def PlotHistogram(dsetMgr, histoName, opts):
# Get kistogram argumetns
kwargs = GetHistoKwargs(histoName, opts)
saveName = histoName.replace(opts.folder + "/", "")
# Create the plotting object (Data, "FakeB")
p1 = plots.DataMCPlot(dsetMgr, histoName, saveFormats=[])
# Copy dataset manager before changing datasets. Keep only EWK (GenuineB) datasets
datasetMgr = dsetMgr.deepCopy()
datasetMgr.selectAndReorder(aux.GetListOfEwkDatasets())
# Create the MCPlot for the EWKGenuineB histograms
if opts.useMC:
p2 = plots.MCPlot(datasetMgr, histoName, normalizeToLumi=opts.intLumi, saveFormats=[])
else:
histoNameGenuineB = histoName.replace(opts.folder, opts.folder + "EWKGenuineB")
p2 = plots.MCPlot(datasetMgr, histoNameGenuineB, normalizeToLumi=opts.intLumi, saveFormats=[])
# Add the datasets to be included in the plot
myStackList = []
# Data-driven FakeB background
if not opts.useMC:
hFakeB = p1.histoMgr.getHisto("FakeB").getRootHisto()
hhFakeB = histograms.Histo(hFakeB, "FakeB", legendLabel="Fake-b")
hhFakeB.setIsDataMC(isData=False, isMC=True)
myStackList.append(hhFakeB)
else:
hQCD = p1.histoMgr.getHisto("QCD").getRootHisto()
hhQCD = histograms.Histo(hQCD, "QCD", legendLabel="QCD")
hhQCD.setIsDataMC(isData=False, isMC=True)
myStackList.append(hhQCD)
# EWK GenuineB background (Replace all EWK histos with GenuineB histos)
ewkHistoList = []
# For-loop: All EWK datasets
for dataset in aux.GetListOfEwkDatasets():
h = p2.histoMgr.getHisto(dataset).getRootHisto()
hh = histograms.Histo(h, dataset, plots._legendLabels[dataset])
hh.setIsDataMC(isData=False, isMC=True)
myStackList.append(hh)
# Collision data
hData = p1.histoMgr.getHisto("Data").getRootHisto()
hhData = histograms.Histo(hData, "Data")
hhData.setIsDataMC(isData=True, isMC=False)
myStackList.insert(0, hhData)
# Signal
hSignal = p1.histoMgr.getHisto(opts.signal).getRootHisto()
hhSignal = histograms.Histo(hSignal, opts.signal, plots._legendLabels[opts.signal])
hhSignal.setIsDataMC(isData=False, isMC=True)
myStackList.insert(1, hhSignal)
# Create the final plot by passing the histogram list
p3 = plots.DataMCPlot2(myStackList, saveFormats=[])
p3.setLuminosity(opts.intLumi)
p3.setDefaultStyles()
# Apply blinding of data in Signal Region (After creating the plot)
if "blindingRangeString" in kwargs:
startBlind = float(kwargs["blindingRangeString"].split("-")[1])
endBlind = float(kwargs["blindingRangeString"].split("-")[0])
plots.partiallyBlind(p3, maxShownValue=startBlind, minShownValue=endBlind, invert=True, moveBlindedText=kwargs["moveBlindedText"])
# Draw and save the plot
plots.drawPlot(p3, saveName, **kwargs)
SavePlot(p3, saveName, os.path.join(opts.saveDir, opts.optMode), saveFormats = [".png", ".pdf"])
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):
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setLogX(opts.logX)
# If you want BOTH pads (main and ratio) in log scale
if 0:
style.setLogY(opts.logY)
# Overwrite default legends
plots._legendLabels["MCStatError"] = "Bkg. stat."
plots._legendLabels["MCStatSystError"] = "Bkg. stat.#oplussyst."
plots._legendLabels["BackgroundStatError"] = "Bkg. stat. unc"
plots._legendLabels["BackgroundStatSystError"] = "Bkg. stat.#oplussyst. unc."
# Define optimisatio modes to run on
optModes = [""] #["", "OptChiSqrCutValue50", "OptChiSqrCutValue100"]
if opts.optMode != None:
optModes = [opts.optMode]
# Inform user of EWK datasets used
Print("The EWK datasets used are the following:", True)
for i,d in enumerate(aux.GetListOfEwkDatasets(), 1):
Print(ShellStyles.NoteStyle() + d + ShellStyles.NormalStyle(), i==0)
# For-loop: All opt Mode
for opt in optModes:
opts.optMode = opt
# Setup & configure the dataset manager
dsetMgr1 = GetDatasetsFromDir(opts, False)
dsetMgr2 = GetDatasetsFromDir(opts, True)
# Setup the dataset managers
dsetMgr1.updateNAllEventsToPUWeighted()
dsetMgr2.updateNAllEventsToPUWeighted()
# Load luminosities
dsetMgr1.loadLuminosities() # from lumi.json
# dsetMgr2.loadLuminosities()
# Print PSets. Perhaps i can use this to ensure parameters are matching!
if 0:
dsetMgr1.printSelections()
dsetMgr2.printSelections()
PrintPSet("FakeBMeasurement", dsetMgr1)
PrintPSet("TopSelectionBDT" , dsetMgr2)
# Remove datasets with overlap?
removeList = ["QCD-b"]
dsetDY = "DYJetsToQQ_HT180"
dsetZJ = "ZJetsToQQ_HT600toInf"
dsetRM = dsetZJ # datasets with overlap
removeList.append(dsetRM)
# Set/Overwrite cross-sections. Remove all but 1 signal mass
for d in dsetMgr1.getAllDatasets():
if "ChargedHiggs" in d.getName():
dsetMgr1.getDataset(d.getName()).setCrossSection(1.0) # ATLAS 13 TeV H->tb exclusion limits
if d.getName() != opts.signal:
removeList.append(d.getName())
# Print useful information?
if opts.verbose:
dsetMgr1.PrintCrossSections()
dsetMgr1.PrintLuminosities()
dsetMgr2.PrintCrossSections()
dsetMgr2.PrintLuminosities()
# Merge histograms
plots.mergeRenameReorderForDataMC(dsetMgr1)
# Get the luminosity
if opts.intLumi < 0:
opts.intLumi = dsetMgr1.getDataset("Data").getLuminosity()
# Custom Filtering of datasets
for i, d in enumerate(removeList, 1):
msg = "Removing datasets %s from dataset manager" % (ShellStyles.NoteStyle() + d + ShellStyles.NormalStyle())
Verbose(msg, i==1)
dsetMgr1.remove(filter(lambda name: d == name, dsetMgr1.getAllDatasetNames()))
# Print dataset information
dsetMgr1.PrintInfo()
dsetMgr2.PrintInfo()
# Replace MC datasets with data-driven
if not opts.useMC:
replaceQCD(dsetMgr1, dsetMgr2, "FakeBMeasurementTrijetMass", "FakeB") #dsetMgr1 now contains "FakeB" pseudo-dataset
# Definitions
allHistos = dsetMgr2.getAllDatasets()[0].getDirectoryContent(opts.folder)
histoPaths = []
ignoreKeys = ["MCEWK", "Purity", "BJetPt", "BJetEta", "BtagDiscriminator", "METPhi", "MHT", "NBjets", "Njets", "_Vs_", "JetEta"]
# For-loop: All histograms in directory
for h in allHistos:
bKeep = True
for k in ignoreKeys:
if k in h:
bKeep = False
continue
if bKeep:
histoPaths.append(os.path.join(opts.folder, h))
# For-loop: All histograms in list
for i, hName in enumerate(histoPaths, 1):
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (len(histoPaths)), hName)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistogram(dsetMgr1, hName, opts)
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 "_" + 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.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):
optModes = [""]
if opts.optMode != None:
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(fname="lumi.json")
# Get Luminosity
if opts.intLumi < 0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = datasetsMgr.loadLumi()
# Set/Overwrite cross-sections
datasetsToRemove = []
for d in datasetsMgr.getAllDatasets():
mass = "M_%s" % (opts.signalMass)
if mass in d.getName():
if ("%s" % opts.signalMass) != d.getName().split("M_")[-1]:
datasetsMgr.remove(d.getName())
else:
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
else:
#datasetsToRemove.append(d.getName())
datasetsMgr.remove(d.getName())
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# # Custom Filtering of datasets
# for i, d in enumerate(datasetsToRemove, 0):
# msg = "Removing dataset %s" % d
# Verbose(ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(), i==0)
# datasetsMgr.remove(filter(lambda name: d == name, datasetsMgr.getAllDatasetNames()))
if opts.verbose:
datasetsMgr.PrintInfo()
# Merge EWK samples
if opts.mergeEWK:
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
# Do Data-MC histograms with DataDriven QCD
folder = opts.folder
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths = [os.path.join(folder, h) for h in histoList]
keepList = ["LdgTetrajetMass_AfterAllSelections"]
#keepList = ["LdgTetrajetMass_AfterStandardSelections"]
myHistos = []
for h in histoPaths:
if h.split("/")[-1] not in keepList:
continue
else:
myHistos.append(h)
for i, h in enumerate(myHistos, 1):
PlotHistograms(datasetsMgr, h)
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(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
def main(opts):
#optModes = ["", "OptChiSqrCutValue50", "OptChiSqrCutValue100"]
optModes = [""]
if opts.optMode != None:
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
datasetsMgr_matched = GetDatasetsFromDir(opts)
datasetsMgr_matched.updateNAllEventsToPUWeighted()
datasetsMgr_matched.loadLuminosities() # from lumi.json
plots.mergeRenameReorderForDataMC(datasetsMgr)
datasetsMgr.remove(filter(lambda name: "QCD_b" in name, datasetsMgr.getAllDatasetNames())) #soti
datasetsMgr_matched.remove(filter(lambda name: "QCD" in name, datasetsMgr_matched.getAllDatasetNames())) #soti
# Set/Overwrite cross-sections
datasetsToRemove = ["QCD-b"]#, "QCD_HT50to100", "QCD_HT100to200"]#, "QCD_HT200to300"]#, "QCD_HT300to500"]
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0) # ATLAS 13 TeV H->tb exclusion limits
#if d.getName() != opts.signal:
if "M_650" in d.getName(): #soti fixmi
datasetsToRemove.append(d.getName())
if "M_800" in d.getName():
datasetsToRemove.append(d.getName())
if "M_200" in d.getName():
datasetsToRemove.append(d.getName())
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Custom Filtering of datasets
for i, d in enumerate(datasetsToRemove, 0):
msg = "Removing dataset %s" % d
Print(ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(), i==0)
datasetsMgr.remove(filter(lambda name: d in name, datasetsMgr.getAllDatasetNames()))
datasetsMgr_matched.remove(filter(lambda name: d in name, datasetsMgr_matched.getAllDatasetNames())) #soti
if opts.verbose:
datasetsMgr.PrintInfo()
# 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()))
#datasetsMgr.merge("QCD", GetListOfQCDatasets())
#plots._plotStyles["QCD"] = styles.getQCDLineStyle()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
# Get Luminosity
#intLumi = datasetsMgr.getDataset("Data").getLuminosity() Soti
intLumi = 35920
# Re-order datasets (different for inverted than default=baseline)
newOrder = []
# For-loop: All MC datasets
for d in datasetsMgr.getMCDatasets():
newOrder.append(d.getName())
# Move signal to top
# if opts.signal in newOrder:
# s = newOrder.pop( newOrder.index(opts.signal) )
# newOrder.insert(0, s)
print len(newOrder), "newOrder"
signalMass = ["M_300", "M_500", "M_1000"]
for d in datasetsMgr.getMCDatasets():
for m in signalMass:
if m in d.getName():
s = newOrder.pop( newOrder.index(d.getName()) )
newOrder.insert(0, s)
#datasetsMgr.selectAndReorder(newOrder)
print len(newOrder), "newOrder"
# Add Data to list of samples!
if not opts.onlyMC:
newOrder.insert(0, "Data")
# Apply new dataset order!
datasetsMgr.selectAndReorder(newOrder)
# Merge EWK samples
if opts.mergeEWK:
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
# Do Data-MC histograms with DataDriven QCD
folder = opts.folder
histoList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths1 = [os.path.join(folder, h) for h in histoList]
histoPaths2 = [h for h in histoPaths1]# if "jet" not in h.lower()]
nHistos = len(histoPaths2)
# For-loop: All histograms
for i, h in enumerate(histoPaths2, 1):
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (nHistos), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistograms(datasetsMgr, datasetsMgr_matched, h, intLumi)
ROOT.gStyle.SetNdivisions(10, "X")
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(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
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):
#optModes = ["", "OptChiSqrCutValue50", "OptChiSqrCutValue100"]
optModes = [""]
if opts.optMode != None:
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 0:
datasetsMgr.printSelections()
sys.exit()
# Define datasets to remove by default
QCD_list = ["QCD_HT700to1000", "QCD_HT50to100", "QCD_HT500to700", "QCD_HT300to500",
"QCD_HT200to300", "QCD_HT2000toInf", "QCD_HT1500to2000", "QCD_HT100to200", "QCD_HT1000to1500"]
QCDExt_list = [x+"_ext1" for x in QCD_list]
datasetsToRemove = ["QCD-b"]
# datasetsToRemove.extend(QCD_list)
# datasetsToRemove.extend(QCDExt_list)
# 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()))
#datasetsMgr.remove(filter(lambda name: "DYJetsToQQ" in name, datasetsMgr.getAllDatasetNames()))
# Set/Overwrite cross-sections
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0) # ATLAS 13 TeV H->tb exclusion limits
if d.getName() != opts.signal:
datasetsToRemove.append(d.getName())
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Custom Filtering of datasets
for i, d in enumerate(datasetsToRemove, 0):
msg = "Removing dataset %s" % d
Verbose(ShellStyles.WarningLabel() + msg + ShellStyles.NormalStyle(), i==0)
datasetsMgr.remove(filter(lambda name: d == name, datasetsMgr.getAllDatasetNames()))
if opts.verbose:
datasetsMgr.PrintInfo()
# Re-order datasets (different for inverted than default=baseline)
newOrder = ["Data"]
for i, d in enumerate(datasetsMgr.getAllDatasets(), 0):
if d.isData():
continue
else:
newOrder.append(d.getName())
# Re-arrange dataset order?
if 0:
s = newOrder.pop( newOrder.index("noTop") )
newOrder.insert(len(newOrder), s) #after "Data"
# Move signal to top
if opts.signal in newOrder:
s = newOrder.pop( newOrder.index(opts.signal) )
newOrder.insert(1, s)
datasetsMgr.selectAndReorder(newOrder)
# Merge EWK samples
if opts.mergeEWK:
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
datasetsMgr.PrintInfo()
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(True)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
# Do Data-MC histograms with DataDriven QCD
folder = opts.folder
histoList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths = [os.path.join(folder, h) for h in histoList]
ignoreList = ["Aplanarity", "Planarity", "Sphericity", "FoxWolframMoment", "Circularity", "ThirdJetResolution", "Centrality", "_Vs_"]
myHistos = []
for h in histoPaths:
skip = False
# Skip unwanted histos
for i in ignoreList:
if i in h:
skip = True
if skip:
continue
else:
myHistos.append(h)
for i, h in enumerate(myHistos, 1):
# Plot the histograms!
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (len(myHistos)), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
DataMCHistograms(datasetsMgr, h)
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("")
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):
optModes = [""]
if opts.optMode != None:
optModes = [opts.optMode]
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setLogX(opts.logX)
style.setLogY(opts.logY)
style.setLogZ(opts.logZ)
# 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()
datasetsMgr.PrintInfo()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Print merged datasets and MC samples
if 0:
datasetsMgr.PrintInfo()
# Get Luminosity
if opts.intLumi < 0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = 1.0
# Set/Overwrite cross-sections. Remove all but 1 signal mass
removeList = []
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if d.getName() != opts.signal:
removeList.append(d.getName())
# Custom filtering of datasets
for i, d in enumerate(removeList, 1):
msg = "Removing datasets %s from dataset manager" % (
ShellStyles.NoteStyle() + d + ShellStyles.NormalStyle())
Verbose(msg, i == 1)
datasetsMgr.remove(
filter(lambda name: d == name,
datasetsMgr.getAllDatasetNames()))
# Merge EWK samples
if opts.mergeEWK == "EWK":
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
if 1:
datasetsMgr.PrintInfo()
# Get list of histogram paths
folder = opts.folder
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths = [os.path.join(folder, h) for h in histoList]
# For-loop: All histograms
for i, h in enumerate(histoPaths, 1):
if "_Vs_" in h:
continue
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format(
"Histogram", "%i" % i, "/", "%s:" % (len(histoPaths)), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
i == 1)
PlotHistograms(datasetsMgr, h)
break
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
def main(opts):
optModes = [""]
if opts.optMode != None:
optModes = [opts.optMode]
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setLogX(opts.logX)
style.setLogY(opts.logY)
style.setLogZ(opts.logZ)
style.setWide(True, 0.15)
# style.setPadRightMargin()#0.13)
# 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()
datasetsMgr.PrintInfo()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Print merged datasets and MC samples
if 0:
datasetsMgr.PrintInfo()
# Get Luminosity
if opts.intLumi < 0:
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
# Merge EWK samples
if opts.dataset == "EWK":
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Re-order datasets (different for inverted than default=baseline)
newOrder = []
for d in datasetsMgr.getAllDatasets():
if d.getName() == opts.dataset:
newOrder.append(d.getName())
# Sanity check on selected dataset
nDatasets = len(newOrder)
if nDatasets < 1:
msg = "Please select a valid dataset. Dataset \"%s\" does not exist!" % (
opts.dataset)
Print(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(),
True)
datasetsMgr.PrintInfo()
sys.exit()
if nDatasets > 1:
msg = "Please select only 1 valid dataset. Requested %i datasets for plotting!" % (
nDatasets)
Print(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(),
True)
datasetsMgr.PrintInfo()
sys.exit()
# Select only given dataset
datasetsMgr.selectAndReorder(newOrder)
# Print dataset information
msg = "Plotting for single dataset \"%s\". Integrated luminosity is %.2f 1/fb" % (
opts.dataset, opts.intLumi)
Print(ShellStyles.NoteStyle() + msg + ShellStyles.NormalStyle(), True)
datasetsMgr.PrintInfo()
# Get list of histogram paths
folder = opts.folder
histoList = datasetsMgr.getDataset(
datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths = [os.path.join(folder, h) for h in histoList]
# For-loop: All histograms
for h in histoPaths:
if "_vs_" not in h.lower():
continue
Plot2dHistograms(datasetsMgr, h)
Print(
"All plots saved under directory %s" %
(ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) +
ShellStyles.NormalStyle()), True)
return
def main(opts):
optModes = [""]
if opts.optMode != None:
optModes = [opts.optMode]
# Apply TDR style
global style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(opts.gridX)
style.setGridY(opts.gridY)
style.setLogX(opts.logX)
style.setLogY(opts.logY)
# 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()
datasetsMgr.PrintInfo()
# Merge histograms (see NtupleAnalysis/python/tools/plots.py)
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Print merged datasets and MC samples
if 0:
datasetsMgr.PrintInfo()
# Get Luminosity
if opts.intLumi == 0.0:
if "Data" in datasetsMgr.getAllDatasetNames():
opts.intLumi = datasetsMgr.getDataset("Data").getLuminosity()
else:
opts.intLumi = GetLumi(datasetsMgr)
# Set/Overwrite cross-sections. Remove all but 1 signal mass
signalsList = []
for s in opts.signalList:
signalsList.append("ChargedHiggs_HplusTB_HplusToTB_M_%s" % s)
# Remove datasets with overlap?
removeList = ["QCD-b"]
dsetDY = "DYJetsToQQ_HT180"
dsetZJ = "ZJetsToQQ_HT600toInf"
dsetRM = dsetZJ # datasets with overlap
removeList.append(dsetRM)
for d in datasetsMgr.getAllDatasets():
if "ChargedHiggs" in d.getName():
datasetsMgr.getDataset(d.getName()).setCrossSection(1.0)
if d.getName() not in signalsList:
removeList.append(d.getName())
# Custom filtering of datasets
for i, d in enumerate(removeList, 1):
msg = "Removing datasets %s from dataset manager" % (ShellStyles.NoteStyle() + d + ShellStyles.NormalStyle())
Verbose(msg, i==1)
datasetsMgr.remove(filter(lambda name: d == name, datasetsMgr.getAllDatasetNames()))
# Merge EWK samples
if opts.mergeEWK == "EWK":
datasetsMgr.merge("EWK", aux.GetListOfEwkDatasets())
plots._plotStyles["EWK"] = styles.getAltEWKStyle()
# Print dataset information
if 1:
datasetsMgr.PrintInfo()
# Get list of histogram paths
folder = opts.folder
histoList = datasetsMgr.getDataset(datasetsMgr.getAllDatasetNames()[0]).getDirectoryContent(folder)
histoPaths = [os.path.join(folder, h) for h in histoList if "AllSelections" in h]
myHistos = []
skipStrings = ["StandardSelections", "JetPt", "JetEta", "JetEtaPhi", "BJetPt", "BJetEta",
"BtagDiscriminator", "Eta", "LdgTrijetTopMassWMassRatio", "Subldg", "Dijet",
"MET", "MHT", "MinDeltaPhiJetMHT", "MaxDeltaPhiJetMHT", "MinDeltaRJetMHT", "MinDeltaRJetMHTReversed"]
for h in histoPaths:
bSkip = False
for s in skipStrings:
if s in h:
bSkip = True
break
if bSkip:
continue
else:
myHistos.append(h)
# For-loop: All histograms
for i, h in enumerate(myHistos, 1):
msg = "{:<9} {:>3} {:<1} {:<3} {:<50}".format("Histogram", "%i" % i, "/", "%s:" % (len(myHistos)), h)
Print(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), i==1)
PlotHistograms(datasetsMgr, h, signalsList, opts.cutDir)
Print("All plots saved under directory %s" % (ShellStyles.NoteStyle() + aux.convertToURL(opts.saveDir, opts.url) + ShellStyles.NormalStyle()), True)
return
