def findSample(args):
# signal
if 'HTo2XTo4Mu' in args.NAME or 'HTo2XTo2Mu2J' in args.NAME:
samples = getattr(DH, 'get' + args.NAME + 'Samples')()
SIGNALPOINT = tuple(args.SIGNALPOINT)
DKEY = 'PAT'
if args.GENONLY: DKEY = 'GEN'
if args.AODONLY: DKEY = 'AOD'
DATASET = DEFAULT_DATASETS[args.NAME][DKEY]
for data in samples:
if data.signalPoint() == SIGNALPOINT and DATASET in data.datasets:
return data
else:
print '[DATAHANDLER ERROR]: No {} dataset found with signal point {} and process {}'.format(
args.NAME, SIGNALPOINT, DATASET)
exit()
# DoubleMuon data
elif args.NAME.startswith('DoubleMuon'):
samples = DH.getDoubleMuonDataSamples()
for data in samples:
if data.name == args.NAME:
return data
# Cosmics data
elif args.NAME.startswith('Cosmics'):
samples = DH.getCosmicsDataSamples()
for data in samples:
if data.name == args.NAME:
return data
# NoBPTX data
elif args.NAME.startswith('NoBPTX'):
samples = DH.getNoBPTXDataSamples()
for data in samples:
if data.name == args.NAME:
return data
# background
else:
samples = DH.getBackgroundSamples()
for data in samples:
if data.name == args.NAME:
return data
# if nothing found yet, raise error
print '[DATAHANDLER ERROR]: "' + args.NAME + '" is not a known dataset name; see .dat files in dat/'
exit()
import re, itertools
import DisplacedDimuons.Common.DataHandler as DH
##### Place the output of this script, "SignalMCSamples.dat", in ../dat
##### It is required by the DataHandler library for signal samples
##### and therefore by Tupler and by Analysis
# gets HTo2LongLivedTo4mu and HTo2LongLivedTo2mu2jets dataset strings
AODDatasetStrings = DH.DASQueryList(
'dataset=/HTo2LongLivedTo*_MH-*_MFF-*_CTau-*/escalant-April2019reHLT-AOD-v1*/* instance=prod/phys03'
)
# gets PAT Tuple datasets -- there may be many of these, and the regexes may have to be tweaked to be unique
PATDatasetStrings = DH.DASQueryList(
'dataset=/HTo2LongLivedTo*_MH-*_MFF-*_CTau-*/stempl-MC2016_*reHLT_Apr2019-v4*/* instance=prod/phys03'
)
# matches dataset string with six groups: final state (4mu or 2mu2jets), mH, mX, cTau (digits 1-4), process string (\S*), weird hash ID (\w*)
# apply these to AODDatasetStrings
rxAOD = re.compile(
r'/HTo2LongLivedTo(4mu|2mu2jets)_MH-(\d{1,4})_MFF-(\d{1,4})_CTau-(\d{1,4})mm\S*pythia8/escalant-(\S*)-(\w*)/USER'
)
# matches PAT Tupler dataset string with five groups: final state (4mu or 2mu2jets), mH, mX, cTau (digits 1-4), weird hash ID (\w*)
# apply these to PATDatasetStrings
# rxPAT = re.compile(r'/HTo2LongLivedTo(4mu|2mu2jets)_MH-(\d{1,4})_MFF-(\d{1,4})_CTau-(\d{1,4})mm\S*pythia8/stempl-MC2016_\S*_reHLT\S*_Jun2019-v4-(\w*)/USER')
rxPAT = re.compile(
r'/HTo2LongLivedTo(4mu|2mu2jets)_MH-(\d{1,4})_MFF-(\d{1,4})_CTau-(\d{1,4})mm\S*pythia8/stempl-MC2016_\S*_reHLT_Apr2019-v4-(\w*)/USER'
)
# fill a list with split up metadata that will be sorted
import DisplacedDimuons.Common.DataHandler as DH
FINALSTATE = '4Mu'
SIGNALPOINT = (125, 20, 13)
PROCESS = 'AODSIM-ReHLT_V37-v1'
datasets = getattr(DH, 'getHTo2XTo' + FINALSTATE + 'Samples')()
for data in datasets:
if data.signalPoint() == SIGNALPOINT:
if PROCESS in data.datasets:
fleList = DH.DASQueryList(
'file,lumi,events dataset={} instance={}'.format(
data.datasets[PROCESS], data.instances[PROCESS]))
totalE = 0
for rstring in fleList:
cols = rstring.split()
f = cols[0]
l = map(int, cols[1].strip('[').strip(']').split(','))
e = map(int, cols[2].strip('[').strip(']').split(','))
totalE += sum(e)
print 'HTo2XTo{} : {} : {} :: {} events'.format(
FINALSTATE, SIGNALPOINT, PROCESS, totalE)
else:
raise Exception(
'No HTo2XTo{} dataset found with signal point {} and process {}'
.format(FINALSTATE, SIGNALPOINT, PROCESS))
dest='TRIGGER',
action='store_true',
help='apply trigger to signal')
PARSER.add_argument('--cuts', dest='CUTS', default='', help='cut string')
PARSER.add_argument('--skim',
dest='SKIM',
action='store_true',
help='whether to use the skim file')
# important setSample function
# this function takes the inputs from ARGS and selects the unique matching Dataset from DataHandler
# this Dataset object contains many important parameters: nEvents, negFrac, systFrac, etc
# and they are available to the Analyzer object through the member variable SAMPLE
# so if e.g. a weight in the analyze() function is desired, it is as simple as
# using self.SAMPLE.nEvents or self.SAMPLE.negFrac
SAMPLES = DH.getAllSamples()
def setSample(ARGS):
if ARGS.NAME.startswith('HTo2X') or ARGS.NAME.startswith('reHLT_HTo2X'):
if ARGS.SIGNALPOINT is None:
raise Exception('Need a signal point for HTo2X MC signal.')
ARGS.SAMPLE = SAMPLES[ARGS.NAME + '_' +
Utilities.SPStr(ARGS.SIGNALPOINT)]
else:
ARGS.SAMPLE = SAMPLES[ARGS.NAME]
# Analyzer class, one instance per sample, runs over a tree, calls analysis functions
class Analyzer(object):
# constructor:
for mH, mX, cTau in SIGNALPOINTS:
verbose('GEN ONLY SIGNAL {} : {} {} {}'.format('HTo2XTo2Mu2J', mH, mX, cTau))
bash.call('python runNTupler.py HTo2XTo2Mu2J --signalpoint {mH} {mX} {cTau} --genonly {MODE}'.format(**locals()), shell=True)
# submit all HTo2XTo2Mu2J signal aod only jobs
if Do_2Mu2J_AODOnly:
for mH, mX, cTau in SIGNALPOINTS:
verbose('AOD ONLY SIGNAL {} : {} {} {}'.format('HTo2XTo2Mu2J', mH, mX, cTau))
bash.call('python runNTupler.py HTo2XTo2Mu2J --signalpoint {mH} {mX} {cTau} --aodonly {MODE}'.format(**locals()), shell=True)
# submit all HTo2XTo2Mu2J signal PAT Tuple jobs
if Do_2Mu2J:
for mH, mX, cTau in SIGNALPOINTS:
verbose('SIGNAL {} : {} {} {}'.format('HTo2XTo2Mu2J', mH, mX, cTau))
bash.call('python runNTupler.py HTo2XTo2Mu2J --signalpoint {mH} {mX} {cTau} {MODE}'.format(**locals()), shell=True)
# submit all background MC jobs
if Do_Background:
mcSamples = DH.getBackgroundSamples()
for data in mcSamples:
verbose('BACKGROUND : {}'.format(data.name))
if data.datasets['PAT'] == '_': continue
bash.call('python runNTupler.py {NAME} {MODE}'.format(NAME=data.name, MODE=MODE), shell=True)
# submit all data jobs
if Do_Data:
dataSamples = DH.getDoubleMuonDataSamples()
for data in dataSamples:
verbose('DATA : {}'.format(data.name))
bash.call('python runNTupler.py {NAME} {MODE}'.format(NAME=data.name, MODE=MODE), shell=True)