May 15/25 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, Muhammad JUnaid, Ivan Zhenchuk,
   Nacer Hamdi, Ali Usman, Vijay Kumar, Alicia Postuma
CSULA - Konrad Aniol
Virginia - Richard Trotta

Ali
---
piDelta MM shape study, refining high statistics settings

Q2=3.0, W=3.14 - now 3 t-bins, previously 2 t-bins
- new t-bins: 0-0.33, 0.33-0.42, 0.42-0.72
- low t center SHMS
  - pi+n (brown line) MC has more fluctutations in Delta region than other 2
    MC, might need to run it longer to get adequate statistics in tail
    
- mid t center SHMS
  - pi+n doesn't have very many events in this region, so removed it from fit
  - fit region is 1.11<MM<1.4 GeV
- mid t right SHMS
  - pi+n is included here
- mid t left SHMS
  - of 8 phi-bins, 2 are excluded due to difficult fits in Delta region
  - 4 phi bins are empty, so this leaves just 2 phi bins with fit data:
    -135 to -180, +130 to +180
- full phi coverage is obtained over 3 SHMS settings

- high t center SHMS
  - bkd much larger here than at lower t
  - there is data in all 8 phi bins, but 4 bins need to be excluded due to lack
    of Delta peak
- high t right SHMS
  - MM fits are done for only 2 phi bins, 0 to +45, -45 to 0
  - 2 phi bins excluded due to unclear Delta peak
  - 4 phi bins empty
- high t left SHMS
  - can do fits for all 4 phi bins with data
- full phi coverage is obtained, 4+2+4 phi bins in total over 3 SHMS settings

Q2=2.1, W=2.95
- new t-bins: 0-0.25, 0.25-0.33, 0.33-0.60
- only results for highest t-bin shown
- high t center SHMS
  - fit only 4 phi bins: 0 to +45, +45 to +90, -45 to 0, -90 to -45
- high t right SHMS
  - fit MM data for 2 phi bins: 0 to +45, -45 to 0
- high t left SHMS
  - fit MM data for 4 phi bins:
    +90 to +135, +135 to +180, -180 to -135, -135 to -90
- full phi coverage is obtained, 4+2+4 phi bins in total over 3 SHMS settings

Next steps:
- check that all shape study plots are ok
- then extract asymmetry and recalculate mean kinematics per t-bin
- start systematic studies on MM fit dependence

Nathan
------
PionLT luminosity scans
- last week, noted some weird behavior in #counts surviving cuts.  Looked at
  actual numbers produced by scripts to diagnose
- found that scipy.integrate.* was used in scripts to integrate counts
  - this is not a true integration, rather a function is fit to the histogram
    and the fit is integrated.  Anomalies happen if the fit is not good
  - removed this function, now gets true #events in histogram
    - ***NB*** EXTREMELY IMPORTANT THAT EVERYONE CHECK THEIR SCRIPTS TO MAKE
      SURE scipy.integrate.* IS NOT USED ANYWHERE!!!
    - possibly this could explain some of the weird behavior Richard has noted
      in the past
- scaler yields are unchanged, but all histogram yields need to be rechecked

- Carbon Lumi
  - still need to recheck if Dave Mack's combinatoric ELLT is applied to
    scaler scans
  - Yield-NoTrack, Yield-Track generally look good, but some new outliers
    - investigating further
  - overall, very similar to earlier Lumi analysis, conclusions overall valid

- LH2 Lumi
  - some very large error bars for SHMS scalers but not HMS scalers, which
    doesn't look right, will investigate

- Coin Lumi
  - analysis still has a major bug, looking into it

- *NB* still looking at hodoscope time shift reported last week

Junaid
------
PionLT LT-sep analysis
- calculating average kinematic variables for each t-bin
  - still need to calculate epsilon, theta* per bin
  - the angle variable is P.kin.secondary.th_pq

- also some discussion about units:
  - theta*,phi is in rad for data 
  - theta*,phi is in deg for SIMC (preferred)
  - phi in SIMC goes from -180 to +180, while in hcana it goes 0 to 2pi
- Richard did not make these consistent in recon_hcana, but did it in his
  analysis script
  - *NB* a good future task would be to add this in recon_hcana so SIMC and
    data have same units and ranges

- next step will be sig_UNS cross-section calculation from the Data/MC ratios
  and model

Richard
-------
KaonLT LT-sep analysis
- Q2=3.0, W=2.32
  - 5 iterations with same functional form as Q2=4.4
  - high epsilon, highest 2 t-bins:
    - some very low Data/MC ratios around phi=0, corresponding to right SHMS
      setting
  - overlay of Data and MC histos for kinematic quantities
    - MC has a double hump in W compared to data
  - lowest 2 t-bins have nice ratios

- Q2=4.4, W=2.74
  - Data/MC ratios overall look good
  - some shift between data and MC for Q2, W distributions
  - separated cross sections
    - sigL<sigT in general, but 3rd t-bin has sigL high compared to the others

- planned systematic studies
  - presents a summary of systmatic error types from KaonLT proposal

Vijay
-----
Low Q2 pion absorption correction study
- removed PID cuts from Alicia's script since he is using CoinTime and RF cuts
  instead of Cherenkov
- getting a smaller correction factor than Alicia at low pion momentum

- next step will be to finalize Q2=0.425 yields

Nacer
-----
Low Q2 KaonLT LT-sep analysis
- working on average kinematic variables per t-bin
  - averages are calculated after pion-leakthrough and Dummy subtractions
  - had to replay data because theta* was missing from ROOT trees
- formed R=Data/MC ratios for each t-phi bin
  - then form weighted average R per t-phi bin
  - when combining ratios, if counts=0 set error=100 so it doesn't contribute
    to weighted average

- found an error in boiling factor when computing yields
- also needs to correctly propagate the uncertainty in the beam current/charge
  - for KaonLT BCM calibrations, see Dave Mack's report:
    https://hallcweb.jlab.org/doc-public/ShowDocument?docid=1034

- Lambda: low epsilon, R about 2
  - high epsilon, some random bins are zero, overall R about 2
  - *NB* Richard: suggests to make MM plots for each t-phi bin to better see
    what's going on, maybe the pion subtraction needs adjustment?

- Sigma: R about 8
  
Alicia
------
u-channel analysis
- incorporating Pythia generator for high MM background into SIMC
  - got output files from Pythia into SIMC
  - Sylvester Joosten had added a lot of containerization to SIMC for ANL
    cluster that Alicia removed
  - still getting some code crashes
- hoping to have some plots by next week

Next Meeting
-------------
- Thur May 23 @ 15:30 Eastern/13:30 Regina
  - KaonLT will go first
  ** THIS TIME WILL BE USED AT LEAST UNTIL THE END OF AUGUST **