Jul 14/22 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, Nathan Heinrich, Garth Huber, Ali Usman, Vijay Kumar, Love Preet CSULA - Konrad Aniol Ohio - Jacob Murphy CUA - Tanja Horn, Richard Trotta JLab - Dave Gaskell Richard Updates --------------- - PyRoot-based L/T-separation package - Got the information he needed from Bill's files via Garth's back-up - Changed formatting to fit our efficiency list style a little more - LTSep python package changes - Ver 3.3.0 is latest version - Branches kept are included in runtype - Branch definitions - Grabs required branches - Code runs slightly faster - New version is backwards compatible - Should be able to merge in straightforwardly - Comparison of W distributions between MC and Data for 10.6 GeV HeepCoin - Data always lower than MC - changing up to 4 offsets - W should not move in MC, since it knows it's elastic scattering! - Replay didn't shift W with offsets - That's very odd, doesn't seem to be changing in hcana - clearly there is a bug that Richard will need to track - Richard 2nd updates - One issue is that Richard is applying the offsets to std.kin, instead of the offsets file. *This is dangerous*, and was discussed in the 22-May-17 meeting, as it can lead to errors and inconsistences. *Much better to use offsets file* - Vijay's replay shifts as expeccted, using the offests file. Ask that Vijay and Richard discuss this offline Vijay Updates ------------- - Comparison of various quantities between Heep coin data/MC for 5 energies - No offsets applied - Sees larger shifts in EM, PMx,y,z between Data and MC at 2.7, 4.5, 4.9 GeV, and smaller shifts at 3.6, 3.9 GeV - Consistency between similar beam energies between different run periods - The good news is that 3.6, 3.9 GeV behave similarly, even though taken roughly 6 months apart. i.e. the behavior is reproducible. - *Tanja's quick back of the envelope offsets calculation* - For 3.6, 3.9 GeV, momentum offset in electron arm would account for offsets - Larger offsets in other 3 settings, angle offset (~1 mrad) in addition to small momentum offset - EMiss offset -> Need a momentum offset in SHMS (hadron arm): ~0.3% at 2.7/4.5/4.9GeV - All settings - 0.1% momentum offset in HMS (electron arm) - 2.7/4.5/4.9 - 1 mrad electron angle offset -> 1 mrad = 0.0573 degrees - a possible reason why the offsets appear to differ at 3.6,3.9 GeV is that the sensitivity to the needed offset depends on the acceptance matching between the two spectrometers for that setting - i.e. does the electron arm (HMS) define the cone of proton coincidences within the proton arm (SHMS), or the other way around, or are they matched? - some settings are better matched kinematically than others. Best matching occurs when the two spectrometers are near equal angles and momenta. Ali Updates ----------- - Brief analysis for increasing range of SHMS delta after discussions with Peter B - Aerogel > 1.5 NPE cut, would need to alter if we increase delta - n = 1.011 aerogel tray does not cover full focal plane and may be restricting our negative delta acceptance - Dave/Stephen point out that the bigger difference will really be on adjusting the HMS delta too - SHMS and HMS delta correlated, with negative SHMS delta in coincidence with positive HMS delta for exclusive pi+/K+ events - Our online analysis indicates that we would only gain a few percent of exclusive events - HMS delta should NOT be expanded much - Stephen - Older settings should be re-analysed - Delta vs Delta plots not there previously - GH discusses with PB afterward. He agrees that this is the case for exclusive events, but he is interested in expanding the delta range for SIDIS events in our data sample. So this does not sound like it should be a high priority for us. - CPULT > 100% - as shown at last meeting, CPULT are OK for COIN replay, but some SHMS CPULT are still over 100% - Changes to report file template - Updated error calculation does not resolve problem - For EDTM, should be using time RAW not time - CPULT should use time (not RAW) - will discuss more about this with Jacob tomorrow - When is the aerogel tray changed? - Peter B did not use the run list to determine what aerogel tray is used, but instead looked at the x-y histogram, and placed limits on the edge of the efficient region - he finds some sets of run numbers where he thinks n=1.011 tray is used that differs from Richard's list - the run plan and logbook are unambiguous on when the aerogel was changed. *Ali should independently go through the logbook* to note when aerogel was changed, and see who is right - Trying to make EM, PMx,y,z calculations the same between hcana and SIMC - had meeting with DG to understand what variables to look at in SIMC code - Hopefully able to compare to hcana variables shortly - Dave - PMx/PMy/PMz variables in SIMC are just components in the lab frame - In Heep case, PM is translated relative to the q-vector - PMz/PMy, only calculates in-plane components (i.e. those in the central plane of the two spectrometers), ignores out of plane components (i.e. the fact that the q-vector is tilted vertically) - Thus, the SIMC versions are not really the missing momentum. 99% sure this this not the way it's done in hcana - Now that the inconsistency is identified, we need to fix it - Add new variables to hcana? - Or add them to SIMC? - Dave - Probably easier to change SIMC - New variable that is equivalent to SIMC variable in hcana should be easy too - Two pronged approach makes sense - Tanja suggests to Calculate first by hand and compare - Then decide which one to update - Dave will look into modifying SIMC - *Ali will work on this as a priority* - Singles scripts (Heep-singles and Lumi) are ready. Can replay after calibs done Jacob Updates ------------- - In preparation for cutting settings from run plan (i.e. needing to save time due to slow beam start), looking at how many delta scans needed, using KaonLT 2018 data (10.6GeV) with HMS p=6.6 GeV/c - W vs xfp, W vs xpfp and W vs delta comparisons - after optics calibrations are done, the agreement between data and SIMC is good. Then does predictions for 2022 10.5 GeV delta scan - 3 point scan should be ok: 0 and +/-8% - More stats in -xfp and -delta would be useful -> More in -8% run - Keep overall time the same, shift time from +/4% to +/-8% -> -8% needs ~2x time alloted (roughtly 60% more than 0 setting) - Can cut some settings Junaid updates -------------- - not present. He has adjusted the Heep singles settings in the run plan and distributed a new spreadsheet Nathan Updates -------------- - Prepping for Gordon conference - Plots for HGC calibration - good fits for PMTs 1,2,4. PMT3 fits having a hard time - width for 2nd PE peak is far too broad - PMT3 gain is a bit different, peaks closer together - constrain range the width is allowed to vary in for 2nd PE fit? - can try a manual fit to see if it works better Next meeting ----------- - 22-Aug-04 at 11:30 Eastern, after RC meeting - Before Gordon Conf, after EIC meeting