Aug 15/24 PionLT/KaonLT Analysis Meeting Notes
                 ----------------------------------------------
                            (Notes by Dave and Alicia)

                     Today: PionLT will be discussed first

Present
-------
DG, Junaid, Nathan, Nacer, Zachary, Alicia, Richard, Ali, Pete

Junaid
------
Showed Heep Em/Pm distributions vs. raster y and x.  
- After flipping sign on raster y calibration constants, correlation is greatly
  reduced (XEM2 also had to flip the sign).  
- Still some small remaining correlation.  
  - DG suggests flipping sign on raster x calibration also - might impact
    reconstruction due to "ExtTar" corrections.

Zachary and Nathan
------------------
Lumi scans from 2021.  
- Able to look at C and LH2 for 2nd scan, but only LH2 from 1st scan due to
  report output issues.

- Scan 2:
  - Carbon yields are flat for all quantities (scalers, events w/out tracking,
    tracks).  
  - LH2 scalers look "ok", but other qunatities show anti-boiling.  
  - Total live time looks not plausible for some cases 
  - DG suggest just using simple computer live time for 1st pass.

- Scan 1: LH2 shows similar anti-boiling here.

Nacer
-----
Looking at 8.2 GeV Heep data.  
- Large DATA/SIMC ratio (1.122+/-0.006).  
- Observed strange shape in HMS yptar distribution in data 
  - excess at larger yptar.  
  - Tried to look at that region to see if there are some strange
    backgrounds.  Interestingly, when just looking at events from radiative
    tail, the yptar distribution looks more reasonable.

- DG notes that he's seen similar issues in the DIS data from 2022-2023.
  - Suspect matrix element problem (we are using new matrix elements fit in
    2018).
  - Higher HMS momentum settings that use Jacob's matrix elements look better.

- DATA/SIMC ratio likely not due to this yptar issue 
  - DG is suspicious of the large dead time from the EDTM. 
  - Suggests looking at computer live time again.

Richard
-------
Shows results from his analysis of same Heep data.  
- Has even large DATA/SIMC ratio (~1.2), but this is likely due to smaller
  3/4 efficiency in SHMS (Richard sees ~92%, whicle Nacer sees ~97%).  
- Need to determine the origin of this difference between Richard's and
  Nacer's analyses.

At this point, Dave had to leave.  Alicia takes notes for rest of meeting.

- Richard/Nacer will meet this week to discuss discrepancies between their
  results
  - Suggestion from Dave/Ali to compare EDTM vs CPU live times
  - EDTM LT drops to 90% for the 8.2 GeV run period, otherwise is stable around
    99%
  - Correcting EDTM should bring this point in line with the rest of the data

- Still need to add error bars

Plots of sigL, sigT, sigLT, sigTT vs -t
- Same plots as last week with addition of Q2=4.4, W=2.74 setting
- Currently working on Q2=3.0, W=2.32
- New setting shows same general trends, but with some large outliers from
  previous fit functions

- All settings normalized to W=3.0. W=2.74 may be too far away to normalize
  without better understanding the W-dependence (higher order terms may be
  required)

- Suggestion from Ali to split data into two sets, with mean W=3.0 (3.14, 3.02,
  2.95) and W=2.5 (2.32, 2.74)

- Meta-analysis post-publication of Q2, W-dependence etc may be very
  interesting

Alicia
-------------
- BSA paper still under review
- No update, studying for comprehensive exam 

Ali
--------------
- No update, writing thesis
- Expect next updates in a few weeks


Next Meeting
------------------
- Thur Aug 22 @ 15:00 Eastern/13:00 Regina
- KaonLT will go first