Project

General

Profile

Kaon LT Meetings » LTsep_mtg_22aug26.txt

Garth Huber, 08/26/2022 08:24 PM

 
1
LT Separation Meeting - 26/08/22
2

    
3
Stephen Kay
4
Garth Huber
5
Vijay Kumar
6
Ali Usman
7
Muhammad Junaid
8

    
9
Garth has a set of slides - https://redmine.jlab.org/attachments/download/1593/LT_sep_iterations.pdf
10
See also Bill's slides - https://redmine.jlab.org/attachments/download/1594/Kaon_LT_tutorial_18nov28.pdf
11
Start from Bill's code - https://github.com/billlee77/omega_analysis
12

    
13
- Prior to LT separation, need *final* normalised yields
14
	- Most of the work involved is actually in getting to this point! 
15
	- If things are converging, iteration steps should be quick
16
	- This includes diamond cut already
17
		- One could analyse the variations across only the larger, high epsilon diamond
18
	- All efficincies included
19
		- LT
20
		- FADC DT
21
		- Cryotarget
22
		- Yield corrections
23
		- ALL tested for reliablity over a wide range and applied
24
	- All kinematic offsets determined and finalised
25
	- Need one thing kept constant between iterations
26
		- The normalised yields and distributions
27
- Cross section varies across experimental acceptance
28
- Need to choose a functional form that will reasonably account for this
29
	- Don't know this in advance
30
	- Make a choice, start iteration process with it
31
	- Can check previous analyses for guidance on forms to try
32
	- Can be different for each diamond plot
33
		- So long as it it well understood for each individual diamond
34
- On diamond plot, t bins slice across diagonally
35
	- W and Q2 different for each t bin
36
	- Results quote different Q2/W value for each t bin
37
	- t min varies across the diamond
38
	- t min varies even within a t bin
39
- In SIMC, replace physics_pion.f with physics_iterate.f
40
	- Need to make some good guesses of intial paramter values for your first iteration
41
		- Need some Q2 dependence, t dependence etc
42
	- Will likely have to go back and change these functions a lot
43
- Each Q2/W should be done separately
44
	- Can't expect the procedure to work globally
45
	- *Keep notes organised!*
46
		- *Keep all output!*
47
		- Store each iteration in its own directory
48
- Step 1 - SIMC distributions
49
	- Run SIMC for large #events
50
	- Generate spectrometer and physics variables using functional form and fitpar
51
	- Do this setting by setting for a given Q2/W diamond
52
		- E.g. Left/Centre/Right all different
53
	- Compare to data
54
- Step 2 - Combine Left/Centre/Right settings at high and low epsilon for each W/Q2/t/phi/epsilon bin for data and MC
55
	- Get statistical errors for each bin
56
	- Get yield, error in yield for data and simulation
57
		- Calculate ratio and error in ratio for each bin
58
	- Process enough events to minimise SIMC statistical error!
59
		- Shouldn't be dominated by simulation statistical error!
60
- Step 3 - Calculate average kinematics
61
	- Mean data values of W/Q2/Theta/epsilon for each t bin at high an low epsilon
62
		- Need this for data and MC
63
		- Values will differ between high and low epsilon slghtly
64
			- Will also change slightly as the model is iterated
65
- Step 4a - Inspect and understand data closely
66
	- Examine variation on phi between data and MC within each t bin
67
	- Deviations between data/MC are usually indicated as wiggles in the ratio
68
		- Want R ~ 1 across a broad kinematic range
69
	- Some plots earlier on might look ok, only by examining the ratio very closely in individual bins can some differences be found
70
	- How things vary should guide what changes in the next iteration
71
		- E.g. Is it the interference terms that are off?
72
- Step 4b - Even closer examination
73
	- Each t bin (for one epsilon), binned in theta
74
		- Interference terms depend upon theta
75
			- Interference terms vanish in parallel kinematics
76
			- Explicitly chose a functional form that makes this true
77
		- In lowest bin, should have minimal interference terms
78
		- Wiggles get larger as theta increases 
79
		- Phi distributions for each t bin, subdivided into 8 theta bins
80
	- A lot of plots
81
		- More than you probably want to put in thesis
82
			- BUT a tech note/report is encouraged
83
				- Even here, you probably want to show only a few plots as an example
84
- Step 5 - Calculate unseparated cross section
85
	- Using fitpar for the iteration, evaluate the model at *average* kinematics of the data for each t-bin
86
	- Calculate using formula and fit parameters, calculate the cross section at the average values of W/Q2/Theta and epsilon for each bin
87
		- Read over Blok PRC 78 (2008) 045202 carefully
88
	- Each t/phi bin at both high and low epsilon
89
	- The weight in SIMC is NOT the unseparated two fold cross section
90
		- It is the five fold cross section (d5sigma)
91
	- Fitting procedure iterated until the experimental cross section changed by less than a prescribed amount (~1%)
92
- Step 6 - Fit Rosenbluth Eqn
93
	- Each t bin is fit separately
94
	- Fit result gives L, T, LT, TT cross sections for each t bin
95
- Step 7 - Fit L, T, LT, TT values to get new fitpar
96
	- Compare t bins with each other
97
	- Fit with physics_iterate.f functions to give next iteration model parameters
98
	- Now, we're back at the top. Feed these back into step 1
99
	- Repeat steps 1-7 until separated cros sections are stable
100
	- *Don't re-run SIMC*
101
		- Recalculate weight for each event
102
- Any theses older than Tanja's are likely using different procedures
103
	- Tanja's procedures were improved on earlier ones
104
	- Some slight modifications after this
105
	- Only Marco and Bill used Tanja's procedure
106
- Evaluating if fit equations are ok
107
	- Procedure usually works ok, but for some kinematics, pi-/pi+ sigma wouldn't converge
108
	- Compare fitpar from different Q2/W and see if they were slowly varying
109
	- If not, could use their variation as a pointer to an alternative functional form variety to try
110
	- Things will behave in a consistent manner if the correct solution is found
111
- General pointers
112
	- Read Blok paper carefully
113
	- Review Bill's slides
114
	- A single Q2/W iteration should only take 1-2 hours
115
	- Fitpar should converge in a few iterations to give 0.5<R<2
116
	- The main work is getting R to be acceptably flat
(74-74/564)