Online Monitoring developments Dorota, Robert and Voica LArSoft - PowerPoint PPT Presentation

Online Monitoring developments Dorota, Robert and Voica

LArSoft modules for OnlineMonitoring working on LArSoft dunetpc/dune/Protodune: git branch: feature_vradescu/OnlineMonitor 1. Module: RawEeventDisplay saves in root the event displays per event for each APA in each View (mimics the standard event display in lar) • mean • RMS Z view U view Marco will process this root file for user friendlier model in his tool for web display V. Radecu | U.Oxford/CERN

LArSoft modules for OnlineMonitoring working on LArSoft dunetpc/dune/Protodune: git branch: feature_vradescu/OnlineMonitor 1. Module: RawEeventDisplay saves in root the event displays per event for each APA in each View (mimics the standard event display in lar) 2. Module: OnlineMonitor - can be used for “N" events saves the raw uncompressed ADC counts (-minus Pedestal) in each channel per time ticks • mean • RMS Over 5 events: RMS Chan # same y scale -> one can see how it jumps with noise V. Radecu | U.Oxford/CERN

LArSoft modules for OnlineMonitoring working on LArSoft dunetpc/dune/Protodune: git branch: feature_vradescu/OnlineMonitor 1. Module: RawEeventDisplay saves in root the event displays per event for each APA in each View (mimics the standard event display in lar) 2. Module: OnlineMonitor - can be used for “N" events saves the raw uncompressed ADC counts (-minus Pedestal) in each channel per time ticks • mean Over 1 event: here shown for one APA • RMS U view Mean RMS V view Chan # V. Radecu | U.Oxford/CERN

Wish List Reminder

Profiling exercise To profile, i have used igprof which does not require re-configuration and it works with the default LArSoft optimisation: However, the demangle does not really work and then it’s hard to digest the performance: igprof -t lar -o igprof_ED.gz lar -c jobs/eventdisplayprotodune.fcl ../Dec2016/data detsim_high_noise_cosmics_beam_evs6.root -n1 igprof-analyse --sqlite --demangle -v igprof_ED.gz | sqlite3 igprof_ED.sql3 igprof-navigator -p 8080 igprof_ED.sql3 & using e10:prof using debug:e10 V. Radecu | U.Oxford/CERN

Profiling exercise Also tried on a standard reco fcl file using debug:e10 i gprof -t lar -o igprof_recoStd.gz lar -c protoDUNE_reco.fcl ../Feb2017_v22/pi_detsim.root -n1 igprof-analyse --sqlite --demangle -v igprof_recoStd.gz | sqlite3 igprof_recoStd.sql3 igprof-navigator -p 8080 igprof_recoStd.sql3 & firefox http://neut.cern.ch:8080 while, i’d expect smthg like: For now it does not work, awaiting from experts for more feedback on how to make the demangle working V. Radecu | U.Oxford/CERN

Summary Progress on the metrics filling: decided with a modular structure analysis done in LArSoft (based on art Framework) individual modules that can be loaded depending on the needs: per 1 event, per x events .. input is based on simulated data with different noise scenarios —> helps to identify sensible metrics Test of the full chain propagation ongoing: connect the output of modules (root file) to the web browser (Marco) define scripts how to connect all pieces Profiling tools: started the tests using igprof, however the demangle (human readable reports) not working properly yet V. Radecu | U.Oxford/CERN

event displays for different det_sim (event #2) high noise: medium noise zero exp. noise

Running logs for deconvolution take the 6 events of cosmic muons over beam+ noise: medium noise: time to run simple reconstruction over detsim: =============================================================================================================================== TimeTracker printout (sec) Min Avg Max Median RMS nEvts =============================================================================================================================== Full event 21.4558 26.0452 38.3775 23.8579 5.80397 6 ------------------------------------------------------------------------------------------------------------------------------- reco:rns:RandomNumberSaver 0.000220929 0.00110953 0.00309061 0.000491943 0.00108128 6 reco:ophit:OpHitFinder 0.435359 0.594274 0.963128 0.490928 0.193101 6 reco:opflash:OpFlashFinder 2.02451 3.41731 4.75797 3.41418 0.828333 6 reco:caldata:CalWireDUNE10kt 13.711 14.7992 15.6898 14.8431 0.833141 6 reco:gaushit:GausHitFinder 3.33986 6.77727 16.2454 5.14749 4.39567 6 reco:hitfd:HitFinder35t 0.00839139 0.0160175 0.0342197 0.012183 0.00885606 6 reco:linecluster:LineCluster 0.0197881 0.0655452 0.136079 0.0592817 0.0360106 6 reco:TriggerResults:TriggerResultInserter 5.2742e-05 6.48257e-05 8.4454e-05 5.75355e-05 1.28505e-05 6 end_path:out1:RootOutput 1.0858e-05 3.1749e-05 0.000112482 1.77555e-05 3.62612e-05 6 end_path:out1:RootOutput(write) 0.201762 0.369787 0.623712 0.310109 0.164279 6 =============================================================================================================================== no noise =============================================================================================================================== TimeTracker printout (sec) Min Avg Max Median RMS nEvts =============================================================================================================================== Full event 11.9912 15.2422 23.6715 13.5099 4.04151 6 ------------------------------------------------------------------------------------------------------------------------------- reco:rns:RandomNumberSaver 0.00018271 0.000280978 0.000722385 0.000195173 0.000197536 6 reco:ophit:OpHitFinder 0.33114 0.429604 0.781332 0.349178 0.160397 6 reco:opflash:OpFlashFinder 1.81249 1.95899 2.22249 1.89011 0.151792 6 reco:caldata:CalWireDUNE10kt 6.84066 7.20328 8.79466 6.89365 0.712647 6 reco:gaushit:GausHitFinder 2.58688 5.37734 13.8941 4.27916 3.88823 6 reco:hitfd:HitFinder35t 0.00518069 0.00895443 0.0126466 0.00887943 0.00250608 6 reco:linecluster:LineCluster 0.0170539 0.0377476 0.0941708 0.0258525 0.0264651 6 reco:TriggerResults:TriggerResultInserter 4.3333e-05 4.98865e-05 7.769e-05 4.36245e-05 1.2527e-05 6 end_path:out1:RootOutput 5.294e-06 7.3985e-06 1.5571e-05 5.8955e-06 3.6633e-06 6 end_path:out1:RootOutput(write) 0.183723 0.225156 0.376499 0.192741 0.0688089 6 ===============================================================================================================================