Aug 17/22 PionLT/KaonLT Analysis Meeting Notes ---------------------------------------------- (Notes by GH and SJDK) Please remember to post your slides at: https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings Present: Regina - Stephen Kay, Garth Huber, Vijay Kumar, Ali Usman, Nathan Heinrich, Muhammad Junaid, Love Preet CUA - Richard Trotta, Tanja Horn JLab - Dave Gaskell Richard Trotta Updates ---------------------- - Comparing PM, EM equations between SIMC and hcana - showed some derivations comparing the equations in the two codes - confirmed the PM equations have a flipped minus sign in SIMC versions being used for pmx etc - pmy and pmz look better, pmx seems to have some tail still - However, it is very important to also be sure the coordinates are not rotated, in addition to the sign change - SIMC x is data y - Sign flip and component flip - Z along beam, x points down, y is beam left (SIMC co-ordinates) - SIMC elastics analysis - From W dist, very few inelastics - From hcana, Emiss is -ve, secondary particle causing issues? - shift in EM is almost certainly an offset issue, not real - data EM wider than SIMC. - Vijay saw it was wider than SIMC, but not by as much as Richard's - Richard/Ali at higher momentum - Are the "best" SHMS matrix elements being used here? - Magnetic reconstruction matrix is probably the cause of the broadening - DG: EMiss vs SHMS delta might be a useful plot - Check for unphysical correlations, e.g. that EM is not same for all delta, or wiggles in correlations where they are expected to be smooth - Will post HeeP plots and write up on redmine - Strongly looks like some magnetic optics reconstruction effect - Will add EMiss/PMiss vs delta/xp/yp etc, look for wiggles and correlations - Will also check with Peter B if he saw anything like this in his analysis of KaonLT data Ali Usman Updates ----------------- - In switching matrix elements, Online_PionLT uses a different SHMS matrix element to the one Richard/Ali are using - KaonLT was using Holly's matrix elements - Online_PionLT is using Mark Jones updated matrix elements - Ali is replaying data with Mark Jones' matrix elements - These matrix elements are expliticly labelled by MJ as the ones that should be used for the 2021 PionLT data - A little unclear which matrix elements are "newer" though - Waiting on replays to finish - What about the HMS matrix elements? - Some high HMS momenta in the settings Ali is looking at - Haven't checked matrix elements here yet, will follow up with Jacob - There are two lower momentum HMS matrix elements available, Jacob recommends a certain one - Compare with cdaq, see the one that is being used currently in hcana.param - Haven't tried plotting physics data yet, just looking at HeeP - Richard/Ali discussed implementing efficiency script to the data - When calculating normalized yields, should we apply it averaged to all fo the data in a setting, or run-by-run? - Yes, do it run by run - In Fpi-2, we calculated an "effective beam charge" Qeff=Q*effic*livetime, run by run, so the charge was scaled down, rather than scaling up the observed counts - also propagate the statistical and random errors run-by-run - normalization systematic uncertainties should be dealt with at the end Vijay Kumar Updates ------------------- - HeeP study for 2.7, 3.6, 3.8, 4.5 and 4.9 GeV - Low energy PionLT and KaonLT settings - Offsets for each energy and angle offsets for each - Large momentum offset at low SHMS momentum - Small offset at moderate momentum - Increases again at large momentum - Similar in the energy offsets too - Two points with 2.583 GeV/c SHMS momentum - Two different offsets at the same momentum - These two points are both from KaonLT (December 2018), 3.8 GeV and 4.9 GeV - Implies SHMS is not reproducible, which is not the case - Fix certain parameters, let others vary, based upon physics knowledge - Can we find some consistent set of offsets? - Can't just ignore two points because they don't fit the trend - Do a new offset iteration on all five settings, try and get something that is globally self consistent - DG: In the past, could fit bulk of HMS settings with a single angle offset - Single momentum offset until they reached higher momenta and saturation effects came in - no non-linearity in dipoles should exist at these momenta, don't understand why the offset should be varying a low momenta - i.e. the offset should depend on the physics of the spectrometer (saturation) and should not be random, just to get agreement between data and MC - Should be able to get a consistent set, it doesn't have to be the "optimal one" but rather an "acceptable and justifiable one" - Vijay has an offset solution, but not the correct one yet - Consider the five settings globally, not in isolation from each other Nathan Heinrich Updates ----------------------- - Prepping for comprehensive, no report Muhammad Junaid Updates ----------------------- - Working on runplan - Comprehensive prep, no report - Status page up to date? - Stephen, working on adding the Q2 = 2.12 setting Next Meeting - Aug 31 @ 11:30 Eastern/09:30 Regina time - GH: this will likely be the last meeting at that time, and after the run is over we should reconsider the meeting schedule - KaonLT/PionLT related, but at different analysis stages, should we divide the meetings? i.e. everyone is invited to both, but who gets to present will be different - KaonLT meeting one week, PionLT the next? Richard in particular thinks this is a good idea At end, Vijay shows a New Plot ------------------------------ - starting to look at possible t-phi binning for the physics data - One binning for pion data, different one for kaon - Don't look separately at L,R,C Add together left/center/right, see what binning looks like - Vijay asks to have a separate meeting where GH explains in more detail how the SIMC model optimization is done - GH: next week would be good - the way it is done in practice probably differs a bit from how its explained in theses,