# Copyright (C) 2002-2020 CERN for the benefit of the ATLAS collaboration
from AthenaConfiguration.AthConfigFlags import AthConfigFlags
from AthenaCommon.Logging import logging
log=logging.getLogger('TriggerConfigFlags')
def createTriggerFlags():
flags = AthConfigFlags()
# enables L1 simulation
flags.addFlag('Trigger.doLVL1', lambda prevFlags: prevFlags.Input.isMC)
# enables L1 topological trigger simulation
flags.addFlag('Trigger.doL1Topo', True )
# writes additional info from Topo simulation
flags.addFlag('Trigger.writeL1TopoValData', True )
# need proper documentation
flags.addFlag('Trigger.useL1CaloCalibration', False)
# need proper documentation
flags.addFlag('Trigger.useRun1CaloEnergyScale', False)
# enable HLT part of the trigger
flags.addFlag('Trigger.doHLT', True)
# changes decoding of L1 so that allways all configured chains are enabled, testing mode
flags.addFlag("Trigger.L1Decoder.forceEnableAllChains", False)
# Enable Run-3 LVL1 simulation and/or decoding
flags.addFlag('Trigger.enableL1Phase1', False)
# Enable usage of new L1 menu
flags.addFlag('Trigger.readLVL1FromJSON', True)
# Enable Run-2 L1Calo simulation and/or decoding (possible even if enablePhase1 is True)
flags.addFlag('Trigger.enableL1CaloLegacy', True)
# Enable Inner Detector
flags.addFlag('Trigger.doID', True)
# Enable muon system
flags.addFlag('Trigger.doMuon', True)
# Enable calorimeters
flags.addFlag('Trigger.doCalo', True)
# if 1, Run1 decoding version is set; if 2, Run2; if 3, Run 3
def EDMDecodingVersion(flags):
log.debug("Attempting to determine EDMDecodingVersion.")
version = 3
if flags.Input.Format=="BS":
log.debug("EDMDecodingVersion: Input format is ByteStream")
inputFileName = flags.Input.Files[0]
if not inputFileName and flags.Common.isOnline():
log.debug("EDMDecodingVersion: Online reconstruction, no input file. Return default version, i.e. AthenaMT.")
return version
log.debug("EDMDecodingVersion: Checking ROD version.")
import eformat
from libpyeformat_helper import SubDetector
bs = eformat.istream(inputFileName)
rodVersionM = -1
rodVersionL = -1
# Find the first HLT ROBFragment in the first event
for robf in bs[0]:
if robf.rob_source_id().subdetector_id()==SubDetector.TDAQ_HLT:
rodVersionM = robf.rod_minor_version() >> 8
rodVersionL = robf.rod_minor_version() & 0xFF
log.debug("EDMDecodingVersion: HLT ROD minor version from input file is {:d}.{:d}".format(rodVersionM, rodVersionL))
break
if rodVersionM >= 1:
version = 3
return version
log.info("EDMDecodingVersion: Could not determine ROD version -- falling back to run-number-based determination")
# Use run number to determine decoding version
runNumber = flags.Input.RunNumber[0]
log.debug("EDMDecodingVersion: Read run number {}.".format(runNumber))
boundary_run12 = 230000
boundary_run23 = 368000
if runNumber <= 0:
log.warning("EDMDecodingVersion: Cannot determine decoding version because run number {} is invalid. Leaving the default version.".format(runNumber))
elif runNumber < boundary_run12:
# Run-1 data
version = 1
elif runNumber < boundary_run23:
# Run-2 data
version = 2
else:
# Run-3 data
version = 3
else:
log.debug("EDMDecodingVersion: Input format is POOL -- determine from input file collections.")
# POOL files: decide based on HLT output type present in file
if "HLTResult_EF" in flags.Input.Collections:
version = 1
elif "TrigNavigation" in flags.Input.Collections:
version = 2
elif "HLTNav_Summary" in flags.Input.Collections:
version = 3
elif flags.Input.Format == "POOL":
# If running Trigger on RDO input (without previous trigger result), choose Run-3
version = 3
log.info("Determined EDMDecodingVersion to be {}.".format({1:"Run 1", 2:"Run 2", 3:"AthenaMT"}[version]))
return version
flags.addFlag('Trigger.EDMDecodingVersion', lambda prevFlags: EDMDecodingVersion(prevFlags))
# enables additional algorithms colecting MC truth infrmation (this is only used by IDso maybe we need Trigger.ID.doTruth only?)
flags.addFlag('Trigger.doTruth', False)
# only enable services for analysis and BS -> ESD processing (we need better name)
flags.addFlag('Trigger.doTriggerConfigOnly', False)
# Enables collection and export of detailed monitoring data of the HLT execution
flags.addFlag('Trigger.CostMonitoring.doCostMonitoring', False)
flags.addFlag('Trigger.CostMonitoring.chain', 'HLT_costmonitor_CostMonDS_L1All')
flags.addFlag('Trigger.CostMonitoring.outputCollection', 'HLT_TrigCostContainer')
flags.addFlag('Trigger.CostMonitoring.monitorAllEvents', False)
# enable Bcm inputs simulation
flags.addFlag('Trigger.L1.doBcm', True)
# enable muon inputs simulation
flags.addFlag('Trigger.L1.doMuons', True)
# version of CTP data, int value up to 4
flags.addFlag('Trigger.L1.CTPVersion', 4)
# list of thresholds (not sure if we want to use new flags to generate L1, leaving out for now?)
# partition name used to determine online vs offline BS result writing
import os
flags.addFlag('Trigger.Online.partitionName', os.getenv('TDAQ_PARTITION') or '')
# shortcut to check if job is running in a partition (i.e. partition name is not empty)
flags.addFlag('Trigger.Online.isPartition', lambda prevFlags: len(prevFlags.Trigger.Online.partitionName)>0)
# write BS output file
flags.addFlag('Trigger.writeBS', False)
# Write transient BS before executing HLT algorithms (for running on MC RDO with clients which require BS inputs)
flags.addFlag('Trigger.doTransientByteStream', False)
# list of EDM objects to be written to AOD
flags.addFlag('Trigger.AODEDMSet', 'AODSLIM')
# list of objects to be written to ESD
flags.addFlag('Trigger.ESDEDMSet', 'ESD')
# to allow stroing extra EDM items via preExec
flags.addFlag('Trigger.ExtraEDMList', [])
# tag to be used for condutions used by HLT code
flags.addFlag('Trigger.OnlineCondTag', 'CONDBR2-HLTP-2018-01')
# geometry version used by HLT online
flags.addFlag('Trigger.OnlineGeoTag', 'ATLAS-R2-2016-01-00-01')
# configuration tune for various years of Run2 (why string?)
flags.addFlag('Trigger.run2Config', '2018')
# comissionig options
# one of: 'HltOnly',
# 'Lvl1Only',
# 'FullTrigger',
# 'NoTrigger'
flags.addFlag('Trigger.dataTakingConditions', 'FullTrigger')
# use or not frontier proxies
flags.addFlag('Trigger.triggerUseFrontier', False)
# the configuration source
# see https://twiki.cern.ch/twiki/bin/view/Atlas/TriggerConfigFlag
flags.addFlag('Trigger.triggerConfig', 'FILE')
# name of the trigger menu
flags.addFlag('Trigger.triggerMenuSetup', 'LS2_v1')
# name of the trigger menu
flags.addFlag('Trigger.generateMenuDiagnostics', False)
# version of the menu
from AthenaCommon.AppMgr import release_metadata
flags.addFlag('Trigger.menuVersion',
lambda prevFlags: release_metadata()['release'] )
# generate or not the HLT configuration
flags.addFlag('Trigger.generateHLTMenu', False)
# HLT XML file name
flags.addFlag('Trigger.HLTMenuFile',
lambda prevFlags: 'HLTMenu_'+prevFlags.Trigger.triggerMenuSetup+'_' + prevFlags.Trigger.menuVersion + '.xml')
# generate or not the L1 configuration
flags.addFlag('Trigger.generateL1Menu', False)
def _deriveL1ConfigName(prevFlags):
import re
log = logging.getLogger('TrigConfigSvcCfg')
pattern = re.compile(r'_v\d+|DC14')
menuName=prevFlags.Trigger.triggerMenuSetup
patternPos = pattern.search(menuName)
if patternPos:
menuName=menuName[:patternPos.end()]
else:
log.info('Can\'t find pattern to shorten menu name, either non-existent in name or not implemented.')
return "LVL1config_"+menuName+"_" + prevFlags.Trigger.menuVersion + ".xml"
# L1 XML file name
flags.addFlag('Trigger.LVL1ConfigFile', _deriveL1ConfigName)
# L1 Json file name
flags.addFlag('Trigger.L1MenuFile',
lambda prevFlags: 'L1Menu_'+prevFlags.Trigger.triggerMenuSetup+'_' + prevFlags.Trigger.menuVersion + '.json')
# L1 topo XML file name
def _deriveTopoConfigName(prevFlags):
import re
menuSetup = prevFlags.Trigger.triggerMenuSetup
m = re.match(r'(.*v\d).*', menuSetup)
if m:
menuSetup = m.groups()[0]
return "L1Topoconfig_" + menuSetup + "_" + prevFlags.Trigger.menuVersion + ".xml"
flags.addFlag('Trigger.LVL1TopoConfigFile', _deriveTopoConfigName)
# trigger reconstruction
# enables the correction for pileup in cell energy calibration (should it be moved to some place where other calo flags are defined?)
flags.addFlag('Trigger.calo.doOffsetCorrection', True )
# helper to switch between versions
def __tunes(default, ver2016, ver2017):
from collections import defaultdict
return lambda year: defaultdict( lambda: default,
(('2016', ver2016),
('2017', ver2017)))[year]
# Particle ID tune
flags.addFlag('Trigger.egamma.pidVersion',
lambda prevFlags:
__tunes(default='ElectronPhotonSelectorTools/trigger/rel21_mc16a/',
ver2016='ElectronPhotonSelectorTools/trigger/rel21_mc16a/',
ver2017='ElectronPhotonSelectorTools/trigger/rel21_20170214/')( prevFlags.Trigger.run2Config )
)
# cluster correction version, allowed value is: None or v12phiflip_noecorrnogap
flags.addFlag('Trigger.egamma.clusterCorrectionVersion',
lambda prevFlags:
__tunes(default=None,
ver2016=None,
ver2017='v12phiflip_noecorrnogap')( prevFlags.Trigger.run2Config )
)
# tune of MVA
flags.addFlag('Trigger.egamma.calibMVAVersion',
lambda prevFlags:
__tunes(default='egammaMVACalib/online/v3',
ver2016='egammaMVACalib/online/v3',
ver2017='egammaMVACalib/online/v6')( prevFlags.Trigger.run2Config )
)
# muons
flags.addFlag('Trigger.muon.doEFRoIDrivenAccess', False)
# muon offline reco flags varaint for trigger
def __muon():
from MuonConfig.MuonConfigFlags import createMuonConfigFlags
return createMuonConfigFlags()
flags.addFlagsCategory('Trigger.Offline', __muon, prefix=True)
from TrigInDetConfig.TrigTrackingCutFlags import createTrigTrackingFlags
flags.addFlagsCategory( 'Trigger.InDetTracking', createTrigTrackingFlags )
return flags
# for reference, this flags are skipped as never used or never set in fact, or set identical to de default or used in a very old JO:
# readLVL1Calo, readLVL1Muon, fakeLVL1, useCaloTTL
# doCosmicSim - old JO
# disableRandomPrescale - never used
# doMergedHLTResult - not needed now
# doAlwaysUnpackDSResult - never set
# doTrt - fast TRT or trigger - never used
# doLucid - not sure if ever set to False, identical rec flag exists
# doZdc - never set to anything different than default
# doHLTpersistency - never red
# useOfflineSpacePoints - used only to set
# doNtuple - never red
# configForStartup
# the flags related to trigger DB are redundant of triggerConfig - need to decide if they are needed in this form
# also not defined the Prescale sets yet
# abortOnConfigurationError - never red
# in signatures
# egamma: ringerVersion - not used
# muon: doMuonCalibrationStream - not used
# tau: doTrackingApproach - not used
import unittest
class __YearDependentFlagTest(unittest.TestCase):
def runTest(self):
"""... Check if year dependent flags propagate the info correctly"""
from AthenaConfiguration.AllConfigFlags import ConfigFlags as flags
flags.Trigger.run2Config='2017'
self.assertEqual(flags.Trigger.egamma.clusterCorrectionVersion, "v12phiflip_noecorrnogap", " dependent flag setting does not work")
flags.dump()
class __UseOfOfflineRecoFlagsTest(unittest.TestCase):
def runTest(self):
"""... Check if offline reco flags can be added to trigger"""
from AthenaConfiguration.AllConfigFlags import ConfigFlags as flags
flags.Trigger.Offline.Muon.doMDTs=False
flags.Muon.doMDTs=True
self.assertEqual(flags.Trigger.Offline.Muon.doMDTs, False, " dependent flag setting does not work")
self.assertEqual(flags.Muon.doMDTs, True, " dependent flag setting does not work")
newflags = flags.cloneAndReplace('Muon', 'Trigger.Offline.Muon')
self.assertEqual(flags.Muon.doMDTs, True, " dependent flag setting does not work")
self.assertEqual(newflags.Muon.doMDTs, False, " dependent flag setting does not work")
newflags.dump()
if __name__ == "__main__":
suite = unittest.TestSuite()
suite.addTest(__YearDependentFlagTest())
suite.addTest(__UseOfOfflineRecoFlagsTest())
runner = unittest.TextTestRunner(failfast=False)
runner.run(suite)