|
1
|
Mar 9/23 PionLT/KaonLT Analysis Meeting Notes
|
|
2
|
---------------------------------------------
|
|
3
|
(Notes by GH and SJDK)
|
|
4
|
|
|
5
|
Today: PionLT 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 - Stephen Kay, Garth Huber, Muhammad Junaid, Ali Usman, Alicia Postuma,
|
|
13
|
Vijay Kumar, Nathan Heinrich
|
|
14
|
JLab - Dave Gaskell
|
|
15
|
CUA - Richard Trotta
|
|
16
|
CSULA - Konrad Aniol
|
|
17
|
Ohio - Julie Roche
|
|
18
|
|
|
19
|
Nathan Updates
|
|
20
|
--------------
|
|
21
|
Aerogel calibrations
|
|
22
|
- Petr Stepanov's calibration code used
|
|
23
|
- presentation from Petr on this, 2020-Sept-23
|
|
24
|
- Fits 20 Gaussians with some exponential tail dropping off as BG
|
|
25
|
- Each Gaussian is convoluted with an exponential fall off
|
|
26
|
- Realistically only up to ~5 PE is needed
|
|
27
|
- also note that we should have no pedestal as this is instrinsically
|
|
28
|
removed as part of FADC algorithm
|
|
29
|
- Ran calibration, plotted constants for each PMT vs run number
|
|
30
|
- Looks fairly flat across run numbers
|
|
31
|
- Scale is very large due to large error bars though
|
|
32
|
- Only PMT that looks odd is +3, poor Chi2
|
|
33
|
- Three fit methods in script, could try changing them and testing out?
|
|
34
|
- Will try reducing number of Gaussian's down to 10 or lower rather than 20
|
|
35
|
- GH: Maybe 6?
|
|
36
|
|
|
37
|
- Compare to old algorithm (Peter B?)
|
|
38
|
- Old algorithm plot scales are very compressed compared to Petr's ones
|
|
39
|
- Petr's is better, but a lot of free parameters, which inflates errors
|
|
40
|
- the calibration parameters from the two algorithms seem not consistent
|
|
41
|
each other, NH needs to better understand exactly what is calculated
|
|
42
|
|
|
43
|
- Add some cuts to clean up protons?
|
|
44
|
- protons would add background events to left of 1pe peak (delta
|
|
45
|
knockout) which could throw off the calibration
|
|
46
|
- NH: Petr's script requires some cuts before hand to begin with, wants a
|
|
47
|
slightly cleaned up file, but not yet familiar with exactly what is done
|
|
48
|
- DG: Did kaons fire aero in PionLT?
|
|
49
|
- If just above threshold, might affect statistical sample
|
|
50
|
- GH: Kaons do fire the aerogel (or should) all the time
|
|
51
|
If want to reject kaons, then should place a cut on HGC in addition
|
|
52
|
|
|
53
|
- RT: Applying any geometry cuts on the tray? i.e. for some indices the
|
|
54
|
aerogel does not fully fill the tray
|
|
55
|
- Old algorithm isn't, Petr's is
|
|
56
|
- Should apply in both
|
|
57
|
|
|
58
|
- *Aerogel calib to do:*
|
|
59
|
- Need to check cuts in Petr's code
|
|
60
|
- Be sure protons are removed
|
|
61
|
- tray size cut
|
|
62
|
- Reduce #Gaussians in fit, hopefully get smaller error bars
|
|
63
|
- understand why params from two codes are so different
|
|
64
|
- read up on FADC operation, pedestal suppression (e.g. C Yero thesis)
|
|
65
|
|
|
66
|
-Mystery LD- setting investigation (one with missing events after RefTime cut)
|
|
67
|
- Identical output with different reference time cuts, something's wrong
|
|
68
|
- can't figure out what's wrong with his scripts, asks if someone else can
|
|
69
|
try running from their account to see if get same answer
|
|
70
|
- *Junaid* will try to take a look at this setting next week
|
|
71
|
|
|
72
|
- DG: Do you apply timing cuts to trigger aparatus channels separately or just
|
|
73
|
to coincidence after it has formed?
|
|
74
|
- Yes, does have both
|
|
75
|
- Carlos had issues because they didn't have cuts on T1/T3, just the coin
|
|
76
|
time
|
|
77
|
- NH will *send email* to CY to confirm what he's doing is correct
|
|
78
|
|
|
79
|
Junaid Updates
|
|
80
|
--------------
|
|
81
|
- No report, Busy with assignments
|
|
82
|
|
|
83
|
Richard Updates
|
|
84
|
---------------
|
|
85
|
SIMC to hcana conversion scripts (EM, PM, etc)
|
|
86
|
- Finishing up scripts
|
|
87
|
- Separate versions for production/HeeP, since different output variables
|
|
88
|
|
|
89
|
Running more SIMC stats to reduce errors seen last week
|
|
90
|
- GH: This is fine, but What we actually need from SIMC is a functional form
|
|
91
|
for L/T/LT/TT and how they vary with kinematics
|
|
92
|
- This is different to what is in SIMC now
|
|
93
|
- Garth sent physics_iterate.f routine as an example on how to do this
|
|
94
|
- Should be more straightforward for Vijay looking at the Pion
|
|
95
|
|
|
96
|
- GH: I'm happy to have a separate meeting to discuss this more
|
|
97
|
- Need to pick *a* functional form separate for each before you do any
|
|
98
|
iterations
|
|
99
|
- *Then* run SIMC for large # of events
|
|
100
|
|
|
101
|
- DG - Recent inclusive data had the sign of the raster correction wrong, not
|
|
102
|
sure how far back this goes
|
|
103
|
- this would adversely affect the reconstruction resolution
|
|
104
|
- PionLT almost certainly affected, possibly even also for KaonLT
|
|
105
|
- RT *will check*
|
|
106
|
|
|
107
|
- Will be doing a new replay round soon
|
|
108
|
- SK: Make sure to save it to /cache! Want backed up runs at this point!
|
|
109
|
|
|
110
|
Ali Updates
|
|
111
|
-----------
|
|
112
|
Looking at pion PID in KaonLT data
|
|
113
|
- Setting up scripts
|
|
114
|
- Hopefully plots/step by step guide next week
|
|
115
|
- AU/RT should *meet* next week to discuss PID, since their
|
|
116
|
foreground/background are linked
|
|
117
|
|
|
118
|
Vijay Updates
|
|
119
|
-------------
|
|
120
|
LT separation code
|
|
121
|
- Lots of errors when updated average kinematic script
|
|
122
|
- Fixing/debugging this
|
|
123
|
- RT: There's a lot to go through, use diff/grep to check between scripts
|
|
124
|
|
|
125
|
Alicia Updates
|
|
126
|
--------------
|
|
127
|
- Very busy with two classes, no report
|
|
128
|
|
|
129
|
Next Meeting
|
|
130
|
------------
|
|
131
|
New Meeting Time after USA savings time starts:
|
|
132
|
- Thur Mar 16 @ 16:00 Eastern/14:00 Regina/13:00 Pacific
|
|
133
|
- KaonLT will go first
|