Prsentation PowerPoint

1
Effect of CH seed on 0.5 cone jets (Update)
Emmanuel Busato Jet/Met meeting feb 27 2003
1. Data sample 2. Comparison old/new seeding
with new seeding parameter dropCHseedsigma
3.5 3. What's the best value for
dropCHseedsigma ? 4. Conclusion
?
2
Data sample
? Run 172603 (feb-16-2003) ? Reconstrution
version p13.06.01 ? Quality cuts
0.05ltemflt0.95 chflt0.4 hotflt10 n90gt1 ? No
trigger selection
3
Comparison old/new seeding with new seeding
parameter 3.5 ?
4
An example
Before split/merge
Old preclustering 1 normal seed 1 CH
seed 2 low pT clusters
reconstruted with cone algo, individually
above threshold
New preclustering 1 normal seed
1 low pT cluster
reconstruted with cone
algo, above threshod
After split/merge
Splitting Assign shared cells to nearest
proto-jets clusters below threshold
(or fail quality cut)
No split/merge
No jets left
One jet left
5
What does new jets look like ?
Jet by jet comparison
For each event
Match jets reconstructed with current
preclustering to jets reconstructed with modified
preclustering ? ? R lt 0.5
? if more than one jet take the
nearest
Jets found in both cases
6
Jets found with old preclustering ONLY
Number of these jets N1 (1614 for 3.5 ? )
Properties of jets not found with new
preclustering - low pT - high f90 - mostly
in ICR but also in CC - big number of merge
7
Jets found with new preclustering ONLY
Number of these jets N2 (436 for 3.5 ?)
Properties of jets not found with old
preclustering - number of merge is lower -
eta ?
Could be noise jets but could also be good jets
at low pT
8
Summary
New seeding seems to improve things ? jets
found with old preclustering only seem to be
noise jets ? if jets found with new
preclustering only are good jets very
good (reduction of the effect of noise on low pT
good jets) if noise jets we gain a
factor of N1/N2 1614/436 ? 3.7 when
dropCHseedsigma 3.5
9
What's the best value for dropCHseedsigma ?
Jets found with old preclustering ONLY
dropCHseedsigma 8 ?
N11619
Jets found with new preclustering ONLY
N2334
10
dropCHseedsigma 1000 ?
N1 1730
N2342
11
nb1
nb2
12
Conclusion

I think we should include this modification now
with a high value for dropCHseedsigma