May 31/22 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, Nathan Heinrich, Ali Usman, Muhammad Junaid, Love Preet, Vijay Kumar (arrived late due to doctor appointment) Ohio - Jacob Murphy, Julie Roche CUA - Richard Trotta JLab - Dave Gaskell FIU - Pete Markowitz Richard Trotta -------------- - AU has been very helpful fixing some scripts - Charge corrections/normalisations for data were off - GH - Lumi scan Q, are charge normalizations not done here? - RT - Had things a bit mixed up when doing SIMC comparison, charge corrections were done in Lumi Scan? - SIMC data looks better, recompiled SIMC - SIMC kept producing somewhat odd results, not sure why this should have been needed - SIMC Blue, red is data (not charge corrected) - HeeP Coin data - *Shouldn't be hard coding charge normalisations/corrections* - Efficiencies work - Script to compile all relevant info to one csv file - AU needs aerogel positions still for cuts in report file - Script to get positions written, just checking a broad range - 2 week plan - Better HeeP data/MC comparison with fixed charge normalisations and SIMC fixes - Aerogel positions - Report file part of efficiency script should be done - Hodoscope calibrations (hopefully!) for 2018, 10.6 GeV data Ali Usman --------- - HeePCoin 8.2 GeV data - Red data, blue SIMC, charge normalised, dummy subtracted and SIMC weights - Despite changes, still some shifts apparent - No SIMC or data re-normalization needed *GOOD* - No offsets, no efficiencies applied yet - DG - Could just grep some values and use them for now? - Could just do a quick and dirty correction and see how it looks - DG - What is spectrometer central momentum? - ~4.3-4.6 GeV for both - x/y seem to be flipped between SIMC/Data - Just plot the wrong way around for now? - GH: *AU and VK need to check the equations in hcana and SIMC* to confirm the differences. For SIMC, look at event.f - also sees that Pmiss and Emiss distributions for MC are narrower than Data - *Reconstructed variables (Pmiss, Emiss) in SIMC not calculated in the same way as hcana* - this affects not only the variable mis-matches, but also the shapes of the distributions themselves, as the resolutions can work out better or worse depending on how the calculation is done - DG fairly certain on this. He tried looking at hcana code but no one-to-one correspondence between the hcana and SIMC equations was obvious - SK - I agree that this appears to be the case - Should be addressed long term - Standardise calculation in hcana/SIMC - This needs to be on the agenda long term - Peter B looked at HeeP and recalculated values himself, since he didn't trust that hcana and SIMC calculated them the same way - 2 week plan - Finish report files for Richard - Debug some of the HeeP issues - Implement some efficiencies - Hodoscope calibrations too (different energies) Vijay Kumar (Slides presented by GH) ----------- - Data/SIMC comparisons, HeePCoin - Old -> Improved - Used Jochen's Heepcheck program, 0.13% HMS momentum offset applied - Mean quoted on plots, histogram mean? - Found bug last week relating to how weight was applied - Was applying aeverage weight to all events, NOT weighting each individual event - Scripts now use weight correctly, distributions now have similar shape - Flipped pmz - AU Q - Is pmz flipped in the plot or when calculating it? - GH - Assume probably just when filling histogram - 2 week plan - Need to do 4.8 GeV and other Heep COIN - include Effs - should help track pmx,y,z differences between hcana and SIMC Pete Markowitz -------------- - Hall C UG wants two talks - One on PionLT/KaonLT talk - One on the Pion form factor - 20+5 talks, Thursday afternoon June 16th - Vijay -> KaonLT - Junaid -> PionLT - Circulate 1st draft of slides by 10th of June - Dry run: 14th of June at 15:30 Eastern Jacob Murphy ------------ - Hodoscope calibrations are at a pretty good point going in to the upcoming experiment - will start upcoming run with best params from fall/winter run - will need to redo calib w/defocused data soon after data taking starts - EDTM DeadTime calcs - wrote up a report which has been circulated around a little bit - will make improvements and distribute to whole PionLT/KaonLT group - How event type 3 is handled is a bit odd - Need to try and make sense of the event types and how they are actually determined - the issue seems to be when prescale factors are checked by the DAQ - event type 3 is BOTH singles - the issue is when an event is accepted in both singles triggers at the same time, but only one is saved - this will happen for every EDTM event, unlikely to happen for physics events - *event type preference order: 6,5,3,1,2,4* - COIN physics events are fine, EDTM goes through COIN trigger - but COINs are turned off during Heep singles and Lumi in this case one arm will take precedence - *need to set EDTM rate for lowest rate arm* - DG - Should we try running singles completely on their own? Try and disentangle this - JR - Would be nice to try to disentangle this further - One trigger at a time would double the time it takes to do HeeP singles though. - Can we switch the singles DAQs? - Would need to switch, then switch back to COIN DAQ - Reset all timing windows etc each time too, so it's not very practical - GH - Are we not ok if we only prescale one arm? - JM - Didn't consider EDTM when considering this - *Any time you have a PS=0 in SHMS, you will *never* see EDTMs in HMS* - Presentation on all of this info back in February https://redmine.jlab.org/attachments/download/1401/PionLT_EDTM_Study2_Analysis.pdf - Problem has always been there, but never understood - 6 GeV data handled very differently - CPULT can still be calculated without too much issue - Physics/Coin runs, EDTM value is fine - Single arm EDTM LT is the issue - PionLT summer run plan optimisations discussions - *Take another look at 10 GeV run plan, HeepSing and Lumi*, do any of these present an issue? - Skip some lumi scans and do some as single arm? - P > 5 GeV/c in HMS on one lumi scan? - Want carbon optics and delta scan for anything with HMS P > 5 GeV/c - DG - No LUMI just a white spectrum, shouldn't care - but *delta scan needed for every physics setting with HMS P > 5 GeV/c* - discussion on sieve runs - can do rastered, 1x1 is preferred over 2x2 - note Run 4819 (KaonLT) used 2x2 and sieve plots look fine - Need to focus on decisions wrt 10 GeV asap - GH, JM, NH, MJ will discuss again at tomorrow's Run Plan meeting - 2 week plan - finish 10 GeV run plan - some EDTM studies w/ event types - revisit report on EDTM, make implications clearer Junaid ------ - Finished lumi scans part of summer run plan - Arrived at several settings - sent comments on NH, JM plans. 2nd draft posted on RedMine - Shift Worker instructions circulated for review - Not on Redmine yet - JM will post on Redmine - DC/Cal calibrations - path issues fixed, deciding on runs to do - 2 week plan - Detector calibrations - Hall A/C Presentation Nathan ------ - Run plan draft on the Redmine - Spreadsheet on there too - Trying to get the HCG/NGC calibration working - Issues running it for Nathan - Vijay checked and it ran OK for him - Need to be 100% sure the conditions are identical - Trying to pull script out of TProof framework - Script seems to have open brackets and errors included, is version the same as Vijay's? - Non TProof version probably best - Could just use TChain or just hadd would work ok Shifts ------ - GH will circulate another reminder for people to sign up Comment from Tanja ------------------ - Redmine - Would be nice to add an analysis task/timeline to the PionLT redmine Next Meeting ------------ 15:30 EDT on the 14th of June - main agenda item will be Vijay and Junaid dry runs