---+++ CMSSW Mixing Module - Known issues The modul loaded by the --pileup AVE_35_BX_25ns cmsDriver.py option: <verbatim>process.load('SimGeneral.MixingModule.mix_POISSON_average_cfi')</verbatim> Source available at: http://cmslxr.fnal.gov/source/SimGeneral/MixingModule/python/mix_POISSON_average_cfi.py Extra commands in the generated cfg file (these are implicit in the flat configuration): <verbatim> process.mix.bunchspace = cms.int32(25) process.mix.minBunch = cms.int32(-12) process.mix.maxBunch = cms.int32(3) </verbatim> The modul loaded by the --pileup Flat_20_50 cmsDriver.py option: <verbatim>process.load('SimGeneral.MixingModule.mix_Flat_20_50_cfi')</verbatim> Source available at: http://cmslxr.fnal.gov/source/SimGeneral/MixingModule/python/mix_Flat_20_50_cfi.py The differecences between the two mixing modules (extracted using the python interactive mode) (diff -ub poisson flat): <style type="text/css"> code { color: white; font-weight: 600; } code pre { color: inherit; display: inline; } .redbg { background: rgb(220, 30, 15); } .lightgreenbg { background: rgb(30, 230, 180); } .yellowbg { background: rgb(235, 205, 10); </style> <code> <div class="lightgreenbg"> 708a709 </div> <div class="yellowbg"> <pre>> OOT_type = cms.untracked.string('Poisson'), # property of process.mix.input</pre> </div> <div class="lightgreenbg"> 717a719 </div> <div class="yellowbg"> <pre>> manage_OOT = cms.untracked.bool(True), # property of process.mix.input</pre> </div> <div class="lightgreenbg"> 719c721,743 </div> <div class="redbg"> <pre>< averageNumber = cms.double(35.0) # property of process.mix.input.nbPileupEvents</pre> </div> <pre style="color: black;">---</pre> <div class="yellowbg">> histoFileName = cms.untracked.string('histProbFunction.root'), # property of process.mix.input.nbPileupEvents <pre>> probFunctionVariable = cms.vint32(0, 1, 2, 3, 4, # property of process.mix.input.nbPileupEvents </pre> <pre>> 5, 6, 7, 8, 9, </pre> <pre>> 10, 11, 12, 13, 14, </pre> <pre>> 15, 16, 17, 18, 19, </pre> <pre>> 20, 21, 22, 23, 24, </pre> <pre>> 25, 26, 27, 28, 29, </pre> <pre>> 30, 31, 32, 33, 34, </pre> <pre>> 35, 36, 37, 38, 39, </pre> <pre>> 40, 41, 42, 43, 44, </pre> <pre>> 45, 46, 47, 48, 49, </pre> <pre>> 50), </pre> <pre>> probValue = cms.vdouble(0, 0, 0, 0, 0, # property of process.mix.input.nbPileupEvents </pre> <pre>> 0, 0, 0, 0, 0, </pre> <pre>> 0, 0, 0, 0, 0, </pre> <pre>> 0, 0, 0, 0, 0, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065, 0.032258065, 0.032258065, 0.032258065, 0.032258065, </pre> <pre>> 0.032258065) </pre> </div> <div class="lightgreenbg"> 722c746 </div> <div class="redbg"> <pre>< type = cms.string('poisson') # property of process.mix.input</pre> </div> <pre style="color: black;">---</pre> <div class="yellowbg"> <pre>> type = cms.string('probFunction') # property of process.mix.input</pre> </div> </code> It seems that management of the out of time pileup is necesarry and it's mode should be set to 'Poisson': <verbatim> process.mix.input.OOT_type = cms.untracked.string('Poisson') process.mix.input.manage_OOT = cms.untracked.bool(True) </verbatim> %ICON{"help"}% Adding this lines to a cmsDriver configuration file still seems to produce errors. ---+++ Known error messages: %ICON{"warning"}% Message: <verbatim>OscarMTMasterThread::stopThread() has not been called to stop Geant4 and join the master thread terminate called without an active exception</verbatim> This is an useless error message. It is known to occur when the probability function used to generate the pileup distribution is not properly normalized. A guess for what it might mean: The mixingmodule starts children processes to generate a given number of one-pileup events at the same time for having them mixed later. Whenever it requires no more pileups mixed, it stops waiting for its children processes and it might cause some kind of error. %ICON{"warning"}% Message: <verbatim> ----- Begin Fatal Exception 17-Feb-2017 11:15:31 CET----------------------- An exception of category 'FatalRootError' occurred while [0] Processing run: 1 lumi: 1 event: 1 [1] Running path 'HLTriggerFirstPath' [2] Calling event method for module HLTGetRaw/'hltGetRaw' [3] Calling produce method for unscheduled module RawDataCollectorByLabel/'rawDataCollector' [4] Calling produce method for unscheduled module SiStripDigiToRawModule/'SiStripDigiToRaw' [5] Calling produce method for unscheduled module MixingModule/'mix' Additional Info: [a] Fatal Root Error: @SUB=TH1F::ComputeIntegral Integral = zero ----- End Fatal Exception ------------------------------------------------- </verbatim> This error seems to come from the pileup mixing, possibly from a bad normalization: http://cmslxr.fnal.gov/source/Mixing/Base/src/PileUp.cc -- Main.AdamHunyadi - 2017-02-21
This topic: CMS
>
WebHome
>
NewPhysics
>
CmsSusy
>
CMSSusyTWiki
>
MixingModuleKnownIssues
Topic revision: r3 - 2017-03-16 - AdamHunyadi
Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki?
Send feedback