May 2/24 PionLT/KaonLT Analysis Meeting Notes
---------------------------------------------
(Notes by GH)
Today: PionLT will be discussed first
Please remember to post your slides at:
https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings
Present
-------
Regina - Garth Huber, Nathan Heinrich, Nacer Hamdi, Ali Usman, Muhammad Junaid,
Vijay Kumar, Zach Sullivan
CUA - Casey Morean, Tanja Horn
JLab - Dave Gaskell
CSULA - Konrad Aniol
York - Stephen Kay
UVa - Richard Trotta
FIU - Pete Markowitz
Nathan
------
BCM calib results for PionLT
- Dave Mack impressed with Nathan's fast progress
- DM loked at NH's 2021 results, calculated some numbers to use
- NH looked at 2022 results, calculating his own numbers to show DM
- BCM4A 2021-22 data
- DM, NH get similar calibration constants
- BCM2 looks good, DM & NH numbers almost identical
- NH shows fit of BCM Rate vs Unser Current and Residuals
- Time vs Residual plot uses time of BCM runs themselves, to check for time
dependence in fit residuals
- residuals vs time look pretty stable, which is good news
- the residuals plot that is concerning is BCM4A, 2021 data
- shows a small (2-3 sigma) trend in residual with time
- two BCM calib for runs 12500-12900 very consistent, so also are two BCM
calib for runs 13700-14100, but a noticeable change between these two sets
- no obvious reason why these 2 sets are different, or even where to put a
time break between them
- pushed calib param to HallC_replay
- BCM4A uses error weighted average of these 4 2021 points
- also takes weighted average of 4 2022 points
- obvious break between these two sets is the end of 2021/start of 2022
running
- Conclusion:
1) can use BCM2 or BCM4A, they have similar saturation @ high current
- DM adds a quadratic term above 50uA
- DG: Nathan should use the quadratic correction already in hcana, it was
determined with a current source and should be still valid
- Nathan will apply DM's quadratic correction and see if it works
- after this, Nathan will replay all Lumi data
2) Nathan has not calibrated BCM4B
- 2022 gain is very large, saturates at high current, but it might be very
useful for low current LD2 data
- this will be left for a future student looking at LD2 data, we need to be
sure to leave an obvious note so this is not forgotten!
3) BCM4C should not be used at all
- less stable than BCM4A and similar gain
- everything seems sensible so far, we look forward to seeing the Lumi results
after the replay, to confirm the calibration and non-linear correction
Junaid
------
PionLT Heep OOP offset results
- followed Garth's method for KaonLT data
- get similar offsets
- phi_HMS=+2.28mr, phi_SHMS=-0.41mr
- error bars seem very small, and this could affect the fitting results
- GH needs to review his notes and let Junaid know what errors to use
- GH gave Junaid a set of in-plane Heep offfsets, not yet implemented
Vijay
-----
Low Q2 PionLT L/T-separations
- changed MM lower cut from 0.93 to 0.91, as per last week's feedback
- showed cointime for all 3 epsilon, and diamond cuts
- did 15 iterations since last week
- cross sections are stable now, which is good, but there are some anomalies
in the phi-distributions and Data/MC ratios
- most likely, the fitting function needs to be adjusted to get a better fit
- e.g. sigT is just a simple linear dependence in t right now
- Garth and Vijay need to discuss this early next week
Richard
-------
KaonLT L/T-separations
- concentrating now on Q2=3, W=3.14 data, after it's working better will move
next to high stats Q2=2.1 setting, and the lower statistics high Q2 settings
will be last
- tried customizing t-binning a bit more
- observes a significant Q2-shift with t-bin, and this influences the
parameterization
- still using previous equation for t_ave=0.1112+0.0066*log(Q2) used in sigT,
will need to revise this for KaonLT kinematics
- GH thinks that the function for sigT is too complcated, should try a simpler
function, as Vijay has done
- Richard is giving a talk on Meson Structure @ EIC in Nanjing, May 14-17
- 25+10 min talk
- has old slides to use
- needs to circulate a draft soon
Ali
---
piDelta BSA analysis
- looking at MM fits again after fixing recon_hcana to use m_pi instead of m_p
- Q2=5.5 will have one t-bin, -t<1.05, excluding 1.05-1.60 region
- gets reasonable phi coverage now
- varied t-cut and then looked at MM distribution, to see what t-cut to use
- Center setting: all 8 phi-bins are filled
- shape study fits generally consistent w/shape of data
- data have some fluctuations (statistics?) and the MC smoothly goes
through the trend
- Right setting: data fill 4 phi-bins, 2 have lower statistics, but MM fit
still does a reasonable job
- Left setting: data fill the other 4 phi-bins, 2 with low statistics
- fit fails for the 2 low stats phi bins also because MM range is limited
- however, these bins are filled w/center data, so overall will be OK
- will redo shape study for other (Q2,W) kinematics
- Garth: the SIMC resolution tuning was done when recon_hcana had wrong mass
- did you rerun your scripts to check that things are OK?
- yes, checked that MC did not over-estimate tail for any t-phi bin
Nacer
-----
Low Q2 KaonLT analysis
- looked at Lumi scripts written by Richard
- had to do some path tweaks to get them to work in his account
- Richard will point Nacer to special Lumi runs done later by Mark Jones &
Peter Bosted @ 10.6 GeV
- this exercise is mostly for Nacer's overall understanding, next step would
be to look more closely at the Heep data taken with 3.8, 4.9 GeV beam
Hall A/C Summer Workshop discussion
-----------------------------------
- it seems they are looking for talks about "publication ready" results
- Richard, Vijay, Ali results not yet at this stage, maybe January workshop
would be better
- Alicia has done further BSA work since January update, and plots she shows
should be the same ones as in the paper
- Tanja will reply
- workshop is July 15-16
Next Meeting
------------
- Thur May 9 @ 15:00 Eastern/13:00 Regina
- KaonLT will go first