def addDatasetsFromMulticrab(self, directory, *args, **kwargs):
'''
kwargs for 'includeOnlyTasks' or 'excludeTasks' to set the datasets over which this analyzer is processed, default is all datasets
'''
Verbose("addDatasetsFromMulticrab()", True)
blacklist = []
if "blacklist" in kwargs.keys():
if isinstance(kwargs["blacklist"], str):
blacklist.append(kwargs["blacklist"])
elif isinstance(kwargs["blacklist"], list):
blacklist.extend(kwargs["blacklist"])
else:
raise Exception("Unsupported input format!")
del kwargs["blacklist"]
# dataset._optionDefaults["input"] = "miniaod2tree*.root"
dataset._optionDefaults["input"] = "histograms-*.root"
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=directory, *args, **kwargs)
dsets = dsetMgrCreator.getDatasetPrecursors()
dsetMgrCreator.close()
for dset in dsets:
isOnBlackList = False
for item in blacklist:
if dset.getName().startswith(item):
isOnBlackList = True
if isOnBlackList:
print "Ignoring dataset because of blacklist options: '%s' ..."%dset.getName()
else:
self.addDataset(dset.getName(), dset.getFileNames(), dataVersion=dset.getDataVersion(), lumiFile=dsetMgrCreator.getLumiFile())
return
def createDsetMgr(self,
multicrabDir,
era,
searchMode,
optimizationMode=None,
systematicVariation=None):
self._era = era
self._searchMode = searchMode
self._optimizationMode = optimizationMode
self._systematicVariation = systematicVariation
# Construct info string of module
self._moduleInfoString = "%s_%s_%s" % (era, searchMode,
optimizationMode)
# Obtain dataset manager
self._dsetMgrCreator = dataset.readFromMulticrabCfg(
directory=multicrabDir)
self._dsetMgr = self._dsetMgrCreator.createDatasetManager(
dataEra=era,
searchMode=searchMode,
optimizationMode=optimizationMode,
systematicVariation=systematicVariation)
# Do the usual normalisation
self._dsetMgr.updateNAllEventsToPUWeighted()
self._dsetMgr.loadLuminosities()
plots.mergeRenameReorderForDataMC(self._dsetMgr)
self._dsetMgr.merge("EWK", _generalOptions["ewkDatasetsForMerging"])
# Obtain luminosity
self._luminosity = self._dsetMgr.getDataset("Data").getLuminosity()
def addDatasetsFromMulticrab(self, directory, *args, **kwargs):
blacklist = []
if "blacklist" in kwargs.keys():
if isinstance(kwargs["blacklist"], str):
blacklist.append(kwargs["blacklist"])
elif isinstance(kwargs["blacklist"], list):
blacklist.extend(kwargs["blacklist"])
else:
raise Exception("Unsupported input format!")
del kwargs["blacklist"]
# dataset._optionDefaults["input"] = "miniaod2tree*.root"
dataset._optionDefaults["input"] = "histograms-*.root"
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=directory,
*args,
**kwargs)
dsets = dsetMgrCreator.getDatasetPrecursors()
dsetMgrCreator.close()
for dset in dsets:
isOnBlackList = False
for item in blacklist:
if dset.getName().startswith(item):
isOnBlackList = True
if isOnBlackList:
print "Ignoring dataset because of blacklist options: '%s' ..." % dset.getName(
)
else:
self.addDataset(dset.getName(),
dset.getFileNames(),
dataVersion=dset.getDataVersion(),
lumiFile=dsetMgrCreator.getLumiFile())
def main():
style = tdrstyle.TDRStyle()
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
parser = OptionParser(usage="Usage: %prog [options]",add_help_option=True,conflict_handler="resolve")
myModuleSelector.addParserOptions(parser)
parser.add_option("-m", "--mcrab", dest="mcrab", action="store", help="Path to the multicrab directory for input")
parser.add_option("-d", "--dataset", dest="dataset", action="store", help="Name of the dataset to be plotted")
parser.add_option("-e", "--error", dest="errorlevel", default=0.10, action="store", help="Maximum relative uncertainty per bin (default=10%%)")
(opts, args) = parser.parse_args()
if opts.mcrab == None:
raise Exception("Please provide input multicrab directory with -m")
if not os.path.exists(opts.mcrab):
raise Exception("The input root file '%s' does not exist!"%opts.mcrab)
if opts.dataset == None:
raise Exception("Please provide dataset name with -d")
# Find module names
#myNames = findModuleNames(opts.mcrab, "BTagEfficiency")
# Get dataset managers
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts)
myModuleSelector.printSelectedCombinationCount()
#for n in myNames:
results = []
#for era in myModuleSelector.getSelectedEras(): #degug
for era in ["Run2016"]:
msg = "=== FIXME:\n\tRunning only over era \"%s\". The function myModuleSelector.getSelectedEras does not seem to work. Press any key to proceed" % (era)
raw_input(msg)
for searchMode in myModuleSelector.getSelectedSearchModes():
for optimizationMode in myModuleSelector.getSelectedOptimizationModes():
dsetMgr = dsetMgrCreator.createDatasetManager(dataEra=era,searchMode=searchMode,optimizationMode=optimizationMode)
#datasets = dataset.getDatasetsFromMulticrabDirs(dirs=[opts.mcrab], analysisName=n)
dsetMgr.loadLuminosities()
dsetMgr.updateNAllEventsToPUWeighted()
plots.mergeRenameReorderForDataMC(dsetMgr)
lumi = dsetMgr.getDataset("Data").getLuminosity()
for dset in dsetMgr.getMCDatasets():
print dset.name
if dset.name.startswith(opts.dataset):
results.extend(doPlot("btageff_%s_%s_%s"%(era, searchMode, optimizationMode), dset, opts.errorlevel, optimizationMode, lumi))
#dsetMgr.close()
print "\nFigures generated"
for item in results:
print item
# Write results to a json file
filename = "btageff_%s.json"%opts.dataset
with open(filename, 'w') as outfile:
json.dump(results, outfile)
print "Written results to %s"%filename
def addDatasetsFromMulticrab(self, directory, *args, **kwargs):
'''
kwargs for 'includeOnlyTasks' or 'excludeTasks' to set the datasets over which this analyzer is processed, default is all datasets
'''
Verbose("addDatasetsFromMulticrab()", True)
blacklist = []
if "blacklist" in kwargs.keys():
if isinstance(kwargs["blacklist"], str):
blacklist.append(kwargs["blacklist"])
elif isinstance(kwargs["blacklist"], list):
blacklist.extend(kwargs["blacklist"])
else:
raise Exception("Unsupported input format!")
del kwargs["blacklist"]
whitelist = []
if "whitelist" in kwargs.keys():
if isinstance(kwargs["whitelist"], str):
whitelist.append(kwargs["whitelist"])
elif isinstance(kwargs["whitelist"], list):
whitelist.extend(kwargs["whitelist"])
else:
raise Exception("Unsupported input format!")
del kwargs["whitelist"]
# dataset._optionDefaults["input"] = "miniaod2tree*.root"
dataset._optionDefaults["input"] = "histograms-*.root"
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=directory,
*args,
**kwargs)
dsets = dsetMgrCreator.getDatasetPrecursors()
dsetMgrCreator.close()
if len(whitelist) > 0:
for dset in dsets:
isOnWhiteList = False
for item in whitelist:
if dset.getName().startswith(item):
isOnWhiteList = True
if not isOnWhiteList:
blacklist.append(dset.getName())
for dset in dsets:
isOnBlackList = False
for item in blacklist:
if dset.getName().startswith(item):
isOnBlackList = True
if isOnBlackList:
print "Ignoring dataset because of black/whitelist options: '%s' ..." % dset.getName(
)
else:
self.addDataset(dset.getName(),
dset.getFileNames(),
dataVersion=dset.getDataVersion(),
lumiFile=dsetMgrCreator.getLumiFile())
return
def main():
style = tdrstyle.TDRStyle()
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
parser = OptionParser(usage="Usage: %prog [options]",add_help_option=True,conflict_handler="resolve")
myModuleSelector.addParserOptions(parser)
parser.add_option("-m", "--mcrab", dest="mcrab", action="store", help="Path to the multicrab directory for input")
parser.add_option("-d", "--dataset", dest="dataset", action="store", help="Name of the dataset to be plotted")
parser.add_option("-e", "--error", dest="errorlevel", default=0.10, action="store", help="Maximum relative uncertainty per bin (default=10%%)")
(opts, args) = parser.parse_args()
if opts.mcrab == None:
raise Exception("Please provide input multicrab directory with -m")
if not os.path.exists(opts.mcrab):
raise Exception("The input root file '%s' does not exist!"%opts.mcrab)
if opts.dataset == None:
raise Exception("Please provide dataset name with -d")
# Find module names
#myNames = findModuleNames(opts.mcrab, "BTagEfficiency")
# Get dataset managers
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts)
myModuleSelector.printSelectedCombinationCount()
#for n in myNames:
results = []
for era in myModuleSelector.getSelectedEras():
for searchMode in myModuleSelector.getSelectedSearchModes():
for optimizationMode in myModuleSelector.getSelectedOptimizationModes():
dsetMgr = dsetMgrCreator.createDatasetManager(dataEra=era,searchMode=searchMode,optimizationMode=optimizationMode)
#datasets = dataset.getDatasetsFromMulticrabDirs(dirs=[opts.mcrab], analysisName=n)
dsetMgr.loadLuminosities()
dsetMgr.updateNAllEventsToPUWeighted()
plots.mergeRenameReorderForDataMC(dsetMgr)
lumi = dsetMgr.getDataset("Data").getLuminosity()
for dset in dsetMgr.getMCDatasets():
print dset.name
if dset.name.startswith(opts.dataset):
results.extend(doPlot("btageff_%s_%s_%s"%(era, searchMode, optimizationMode), dset, opts.errorlevel, optimizationMode, lumi))
#dsetMgr.close()
print "\nFigures generated"
for item in results:
print item
# Write results to a json file
filename = "btageff_%s.json"%opts.dataset
with open(filename, 'w') as outfile:
json.dump(results, outfile)
print "Written results to %s"%filename
def getDsetCreator(label, mcrabPath, mcrabInfoOutput, enabledStatus=True):
if enabledStatus:
if mcrabPath == "":
mcrabInfoOutput.append("- %s: not present"%(label))
print "- %s: not present"%(label)
else:
mcrabInfoOutput.append("- %s: multicrab dir found (%s)"%(label,mcrabPath))
print "- %s%s: multicrab dir found%s (%s)"%(HighlightStyle(),label,NormalStyle(),mcrabPath)
return dataset.readFromMulticrabCfg(directory=mcrabPath)
else:
mcrabInfoOutput.append("- %s: not considered"%(label))
print "- %s: not considered"%(label)
return None
def getDsetCreator(label, mcrabPath, mcrabInfoOutput, enabledStatus=True):
if enabledStatus:
if mcrabPath == "":
mcrabInfoOutput.append("- %s: not present"%(label))
print "- %s: not present"%(label)
else:
mcrabInfoOutput.append("- %s: multicrab dir found (%s)"%(label,mcrabPath))
print "- %s%s: multicrab dir found%s (%s)"%(HighlightStyle(),label,NormalStyle(),mcrabPath)
return dataset.readFromMulticrabCfg(directory=mcrabPath)
else:
mcrabInfoOutput.append("- %s: not considered"%(label))
print "- %s: not considered"%(label)
return None
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 1:
datasetsMgr.printSelections()
# Print PSets used for FakeBMeasurement
if 0:
datasetsMgr.printSelections()
PrintPSet("BJetSelection", datasetsMgr)
PrintPSet("TopSelectionBDT", datasetsMgr)
PrintPSet("FakeBMeasurement", datasetsMgr)
sys.exit()
sys.exit()
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 1:
datasetsMgr.printSelections()
# Print PSets used for FakeBMeasurement
if 0:
datasetsMgr.printSelections()
PrintPSet("BJetSelection", datasetsMgr)
PrintPSet("TopSelectionBDT", datasetsMgr)
PrintPSet("FakeBMeasurement", datasetsMgr)
sys.exit()
sys.exit()
return
def getDsetCreator(label, mcrabPath, mcrabInfoOutput, enabledStatus=True):
if enabledStatus:
if mcrabPath == "":
msg = "%s not present!" % (label) #fixme: application to H2tb?
mcrabInfoOutput.append(msg)
Print(WarningLabel() + msg)
else:
mcrabInfoOutput.append("- %s: multicrab dir found (%s)" % (label,mcrabPath) )
msg = "%s multicrab dir found in %s" % (label, mcrabPath)
Verbose(NoteStyle() + msg + NormalStyle()) # HighlightStyle()
return dataset.readFromMulticrabCfg(directory=mcrabPath)
else:
msg = "%s not considered!" % (label)
mcrabInfoOutput.append(msg)
Print(NoteStyle() + msg)
return None
def createDsetMgr(self,
multicrabDir,
era,
searchMode,
optimizationMode=None,
systematicVariation=None):
self._era = era
self._searchMode = searchMode
self._optimizationMode = optimizationMode
self._systematicVariation = systematicVariation
# Construct info string of module
self._moduleInfoString = self.getModuleInfoString()
# Obtain dataset manager
self._dsetMgrCreator = dataset.readFromMulticrabCfg(
directory=multicrabDir)
self._dsetMgr = self._dsetMgrCreator.createDatasetManager(
dataEra=era,
searchMode=searchMode,
optimizationMode=optimizationMode,
systematicVariation=systematicVariation)
# Do the usual normalisation
self._dsetMgr.updateNAllEventsToPUWeighted()
self._dsetMgr.loadLuminosities()
plots.mergeRenameReorderForDataMC(self._dsetMgr)
# Print initial dataset info
if opts.verbose:
self._dsetMgr.PrintInfo()
self._dsetMgr.merge("EWK", opts.ewkDatasets)
# Remove all other datasets
datasetList = ["ZJetsToQQ_HT600toInf", "Charged", "QCD"]
for d in datasetList:
self._dsetMgr.remove(
filter(lambda name: d in name,
self._dsetMgr.getAllDatasetNames()))
# Print final dataset info
if opts.verbose:
self._dsetMgr.PrintInfo()
# Obtain luminosity
self._luminosity = self._dsetMgr.getDataset("Data").getLuminosity()
return
def getDsetCreator(label, mcrabPath, mcrabInfoOutput, enabledStatus=True):
if enabledStatus:
if mcrabPath == "":
msg = "%s not present!" % (label) #fixme: application to H2tb?
mcrabInfoOutput.append(msg)
Print(WarningLabel() + msg)
else:
mcrabInfoOutput.append("- %s: multicrab dir found (%s)" %
(label, mcrabPath))
msg = "%s multicrab dir found in %s" % (label, mcrabPath)
Verbose(NoteStyle() + msg + NormalStyle()) # HighlightStyle()
return dataset.readFromMulticrabCfg(directory=mcrabPath)
else:
msg = "%s not considered!" % (label)
mcrabInfoOutput.append(msg)
Print(NoteStyle() + msg)
return None
def createDsetMgr(self, multicrabDir, era, searchMode, optimizationMode=None, systematicVariation=None):
self._era = era
self._searchMode = searchMode
self._optimizationMode = optimizationMode
self._systematicVariation = systematicVariation
# Construct info string of module
self._moduleInfoString = "%s_%s_%s"%(era, searchMode, optimizationMode)
# Obtain dataset manager
self._dsetMgrCreator = dataset.readFromMulticrabCfg(directory=multicrabDir)
self._dsetMgr = self._dsetMgrCreator.createDatasetManager(dataEra=era,searchMode=searchMode,optimizationMode=optimizationMode,systematicVariation=systematicVariation)
# Do the usual normalisation
self._dsetMgr.updateNAllEventsToPUWeighted()
self._dsetMgr.loadLuminosities()
plots.mergeRenameReorderForDataMC(self._dsetMgr)
self._dsetMgr.merge("EWK", _generalOptions["ewkDatasetsForMerging"])
# Obtain luminosity
self._luminosity = self._dsetMgr.getDataset("Data").getLuminosity()
def main(opts):
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Obtain systematics names
mySystematicsNamesRaw = dsetMgrCreator.getSystematicVariationSources()
mySystematicsNames = getSystematicsNames(mySystematicsNamesRaw, opts)
# Set the primary source
Verbose(
"Setting the primary source (label=%s)" %
(ShellStyles.NoteStyle() + opts.analysisName +
ShellStyles.NormalStyle()), True)
myModuleSelector.setPrimarySource(label=opts.analysisName,
dsetMgrCreator=dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts=None)
# Loop over era/searchMode/optimizationMode combos
myTotalModules = myModuleSelector.getSelectedCombinationCount() * (
len(mySystematicsNames) + 1)
count, nEras, nSearchModes, nOptModes, nSysVars = myModuleSelector.getSelectedCombinationCountIndividually(
)
if nSysVars > 0:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes x %d systematic variations)" % (
count, nEras, nSearchModes, nOptModes, nSysVars)
else:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes)" % (
count, nEras, nSearchModes, nOptModes)
Verbose(msg, True)
# Create pseudo-multicrab creator
msg = "Creating pseudo-dataset \"%s\" inside the pseudo-multicrab directory \"%s\"" % (
opts.newDsetName, opts.mcrab)
Verbose(msg, True)
myOutputCreator = pseudoMultiCrabCreator.PseudoMultiCrabCreator(
opts.newDsetName, opts.mcrab, verbose=opts.verbose)
# Make time stamp for start time
myGlobalStartTime = time.time()
# Initialize
myOutputCreator.initialize(subTitle="", prefix="")
# Get lists of settings
erasList = myModuleSelector.getSelectedEras()
modesList = myModuleSelector.getSelectedSearchModes()
optList = myModuleSelector.getSelectedOptimizationModes()
if 0:
optList.append(
"") #append the default opt mode iff more optimization modes exist
Verbose(
"Found %d eras, %d modes, %d optimisations" %
(len(erasList), len(modesList), len(optList)), True)
iModule = 0
# For-loop: All eras
for era in erasList:
# For-loop: All searchModes
for searchMode in modesList:
# For-loop: AlloptimizationModes
for optimizationMode in optList:
Verbose(
"era = %s, searchMode = %s, optMode = %s" %
(era, searchMode, optimizationMode), True)
# If an optimization mode is defined in options skip the rest
if opts.optMode != None:
if optimizationMode != opts.optMode:
continue
# Nominal module
myModuleInfoString = getModuleInfoString(
opts.analysisName, era, searchMode, optimizationMode)
iModule += 1
# Inform user of what is being processes
msg = "{:<10} {:>3} {:^1} {:<3}: {:<20}".format(
"Module", iModule, "/", myTotalModules, myModuleInfoString)
Print(hs + msg + ns, iModule == 1)
Verbose("Creating dataset manager for nominal module", True)
myStartTime = time.time()
nominalModule = ModuleBuilder(opts, myOutputCreator,
opts.verbose)
nominalModule.createDsetMgr(opts.mcrab, opts.analysisName, era,
searchMode, optimizationMode)
# Build the module
nominalModule.buildModule(opts.dsetSrc)
Verbose("Deleting nominal module", True)
nominalModule.delete()
Verbose("Printing time estimate", True)
printTimeEstimate(myGlobalStartTime, myStartTime, iModule,
myTotalModules)
Verbose("Now do the rest of systematics variations", True)
for syst in mySystematicsNames:
myModuleInfoString = getModuleInfoString(
opts.analysisName, era, searchMode, optimizationMode,
syst)
iModule += 1
msg = "{:<10} {:>3} {:^1} {:<3}: {:<20}".format(
"Module", iModule, "/", myTotalModules,
myModuleInfoString)
Print(hs + msg + ns, False)
Verbose(
"Creating dataset manager for systematics module %s" %
(syst), True)
myStartTime = time.time()
systModule = ModuleBuilder(opts, myOutputCreator,
opts.verbose)
systModule.createDsetMgr(opts.mcrab,
opts.analysisName,
era,
searchMode,
optimizationMode,
systematicVariation=syst)
# Build the module
systModule.buildModule(opts.dsetSrc)
Verbose("Deleting nominal module", True)
systModule.delete()
Verbose("Printing time estimate", True)
printTimeEstimate(myGlobalStartTime, myStartTime, iModule,
myTotalModules)
Verbose(
"New dataset %s added to pseud-multicrab %s" %
(hs + opts.newDsetName + ns, hs + opts.mcrab + ns), True)
# Print some timing statistics
Verbose(
"Average processing time per module was %.1f seconds" %
getAvgProcessTimeForOneModule(myGlobalStartTime, myTotalModules), True)
Verbose(
"Total elapsed time was %.1f seconds" %
getTotalElapsedTime(myGlobalStartTime), True)
# Create rest of pseudo multicrab directory
myOutputCreator.finalize(silent=False)
# Update the multicrab.cfg file
updateMulticrabCfg()
return
def createDsetMgr(self,
multicrabDir,
analysis,
era,
searchMode,
optimizationMode=None,
systematicVariation=None):
# Store variables
self._analysis = analysis
self._era = era
self._searchMode = searchMode
self._optimizationMode = optimizationMode
self._systematicVariation = systematicVariation
# Construct info string of module
self._moduleInfoString = self.getModuleInfoString()
# Obtain dataset manager
self._dsetMgrCreator = dataset.readFromMulticrabCfg(
directory=multicrabDir)
self._dsetMgr = self._dsetMgrCreator.createDatasetManager(
dataEra=era,
searchMode=searchMode,
optimizationMode=optimizationMode,
systematicVariation=systematicVariation)
# Do the usual normalisation
self._dsetMgr.updateNAllEventsToPUWeighted()
self._dsetMgr.loadLuminosities()
# Merge, Rename, Reorder datasets
plots.mergeRenameReorderForDataMC(self._dsetMgr)
# Print initial dataset info
if opts.verbose:
self._dsetMgr.PrintInfo()
self.Verbose(
"Merging datasets to get new dataset: [%s] -> %s" %
(ts + ", ".join(opts.dsetsToMerge) + ns,
ss + opts.newDsetName + ns), True)
self._dsetMgr.merge(opts.newDsetName,
opts.dsetsToMerge,
keepSources=True)
# Remove all other datasets
datasetList = opts.dsetsToRemove
self.Verbose(
"Removing datasets: [%s]" %
(es + ", ".join(opts.dsetsToRemove) + ns), True)
for i, d in enumerate(datasetList, 1):
self._dsetMgr.remove(
filter(lambda name: d in name,
self._dsetMgr.getAllDatasetNames()))
# Print final dataset info
if opts.verbose:
self._dsetMgr.PrintInfo()
# Obtain luminosity
self._luminosity = self._dsetMgr.getDataset("Data").getLuminosity()
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(True) #opts.gridX)
style.setGridY(True) #opts.gridY)
style.setLogX(False) #opts.logX)
style.setLogY(False) #opts.logY)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 0:
datasetsMgr.printSelections()
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Custom Filtering of datasets
if 0:
datasetsMgr.remove(filter(lambda name: "Charged" in name and not "M_500" in name, datasetsMgr.getAllDatasetNames()))
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames() and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print("Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..", True)
datasetsMgr.remove(filter(lambda name: "ZJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
datasetsMgr.PrintInfo()
# Get Luminosity
if opts.intLumi < 0.0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Merge EWK samples
if opts.dataset == "EWK":
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Get all histogram names in the given ROOT folder
histoNames = datasetsMgr.getAllDatasets()[0].getDirectoryContent(opts.folder)
# histoList = [os.path.join(opts.folder, h) for h in histoNames if "_" + opts.region in h and opts.refBdisc in h]
histoList = [os.path.join(opts.folder, h) for h in histoNames if opts.refBdisc in h]
# For-loop: All histos in CR of interest
nHistos = len(histoList)
for i, h in enumerate(histoList, 1):
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i,"/", "%s:" % (nHistos), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistograms(datasetsMgr, h)
# Inform user where the plots where saved
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(True) #opts.gridX)
style.setGridY(True) #opts.gridY)
style.setLogX(False) #opts.logX)
style.setLogY(False) #opts.logY)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 0:
datasetsMgr.printSelections()
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Custom Filtering of datasets
if 0:
datasetsMgr.remove(filter(lambda name: "Charged" in name and not "M_500" in name, datasetsMgr.getAllDatasetNames()))
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames() and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print("Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..", True)
datasetsMgr.remove(filter(lambda name: "ZJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
datasetsMgr.PrintInfo()
# Get Luminosity
if opts.intLumi < 0.0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Merge EWK samples
if opts.dataset == "EWK":
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Get all histogram names in the given ROOT folder
histoNames = datasetsMgr.getAllDatasets()[0].getDirectoryContent(opts.folder)
histoList_ = [os.path.join(opts.folder, h) for h in histoNames if "_SR" in h]
histoList = [h for h in histoList_ if "Pt" in h] + [h for h in histoList_ if "Eta" in h]
# For-loop: All histos in SR
nHistos = len(histoList)
for i, h in enumerate(histoList, 1):
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i,"/", "%s:" % (nHistos), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
# print "*** h = ", h
PlotHistograms(datasetsMgr, h)
#break
# Inform user where the plots where saved
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
dest="multicrabDir",
action="store",
help="Multicrab directory")
# Add here parser options, if necessary, following line is an example
#parser.add_option("--showcard", dest="showDatacard", action="store_true", default=False, help="Print datacards also to screen")
# Parse options
(opts, args) = parser.parse_args()
# Obtain multicrab directory
myMulticrabDir = "../FakeTauComparison_140603_182139/"
if opts.multicrabDir != None:
myMulticrabDir = opts.multicrabDir
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=myMulticrabDir)
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts)
# Apply TDR style
style = tdrstyle.TDRStyle()
histograms.createLegend.moveDefaults(dx=-0.1, dh=-0.15)
myDisplayStatus = True
# Loop over era/searchMode/optimizationMode options
for era in myModuleSelector.getSelectedEras():
for searchMode in myModuleSelector.getSelectedSearchModes():
for optimizationMode in myModuleSelector.getSelectedOptimizationModes(
):
# Construct info string of module
def main(opts):
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Obtain systematics names
mySystematicsNamesRaw = dsetMgrCreator.getSystematicVariationSources()
mySystematicsNames = []
for i, item in enumerate(mySystematicsNamesRaw, 0):
Print(
"Using systematic %s" %
(ShellStyles.NoteStyle() + item + ShellStyles.NormalStyle()),
i == 0)
mySystematicsNames.append("%sPlus" % item)
mySystematicsNames.append("%sMinus" % item)
if opts.test:
mySystematicsNames = []
# Set the primary source
Verbose(
"Setting the primary source (label=%s)" %
(ShellStyles.NoteStyle() + opts.analysisName +
ShellStyles.NormalStyle()), True)
myModuleSelector.setPrimarySource(
label=opts.analysisName,
dsetMgrCreator=dsetMgrCreator) #fixme: what is label for?
# Select modules
myModuleSelector.doSelect(opts=None) #fixme: (opts=opts)
# Loop over era/searchMode/optimizationMode combos
myTotalModules = myModuleSelector.getSelectedCombinationCount() * (
len(mySystematicsNames) + 1) * len(opts.shape)
Verbose("Found %s modules in total" % (myTotalModules), True)
count, nEras, nSearchModes, nOptModes, nSysVars = myModuleSelector.getSelectedCombinationCountIndividually(
)
if nSysVars > 0:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes x %d systematic variations)" % (
count, nEras, nSearchModes, nOptModes, nSysVars)
else:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes)" % (
count, nEras, nSearchModes, nOptModes)
Verbose(msg, True)
# Create pseudo-multicrab creator
msg = "Will create pseudo-dataset %s inside the pseudo-multicrab directory" % (
ShellStyles.NoteStyle() + opts.analysisName +
ShellStyles.NormalStyle())
Verbose(msg, True)
myOutputCreator = pseudoMultiCrabCreator.PseudoMultiCrabCreator(
opts.analysisName, opts.mcrab, verbose=opts.verbose)
# Make time stamp for start time
myGlobalStartTime = time.time()
iModule = 0
# For-loop: All Shapes
for iShape, shapeType in enumerate(opts.shape, 1):
msg = "Shape %d/%d:%s %s" % (iShape, len(
opts.shape), ShellStyles.NormalStyle(), shapeType)
Print(ShellStyles.CaptionStyle() + msg, True)
# Initialize
myOutputCreator.initialize(
subTitle=shapeType,
prefix="") #fixeme: remove shapeType from sub-directory name?
# Get lists of settings
erasList = myModuleSelector.getSelectedEras()
modesList = myModuleSelector.getSelectedSearchModes()
optList = myModuleSelector.getSelectedOptimizationModes()
if 0:
optList.append(
""
) #append the default opt mode iff more optimization modes exist
# For-Loop over era, searchMode, and optimizationMode options
for era in erasList:
for searchMode in modesList:
for optimizationMode in optList:
Verbose(
"era = %s, searchMode = %s, optMode = %s" %
(era, searchMode, optimizationMode), True)
# If an optimization mode is defined in options skip the rest
if opts.optMode != None:
if optimizationMode != opts.optMode:
continue
# Obtain normalization factors
myNormFactors = importNormFactors(era, searchMode,
optimizationMode,
opts.mcrab)
# Nominal module
myModuleInfoString = getModuleInfoString(
era, searchMode, optimizationMode)
iModule += 1
# Inform user of what is being processes
msg = "Module %d/%d:%s %s/%s" % (
iModule, myTotalModules, ShellStyles.NormalStyle(),
myModuleInfoString, shapeType)
Print(ShellStyles.CaptionStyle() + msg, True)
# Keep time
myStartTime = time.time()
Verbose("Create dataset manager with given settings", True)
nominalModule = ModuleBuilder(opts, myOutputCreator,
opts.verbose)
nominalModule.createDsetMgr(opts.mcrab, era, searchMode,
optimizationMode)
if (iModule == 1):
if opts.verbose:
nominalModule.debug()
doQCDNormalizationSyst = False #FIXME
if not doQCDNormalizationSyst:
msg = "Disabling systematics"
Verbose(ShellStyles.WarningLabel() + msg, True) #fixme
# Build the module
nominalModule.buildModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors[opts.normFactorKey],
doQCDNormalizationSyst, opts.normDataSrc,
opts.normEwkSrc)
if len(mySystematicsNames) > 0:
Print(
"Adding QCD normalization systematics (iff also other systematics present) ",
True)
nominalModule.buildQCDNormalizationSystModule(
opts.dataSrc, opts.ewkSrc)
# FIXME: add quark gluon weighting systematics!
if 0:
Print("Adding Quark/Gluon weighting systematics", True)
nominalModule.buildQCDQuarkGluonWeightingSystModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors["FakeWeightingUp"],
myNormFactors["FakeWeightingDown"], False,
opts.normDataSrc, opts.normEwkSrc)
Verbose("Deleting nominal module", True)
nominalModule.delete()
Verbose("Printing time estimate", True)
printTimeEstimate(myGlobalStartTime, myStartTime, iModule,
myTotalModules)
Verbose("Now do the rest of systematics variations", True)
for syst in mySystematicsNames:
iModule += 1
msg = "Analyzing systematics variations %d/%d: %s/%s/%s" % (
iModule, myTotalModules, myModuleInfoString, syst,
shapeType)
Print(
ShellStyles.CaptionStyle() + msg +
ShellStyles.NormalStyle(), True)
myStartTime = time.time()
systModule = ModuleBuilder(opts, myOutputCreator)
# Create dataset manager with given settings
systModule.createDsetMgr(opts.mcrab,
era,
searchMode,
optimizationMode,
systematicVariation=syst)
# Build asystematics module
systModule.buildModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors[opts.normFactorKey], False,
opts.normDataSrc, opts.normEwkSrc)
printTimeEstimate(myGlobalStartTime, myStartTime,
iModule, myTotalModules)
systModule.delete()
Verbose("Pseudo-multicrab ready for %s" % shapeType, True)
# Print some timing statistics
Print(
"Average processing time per module was %.1f seconds" %
getAvgProcessTimeForOneModule(myGlobalStartTime, myTotalModules), True)
Print(
"Total elapsed time was %.1f seconds" %
getTotalElapsedTime(myGlobalStartTime), False)
# Create rest of pseudo multicrab directory
myOutputCreator.finalize(silent=False)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 0:
datasetsMgr.printSelections()
# PrintPSet("BJetSelection", datasetsMgr)
# PrintPSet("TopSelectionBDT", datasetsMgr)
# PrintPSet("FakeBMeasurement", datasetsMgr)
sys.exit()
# Print dataset info?
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Set signal cross-section
if 0:
datasetsMgr.getDataset(opts.signal).setCrossSection(1.0)
# Remove unwanted datasets
if 0:
datasetsMgr.remove(filter(lambda name: "QCD-b" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get Luminosity
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Get all the histograms and their paths (e.g. ForFakeBMeasurement/Baseline_DeltaRLdgTrijetBJetTetrajetBJet_AfterCRSelections)
hList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(opts.folder)
hPaths = [os.path.join(opts.folder, h) for h in hList]
# Create two lists of paths: one for "Baseline" (SR) and one for "Inverted" (CR)
path_SR = [] # baseline, _AfterAllSelections
path_CR1 = [] # baseline, _AfterCRSelections
path_VR = [] # inverted, _AfterAllSelections
path_CR2 = [] # inverted, _AfterCRSelections
# For-loop: All histogram paths
for p in hPaths:
if "AfterStandardSelections" in p:
continue
if "Baseline" in p:
if "AllSelections" in p:
path_SR.append(p)
if "CRSelections" in p:
path_CR1.append(p)
if "Inverted" in p:
if "AllSelections" in p:
path_VR.append(p)
if "CRSelections" in p:
path_CR2.append(p)
# For-loop: All histogram pairs
for hVR, hCR2, hCR1 in zip(path_VR, path_CR2, path_CR1):
break
if "IsGenuineB" in hVR:
continue
PlotComparison(datasetsMgr, hVR, hCR2, "VRvCR2")
# For-loop: All histogram pairs
counter = 1
for hCR1, hCR2 in zip(path_CR1, path_CR2):
if "IsGenuineB" in hCR1:
continue
hName = hCR1.replace("_AfterCRSelections", "_CR1vCR2").replace("ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)), hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), counter==1)
PlotComparison(datasetsMgr, hCR1, hCR2, "CR1vCR2")
counter+=1
# For-loop: All histogram pairs
for hSR, hVR in zip(path_SR, path_VR):
break
# Print("UNBLINDING SR! Are you nuts ? BREAK!", False)
if "IsGenuineB" in hSR:
continue
PlotComparison(datasetsMgr, hSR, hVR, "SRvVR")
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
# Print PSets used for FakeBMeasurement
if 0:
datasetsMgr.printSelections()
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ZJetsToQQ_HT600toInf" in d.getName():
datasetsMgr.remove(d.getName())
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if d.getName() not in opts.signal:
if not opts.acceptance:
datasetsMgr.remove(d.getName())
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
datasetsMgr.PrintInfo()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get Luminosity
# For acceptance plot use all available masses
if opts.acceptance:
opts.signal = []
opts.signalMasses = []
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
#dName = d.getName().replace("_ext1", "").split("M_")
dName = d.getName().split("M_")
m = int(dName[-1])
opts.signal.append(d.getName())
opts.signalMasses.append(m)
if opts.intLumi < 0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Merge EWK samples
if 1:
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print post EWK-merge dataset summary
datasetsMgr.PrintInfo()
# Get the efficiency graphs
hGraphList = []
histoName = os.path.join(opts.folder, "counter")
hGraphList, _kwargs = GetHistoGraphs(datasetsMgr, opts.folder,
histoName)
# Plot the histo graphs
PlotHistoGraphs(hGraphList, _kwargs)
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Get the PSets:
if 0:
datasetsMgr.printSelections()
#PrintPSet("BJetSelection", datasetsMgr, depth=150)
#PrintPSet("fakeBMeasurement", datasetsMgr, depth=150)
sys.exit()
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames() and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print("Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..", True)
datasetsMgr.remove(filter(lambda name: "ZJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get luminosity if a value is not specified
if opts.intLumi < 0:
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Remove datasets
removeList = ["QCD-b", "Charged"]
if not opts.useMC:
removeList.append("QCD")
for i, d in enumerate(removeList, 0):
msg = "Removing dataset %s" % d
Verbose(ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(), i==0)
datasetsMgr.remove(filter(lambda name: d in name, datasetsMgr.getAllDatasetNames()))
# Print summary of datasets to be used
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
# Print dataset information
datasetsMgr.PrintInfo()
# List of TDirectoryFile (_CRone, _CRtwo, _VR, _SR)
tdirs = ["LdgTetrajetMass_", "LdgTetrajetPt_", "LdgTrijetPt_", "LdgTrijetMass_", "TetrajetBJetPt_", "TetrajetBJetEta_"]
region = ["CRone", "CRtwo", "SR", "VR"]
hList = []
for d in tdirs:
for r in region:
hList.append(d + r)
# Get the folders with the binned histograms
folderList_ = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(opts.folder)
folderList = [h for h in folderList_ if h in hList]
# For-loop: All folders
histoPaths = []
for f in folderList:
folderPath = os.path.join(opts.folder, f)
histoList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folderPath)
pathList = [os.path.join(folderPath, h) for h in histoList]
histoPaths.extend( pathList )
# Get all the bin labels
binLabels = GetBinLabels("CRone", histoPaths)
for i, t in enumerate(tdirs, 1):
myList = []
for p in histoPaths:
if t in p:
myList.append(p)
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (len(tdirs)), t.replace("_", ""))
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistograms(datasetsMgr, myList, binLabels, opts)
# Save the plots
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(True)
style.setGridY(True)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Get the PSets:
if 0:
datasetsMgr.printSelections()
#PrintPSet("BJetSelection", datasetsMgr, depth=150)
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames(
) and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print(
"Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..",
True)
datasetsMgr.remove(
filter(lambda name: "ZJetsToQQ" in name,
datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get luminosity if a value is not specified
if opts.intLumi < 0:
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Remove datasets
removeList = ["QCD-b"] #, "Charged"]
if not opts.useMC:
removeList.append("QCD")
for i, d in enumerate(removeList, 0):
msg = "Removing dataset %s" % d
Verbose(
ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(),
i == 0)
datasetsMgr.remove(
filter(lambda name: d in name,
datasetsMgr.getAllDatasetNames()))
# Print summary of datasets to be used
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
# Print dataset information
datasetsMgr.PrintInfo()
# Do the fit on the histo after ALL selections (incl. topology cuts)
folderList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
opts.folder)
#folderList1 = [h for h in folderList if "TetrajetPt" in h]
#folderList1 = [h for h in folderList if "TetrajetMass" in h]
#folderList1 = [h for h in folderList if "MET" in h]
#folderList1 = [h for h in folderList if "TetrajetBJetPt" in h]
folderList1 = [h for h in folderList if "QGLR" in h]
folderList2 = [
h for h in folderList1
if "CRtwo" in h or "VR" in h or "SR" in h or "CRone" in h
]
# For-loop: All folders
histoPaths = []
for f in folderList2:
folderPath = os.path.join(opts.folder, f)
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
folderPath)
pathList = [os.path.join(folderPath, h) for h in histoList]
histoPaths.extend(pathList)
binLabels = GetBinLabels("CRone", histoPaths)
PlotHistosAndCalculateTF(datasetsMgr, histoPaths, binLabels, opts)
return
def main(count, runRange, dsetList, opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(True)
style.setGridY(True)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Remove datasets
removeList = ["QCD-b", "Charged", "QCD", "ZJetsToQQ_HT600toInf"]
opts.intLumi = 0.0
# For-loop: All datasets in the manager
for d in datasetsMgr.getAllDatasets():
if d.isMC():
continue
# Inclusive data
if len(dsetList) == 1 and dsetList[0] == "Run2016":
Verbose("Inclusive data. Will not remove anything", True)
opts.intLumi += GetLumi(datasetsMgr)
break
# Special combinations
for rr in dsetList:
if rr not in d.getName():
Verbose(
"\"%s\" is not in dataset name \"%s\"" %
(rr, d.getName()), False)
if d.getName() not in removeList:
# Ensure dataset to be removed is not in the dsetList
if not any(rr in d.getName() for rr in dsetList):
removeList.append(d.getName())
else:
Verbose(
"\"%s\" is in dataset name \"%s\"" % (rr, d.getName()),
False)
# Get luminosity if a value is not specified
opts.intLumi += d.getLuminosity()
# For-loop: All dataset names to be removed
for i, d in enumerate(removeList, 0):
Verbose(
ShellStyles.HighlightAltStyle() + "Removing dataset %s" % d +
ShellStyles.NormalStyle(), False)
datasetsMgr.remove(
filter(lambda name: d in name,
datasetsMgr.getAllDatasetNames()))
# Inform user of dataset and corresponding integrated lumi
Print("%d) %s (%.1f 1/pb)" % (count, runRange, opts.intLumi),
count == 1)
#Print("%d) %s (%.1f 1/pb)" % (count, ", ".join(dsetList), opts.intLumi), count==1)
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Print dataset information
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
# Do the fit on the histo after ALL selections (incl. topology cuts)
folderList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
opts.folder)
folderList1 = [h for h in folderList if opts.histoKey in h]
folderList2 = [
h for h in folderList1 if "VR" in h or "SR" in h or "CRone" in h
or "CRtwo" in h or "CRthree" in h or "CRfour" in h
]
# For-loop: All folders
histoPaths = []
for f in folderList2:
folderPath = os.path.join(opts.folder, f)
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
folderPath)
pathList = [os.path.join(folderPath, h) for h in histoList]
histoPaths.extend(pathList)
binLabels = GetBinLabels("CRone", histoPaths)
PlotHistosAndCalculateTF(runRange, datasetsMgr, histoPaths, binLabels,
opts)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 0:
datasetsMgr.printSelections()
# Print PSets used for FakeBMeasurement
if 0:
datasetsMgr.printSelections()
PrintPSet("BJetSelection", datasetsMgr)
PrintPSet("TopSelectionBDT", datasetsMgr)
PrintPSet("FakeBMeasurement", datasetsMgr)
sys.exit()
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
datasetsMgr.PrintInfo()
# Filter the datasets
datasetsMgr.remove(filter(lambda name: "Charged" in name, datasetsMgr.getAllDatasetNames()))
# datasetsMgr.remove(filter(lambda name: "Charged" in name and not "M_500" in name, datasetsMgr.getAllDatasetNames()))
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames() and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print("Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..", True)
datasetsMgr.remove(filter(lambda name: "ZJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get Luminosity
if opts.intLumi < 0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Re-order datasets (different for inverted than default=baseline)
if 0:
newOrder = ["Data"]
newOrder.extend(aux.GetListOfEwkDatasets())
datasetsMgr.selectAndReorder(newOrder)
# Print post-merged data dataset summary
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print post EWK-merge dataset summary
if 1:
datasetsMgr.PrintInfo()
# Get all histograms from the in the selected folder inside the ROOT files
allHistos = datasetsMgr.getAllDatasets()[0].getDirectoryContent(opts.folder)
hList = [h for h in allHistos if "CRSelections" in h and "_Vs" not in h]
hList.extend([h for h in allHistos if "AllSelections" in h and "_Vs" not in h])
# hList.extend([h for h in allHistos if "StandardSelections" in h and "_Vs" not in h])
# Create a list with strings included in the histogram names you want to plot
myHistos = ["LdgTrijetPt", "LdgTrijetMass", "TetrajetBJetPt", "TetrajetBJetEta", "LdgTetrajetPt", "LdgTetrajetMass", "MVAmax2", "MVAmax1", "HT", "MET"]
# For-loop: All histos
for i, h in enumerate(myHistos, 1):
hGraphList = []
for b in ["Baseline_", "Inverted_"]:
for r in [ "_AfterCRSelections", "_AfterAllSelections"]:
histoName = b + h + r
hgQCD, kwargs = GetPurityHistoGraph(datasetsMgr, opts.folder, histoName)
# Do not draw SR in multigraph plot!
if GetControlRegionLabel(histoName) != "SR":
hGraphList.append(hgQCD)
# Plot individual purity graphs?
if 0:
PlotHistoGraph(hgQCD, kwargs)
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (len(myHistos)), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistoGraphs(hGraphList, kwargs)
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if 0:
datasetsMgr.printSelections()
# PrintPSet("BJetSelection", datasetsMgr)
# PrintPSet("TopSelectionBDT", datasetsMgr)
# PrintPSet("FakeBMeasurement", datasetsMgr)
sys.exit()
# Print dataset info?
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Set signal cross-section
if 0:
datasetsMgr.getDataset(opts.signal).setCrossSection(1.0)
# Remove unwanted datasets
if 0:
datasetsMgr.remove(
filter(lambda name: "QCD-b" in name,
datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get Luminosity
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Get all the histograms and their paths (e.g. ForFakeBMeasurement/Baseline_DeltaRLdgTrijetBJetTetrajetBJet_AfterCRSelections)
hList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
opts.folder)
hPaths = [os.path.join(opts.folder, h) for h in hList]
# Create two lists of paths: one for "Baseline" (SR) and one for "Inverted" (CR)
path_SR = [] # baseline, _AfterAllSelections
path_CR1 = [] # baseline, _AfterCRSelections
path_VR = [] # inverted, _AfterAllSelections
path_CR2 = [] # inverted, _AfterCRSelections
# For-loop: All histogram paths
for p in hPaths:
if "AfterStandardSelections" in p:
#print p
continue
if "Baseline" in p:
if "AllSelections" in p:
path_SR.append(p)
if "CRSelections" in p:
path_CR1.append(p)
if "Inverted" in p:
if "AllSelections" in p:
path_VR.append(p)
if "CRSelections" in p:
path_CR2.append(p)
# For-loop: All histogram pairs
for hVR, hCR2, hCR1 in zip(path_VR, path_CR2, path_CR1):
break # not needed now
if "IsGenuineB" in hVR:
continue
PlotComparison(datasetsMgr, hVR, hCR2, "VRvCR2")
# For-loop: All histogram pairs
counter = 1
for hCR1, hCR2 in zip(path_CR1, path_CR2):
if "IsGenuineB" in hCR1:
continue
hName = hCR1.replace("_AfterCRSelections", "_CR1vCR2").replace(
"ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format(
"Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)),
hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
counter == 1)
PlotComparison(datasetsMgr, hCR1, hCR2, "CR1vCR2")
counter += 1
# For-loop: All histogram pairs
for hSR, hVR in zip(path_SR, path_VR):
break
Print("UNBLINDING SR! Are you nuts ? BREAK!", False)
if "IsGenuineB" in hSR:
continue
PlotComparison(datasetsMgr, hSR, hVR, "SRvVR")
# # For-loop: All histogram pairs
# for hSR, hCR1 in zip(path_SR, path_CR1):
# #break
# Print("UNBLINDING SR! Are you nuts ? BREAK!", False)
# raw_input("Press any key to continue")
# if "IsGenuineB" in hSR:
# continue
# PlotComparison(datasetsMgr, hSR, hCR1, "SRvCR1")
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
treatNegativeBins(qcdMt, "QCD_inverted mT")
# Do calculation
manager.calculateCombinedNormalizationCoefficient(qcdMt, ewkFakeTausMt)
#===== Save normalization
outFileName = "QCDNormalizationFactors_%s_%s.py"%(HISTONAME, moduleInfoString)
outFileFullName = os.path.join(argv[1],outFileName)
manager.writeScaleFactorFile(outFileFullName, moduleInfoString)
if __name__ == "__main__":
# Check parameters
if len(sys.argv) < 2:
usage()
# Find out the era/search mode/optimization mode combinations and run each of them
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=sys.argv[1])
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
myModuleSelector.doSelect(None)
#myModuleSelector.printSelectedCombinationCount()
for era in myModuleSelector.getSelectedEras():
for searchMode in myModuleSelector.getSelectedSearchModes():
for optimizationMode in myModuleSelector.getSelectedOptimizationModes():
print ShellStyles.HighlightStyle()+"\nCalculating normalization for module %s/%s/%s%s"%(era, searchMode, optimizationMode, ShellStyles.NormalStyle())
# Construct info string for the module
moduleInfoString = "%s_%s"%(era, searchMode)
if len(optimizationMode) > 0:
moduleInfoString += "_%s"%(optimizationMode)
# Create dataset manager
dsetMgr = dsetMgrCreator.createDatasetManager(dataEra=era,searchMode=searchMode,optimizationMode=optimizationMode)
main(sys.argv, dsetMgr, moduleInfoString)
dsetMgrCreator.close()
def main(opts):
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setOptStat(False)
# If user does not define optimisation mode do all of them
if opts.optMode == None:
optModes = optList
else:
optModes = [opts.optMode]
# Define genuineB and fakeB datasets
mergeMap = {"TT": "GenuineB",
"QCD_HT50to100" : "FakeB", #v. poor stats!
"QCD_HT200to300" : "FakeB", #v. poor stats!
"QCD_HT100to200" : "FakeB", #v. poor stats!
"QCD_HT200to300" : "FakeB", #v. poor stats!
"QCD_HT300to500" : "FakeB",
"QCD_HT500to700" : "FakeB",
"QCD_HT700to1000" : "FakeB",
"QCD_HT1000to1500": "FakeB",
"QCD_HT1500to2000": "FakeB",
"QCD_HT2000toInf" : "FakeB",
"WJetsToQQ_HT_600ToInf": "FakeB"
}
# Definitions
results = []
# For-loop: All optimisation modes
for index, opt in enumerate(optModes, 1):
opts.optMode = opt
# Definitions
genuineB = None
fakeB = None
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
# Print datasets info ?
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Remove unwanted datasets
removeList = ["QCD_HT50to100", "QCD_HT100to200", "QCD_HT200to300", "QCD_HT200to300_ext1", "WJetsToQQ_HT_600ToInf"]
for k, d in enumerate(datasetsMgr.getAllDatasets(), 1):
if d.getName() in removeList:
datasetsMgr.remove(d.getName())
Verbose(ShellStyles.ErrorStyle() + "Removing dataset %s" % d.getName() + ShellStyles.NormalStyle(), k==1)
if index == 1:
datasetsMgr.PrintInfo()
# Merge datasets into two groups: Genuine-B and Fake-B
datasetsMgr.mergeMany(mergeMap, addition=False)
for d in datasetsMgr.getAllDatasets():
if d.getName() == "GenuineB":
genuineB = d
if d.getName() == "FakeB":
fakeB = d
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get integrated luminosity
intLumi = datasetsMgr.getDataset("Data").getLuminosity()
datasetsMgr.remove(filter(lambda name: "Data" in name, datasetsMgr.getAllDatasetNames()))
# Print dataset information
if index == 1:
datasetsMgr.PrintInfo()
# Do the plot
name = GetPlotName(opts)
myResults = doPlot(name, genuineB, fakeB, opts.errorlevel, opts.optMode, intLumi)
# Save results and msgs
results.extend(myResults)
# For-loop: All points
if opts.verbose:
for item in results:
print item
# Print path of all saved plots
msg = "All plots saved under directory %s" % (aux.convertToURL(opts.saveDir, opts.url))
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), True)
# Write results to a json file
with open(opts.json, 'w') as outfile:
json.dump(results, outfile)
msg = "Wrote results to \"%s\"" % (opts.json)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), True)
return
def main():
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
# Obtain multicrab directory
myMulticrabDir = "."
if opts.mcrab != None:
myMulticrabDir = opts.mcrab
if not os.path.exists("%s/multicrab.cfg" % myMulticrabDir):
msg = "No multicrab directory found at path '%s'! Please check path or specify it with --mcrab!" % (
myMulticrabDir)
raise Exception(ShellStyles.ErrorLabel() + msg +
ShellStyles.NormalStyle())
if len(opts.shape) == 0:
raise Exception(
ShellStyles.ErrorLabel() +
"Provide a shape identifierwith --shape (for example MT)!" +
ShellStyles.NormalStyle())
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=myMulticrabDir)
# Obtain systematics names
mySystematicsNamesRaw = dsetMgrCreator.getSystematicVariationSources()
mySystematicsNames = []
for item in mySystematicsNamesRaw:
mySystematicsNames.append("%sPlus" % item)
mySystematicsNames.append("%sMinus" % item)
if opts.test:
mySystematicsNames = [] #[mySystematicsNames[0]] #FIXME
# Set the primary source
myModuleSelector.setPrimarySource(label=opts.analysisName,
dsetMgrCreator=dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts=None) #FIXME: (opts=opts)
# Loop over era/searchMode/optimizationMode combos
myDisplayStatus = True
myTotalModules = myModuleSelector.getSelectedCombinationCount() * (
len(mySystematicsNames) + 1) * len(opts.shape)
Verbose("Found %s modules in total" % (myTotalModules), True)
count, nEras, nSearchModes, nOptModes, nSysVars = myModuleSelector.getSelectedCombinationCountIndividually(
)
if nSysVars > 0:
msg = "Will run over %d modules (%d eras x %d searchModes x %d optimizationModes x %d systematic variations)" % (
count, nEras, nSearchModes, nOptModes, nSysVars)
else:
msg = "Will run over %d modules (%d eras x %d searchModes x %d optimizationModes)" % (
count, nEras, nSearchModes, nOptModes)
Print(msg, True)
# Create pseudo-multicrab creator
myOutputCreator = pseudoMultiCrabCreator.PseudoMultiCrabCreator(
opts.analysisName, myMulticrabDir)
# Make time stamp for start time
myGlobalStartTime = time.time()
iModule = 0
# For-loop: All Shapes
for shapeType in opts.shape:
# Initialize
myOutputCreator.initialize(shapeType, prefix="")
msg = "Creating dataset for shape \"%s\"%s" % (
shapeType, ShellStyles.NormalStyle())
Verbose(ShellStyles.HighlightStyle() + msg, True)
# Get lists of settings
erasList = myModuleSelector.getSelectedEras()
modesList = myModuleSelector.getSelectedSearchModes()
optList = myModuleSelector.getSelectedOptimizationModes()
optList.append("") #append the default opt mode!
# For-Loop over era, searchMode, and optimizationMode options
for era in erasList:
for searchMode in modesList:
for optimizationMode in optList:
Verbose(
"era = %s, searchMode = %s, optMode = %s" %
(era, searchMode, optimizationMode), True)
# If an optimization mode is defined in options skip the rest
if opts.optMode != None:
if optimizationMode != opts.optMode:
continue
# Obtain normalization factors
myNormFactors = importNormFactors(era, searchMode,
optimizationMode,
opts.mcrab)
# Nominal module
myModuleInfoString = getModuleInfoString(
era, searchMode, optimizationMode)
iModule += 1
# Inform user of what is being processes
msg = "Module %d/%d:%s %s/%s" % (
iModule, myTotalModules, ShellStyles.NormalStyle(),
myModuleInfoString, shapeType)
Print(ShellStyles.CaptionStyle() + msg, True)
# Keep time
myStartTime = time.time()
Verbose("Create dataset manager with given settings", True)
nominalModule = ModuleBuilder(opts, myOutputCreator)
nominalModule.createDsetMgr(myMulticrabDir, era,
searchMode, optimizationMode)
if (iModule == 1):
if opts.verbose:
nominalModule.debug()
doQCDNormalizationSyst = False #FIXME
if not doQCDNormalizationSyst:
msg = "Disabling systematics"
Print(ShellStyles.WarningLabel() + msg, True)
nominalModule.buildModule(opts.dataSrc, opts.ewkSrc,
myNormFactors["nominal"],
doQCDNormalizationSyst,
opts.normDataSrc,
opts.normEwkSrc)
if len(mySystematicsNames) > 0:
Print(
"Adding QCD normalization systematics (iff also other systematics present) ",
True)
nominalModule.buildQCDNormalizationSystModule(
opts.dataSrc, opts.ewkSrc)
# FIXME: add quark gluon weighting systematics!
if 0:
Print("Adding Quark/Gluon weighting systematics", True)
nominalModule.buildQCDQuarkGluonWeightingSystModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors["FakeWeightingUp"],
myNormFactors["FakeWeightingDown"], False,
opts.normDataSrc, opts.normEwkSrc)
Verbose("Deleting nominal module", True)
nominalModule.delete()
Verbose("Printing time estimate", True)
printTimeEstimate(myGlobalStartTime, myStartTime, iModule,
myTotalModules)
Verbose("Now do the rest of systematics variations", True)
for syst in mySystematicsNames:
iModule += 1
msg = "Analyzing systematics variations %d/%d: %s/%s/%s" % (
iModule, myTotalModules, myModuleInfoString, syst,
shapeType)
Print(
ShellStyles.CaptionStyle() + msg +
ShellStyles.NormalStyle(), True)
myStartTime = time.time()
systModule = ModuleBuilder(opts, myOutputCreator)
# Create dataset manager with given settings
systModule.createDsetMgr(myMulticrabDir,
era,
searchMode,
optimizationMode,
systematicVariation=syst)
# Build asystematics module
systModule.buildModule(opts.dataSrc, opts.ewkSrc,
myNormFactors["nominal"], False,
opts.normDataSrc,
opts.normEwkSrc)
printTimeEstimate(myGlobalStartTime, myStartTime,
iModule, myTotalModules)
systModule.delete()
Verbose("Pseudo-multicrab ready for %s" % shapeType, True)
# Create rest of pseudo multicrab directory
myOutputCreator.silentFinalize()
# Print some timing statistics
Print(
"Average processing time per module was %.1f s" %
getAvgProcessTimeForOneModule(myGlobalStartTime, myTotalModules), True)
Print(
"Total elapsed time was %.1f s" %
getTotalElapsedTime(myGlobalStartTime), False)
msg = "Created pseudo-multicrab %s for shape type \"%s\"" % (
myOutputCreator.getDirName(), shapeType)
Print(ShellStyles.SuccessLabel() + msg, True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Get the PSets:
if 0:
datasetsMgr.printSelections()
#PrintPSet("BJetSelection", datasetsMgr, depth=150)
#PrintPSet("fakeBMeasurement", datasetsMgr, depth=150)
sys.exit()
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames(
) and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print(
"Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..",
True)
datasetsMgr.remove(
filter(lambda name: "ZJetsToQQ" in name,
datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get luminosity if a value is not specified
if opts.intLumi < 0:
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Remove datasets
removeList = ["QCD-b", "Charged"]
if not opts.useMC:
removeList.append("QCD")
for i, d in enumerate(removeList, 0):
msg = "Removing dataset %s" % d
Verbose(
ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(),
i == 0)
datasetsMgr.remove(
filter(lambda name: d in name,
datasetsMgr.getAllDatasetNames()))
# Print summary of datasets to be used
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
# Print dataset information
datasetsMgr.PrintInfo()
# List of TDirectoryFile (_CRone, _CRtwo, _VR, _SR)
tdirs = [
"LdgTrijetPt_", "LdgTrijetMass_", "TetrajetBJetPt_",
"TetrajetBJetEta_", "LdgTetrajetPt_", "LdgTetrajetMass_"
]
region = ["CRone", "CRtwo"]
hList = []
for d in tdirs:
for r in region:
hList.append(d + r)
# Get the folders with the binned histograms
folderList_ = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
opts.folder)
folderList = [h for h in folderList_ if h in hList]
# For-loop: All folders
histoPaths = []
for f in folderList:
folderPath = os.path.join(opts.folder, f)
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(
folderPath)
pathList = [os.path.join(folderPath, h) for h in histoList]
histoPaths.extend(pathList)
# Get all the bin labels
binLabels = GetBinLabels("CRone", histoPaths)
for i, t in enumerate(tdirs, 1):
myList = []
for p in histoPaths:
if t in p:
myList.append(p)
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format(
"Histogram", "%i" % i, "/", "%s:" % (len(tdirs)),
t.replace("_", ""))
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
i == 1)
PlotHistograms(datasetsMgr, myList, binLabels, opts)
# Save the plots
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
def main(opts):
# Object for selecting data eras, search modes, and optimization modes
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Obtain systematics names
mySystematicsNamesRaw = dsetMgrCreator.getSystematicVariationSources()
mySystematicsNames = getSystematicsNames(mySystematicsNamesRaw, opts)
# Set the primary source
Verbose(
"Setting the primary source (label=%s)" %
(ShellStyles.NoteStyle() + opts.analysisName +
ShellStyles.NormalStyle()), True)
myModuleSelector.setPrimarySource(label=opts.analysisName,
dsetMgrCreator=dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts=None) #fixme: (opts=opts)?
# Loop over era/searchMode/optimizationMode combos
myTotalModules = myModuleSelector.getSelectedCombinationCount() * (
len(mySystematicsNames) + 1) * len(opts.shape)
Verbose("Found %s modules in total" % (myTotalModules), True)
count, nEras, nSearchModes, nOptModes, nSysVars = myModuleSelector.getSelectedCombinationCountIndividually(
)
if nSysVars > 0:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes x %d systematic variations)" % (
count, nEras, nSearchModes, nOptModes, nSysVars)
else:
msg = "Running over %d modules (%d eras x %d searchModes x %d optimizationModes)" % (
count, nEras, nSearchModes, nOptModes)
Verbose(msg, True)
# Create pseudo-multicrab creator
msg = "Will create pseudo-dataset %s inside the pseudo-multicrab directory" % (
ShellStyles.NoteStyle() + opts.analysisName +
ShellStyles.NormalStyle())
Verbose(msg, True)
myOutputCreator = pseudoMultiCrabCreator.PseudoMultiCrabCreator(
opts.analysisName, opts.mcrab, verbose=opts.verbose)
# Make time stamp for start time
myGlobalStartTime = time.time()
iModule = 0
# For-loop: All Shapes
for iShape, shapeType in enumerate(opts.shape, 1):
msg = "Shape %d/%d:%s %s" % (iShape, len(
opts.shape), ShellStyles.NormalStyle(), shapeType)
Verbose(ShellStyles.HighlightAltStyle() + msg, True)
# Initialize
myOutputCreator.initialize(
subTitle=shapeType,
prefix="") #fixme: remove shapeType from sub-directory name?
# Get lists of settings
erasList = myModuleSelector.getSelectedEras()
modesList = myModuleSelector.getSelectedSearchModes()
optList = myModuleSelector.getSelectedOptimizationModes()
if 0:
optList.append(
""
) #append the default opt mode iff more optimization modes exist
# For-loop: All eras
for era in erasList:
# For-loop: All searchModes
for searchMode in modesList:
# For-loop: AlloptimizationModes
for optimizationMode in optList:
Verbose(
"era = %s, searchMode = %s, optMode = %s" %
(era, searchMode, optimizationMode), True)
# If an optimization mode is defined in options skip the rest
if opts.optMode != None:
if optimizationMode != opts.optMode:
continue
# Obtain normalization factors for given Era, SearchMode, and OptimizationMode!
myNormFactors = importNormFactors(era, searchMode,
optimizationMode,
opts.mcrab)
# Nominal module
myModuleInfoString = getModuleInfoString(
era, searchMode, optimizationMode)
iModule += 1
# Inform user of what is being processes
if optimizationMode != "":
msg = "Module %d/%d: %s_%s_%s_%s" % (
iModule, myTotalModules, opts.analysisName,
searchMode, era, optimizationMode)
else:
msg = "Module %d/%d: %s_%s_%s" % (
iModule, myTotalModules, opts.analysisName,
searchMode, era)
Print(
ShellStyles.HighlightAltStyle() + msg +
ShellStyles.NormalStyle(), iModule == 1)
# Keep time
myStartTime = time.time()
Verbose("Create dataset manager with given settings", True)
nominalModule = ModuleBuilder(opts, myOutputCreator,
opts.verbose)
nominalModule.createDsetMgr(opts.mcrab, era, searchMode,
optimizationMode)
if (iModule == 1):
if opts.verbose:
nominalModule.debug()
# Build the module
doFakeBNormalisationSyst = False
nominalModule.buildModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors[opts.normFactorKey],
doFakeBNormalisationSyst, opts.normDataSrc,
opts.normEwkSrc)
# Do TF variations named "SystVarUp" and "SystVarDown" (i.e. (Get results using TF+Error and TF-Error instead of TF)
if len(mySystematicsNames) > 0:
Verbose(
"Adding FakeB normalization systematics (iff also other systematics present) ",
True)
#nominalModule.buildTransferFactorVarSystModule(opts.dataSrc, opts.ewkSrc, myNormFactors["SystVarUp"], myNormFactors["SystVarDown"])
nominalModule.buildTransferFactorVarSystModule(
opts.dataSrc, opts.ewkSrc, myNormFactors)
Verbose("Deleting nominal module", True)
nominalModule.delete()
Verbose("Printing time estimate", True)
printTimeEstimate(myGlobalStartTime, myStartTime, iModule,
myTotalModules)
Verbose("Now do the rest of systematics variations", True)
for syst in mySystematicsNames:
iModule += 1
msg = "Module %d/%d: %s/%s" % (
iModule, myTotalModules, myModuleInfoString, syst)
print
Print(
ShellStyles.HighlightAltStyle() + msg +
ShellStyles.NormalStyle(), False)
myStartTime = time.time()
systModule = ModuleBuilder(opts, myOutputCreator)
# Create dataset manager with given settings
systModule.createDsetMgr(opts.mcrab,
era,
searchMode,
optimizationMode,
systematicVariation=syst)
# Build systematics module
Verbose(
"Building systematics module (opts.normFactorKey = %s)"
% (opts.normFactorKey), True)
systModule.buildModule(
opts.dataSrc, opts.ewkSrc,
myNormFactors[opts.normFactorKey], False,
opts.normDataSrc, opts.normEwkSrc)
printTimeEstimate(myGlobalStartTime, myStartTime,
iModule, myTotalModules)
systModule.delete()
print
Verbose("Pseudo-multicrab ready for %s" % shapeType, True)
# Print some timing statistics
Verbose(
"Average processing time per module was %.1f seconds" %
getAvgProcessTimeForOneModule(myGlobalStartTime, myTotalModules), True)
Print(
"Total elapsed time was %.1f seconds" %
getTotalElapsedTime(myGlobalStartTime), True)
# Create rest of pseudo multicrab directory
myOutputCreator.finalize(silent=False)
return
myMulticrabDir = "."
if opts.multicrabDir != None:
myMulticrabDir = opts.multicrabDir
if not os.path.exists("%s/multicrab.cfg"%myMulticrabDir):
raise Exception(ShellStyles.ErrorLabel()+"No multicrab directory found at path '%s'! Please check path or specify it with --mdir!"%(myMulticrabDir)+ShellStyles.NormalStyle())
if len(opts.shape) == 0:
raise Exception(ShellStyles.ErrorLabel()+"Provide a shape identifierwith --shape (for example MT)!"+ShellStyles.NormalStyle())
# Set EWK source depending on calculation mode
if opts.qcdonly:
_generalOptions["EWKsource"] = _generalOptions["ewkSourceForQCDOnly"]
else:
_generalOptions["EWKsource"] = _generalOptions["ewkSourceForQCDPlusFakeTaus"]
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=myMulticrabDir)
# Obtain systematics names
mySystematicsNamesRaw = dsetMgrCreator.getSystematicVariationSources()
mySystematicsNames = []
for item in mySystematicsNamesRaw:
mySystematicsNames.append("%sPlus"%item)
mySystematicsNames.append("%sMinus"%item)
if opts.test:
mySystematicsNames = [mySystematicsNames[0]]
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
# Select modules
myModuleSelector.doSelect(opts)
#dsetMgrCreator.close()
# Loop over era/searchMode/optimizationMode combos
myDisplayStatus = True
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
# Print PSets used for FakeBMeasurement
if 0:
datasetsMgr.printSelections()
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ZJetsToQQ_HT600toInf" in d.getName():
datasetsMgr.remove(d.getName())
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if d.getName() not in opts.signal:
if not opts.acceptance:
datasetsMgr.remove(d.getName())
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
datasetsMgr.PrintInfo()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get Luminosity
# For acceptance plot use all available masses
if opts.acceptance:
opts.signal = []
opts.signalMasses = []
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
#dName = d.getName().replace("_ext1", "").split("M_")
dName = d.getName().split("M_")
m = int(dName[-1])
opts.signal.append(d.getName())
opts.signalMasses.append(m)
if opts.intLumi < 0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Merge EWK samples
if 1:
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print post EWK-merge dataset summary
datasetsMgr.PrintInfo()
# Get the efficiency graphs
hGraphList = []
histoName = os.path.join(opts.folder, "counter")
hGraphList, _kwargs = GetHistoGraphs(datasetsMgr, opts.folder, histoName)
# Plot the histo graphs
PlotHistoGraphs(hGraphList, _kwargs)
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def addDatasetsFromMulticrab(self, directory, *args, **kwargs):
'''
kwargs for 'includeOnlyTasks' or 'excludeTasks' to set the datasets over which this analyzer is processed, default is all datasets
'''
Verbose("addDatasetsFromMulticrab()", True)
blacklist = []
if "blacklist" in kwargs.keys():
if isinstance(kwargs["blacklist"], str):
blacklist.append(kwargs["blacklist"])
elif isinstance(kwargs["blacklist"], list):
blacklist.extend(kwargs["blacklist"])
else:
raise Exception("Unsupported input format!")
del kwargs["blacklist"]
whitelist = []
if "whitelist" in kwargs.keys():
if isinstance(kwargs["whitelist"], str):
whitelist.append(kwargs["whitelist"])
elif isinstance(kwargs["whitelist"], list):
whitelist.extend(kwargs["whitelist"])
else:
raise Exception("Unsupported input format!")
del kwargs["whitelist"]
dataset._optionDefaults[
"input"] = "histograms-*.root" #"miniaod2tree*.root"
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=directory,
*args,
**kwargs)
dsets = dsetMgrCreator.getDatasetPrecursors()
# Check if data datasets included
for i, d in enumerate(dsets, 1):
if d.isData():
self.datasetsData.append(d)
dsetMgrCreator.close()
# Create a manager with data datasets (to enable PU reweighting even if not running with data)
if len(self.datasetsData) < 1:
dsetMgrCreator_tmp = dataset.readFromMulticrabCfg(
directory=directory)
dsets_tmp = dsetMgrCreator_tmp.getDatasetPrecursors()
for i, d in enumerate(dsets_tmp, 1):
if d.isData():
self.datasetsData.append(d)
dsetMgrCreator_tmp.close()
self.Print(
"Pileup reweighting will be done according to the sum of:\n\t%s" %
(sh_Note + "\n\t".join([d.getName()
for d in self.datasetsData]) + sh_Normal),
True)
if len(whitelist) > 0:
for dset in dsets:
isOnWhiteList = False
for item in whitelist:
if dset.getName().startswith(item):
isOnWhiteList = True
if not isOnWhiteList:
blacklist.append(dset.getName())
for dset in dsets:
isOnBlackList = False
for item in blacklist:
if dset.getName().startswith(item):
isOnBlackList = True
if isOnBlackList:
self.Print(
"Ignoring dataset because of black/whitelist options: '%s' ..."
% dset.getName(), True)
else:
self.addDataset(dset.getName(),
dset.getFileNames(),
dataVersion=dset.getDataVersion(),
lumiFile=dsetMgrCreator.getLumiFile())
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(False)
style.setGridY(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab1)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Get the datasets from the directory
datasetsMgr1 = GetDatasetsFromDir(opts.mcrab1, opts)
datasetsMgr2 = GetDatasetsFromDir(opts.mcrab2, opts)
datasetsMgr3 = GetDatasetsFromDir(opts.mcrab3, opts)
# Setup the dataset managers
datasetsMgr1.updateNAllEventsToPUWeighted()
datasetsMgr1.loadLuminosities() # from lumi.json
datasetsMgr2.updateNAllEventsToPUWeighted()
datasetsMgr2.loadLuminosities() # from lumi.json
datasetsMgr3.updateNAllEventsToPUWeighted()
datasetsMgr3.loadLuminosities() # from lumi.json
# Print dataset info?
if opts.verbose:
datasetsMgr1.PrintCrossSections()
datasetsMgr1.PrintLuminosities()
datasetsMgr2.PrintCrossSections()
datasetsMgr2.PrintLuminosities()
datasetsMgr2.PrintCrossSections()
datasetsMgr2.PrintLuminosities()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr1)
plots.mergeRenameReorderForDataMC(datasetsMgr2)
plots.mergeRenameReorderForDataMC(datasetsMgr3)
# Get Luminosity
lumi1 = datasetsMgr1.getDataset("Data").getLuminosity()
lumi2 = datasetsMgr2.getDataset("Data").getLuminosity()
lumi3 = datasetsMgr3.getDataset("Data").getLuminosity()
if lumi1 != lumi2 != lumi3:
raise Exception("Lumi1 (=%.2f) != Lumi2 (=%.2f) != Lumi3 (=%.2f" %
(lumi1, lumi2, lumi3))
else:
opts.intLumi = datasetsMgr1.getDataset("Data").getLuminosity()
# Merge EWK samples
datasetsMgr1.merge("EWK", aux.GetListOfEwkDatasets())
datasetsMgr2.merge("EWK", aux.GetListOfEwkDatasets())
datasetsMgr3.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr1.PrintInfo()
if 0:
datasetsMgr2.PrintInfo()
datasetsMgr3.PrintInfo()
# Get all the histograms and their paths (e.g. ForFakeBMeasurement/Baseline_DeltaRLdgTrijetBJetTetrajetBJet_AfterCRSelections)
hList = datasetsMgr1.getDataset(
datasetsMgr1.getAllDatasetNames()[0]).getDirectoryContent(
opts.folder)
hPaths = [os.path.join(opts.folder, h) for h in hList]
# Create a smaller list with only histos of interest
hListS = []
for h in hList:
if "StandardSelections" in h:
continue
if "IsGenuineB" in h:
continue
if "_Bjet" in h:
continue
if "_Jet" in h:
continue
if "_SubLdg" in h:
continue
if "_Njets" in h:
continue
if "_NBjets" in h:
continue
if "_Delta" in h:
continue
if "Dijet" in h:
continue
if "Bdisc" in h:
continue
#if "MVA" in h:
# continue
if "MET" in h:
continue
if "HT" in h:
continue
# Otherwise keep the histogram
hListS.append(h)
hPathsS = [os.path.join(opts.folder, h) for h in hListS]
# Create two lists of paths: one for "Baseline" (SR) and one for "Inverted" (CR)
path_SR = [] # baseline, _AfterAllSelections
path_CR1 = [] # baseline, _AfterCRSelections
path_VR = [] # inverted, _AfterAllSelections
path_CR2 = [] # inverted, _AfterCRSelections
# For-loop: All histogram paths
for p in hPathsS: #hPaths:
if "Baseline" in p:
if "AllSelections" in p:
path_SR.append(p)
if "CRSelections" in p:
path_CR1.append(p)
if "Inverted" in p:
if "AllSelections" in p:
path_VR.append(p)
if "CRSelections" in p:
path_CR2.append(p)
counter = 1
# For-loop: All histogram pairs
for hCR1, hCR2 in zip(path_CR1, path_CR2):
if "IsGenuineB" in hCR1:
continue
#hName = hCR1.replace("_AfterCRSelections", "_CR1vCR2").replace("ForFakeBMeasurement/Baseline_", "")
hName = hCR1.replace("_AfterCRSelections",
" (CR1 and R2)").replace(
"ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format(
"Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)),
hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
counter == 1)
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hCR1,
hCR2, "CR1")
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hCR1,
hCR2, "CR2") #iro
counter += 1
# WARNING! This unblinds the Signal Region (SR)
for hSR, hVR in zip(path_SR, path_VR):
if "IsGenuineB" in hSR:
continue
if 1:
continue
#hName = hCR1.replace("_AfterCRSelections", "_SRvVR").replace("ForFakeBMeasurement/Baseline_", "")
hName = hCR1.replace("_AfterCRSelections", " (SR and VR)").replace(
"ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format(
"Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)),
hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
counter == 1)
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hSR, hVR,
"SR")
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hSR, hVR,
"VR")
counter += 1
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
#===== Save normalization
outFileName = "QCDNormalizationFactors_%s_%s.py" % (HISTONAME,
moduleInfoString)
print argv[1], outFileName
outFileFullName = os.path.join(argv[1], outFileName)
manager.writeScaleFactorFile(outFileFullName, moduleInfoString)
if __name__ == "__main__":
# Check parameters
if len(sys.argv) < 2:
usage()
# Find out the era/search mode/optimization mode combinations and run each of them
myModuleSelector = analysisModuleSelector.AnalysisModuleSelector()
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=sys.argv[1])
myModuleSelector.setPrimarySource("analysis", dsetMgrCreator)
myModuleSelector.doSelect(None)
#myModuleSelector.printSelectedCombinationCount()
for era in myModuleSelector.getSelectedEras():
for searchMode in myModuleSelector.getSelectedSearchModes():
for optimizationMode in myModuleSelector.getSelectedOptimizationModes(
):
print ShellStyles.HighlightStyle(
) + "\nCalculating normalization for module %s/%s/%s%s" % (
era, searchMode, optimizationMode,
ShellStyles.NormalStyle())
# Construct info string for the module
moduleInfoString = "%s_%s" % (era, searchMode)
if len(optimizationMode) > 0:
moduleInfoString += "_%s" % (optimizationMode)
def main(opts):
# Suppress warnings about weight being re-applied
ROOT.gErrorIgnoreLevel = ROOT.kError
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
opts.optMode = ""
mcrabName = opts.mcrab
RunEra = mcrabName.split("_")[1]
# Setup ROOT and style
ROOT.gROOT.SetBatch(opts.batchMode)
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(True)
style.setGridY(True)
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
# Remove some QCD samples (the cross sections of these samples are not calculated)
if 0:
msg = "Removing following samples:"
Print(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(), True)
for d in getDatasetsToExclude():
Print(d, False)
datasetsMgr.remove(d)
# Get run-range
minRunRange, maxRunRange, runRange = GetRunRange(datasetsMgr)
# Get int lumi
intLumi = GetLumi(datasetsMgr)
# Update to PU & load luminosities
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities()
#datasetsMgr.normalizeMCByLuminosity()
# Print luminisoties and cross-sections
datasetsMgr.PrintLuminosities()
datasetsMgr.PrintCrossSections()
# Default merging & ordering: "Data", "QCD", "SingleTop", "Diboson"
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get datasets
datasetsMgr.mergeMC()
dataset_Data = datasetsMgr.getDataDatasets()
dataset_MC = datasetsMgr.getMCDatasets()
# Define lists of Triggers to be plotted and Variables
xVars = ["pt6thJet", "eta6thJet", "phi6thJet", "Ht", "nBTagJets", "pu", "JetMulti", "BJetMulti"]
trgList = ["1BTag", "2BTag", "OR", "OR_PFJet450"]
if opts.fast:
trgList = ["OR_PFJet450"]
xVars = ["pt6thJet", "Ht"]
nPlots = len(trgList)*len(xVars)
counter = 0
# For-loop: All signal triggers
for i, trg in enumerate(trgList, 1):
# For-loop: All x-variables
for j, xVar in enumerate(xVars, 1):
counter+=1
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % counter, "/", "%s:" % (nPlots), "%s Vs %s" % (trg, xVar))
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), counter==1)
# Define names
hNumerator = "hNum_%s_RefTrg_OfflineSel_Signal%s" % (xVar, trg)
hDenominator = "hDen_%s_RefTrg_OfflineSel" % (xVar)
plotName = "Eff_%s_%s" % (xVar, trg)
# Get Efficiency Plots
_kwargs = GetHistoKwargs(xVar, opts)
eff_Data = GetEfficiency(datasetsMgr, dataset_Data, hNumerator, hDenominator , **_kwargs)
eff_MC = GetEfficiency(datasetsMgr, dataset_MC, hNumerator, hDenominator, **_kwargs)
# Apply Styles
styles.dataStyle.apply(eff_Data)
styles.mcStyle.apply(eff_MC)
# Create the plot
p = plots.ComparisonPlot(histograms.HistoGraph(eff_Data, "eff_Data", "p", "P"),
histograms.HistoGraph(eff_MC, "eff_MC" , "p", "P"),
saveFormats=[])
# Define the legend entries
p.histoMgr.setHistoLegendLabelMany(
{
"eff_Data": "Data",
"eff_MC" : "Simulation"
}
)
# Draw and save the plot
p.setLuminosity(intLumi)
plots.drawPlot(p, plotName, **_kwargs)
# Draw
histograms.addText(0.65, 0.06, "Runs "+ runRange, 17)
histograms.addText(0.65, 0.10, "2016", 17)
# Save the canvas to a file
SavePlot(p, plotName, os.path.join(opts.saveDir, opts.optMode), saveFormats=[".pdf", ".png", ".C"] )
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def addDatasetsFromMulticrab(self, directory, *args, **kwargs):
'''
kwargs for 'includeOnlyTasks' or 'excludeTasks' to set the datasets over which this analyzer is processed, default is all datasets
'''
Verbose("addDatasetsFromMulticrab()", True)
blacklist = []
if "blacklist" in kwargs.keys():
if isinstance(kwargs["blacklist"], str):
blacklist.append(kwargs["blacklist"])
elif isinstance(kwargs["blacklist"], list):
blacklist.extend(kwargs["blacklist"])
else:
raise Exception("Unsupported input format!")
del kwargs["blacklist"]
whitelist = []
if "whitelist" in kwargs.keys():
if isinstance(kwargs["whitelist"], str):
whitelist.append(kwargs["whitelist"])
elif isinstance(kwargs["whitelist"], list):
whitelist.extend(kwargs["whitelist"])
else:
raise Exception("Unsupported input format!")
del kwargs["whitelist"]
dataset._optionDefaults["input"] = "histograms-*.root" #"miniaod2tree*.root"
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=directory, *args, **kwargs)
dsets = dsetMgrCreator.getDatasetPrecursors()
# Check if data datasets included
for i, d in enumerate(dsets, 1):
if d.isData():
self.datasetsData.append(d)
dsetMgrCreator.close()
# Create a manager with data datasets (to enable PU reweighting even if not running with data)
if len(self.datasetsData) < 1:
dsetMgrCreator_tmp = dataset.readFromMulticrabCfg(directory=directory)
dsets_tmp = dsetMgrCreator_tmp.getDatasetPrecursors()
for i, d in enumerate(dsets_tmp, 1):
if d.isData():
self.datasetsData.append(d)
dsetMgrCreator_tmp.close()
self.Print("Pileup reweighting will be done according to the sum of:\n\t%s" % (sh_Note + "\n\t".join([d.getName() for d in self.datasetsData]) + sh_Normal), True)
if len(whitelist) > 0:
for dset in dsets:
isOnWhiteList = False
for item in whitelist:
if item==dset.getName() or ( item.endswith('*') and dset.getName().startswith(item[:-1]) ):
isOnWhiteList = True
if not isOnWhiteList:
blacklist.append(dset.getName())
for dset in dsets:
isOnBlackList = False
for item in blacklist:
if item==dset.getName() or ( item.endswith('*') and dset.getName().startswith(item[:-1]) ):
isOnBlackList = True
if isOnBlackList:
self.Print("Ignoring dataset because of black/whitelist options: '%s' ..." % dset.getName(), True)
else:
self.addDataset(dset.getName(), dset.getFileNames(), dataVersion=dset.getDataVersion(), lumiFile=dsetMgrCreator.getLumiFile())
return
def main(opts):
# Suppress warnings about weight being re-applied
ROOT.gErrorIgnoreLevel = ROOT.kError
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
else:
pass
optModes = optList
else:
optModes = [opts.optMode]
opts.optMode = ""
mcrabName = opts.mcrab
RunEra = mcrabName.split("_")[1]
# Setup ROOT and style
ROOT.gROOT.SetBatch(opts.batchMode)
style = tdrstyle.TDRStyle()
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
if opts.paper:
style.setGridX(False)
style.setGridY(False)
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
# Remove some QCD samples (the cross sections of these samples are not calculated)
if 0:
msg = "Removing following samples:"
Print(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(), True)
for d in getDatasetsToExclude():
Print(d, False)
datasetsMgr.remove(d)
# Get run-range
minRunRange, maxRunRange, runRange = GetRunRange(datasetsMgr)
# Get int lumi
intLumi = GetLumi(datasetsMgr)
# Update to PU & load luminosities
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities()
#datasetsMgr.normalizeMCByLuminosity()
# Print luminisoties and cross-sections
datasetsMgr.PrintLuminosities()
datasetsMgr.PrintCrossSections()
# Default merging & ordering: "Data", "QCD", "SingleTop", "Diboson"
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get datasets
datasetsMgr.mergeMC()
dataset_Data = datasetsMgr.getDataDatasets()
dataset_MC = datasetsMgr.getMCDatasets()
# Define lists of Triggers to be plotted and Variables
xVars = ["pt6thJet", "eta6thJet", "phi6thJet", "Ht", "nBTagJets", "pu", "JetMulti", "BJetMulti"]
trgList = ["1BTag", "2BTag", "OR", "OR_PFJet450"]
if opts.fast:
trgList = ["OR_PFJet450"]
xVars = ["pt6thJet", "Ht"]
nPlots = len(trgList)*len(xVars)
counter = 0
# For-loop: All signal triggers
for i, trg in enumerate(trgList, 1):
# For-loop: All x-variables
for j, xVar in enumerate(xVars, 1):
counter+=1
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % counter, "/", "%s:" % (nPlots), "%s Vs %s" % (trg, xVar))
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), counter==1)
# Define names
hNumerator = "hNum_%s_RefTrg_OfflineSel_Signal%s" % (xVar, trg)
hDenominator = "hDen_%s_RefTrg_OfflineSel" % (xVar)
plotName = "Eff_%s_%s" % (xVar, trg)
# Get Efficiency Plots
_kwargs = GetHistoKwargs(xVar, opts)
eff_Data = GetEfficiency(datasetsMgr, dataset_Data, hNumerator, hDenominator , **_kwargs)
eff_MC = GetEfficiency(datasetsMgr, dataset_MC, hNumerator, hDenominator, **_kwargs)
# Apply Styles
styles.dataStyle.apply(eff_Data)
styles.mcStyle.apply(eff_MC)
# Create the plot
p = plots.ComparisonPlot(histograms.HistoGraph(eff_Data, "eff_Data", "p", "P"),
histograms.HistoGraph(eff_MC, "eff_MC" , "p", "P"),
saveFormats=[])
# Define the legend entries
p.histoMgr.setHistoLegendLabelMany(
{
"eff_Data": "Data",
"eff_MC" : "Simulation"
}
)
# Draw and save the plot
p.setLuminosity(intLumi)
plots.drawPlot(p, plotName, **_kwargs)
# Draw
histograms.addText(0.65, 0.06, "Runs "+ runRange, 17)
histograms.addText(0.65, 0.10, "2016", 17)
# Save the canvas to a file
SavePlot(p, plotName, os.path.join(opts.saveDir, opts.optMode), saveFormats=[".pdf", ".png", ".C"] )
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(True)
style.setGridY(True)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
if len(optList) < 1:
optList.append("")
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
datasetsMgr = GetDatasetsFromDir(opts)
datasetsMgr.updateNAllEventsToPUWeighted()
datasetsMgr.loadLuminosities() # from lumi.json
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Get the PSets:
if 0:
datasetsMgr.printSelections()
#PrintPSet("BJetSelection", datasetsMgr, depth=150)
# ZJets and DYJets overlap!
if "ZJetsToQQ_HT600toInf" in datasetsMgr.getAllDatasetNames() and "DYJetsToQQ_HT180" in datasetsMgr.getAllDatasetNames():
Print("Cannot use both ZJetsToQQ and DYJetsToQQ due to duplicate events? Investigate. Removing ZJetsToQQ datasets for now ..", True)
datasetsMgr.remove(filter(lambda name: "ZJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Get luminosity if a value is not specified
if opts.intLumi < 0:
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Remove datasets
removeList = ["QCD-b", "Charged"]
if not opts.useMC:
removeList.append("QCD")
for i, d in enumerate(removeList, 0):
msg = "Removing dataset %s" % d
Verbose(ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(), i==0)
datasetsMgr.remove(filter(lambda name: d in name, datasetsMgr.getAllDatasetNames()))
# Print summary of datasets to be used
if 0:
datasetsMgr.PrintInfo()
# Merge EWK samples
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
# Print dataset information
datasetsMgr.PrintInfo()
# Do the fit on the histo after ALL selections (incl. topology cuts)
folderList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(opts.folder)
#folderList1 = [h for h in folderList if "TetrajetMass" in h]
#folderList1 = [h for h in folderList if "TetrajetBJetPt" in h]
#folderList1 = [h for h in folderList if "TetrajetBJetEta" in h]
folderList1 = [h for h in folderList if opts.histoKey in h]
folderList2 = [h for h in folderList1 if "VR" in h or "SR" in h or "CRone" in h or "CRtwo" in h or "CRthree" in h or "CRfour" in h]
# For-loop: All folders
histoPaths = []
for f in folderList2:
folderPath = os.path.join(opts.folder, f)
histoList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folderPath)
pathList = [os.path.join(folderPath, h) for h in histoList]
histoPaths.extend( pathList )
binLabels = GetBinLabels("CRone", histoPaths)
PlotHistosAndCalculateTF(datasetsMgr, histoPaths, binLabels, opts)
return
def main(opts):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(False)
style.setGridY(False)
# Obtain dsetMgrCreator and register it to module selector
dsetMgrCreator = dataset.readFromMulticrabCfg(directory=opts.mcrab1)
# Get list of eras, modes, and optimisation modes
erasList = dsetMgrCreator.getDataEras()
modesList = dsetMgrCreator.getSearchModes()
optList = dsetMgrCreator.getOptimizationModes()
sysVarList = dsetMgrCreator.getSystematicVariations()
sysVarSrcList = dsetMgrCreator.getSystematicVariationSources()
# If user does not define optimisation mode do all of them
if opts.optMode == None:
optModes = optList
else:
optModes = [opts.optMode]
# For-loop: All optimisation modes
for opt in optModes:
opts.optMode = opt
# Get the datasets from the directory
datasetsMgr1 = GetDatasetsFromDir(opts.mcrab1, opts)
datasetsMgr2 = GetDatasetsFromDir(opts.mcrab2, opts)
datasetsMgr3 = GetDatasetsFromDir(opts.mcrab3, opts)
# Setup the dataset managers
datasetsMgr1.updateNAllEventsToPUWeighted()
datasetsMgr1.loadLuminosities() # from lumi.json
datasetsMgr2.updateNAllEventsToPUWeighted()
datasetsMgr2.loadLuminosities() # from lumi.json
datasetsMgr3.updateNAllEventsToPUWeighted()
datasetsMgr3.loadLuminosities() # from lumi.json
# Print dataset info?
if opts.verbose:
datasetsMgr1.PrintCrossSections()
datasetsMgr1.PrintLuminosities()
datasetsMgr2.PrintCrossSections()
datasetsMgr2.PrintLuminosities()
datasetsMgr2.PrintCrossSections()
datasetsMgr2.PrintLuminosities()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr1)
plots.mergeRenameReorderForDataMC(datasetsMgr2)
plots.mergeRenameReorderForDataMC(datasetsMgr3)
# Get Luminosity
lumi1 = datasetsMgr1.getDataset("Data").getLuminosity()
lumi2 = datasetsMgr2.getDataset("Data").getLuminosity()
lumi3 = datasetsMgr3.getDataset("Data").getLuminosity()
if lumi1 != lumi2 != lumi3:
raise Exception("Lumi1 (=%.2f) != Lumi2 (=%.2f) != Lumi3 (=%.2f" % (lumi1, lumi2, lumi3))
else:
opts.intLumi = datasetsMgr1.getDataset("Data").getLuminosity()
# Merge EWK samples
datasetsMgr1.merge("EWK", aux.GetListOfEwkDatasets())
datasetsMgr2.merge("EWK", aux.GetListOfEwkDatasets())
datasetsMgr3.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr1.PrintInfo()
if 0:
datasetsMgr2.PrintInfo()
datasetsMgr3.PrintInfo()
# Get all the histograms and their paths (e.g. ForFakeBMeasurement/Baseline_DeltaRLdgTrijetBJetTetrajetBJet_AfterCRSelections)
hList = datasetsMgr1.getDataset(datasetsMgr1.getAllDatasetNames()[0]).getDirectoryContent(opts.folder)
hPaths = [os.path.join(opts.folder, h) for h in hList]
# Create a smaller list with only histos of interest
hListS = []
for h in hList:
if "StandardSelections" in h:
continue
if "IsGenuineB" in h:
continue
if "_Bjet" in h:
continue
if "_Jet" in h:
continue
if "_SubLdg" in h:
continue
if "_Njets" in h:
continue
if "_NBjets" in h:
continue
if "_Delta" in h:
continue
if "Dijet" in h:
continue
if "Bdisc" in h:
continue
#if "MVA" in h:
# continue
if "MET" in h:
continue
if "HT" in h:
continue
# Otherwise keep the histogram
hListS.append(h)
hPathsS = [os.path.join(opts.folder, h) for h in hListS]
# Create two lists of paths: one for "Baseline" (SR) and one for "Inverted" (CR)
path_SR = [] # baseline, _AfterAllSelections
path_CR1 = [] # baseline, _AfterCRSelections
path_VR = [] # inverted, _AfterAllSelections
path_CR2 = [] # inverted, _AfterCRSelections
# For-loop: All histogram paths
for p in hPathsS: #hPaths:
if "Baseline" in p:
if "AllSelections" in p:
path_SR.append(p)
if "CRSelections" in p:
path_CR1.append(p)
if "Inverted" in p:
if "AllSelections" in p:
path_VR.append(p)
if "CRSelections" in p:
path_CR2.append(p)
counter = 1
# For-loop: All histogram pairs
for hCR1, hCR2 in zip(path_CR1, path_CR2):
if "IsGenuineB" in hCR1:
continue
#hName = hCR1.replace("_AfterCRSelections", "_CR1vCR2").replace("ForFakeBMeasurement/Baseline_", "")
hName = hCR1.replace("_AfterCRSelections", " (CR1 and R2)").replace("ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)), hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), counter==1)
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hCR1, hCR2, "CR1")
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hCR1, hCR2, "CR2") #iro
counter+=1
# WARNING! This unblinds the Signal Region (SR)
for hSR, hVR in zip(path_SR, path_VR):
if "IsGenuineB" in hSR:
continue
if 1:
continue
#hName = hCR1.replace("_AfterCRSelections", "_SRvVR").replace("ForFakeBMeasurement/Baseline_", "")
hName = hCR1.replace("_AfterCRSelections", " (SR and VR)").replace("ForFakeBMeasurement/Baseline_", "")
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % counter, "/", "%s:" % (len(path_CR1)), hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), counter==1)
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hSR, hVR, "SR")
PlotComparison(datasetsMgr1, datasetsMgr2, datasetsMgr3, hSR, hVR, "VR")
counter+=1
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
