Oct 3/25 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, Alicia Postuma, Nermin Sadoun, Nathan Heinrich, Vijay Kumar, Nacer Hamdi, Muhammad Junaid Virginia - Richard Trotta Ohio - Julie Roche JLab - Dave Gaskell CUA - Tanja Horn, Chi Kin Tam, BC Manoj FIU - Pete Markowitz Nathan ------ PionLT Coin Lumi analysis - final analysis of Coin Yield vs Coin Rate (up to ~2 kHz) - fit individual slopes for each setting, and propagated errors to obtain a weighted average correction of 3.1%/kHz +/- 0.07%(stat) +/- 0.35%(syst) where the systematic uncertainty comes from the variation in slope with setting, about a 15% uncertainty in the correction overall - also looked at Coin Yield vs HMS Rate (up to 180 kHz) to see if it's better 0.0355%/kHz +/- 0.009%(stat) +/- 0.024%(syst) the combined error is almost 100% of the correction! Definitely it's better to use the correction vs Coin Rate Junaid ------ PionLT Q2=3.85 W=2.62 LT-sep - Data/MC ratios had phi-oscillations last week, so modified LT function to sigLT=(p9/Q2+p10/|t|)*sin(thetacm) - after 2 iterations: low epsilon has oscillations in higher t-bins high epsilon has oscillations across a wider range of t-bins - then tried sigLT=(p9/Q2+exp(-10*|t|)*p11/(p12+|t|))*sin(thetacm) - initially started with p10=p12=0 ratios much better behaved - then allowed p10,p12 to float ended up with more oscillations at high epsilon now doing 3rd iteration to see what happens - Garth suggests to move on to one of the other Q2=3.85 settings, as this seems faiarly good and you may learn lessons there to apply to this setting - Richard agrees this is ready to move to next setting BC Manoj -------- - setting up iFarm, got help from Junaid - will look at Heep delta-scans for optics Alicia ------ BSA changes to response to referees - doing some restructuring of response to put "what we did" at beginning of discussion, followed by more details - planning to resubmit on Monday Vijay ----- New Q2=0.375, 0.425 results - previously sigL,T were very simple, with no Q2-dependence - now added (1/Q4) dependence to both - made a spreadsheet comparing LT-sep values for both function fits, differences are larger than statistical uncertainties at lowest -t, similar to statistics at higher -t, so we might need some further improvement - GH made a series of plots comparing Vijay's data to Ackermann - Q6 and Wfac scaling applied to Ackermann data to approximately bring to same kinematics - generally good agreement for sigL, but our sigT values are much higher - some discrepancies between different parts of Ackermann data, not clear exactly how they have binned/combined different parts of their dataset Next steps - need to apply pion absorption corr to pi+ data, and proton absorption corr to Heep data - also will make raw CoinTime plots like Nathan and investigate Coin Blocking correction Richard ------- Working on a paper "The Fit-Finder Algorithm for LT-separation Analysis" - algorithm to get best functional form from separated cross sections - testing the algorithm on different datasets for purposes of paper - looking at Fpi-2 Q2=1.6, 2.45; W=2.22 separated cross sections - Vijay and Nacer will send their latest results KaonLT LT-sep - currently replaying Q2=3.0, 4.4, 5.5 data Chi Kin ------- KaonLT LT-sep - Code Block regarding BCM current file was commented out in forming replay file, this has been fixed - used new database file, Expt Diamonds now agree much closer to SIMC diamonds - Q2=3.0, W=3.14 analysis - chose 4 t-bins, 15 phi-bins, will compare to Richard's results - did pion subtraction for all t-phi bins, in some cases sees a Sigma peak which is also subtracted - Iter 00: start with Richard's recommended parameters - Iter 01: Data/SIMC ratios move much closer to 1 - Iter 02: ratios blow up, but come back after 4 iterations - looked at trend in L,T vs iteration number, found that sigT fit failed at Iter 04, will be fixed - comparsion of data and MC focal plane and physics histos - no MM cuts yet on data, so sees differences w/SIMC - will include pion subtraction, etc Nacer ----- KaonLT low Q2 LT-sep - will try different models for L,T - first step is to recheck MM plots for each (t,phi) bin - needed to replay data as didn't have MM for each (t,phi) before - then will try to add a functional form for LT,TT, right now they are both zero in the parameterization, but non-zero in the Rosenbluth fit Nermin ------ Like BC Manoj, setting up environment on iFarm, getting info from Junaid Garth ----- - still needs to look at Gabi's HeepCheck code modifications and generate new offsets - in about a month, is willing to give the new students a tutorial on the PionLT run plan, describing how an LT-separation expt is structured, so they can better understand how everything fits together Next Meeting ------------ - Fri Oct 10 @ 11:00 Eastern/9:00 Regina - KaonLT will go first - We will see if there's time on Thursday for a short meeting, in order to keep the Friday meetings from getting too long Please fill out the poll at: https://www.when2meet.com/?32764634-ro4er