetsuro)

Materials, in no particular order, for 16 July 2003 meeting:

• Dima's new EV5 note
  
• Tetsuro's page
• My 3P study on kp21 monitors. see explanation below.
• |Z3p-Zpi| vs |Zmu-Zpi| RFLAG=4, piscat monitors (shows that relative-area algorithm assigning 3P,mu is correct)

• For my study, I find that min(chi2_i(2P)/chi2_i(3P))>1.5 is good for selecting potential 3P candidates (page 1, lower left). chi2_i(NP) = chisquare for i^th end for N pulse fit.
  
  
• I compared energy and timing of the 3d pulse in 3P events with that of mu pulse in 2P events.
  
  
• Fits to DT=T1-T2 (Ti = time of i^th end), show three components in 3P events. (page 1, left column)
  1. Narrow, sigma = ~1ns (true pulse correlated with pion)
  2. wider, sigma = ~6ns (true pulse UNcorrelated with pion)
  3. widest, sigma = ~45ns (junk)
  
  
• First two components are in 2P events (page 1, right column) although width of narrow component is larger(1.1 vs 0.9 ns) than for 3P events.
  
  
• I isolate the narrow and wide components with |DT|< 2ns and 2 < |DT| < 16 ns cuts, respectively. page 2 shows energy vs time after chi-square ratio cuts and DT cuts.
  
  
• page 3 shows T3P(top) and Tmu(bottom) distributions. right column is bkgd subtracted dists. left column shows hists that were subtracted.
  
  
• page 4 shows fitted bkgd-subtracted T3P(top) and Tmu(bottom) distributions. Note that the fitted means and sigmas agree with Tetsuro's fig18
  
  
• pages 5 and 6 compare times and energies of the 3d pulse and muon pulse in 3P events
  
  
• page 7 shows E3P vs T3P-Tmu and confirms Tetsuro's fig 13
  
  

  ---

  Agenda for 26 June 2003 meeting:

  1. Telec distribution (Dima)
  2. TDVAR training (Tetsuro)

  Rejection for All BAND and TAIL events.
  1998 from table 3, in section 5.3 of 1998 1/3 pnn1 analysis note.
  2002 from Tetsuro's email.
  

  	1998	2002
  EV5	2.64(.02)	2.26(.04)
  ELVETO	1.68(.01)	2.72(.07)
  TDFOOL	1.003	1.012
  TDVAR	86.7(8.8)	33.9(6.7)
  ALL	388(39)	211(42)
  EV5 and ELVETO	4.46(.04)	6.14(.19)

  
  Materials, in no particular order, for 26 June 2003 meeting:

  • Telec note
    
  • Tetsuro's email describing TDVAR tuning
  • NN training vs epoch
    
  • NN A vs R
    
  • Tmu distributions Tetsuro et al's technote after successive cuts for skim2
    

  ---

  Agenda for 19 June 2003 meeting:

  1. All studies compromised by large peak at Tmu = 5 ns (figs 2,5 of Tetsuro's note) due to loosened CMUPI, PROD cuts (see table 6, p.56 of TN K-029).
     
     Prelim. recommendations:
     1. For 2P fits, use "Reopt" cuts or similar from table 6
     2. For 3P fits mistakenly assigned pi,mu,R (Tetsuro's note, sec 3.2), check chi-squares and (perhaps) revert to 2P fit results
  2. EV5 discussion (see Dima's note and responses)
  3. ELVETO (including ELVETO5)
  4. FITPI 3P events

  Materials, in no particular order, for 19 June 2003 meeting:

  • EV5 questions and Dima's response
    
  • See page 5 of Shaomin's kinematic background study
    
  • Technote K-029 "FITPI: Triple-pulse fitting and other modifications"
    
  • Tetsuro et al's technote on ELVETO, FITPI(3d pulse), etc.
    
  • Dima's EV5 technote