The following is an E-mail message from Naohito Saito describing his studies of the performance of the Muon Identifier:
Date: Wed, 9 Apr 1997 08:36:32 -0400 (EDT)
Reply-To: phenix-spin-l@bnl.gov
Originator: phenix-spin-l@bnl.gov
Sender: phenix-spin-l@bnl.gov
Precedence: bulk
From: saitoh@rikaxp.riken.go.jp
To: Multiple recipients of list <phenix-spin-l@bnl.gov>
Subject: RE: Which gap in the Muon ID to drop
X-Comment: PHENIX Spin Working Group and Workshops
Dear Spin and Muon,
I think Wayne's summary of the problem with one-gap-drop option
is excellent and I want to try to provide some more information
I have.
We have done two studies at RIKEN for pion rejection with MuID.
(1) LVL-1 trigger simulation
(2) Discriminant analysis for offline
Both of them have been done on single particle events with PISA.
(1) LVL-1 trigger simulation from my presentation at Albuquerque, Aug. '95
energy range:2-100Gev
I have done trigger simulation with static road in '95 and
found that 50 GeV/c pion can satisfy the trigger condition
with the probability of 2% for North and 3% for South. At 100 GeV/c,
the probability will be 4%. I assumed 6-gaps BUT ALLOWED the events
with no hit on gap-6. I defined 6-bit word for each trigger road:
"111111" means event with hits in all layers.
"111101" means event with hits in all layers except gap 5..
I allowed one missing gap like: "110111" ,and
two missing gaps but not in series like "110101".
I do not have a breakdown of the probability into each bit pattern.
So 50 GeV/c pion can be rejected at some level by trigger, if
our algorithm can be successfully implemented.
(2) Discriminant analysis by Dr Mao presented at BNL, Nov 1996 by NS
(I do not want to get into "discriminat analysis" vs "simple depth"
here)
energy range 10-100 GeV
His study shows contamination of pion will be the level of 0.4%
with all gap instrumented. If we omit gap-6, that level becomes
higher by a factor of about 1.5. After November, he has also done
the case if we omit gap-5. As Wayne expected, it's alomost same as
all gap instrumented.
So, (2) will be a part of answer to the question raised. I put the
postscript file to show the results from (2) on RHIC cluster as
/phenix/u/saito/outgoing/spin/rejectionp.ps
However, I am still wondering how beam-gas event will damage
our performance. Wayne, could you explain more in detail?
Regards,
Naohito Saito