Nov 9/22 PionLT/KaonLT Analysis Meeting Notes --------------------------------------------- (Notes by GH and SJDK) 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, Stephen Kay, Ali Usman, Alicia Postuma, Love Preet, Vijay Kumar, Nathan Heinrich, Muhammad Junaid CSULA - Konrad Aniol, Johnathan Conrad CUA - Richard Trotta, Tanja Horn JLab - Dave Gaskell FIU - Pete Markowitz SBU - Wenliang (Bill) Li Ohio - Jacob Murphy, Julie Roche Bill LT Software Tutorial - Recorded ------------------------------------ The recording is posted at: http://lichen.phys.uregina.ca/~huberg/kaon/LT_tutorial_20221109_1920x1080.mp4 - LT separation overview - based on c++/ROOT - cross-checked translation of PAW kumac codes used in Fpi-2 analysis - Tanja's kumacs were different than Jochen's (Fpi-1), but separation method was similar - 8 phi bins, 3 u (t in our case) bins - the #bins is based on statistics and kinematic coverage, yours will be different! - each (W,Q2) setting is done separately - ubin_interval: specifies ranges of u-bins for each (W,Q2) setting - Background subtraction - will be less complex in Pion/Kaon case, fewer simulation inputs needed - 10 step flow chart - Efficiency (run by run) - Yield extraction in root_ana folder - Plot combination - BG subtraction (bin-by-bin in Bill's case) - Sum angles - ... - Some files in the git repo are archival, should be moved to separate folder when a HallC version of repo is made - Richard points out that lots of cuts done in Bill's code are already done in our analysis - Rather than removing cut function, Bill recommends just adding in a "blank" cut that does nothing and applying it to the incoming trees - Angles added and dummy target subtracted in the same script - scalers_fpi contains scaler text files run-by-run for each setting - original analysis structure of Fpi-2 PAW kumacs is retained, so everything is in separate files. Bill recommends to NOT consolidate files, since everything is working well and follows structure of previously tested code - analysis.cpp is verified translation of kumac code - In analysis.cpp file, just assign cuts to Dummy cuts, since we will have cuts done elsewhere - the code setup allows run-by-run efficiencies to be applied in the normalization - Existing LT sep code does cuts in a somewhat problematic way - Uses ROOT trim functions to apply cuts - T->Draw("...", "...Cuts...") etc - This is fine for quickly drawing plots, but not ideal for final analysis - it was only done this way because Bill was a grad student, and learning how to do things - now he's more experienced and understands that going forward, this should be changed to an *Event-Based IF loop* - Once analysis makes root ana etc, this is the only input for subsequent steps? - Yes, from that point this is all that's needed - Bill suggests to first look at Heep code to understand the structure of the code - Bill will move setup examples to somewhere on the farm - Stephen will copy to c-kaonlt or similar - Sample data, sample simulations - How are SIMC weights re-calculated? - For each iteration, don't re-run SIMC, just recalculate the weights Jacob Updates ------------- Optics: - meeting with Holly on Thursday - trying to get improved ME for HMS=5.9, 6.8 GeV/c - rewriting some scripts to make them more user friendly - Xsieve resolution is much worse than Ysieve - x resolution for upstream foil isn't great even at -8 - -6.5 in delta - Mark J has seen similar issues - will have more results next week - Going through standard.kinematics - 2022 easier, some missing angles from 2021 - 1 in 5 settings or so has angle off by ~ 0.01 - Should be done by today for 2022, 2021 is already done - Need to set up hallc_replay with PionLT param files - Offline branch exists https://github.com/JeffersonLab/hallc_replay_lt/tree/LTSep_Analysis_2022 - Should set up DB/Param directories in similar manner to KaonLT - https://github.com/JeffersonLab/hallc_replay_lt/tree/LTSep_Analysis_2022/DBASE/COIN/DB_KaonLT - https://github.com/JeffersonLab/hallc_replay_lt/tree/LTSep_Analysis_2022/DBASE/COIN/DB_PionLT - As an example - Determing Pc (central momentum) offsets - Apply FP cuts, electron PID cuts - etottracknorm > 0.8 - npeSum > 2 - abs(delta)<10 - Graphical cut to select elastics along W vs xfp and W vs xpfp - Narrower cut placed to select central momentum region Mark Jones' request - abs(xfp) < 2 - abs(delta) < 2 - Using FP variables and optical matrix, recalculate W using updated central momentum (set by hand) - choose value that gives best proton mass - offset is difference between Optimized Pc and Set Pc - Jacob outlined procedure and sent to Vijay/Ali - Garth Q - SIMC calculation same as hcana calculations, still on the list? - Richard - Yes, looked at but not implemented yet Nathan Updates -------------- - Showed draft slides for PhD committee meeting, received many comments Junaid Updates -------------- - Working on PhD committee slides - draft not shown Vijay Updates ------------- Luminosity scaler study from PionLT (summer 2019) - Carbon lumi data from 2.7 GeV - Using EL_REAL triggers - Both spectrometers behave in similar way, which is good - Dip at low current should be fixable by adjusting the current offset - Working on non-tracked and tracked normalized yields next, to see if they behave similarly - Also going to look at BCM offsets - Stephen/Garth/Dave: - Can't apply a run-by-run current offset - BCM calibration has large uncertainty at low current <10uA - Offset is applied to ALL runs or NONE of them - It is the SAME BCM so cannot apply different current offset to different runs - There is second effect going on which is giving rise to the positive trend, need to also look at the electronic livetimes - Best to adjust one thing at a time - Get the low current sorted first using the BCM current offset - Should be fixable with a current offset adjustment - This is by far the dominant effect at low current - Then investigate the remaining effect at larger I Next Meeting ------------ Wed Nov 16 @ 8:15 Eastern/7:15 Regina/5:15 Pacific - KaonLT will go first