Main INDEX
Monthly INDEX
PREV
NEXT
User name jpchen
Entry time Tue May 11 10:04:51 EDT 1999
Entry number 38
keyword=5/5 analysis meeting minutes
Minutes of 5/5 analysis meeting:
Present: Alexandre Deur, Ioannis Kominis, Marat Rvachev,
Zhengwei Cai, Xiaodong Jiang, Mark Jones, Jian-ping Chen.
Called Temple U, but did not get answer, so the phone conference did not happen.
Ioannis: Showed analysis results on electron shower/preshower. The electron shower blocks #63, 83 and 92 have problems. #63 is one central block, it is a crucial one. The problem for #63 is that it has a lot of events with low ADC pulse height. The high pulse peak is also wider than the normal ones. The low pulse part can be got rid off with a reasonable cut. Showed the effeciency for e and rejection
effeciency for pion. Some discussion on the proper way of doing these.
Also started to look into the hadron shower.
Alexandre:Here a quick summary of my weekly work.
With cut on hadron shower only I was not able get ride of the low energy junks and I was
finding 3 % of good e- with 2 tracks and 99.7 % efficiency (e arm gives 1 % and 99.97 %).
So I did a very tight 2d cut with shower vs cerenkov to make sure I get real-very-good-nice e-.
The efficiency is now 99.88 % and 2 tracks events are 1.05 %. We are close to
the e-arm numbers but it appears even real-very-good-nice e- cut are not
cleaning all these junk events.
I looked at the T0 numbers on the database. They stay the same
for hadron arm but did change at the end for the e arm. For example T0 for wire
number 1 is :
for e arm
1888.3 run 1615
1888.0 run 2195
1888.5 run 2647
1839.7 run 3352 <--
for h arm: 1590.6 run 1615
1586.6 run 2195
1587.0 run 2647
1590.1 run 3073
1586.6 run 3352
I looked also at the origin of these low energy events. on e arm
x_shower vs y_shower plot shows they are mostly concentrated on a same spot, i.e
they are coming from a bad block. If it appears that these events are really
junks like it look like they are easy to cut (2D cut on shower vs preshower).
For h arm I have also plotted x_s2 and y_s2 but nothing strange
show up. (Well it's not Bogdan's opinion but he did not convince me. we can talk
about that).
I am still working on it while Ioannis is looking at shower
block per block. If we have to cut these events it will be difficult to do it
for h arm by 2D cut (shower and cerenkov) since as I was saying at the beginning
these cuts are not cleaning all these events. But if these events are coming
from bad blocks like it seems we can perhaps correct the blocks, avoiding a 2D
plot and getting better results.
It appears if we try to include the 1 % 2 track events, standard
cuts are removing 30 % of these events (0.3 % of the total). As the selection
algorithm can mixed good electron from 3He with good e- from glass (bad choice
of track) we decide to cut these 1 % events and apply a correction as a VDC
efficiency (this rely on assumption that 2 track events vertex is uniformly
distributed on the target).
With cut on hadron shower only I was
not able get ride of the low energy junks and I was finding 3 % of good e- with
2 tracks and 99.7 % efficiency (e arm gives 1 % and 99.97 %). So I did a very
tight 2d cut with shower vs cerenkov to make sure I get real-very-good-nice e-.
The efficiency is now 99.88 % and 2 tracks events are 1.05 %. We are close to
the e-arm numbers but it appears even real-very-good-nice e- cut are not
cleaning all these junk events.
I looked at the T0 numbers on the database. They stay the same
for hadron arm but did change at the end for the e arm. For example T0 for wire
number 1 is :
for e arm
1888.3 run 1615
1888.0 run 2195
1888.5 run 2647
1839.7 run 3352 <--
for h arm: 1590.6 run 1615
1586.6 run 2195
1587.0 run 2647
1590.1 run 3073
1586.6 run 3352
I looked also at the origin of these low energy events. on e arm
x_shower vs y_shower plot shows they are mostly concentrated on a same spot, i.e
they are coming from a bad block. If it appears that these events are really
junks like it look like they are easy to cut (2D cut on shower vs preshower).
For h arm I have also plotted x_s2 and y_s2 but nothing strange
show up. (Well it's not Bogdan's opinion but he did not convince me. we can talk
about that).
I am still working on it while Ioannis is looking at shower
block per block. If we have to cut these events it will be difficult to do it
for h arm by 2D cut (shower and cerenkov) since as I was saying at the beginning
these cuts are not cleaning all these events. But if these events are coming
from bad blocks like it seems we can perhaps correct the blocks, avoiding a 2D
plot and getting better results.
It appears if we try to include the 1 % 2 track events, standard
cuts are removing 30 % of these events (0.3 % of the total). As the selection
algorithm can mixed good electron from 3He with good e- from glass (bad choice
of track) we decide to cut these 1 % events and apply a correction as a VDC
efficiency (this rely on assumption that 2 track events vertex is uniformly
distributed on the target).
Zhengwei: Y_targ optimization without introduing mid-plane symmetry breaking term. But for Phi/Delta
optimization, without introduce midplane symmetry breaking terms, the offset term was too large. So
used P011 etc. Results are nearly ready.
Marat: 2nd round optimization.
Xiaodong: Was at E95001 collaboration meeting. They were making good progress (see minutes on E95001
homepage).
Next week, we will wrap it up on the first pass analysis of the detectors (scintillators/trigger,
shower, VDC, Cherenkov), beam positions and optics study. Then will discuss the plan for next step.
Also discussed about the up-coming collaboration meeting at Kentucky. All students will make presentations at the meeting.