Project

General

Profile

Kaon LT Meetings » mtg_24may23.txt

Garth Huber, 05/23/2024 06:55 PM

 
1
                May 23/24 PionLT/KaonLT Analysis Meeting Notes
2
                ----------------------------------------------
3
                                (Notes by GH)
4

    
5
                     Today: KaonLT will be discussed first
6

    
7
Please remember to post your slides at:
8
https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings
9

    
10
Present
11
-------
12
Regina - Garth Huber, Ali Usman, Nacer Hamdi, Muhammad Junaid, Nathan Heinrich,
13
   Vijay Kumar, Zach Sullivan, Alicia Postuma
14
JLab - Dave Gaskell
15
Virginia - Richard Trotta
16
FIU - Pete Markowitz
17
York - Stephen Kay
18
CSULA - Konrad Aniol
19

    
20
Vijay
21
-----
22
- implementing Garth's final corrections on Thesis before it goes for committee
23
  review
24
- will resume data analysis next week
25
- first task should be to work on systematic uncertainties table, as it's blank
26
  in the draft thesis
27
  - Richard will set up a meeting between Vijay and Nacer to discuss systematic
28
    uncertainties next week
29

    
30
Ali
31
---
32
- also busy with thesis write-up, gave 1 chapter to GH for review
33
- resuming piDelta data analysis
34
  - first task will be to finish MM fitting of high statistics settings, plan
35
    to show results at next meeting
36

    
37
Alicia
38
------
39
- pi+n BSA paper is almost ready, planning to circulate it to this group
40
  tomorrow
41
  - still need to check PRL word count
42
  - the uncertainties in the offsets still need to be decided, we will come
43
    back to this after Junaid's presentation 
44

    
45
Richard
46
-------
47
- KaonLT L/T-separations
48
  - Q2=3.0, W=3.14 -t=0.15-0.5 after 17 iterations presented by Tanja last week
49
  - Q2=2.115, W=2.95 -t=0.15-0.35 0th iteration result shown
50

    
51
- fitting mean Q2 value per bin vs t value to get general idea of
52
  Q2(t)=Aexp(Bt) dependence at high, low epsilon, instead of tave factor in
53
  parameterization
54
- also looking at fits of mean W value per bin vs t value to replace W-fac with
55
  a W=mt+b form
56
  - GH raises a caution: these are the variations in the mean kinematic values
57
    with t-bin, not the variation in the yields with Q2,W.  They are NOT the
58
    same, so don't fool yourself when doing this study
59

    
60
- Ali raises the issue of the synchrotron radiation correction to the beam
61
  energy mentioned by Dave last week
62
  - GH strongly suspects that this correction was not applied to KaonLT higher
63
    energy values, it will affect all >8 GeV data
64
  - it will also affect the offsets, even at low energies, we will discuss more
65
    after Junaid's presentation.  GH is happy to rerun the offset finder
66
    program if Richard supplies corrected data
67

    
68
Nacer
69
-----
70
- KaonLT low Q2 analysis, working on SIMC-Heep comparison
71
  - SIMC input files were corrected for offsets, generated files ready
72
  - now setting up to replay Heep data, trying to solve a problem w/missing BCM
73
    file, probably caused by incorrect pathing, will meet with Ali on this
74

    
75
Nathan
76
------
77
PionLT BCM calibration study
78
- looking at quadratic correction for BCM2
79

    
80
2021 Lumi data: 
81
- if quadratic term included, the Lumi scan looks flatter vs current, but still
82
  not perfect
83
- then allow fit function to refit BCM slope + offset while keeping quadratic
84
  term fixed
85
  - result is a slight over-correction with current, but the best result
86
  - will use this one.  Dave agrees this looks good
87

    
88
2022 Lumi data:
89
- when quadratic term is added, the Lumi scan is better for >60uA, no
90
  difference below this
91
- refitting slope + offset while keeping quadratic fixed gives an over
92
  prediction at high current, and low current is now too low
93
  - this is not as good as just adding the quadratic term to simple linear fit,
94
    so will use that one
95

    
96
  - 2021,22 calibs look very good, just waiting on feedback from Dave Mack
97

    
98
- started looking at Lumi analysis w/ new BCM calibs
99
  - supervising Zach to get him started on Lumi analysis
100

    
101
  - a jump in TotalLT (TLT) is observed at 40uA for Carbon runs 12144-12154
102
    - Dave: consider using CPULT instead of TLT for these runs
103
    - NH: CPULT is not always more reliable, shows Carbon runs 16727-16737
104
      where CPULT~81% but TLT~99%
105
      - need to check if there's a script bug causing this, since the script
106
        assumes 2-arm data, but these runs are 1-arm
107
      - Ali: suggest to compare CPULT from Report File to value from script
108

    
109
Junaid
110
------
111
PionLT Heep-Coin analysis
112
- corrected all beam energy values for synchrotron radiation after Dave's
113
  comment last week, and then reran Heep + SIMC
114
  - only has an affect for >8 GeV, correction is zero below that
115

    
116
- GH generated new offsets from corrected data, Junaid shows comparison
117
  - new offsets have SHMS and HMS momentum offsets near zero, which is good
118
  - comparison of trend in beam offsets is interesting:
119
    - the new beam offsets are typically -0.5x0.1% while the old ones are
120
      0.1x0.1%, so a systematic difference of -0.6x0.1%. 
121
    - however, both sets have the same 0.6x0.1% jump between 9.879 and 10.550
122
      GeV, i.e. 0 to +0.6x0.1% old, and -0.6x0.1% to 0 new
123

    
124
  - Junaid will replay data + SIMC to be sure offsets are applied correctly
125
    - also will compare old and new offset results for 5.984, 7.937, 10.549 GeV
126

    
127
- now back to Alicia's question about offset uncertainties
128
  - based on the difference between the two offset solutions, GH suggests
129
    0.5x0.1% for energy & momentum offsets, 0.5x1mr for angle offsets
130
    - Dave agrees this is sensible
131

    
132
Stephen
133
-------
134
- iFarm transition from RH7 to ALMA9
135
  - Ole Hansen sent out an email on pre-compiled packages:
136
    "As many of you know, we are providing precompiled software under
137
    /group/halla/apps. Much of this software has been built for both RHEL7 and
138
    Alma9 and can be loaded with the "module" command on either
139
    platform. Specifically, the following modules support Alma9 at this time:
140
      analyzer/1.7.12
141
      analyzer/1.7.8
142
      boost/1.84
143
      CLHEP/2.4.6.4
144
      evio/5.3
145
      Geant4/11.1.2
146
      Geant4/11.2.1
147
      hcana/1.1
148
      Pythia/6.4.28
149
      Pythia/8.311
150
      python/2.7.18
151
      Qt/5.15.10
152
      Qt/5.15.13
153
      ROOT/6.26-10
154
      ROOT/6.30-04
155
      xerces-c/3.2.4
156
      group.apps
157

    
158
   "For reference, here is an example of how to set up the Hall C analyzer on
159
    the central systems:
160
      % module purge
161
      % module use /group/halla/modulefiles
162
      % module load hcana
163
      Loading hcana/1.1
164
      Loading requirement: pythia6/6.4.28 pythia8/8.311 root/6.30.04 evio/5.3 analyzer/1.7.12
165

    
166
   "(The pythia requirements appear because ROOT/6.30.04 was compiled with
167
   pythia support.)"
168

    
169
  - Stephen could get hcana to compile, did not yet try LTsep package
170
  - Nathan could not get hcana to compile
171

    
172
Next Meeting
173
------------
174
- We will cancel next week's meeting due to various time conflicts
175

    
176
- Next meet: Thur June 6 @ 15:00 Eastern/13:00 Regina
177
  - PionLT will go first
178

    
179

    
180

    
181

    
182

    
183
  
184

    
185

    
186

    
(432-432/470)