**************************************************
NEW SIMULATION PLOTS FOR AUGUST 2002 TEMPLE REVIEW
**************************************************

==================================================================
JET ALGORITHM PLOTS & ICR: Jovan Mitrevski: 22-Jul-02
-------------------------
o all plots made from MC files with 7.5 min-bias overlaid
  - Rates from weighted sum of: QCD 5,10,20,40
  - Signal samples: ZH-->vvbb, WH-->evbb, ttbar-->e+jets, h-->tau+tau-
o files with "ICR" include ICR energies in jet definitions using
  precision readout
o in most plots sliding windows algorithms 2,1,1; 2,0,1; 3,0,1; 3,-1,1
  are compared. ZH plots also include the current algorithm (1,0,0)
o Plots:
  - jTurnonpassCanv[sig]  = jet trigger turn-on curves
  - jetEffRate[sig]       = eff vs rate for single jet triggers
  - dblJetEffRate[sig]    = eff vs rate for Et(jet1)>cut &
                            Et(jet2)>cut
                            note: this is not an optimal scheme
  - deltaRPassCanvas[sig] = delta-R between trigger cluster and reco
                            jet (JCCB = 0.5 cone) 


==================================================================
JET ALGORITHM PLOTS: Emmanuelle Perez: 5-Aug-02
-------------------
Dear all,

 I updated few simulation plots. 
 Sorry that this comes so late. If too late for the TDR,
 may be this can still be usefull for the talks.
 The updated plots can be found on d0mino:
 /home/perez/R2B_TRIGGER/FIGS_UPDATES

 turnon.gif = update of fig 42 (of version 5.3)
              the (2,1,1) algo is compared to TT's
              (the old fig 42 was 'sick' at low values
              of Pt jet)

 turnon_sliding_vs_nonsliding.gif = update of fig 43

 selectivity_jet20.gif and selectivity_jet40.gif : could
 be shown instead of figs 46. This is done on QCD events
 with 7.5 mb overlaid. The efficiency is calculated for
 events where the highest Pt reco'd jet has Pt > 20 (40) GeV.
 Only 2 sliding algos are shown; the curves for other algos
 are identical and could not be distinguished on the plots.

 selectivity_zh.gif and selectivity_hbb.gif : could
 be shown instead of figs 48.
 It compares sliding vs TT performances, for a single jet
 and a double jet condition. Due to the asymmetry in the pt
 of the two highest pt jets in signal events, a simple 2-jet
 condition (symmetric cut) is not optimal. One does much better
 by requiring Et1 + Et2 > cut. But actually a single jet
 condition is even more powerful.

 
 Concerning the ICR problem : we finally understood the reason why
 the simulated TT's for ICR/MG were doing bad. It is due to the
 fact that the 7.5 mb simulated files were simulated with a 
 quite old version of d0sim, in which the mapping of the TT's
 was not correct when building the L1CalTTChunk.
 As can be seen on the file attached, when looking at 
 Et(trigger) / Et(reco'd) versus eta, on a recent MC file, we
 have the correct behaviour.


 We are on the way of producing few plots on the data.
 Hope that the results will look nice.
 Meanwhile, there are some calo trigger efficiencies (obtained
 on data events) in the D0Note 3949.


 I should leave tomorrow for vacation. I will have some
 limited computing access and will read my emails. Please
 contact Jiri (cc to me) in case you need something.

 Cheers,
Emmanuelle


ps: a typo in the TDR v5.3, table 10 (p71):
    the algo 3,1,1 has not been studied. it should be 3,0,1
    instead (the window size & decl region .. are correct)


==================================================================
JET ALGORITHM PLOTS (DATA): Emmanuelle Perez: 5-Aug-02
---------------------------
Dear all,

 Very first plots showing the performances of a sliding
 window algorithm on some data events can be seen in
 /home/perez/R2B_TRIGGER/FIGS_UPDATES/moriond_c[i].gif
 (i=1,2,3). I do not know if it is still possible to
 include them in the TDR (sorry that I can not do that),
 but I think they are quite fine.
 Anyway:

 moriond_c1.gif  shows that the turn-on curve is indeed sharper
   when triggering with (2,1,1) instead of 0.2x0.2 TT's
  (c2.gif = the same, with a different set of thresholds)

 moriond_c3.gif = Et (trigger jet) / Et (reco'd jet)

 This is done on the Moriond J/psi -> muons sample.
 This is clearly statistics limited.. the code runs
 on reco'd events (not root-tuples), and we need an
 unbiased sample (eg muon events are fine). If you are aware
 of any muon reco'ed sample with more recent data, please 
 let us know.
 Also note that in these plots, we do not use the true
 Et seen by the TT's but re-simulate them from the precision
 readout. For comparisons TT vs sliding, this should
 be fine.

 I hope that we can provide more in the next days.
 Cheers,

Emmanuelle & Jiri.


==================================================================
TAU ALGORITHM PLOTS: Chad Johnson: 6-Aug-02
-------------------
o tau_vs_qcd.*: plot of Et(2x2)/Et(4x4) for H-->tau+tau- and QCD40
		min-bias = 7.5


==================================================================
EM ALGORITHM PLOTS: Alex Hamilton: 7-Aug-02
------------------
o These plots are meant to show the effect on the EM algorithm
  efficiency of changing the EM and hadronic isolation cuts while
  keeping the background rate fixed.
  - the EM isolation is:  Et(ring around RoI) < cut (in GeV)
  - the Had isolation is: Et(RoI)/Et(4x4 had) > cut (fraction cut)
o Plots
  - em_xxx:  em isolation plot at rate xxx in Hz
  - had_xxx: had isolation plot at rate xxx in Hz