|
1
|
Aug 28/25 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, Nathan Heinrich, Vijay Kumar, Nacer Hamdi,
|
|
13
|
Muhammad Junaid
|
|
14
|
CUA - Chi Kin Tam, Tanja Horn, Sameer Jain
|
|
15
|
Ohio - Julie Roche
|
|
16
|
|
|
17
|
Nacer
|
|
18
|
-----
|
|
19
|
KaonLT Q2=0.5 LT-sep
|
|
20
|
- went back from Fpi-2 pi+ to sig_factorized_2007 K+ parameterization
|
|
21
|
- this is based on Q2,W-dependence of Bebek & Brauel data, an update by Tanja
|
|
22
|
of Koltenuk parameterization
|
|
23
|
- this parameterization includes a fixed T/L=2 ratio
|
|
24
|
- has a different W-factor than Fpi-2, retaining Fpi-2 W-fac for now
|
|
25
|
- when using initial params w/no fit (IT00)
|
|
26
|
- Data/MC~2 for both low and high epsilon, good agreement between Data & MC
|
|
27
|
for variable histos except for this normalization factor
|
|
28
|
- found that reweight script does not fully reproduce SIMC output, still
|
|
29
|
investigating
|
|
30
|
- Tanja suggests there may be an issue with tmin dependence in the t-factor
|
|
31
|
i.e. t-fac=exp(p*(t-tmin))
|
|
32
|
- also will look into changing W-fac to the sig_factorized_2007 version
|
|
33
|
|
|
34
|
Junaid
|
|
35
|
------
|
|
36
|
Q2=3.85 LT-sep (pi+)
|
|
37
|
- changed sigL form from Fpi-2 pi+ to P.Bosted parameterization
|
|
38
|
- includes a fixed parameterization of Fpi(Q2) and pole term in sigL form
|
|
39
|
- when using initial params w/no fit (IT00)
|
|
40
|
Data/MC ratios~6
|
|
41
|
- does 1 iteration allowing sigL params to float but keeping sigT frozen, get
|
|
42
|
ratios~10
|
|
43
|
- IT01, instead of fixing sigT, let p1,p2 float
|
|
44
|
- all plots look good, Rosenbluth fits to sig_UNS, fits of params to sigL,T,
|
|
45
|
data-MC shape comparisons
|
|
46
|
- however, the normalization is way off, the Data/MC ratios went up to ~30
|
|
47
|
- Garth: strongly suggests something is wrong in the code, before the
|
|
48
|
iteration SIMC was too small (compared to data) by ~6x.
|
|
49
|
- After the iteration, SIMC went down by 5X instead of up 5X.
|
|
50
|
- Maybe something is backwards in the code, such as a Data/MC ratio being
|
|
51
|
inverted by mistake as MC/Data? For sure, something is not consistent.
|
|
52
|
|
|
53
|
Chi Kin
|
|
54
|
-------
|
|
55
|
KaonLT Q2=3.0, W=3.14 analysis
|
|
56
|
- carefully going through Richard's code and doing checks
|
|
57
|
- found that the MM shifting algorithm needs fixing
|
|
58
|
- RT fit a Gaussian to a too-wide region around the Lambda peak, and the long
|
|
59
|
radiative tail was throwing off the fit
|
|
60
|
- changed to a Gaussian with a flat bkd fit, giving a much better description
|
|
61
|
of the peak value
|
|
62
|
- Garth: the peak shape is not Gaussian due to the radiative tail, suggests
|
|
63
|
to restrict the fit to 1.10<MM<1.12 to fit just the peak region
|
|
64
|
- Nacer: one offset per setting will probably be sufficient, but it is worth
|
|
65
|
confirming this on a run-by-run basis
|
|
66
|
|
|
67
|
- dummy target subtraction
|
|
68
|
- Garth: there is an uncertainty in the thickness ratio between
|
|
69
|
Dummy/Tgt-Vessel that needs to be propagated in the errors
|
|
70
|
- Nacer will send the values he's using
|
|
71
|
|
|
72
|
Nathan
|
|
73
|
------
|
|
74
|
PionLT Lumi studies
|
|
75
|
- had a meeting with Mark Jones on the Reference Time Cuts
|
|
76
|
- Mark recommended tighter ref time cuts
|
|
77
|
- Mark also suggested applying GoodStartTime, GoodDCTrack>0, and HMS/SHMS
|
|
78
|
acceptance cuts when making plots
|
|
79
|
- Nathan also found an oversight, was cutting on HTrig1 instead of PTrig1
|
|
80
|
- CTime.CoinTime_RAW_ROC1,2 distributions are narrower now, with simpler
|
|
81
|
features
|
|
82
|
- both ROCs behave similarly, will be using ROC2
|
|
83
|
- will evaluate a CoinBlockingCorrection from these plots
|
|
84
|
- the region 30-130 has beam pulse structure, outside this region the
|
|
85
|
structure is more random
|
|
86
|
- Garth: suggest first to sort these plots by rate, so you can see how the
|
|
87
|
features change as rate is increased, the coin blocking correction should
|
|
88
|
be negligible at the lowest rate
|
|
89
|
- new set of Lumi yield plots vs current and rate
|
|
90
|
- new tighter cuts (after discuss w/Mark) didn't make any difference
|
|
91
|
- in comparison to plots from before RefTime study, the spread vs Rate is
|
|
92
|
tighter, particularly when plotted vs CoinHMSRate and CoinRate
|
|
93
|
- when investigating outliers, found that CoinTime offset was wrong in some
|
|
94
|
cases
|
|
95
|
- Junaid's new CT offsets use ROC2, while Nathan was using ROC1 by mistake
|
|
96
|
|
|
97
|
Garth
|
|
98
|
-----
|
|
99
|
BSA PLB revised Fig 4
|
|
100
|
- Tanja and Julie found the plot to be confusing
|
|
101
|
- GH will contact Stefan Diehl about standard deviations of CLAS12 data
|
|
102
|
- Julie: instead of all data, one possibility is to just show a selected
|
|
103
|
version, showing how the experiments differ near x=0.25
|
|
104
|
|
|
105
|
Next Meeting
|
|
106
|
-------------
|
|
107
|
- Thur Sept 4 @ 15:30 Eastern/13:30 Regina
|
|
108
|
- PionLT will go first
|
|
109
|
|
|
110
|
_ not everyone will be available, because of teaching conflicts
|
|
111
|
- please fill out the poll to select a new time starting week of Sept 8
|
|
112
|
https://www.when2meet.com/?31948498-x2KTs
|
|
113
|
|
|
114
|
|
|
115
|
|
|
116
|
|
|
117
|
|
|
118
|
|
|
119
|
|
|
120
|
|
|
121
|
|
|
122
|
|
|
123
|
|