def doCounters(datasets):
eventCounter = counter.EventCounter(datasets)
ewkDatasets = [
"WJets", "TTJets",
# "WJets",
"DYJetsToLL", "SingleTop", "Diboson"
]
# if mcOnly:
# eventCounter.normalizeMCToLuminosity(mcOnlyLumi)
# else:
eventCounter.normalizeMCByLuminosity()
print "============================================================"
print "Main counter (MC normalized by collision data luminosity)"
mainTable = eventCounter.getMainCounterTable()
# mainTable.insertColumn(2, counter.sumColumn("EWKMCsum", [mainTable.getColumn(name=name) for name in ewkDatasets]))
# Default
# cellFormat = counter.TableFormatText()
# No uncertainties
cellFormat = counter.TableFormatText(cellFormat=counter.CellFormatText(valueOnly=True))
print mainTable.format(cellFormat)
print eventCounter.getSubCounterTable("TauSelection").format(cellFormat)
print eventCounter.getSubCounterTable("e selection").format(cellFormat)
print eventCounter.getSubCounterTable("mu selection").format(cellFormat)
print eventCounter.getSubCounterTable("jet selection").format(cellFormat)
print eventCounter.getSubCounterTable("angular cuts / Collinear").format(cellFormat)
print eventCounter.getSubCounterTable("bjet selection").format(cellFormat)
print eventCounter.getSubCounterTable("angular cuts / BackToBack").format(cellFormat)
def doCounters(datasets):
eventCounter = counter.EventCounter(datasets)
eventCounter.normalizeMCByLuminosity()
# eventCounter.normalizeMCToLuminosity(73)
print "============================================================"
print "Main counter (MC normalized by collision data luminosity)"
mainTable = eventCounter.getMainCounterTable()
# No uncertainties
cellFormat = counter.TableFormatText(cellFormat=counter.CellFormatText(
valueOnly=True))
print mainTable.format(cellFormat)
print eventCounter.getSubCounterTable("b-tagging").format(cellFormat)
print eventCounter.getSubCounterTable("Jet selection").format(cellFormat)
print eventCounter.getSubCounterTable("Jet main").format(cellFormat)
def main(opts):
datasets = None
if len(opts.files) > 0:
datasets = dataset.getDatasetsFromRootFiles( [(x,x) for x in opts.files], opts=opts, weightedCounters=opts.weighted)
else:
datasets = dataset.getDatasetsFromMulticrabCfg(opts=opts, weightedCounters=opts.weighted)
if os.path.exists(opts.lumifile):
datasets.loadLuminosities(opts.lumifile)
if opts.weighted and opts.PUreweight:
datasets.updateNAllEventsToPUWeighted(era=opts.dataEra)
if opts.mergeData:
datasets.mergeData()
eventCounter = counter.EventCounter(datasets)
print "============================================================"
if opts.printInfo:
print "Dataset info: "
datasets.printInfo()
print
quantity = "events"
if opts.mode == "events":
pass
elif opts.mode in ["xsect", "xsection", "crosssection", "crossSection", "eff"]:
if not opts.PUreweight:
print "Mode '%s' works only with PU reweighting, which you disabled with --noPUreweight" % opts.mode
return 1
eventCounter.normalizeMCByCrossSection()
quantity = "MC by cross section, data by events"
else:
print "Printing mode '%s' doesn't exist! The following ones are available 'events', 'xsect', 'eff'" % opts.mode
return 1
cellFormat = counter.CellFormatText(valueOnly=opts.valueOnly, valueFormat=opts.format)
formatFunc = lambda table: table.format(counter.TableFormatText(cellFormat))
csvSplitter = counter.TableSplitter([" +- ", " +", " -"])
if opts.csv:
formatFunc = lambda table: table.format(counter.TableFormatText(cellFormat, columnSeparator=","), csvSplitter)
if opts.mode == "eff":
cellFormat = counter.CellFormatText(valueFormat="%.4f", valueOnly=opts.valueOnly)
formatFunc = lambda table: counter.counterEfficiency(table).format(counter.TableFormatText(cellFormat))
quantity = "Cut efficiencies"
if opts.csv:
formatFunc = lambda table: counter.counterEfficiency(table).format(counter.TableFormatText(cellFormat, columnSeparator=","), csvSplitter)
if opts.subCounter is not None:
print "============================================================"
print "Subcounter %s %s: " % (opts.subCounter, quantity)
print formatFunc(eventCounter.getSubCounterTable(opts.subCounter))
print
return 0
print "============================================================"
print "Main counter %s: " % quantity
print formatFunc(eventCounter.getMainCounterTable())
print
if not opts.mainCounterOnly:
names = eventCounter.getSubCounterNames()
names.sort()
for name in names:
print "============================================================"
print "Subcounter %s %s: " % (name, quantity)
print formatFunc(eventCounter.getSubCounterTable(name))
print
# print
return 0
def doCounters(datasetsMgr):
'''
Print values for counters and sub-counters (if -s option is invoked)
Options:
--latex (The table formatting is in LaTeX instead of plain text)
--format (The table value-format of strings)
--precision (The table value-precision)
--noError (Don't print statistical errors in tables)
--mergeEWK (Merge all EWK samples into a single sample called EWK)
-i (List of datasets in mcrab to include)
-e (List of datasets in mcrab to exclude)
'''
Verbose("Doing the counters")
# Definitions
eventCounter = counter.EventCounter(datasetsMgr)
ewkDatasets = GetListOfEwkDatasets()
# Normalize MC samples accordingly
if opts.mcOnly and opts.intLumi > -1.0:
eventCounter.normalizeMCToLuminosity(opts.intLumi)
else:
eventCounter.normalizeMCByLuminosity()
# eventCounter.normalizeMCByCrossSection()
# Construct the table
mainTable = eventCounter.getMainCounterTable()
# Only keep selected rows
rows = [
#"ttree: skimCounterAll",
#"ttree: skimCounterPassed",
#"Base::AllEvents",
#"Base::PUReweighting",
#"Base::Prescale",
#"Base::Weighted events with top pT",
#"Base::Weighted events for exclusive samples",
"All events",
"Passed trigger",
"passed METFilter selection ()",
"Passed PV",
"passed e selection (Veto)",
"passed mu selection (Veto)",
"Passed tau selection (Veto)",
#"Passed tau selection and genuine (Veto)",
"passed jet selection ()",
#"passed b-jet selection ()",
#"passed light-jet selection ()",
"Baseline: passed b-jet selection",
"Baseline: b tag SF",
#"passed MET selection (Baseline)",
"passed topology selection (Baseline)",
"passed top selection (Baseline)",
"Baseline: selected events",
"Inverted: passed b-jet veto",
"Inverted: b tag SF",
#"passed MET selection (Inverted)",
"passed topology selection (Inverted)",
"passed top selection (Inverted)",
"Inverted: selected events"
]
mainTable.keepOnlyRows(rows)
# Get number of rows/columns
nRows = mainTable.getNrows()
nColumns = mainTable.getNcolumns()
# Merge EWK into a new column
if opts.mergeEWK:
mainTable.insertColumn(
nColumns,
counter.sumColumn(
"EWK",
[mainTable.getColumn(name=name) for name in ewkDatasets]))
# Additional column (through inter-column operations)
if opts.mergeEWK:
mainTable.insertColumn(
mainTable.getNcolumns(),
counter.subtractColumn("Data-EWK",
mainTable.getColumn(name="Data"),
mainTable.getColumn(name="EWK")))
mainTable.insertColumn(
mainTable.getNcolumns(),
counter.divideColumn("QCD Purity",
mainTable.getColumn(name="Data-EWK"),
mainTable.getColumn(name="Data")))
# Convert "QCD Purity" from Fraction to Percent (%)
mainTable.getColumn(name="QCD Purity").multiply(100)
# Define which columns to keep
columnsToKeep = ["Data", "EWK", "QCD Purity"]
if opts.fractionEWK:
columnsToKeep = [] # columnsToKeep = ["Data", "EWK"]
for d in GetListOfEwkDatasets():
columnName = d + "/EWK"
mainTable.insertColumn(
mainTable.getNcolumns(),
counter.divideColumn(columnName, mainTable.getColumn(name=d),
mainTable.getColumn(name="EWK")))
# Convert EWK Fraction to EWK Percent (%)
mainTable.getColumn(name=columnName).multiply(100)
columnsToKeep.append(columnName)
# Remove all columns that are not needed
if opts.mergeEWK:
mainTable.keepOnlyColumns(columnsToKeep)
# Optional: Produce table in Text or LaTeX format?
if opts.latex:
mainTable.renameRows(GetRowNameLaTeXDict())
cellFormat = counter.TableFormatLaTeX(
counter.CellFormatTeX(
valueOnly=opts.valueOnly,
valueFormat=opts.valueFormat,
withPrecision=opts.withPrecision,
uncertaintyFormat=opts.uncertaintyFormat,
uncertaintyPrecision=opts.uncertaintyPrecision))
else:
cellFormat = counter.TableFormatText(cellFormat=counter.CellFormatText(
valueOnly=opts.valueOnly,
valueFormat=opts.valueFormat,
withPrecision=opts.withPrecision,
uncertaintyFormat=opts.uncertaintyFormat,
uncertaintyPrecision=opts.uncertaintyPrecision))
print mainTable.format(cellFormat)
# Do sub-counters?
subcounters = [
"bjet selection ()",
"e selection (Veto)",
"jet selection ()",
#"light-jet selection ()",
"METFilter selection",
"METFilter selection ()",
"mu selection (Veto)",
"tau selection (Veto)",
"top selection (Baseline)",
"top selection (Inverted)",
"topology selection (Baseline)",
"topology selection (Inverted)"
]
if opts.subcounters:
for sc in subcounters:
Print("\nSub-counter \"%s\"" % (sc), False)
print eventCounter.getSubCounterTable(sc).format(cellFormat)
return
def main(opts):
datasetsMgr = None
# Fixme: This options is currently invalid
if len(opts.files) > 0:
datasetsMgr = dataset.getDatasetsFromRootFiles( [(x,x) for x in opts.files], opts=opts, weightedCounters=opts.weighted)
else:
# Define the cwd as the multicrab dir containing the ROOT files with the counters
multicrabDir = os.getcwd()
# For-loop: All arguments passed during script execution
for d in sys.argv:
# Look for a directory that exists
if os.path.exists(d) and os.path.isdir(d):
multicrabDir = os.path.abspath(d)
Print("The multicrab directory to be used is %s" % (multicrabDir), True)
# Get the datasets
datasetsMgr = dataset.getDatasetsFromMulticrabDirs([multicrabDir],opts=opts, weightedCounters=opts.weighted)
# Optional: Apply include/exclude datasets
datasetsMgr = FilterDatasets(datasetsMgr, opts)
Print("The tasks to be included are:\n\t%s" % ("\n\t".join(d.getName() for d in datasetsMgr.getAllDatasets())), True)
# Optional: Print info on Data and MC samples
if opts.verbose:
datasetsMgr.PrintCrossSections()
datasetsMgr.PrintLuminosities()
# Load the luminosities
if os.path.exists(opts.lumifile):
datasetsMgr.loadLuminosities(opts.lumifile)
# Optional: Apply PU-reweighting
if opts.weighted and opts.PUreweight:
Print("Updating all events to PU-weighted (opts.weighted=%s, opts.PUreweight=%s)\n" % (opts.weighted, opts.PUreweight), True)
datasetsMgr.updateNAllEventsToPUWeighted(era=opts.dataEra)
# Optional: Merge data
Verbose("Merging Data and/or MC datatets (opts.mergeData=%s, opts.mergeData=%s, opts.mergeData=%s)" % (opts.mergeData, opts.mergeMC, opts.mergeForDataMC), True)
if opts.mergeData:
datasetsMgr.mergeData()
if opts.mergeMC:
datasetsMgr.mergeMC()
if opts.mergeForDataMC:
plots.mergeRenameReorderForDataMC(datasetsMgr)
# Optional: Print dataset info
if opts.printInfo:
datasetsMgr.PrintInfo() #datasetsMgr.printInfo()
# Create the event counter
eventCounter = counter.EventCounter(datasetsMgr)
# Proceed differently depending on operation mode (opts.mode= 'events', 'xsect', 'eff')
quantity = "events"
if opts.mode == "events":
if opts.mergeForDataMC:
Print("Normalising the MC histograms to the data luminosity (opts.mergeForDataMC=%s)" % (opts.mergeForDataMC) )
eventCounter.normalizeMCByLuminosity()
else:
pass
elif opts.mode in ["xsect", "xsection", "crosssection", "crossSection", "eff"]:
if not opts.PUreweight:
Print("Mode '%s' works only with PU reweighting, which you disabled with --noPUreweight" % opts.mode)
return 1
Print("Normalising MC by cross-section (opt.mode=%s)" % opts.mode)
eventCounter.normalizeMCByCrossSection()
quantity = "MC by cross section, data by events"
else:
Print("Printing mode '%s' doesn't exist! The following ones are available 'events', 'xsect', 'eff'" % opts.mode)
return 1
# Optional: Produce table in Text or LaTeX format?
if opts.latex:
cellFormat = counter.CellFormatTeX(valueOnly=opts.valueOnly, valueFormat=opts.format)
formatFunc = lambda table: table.format(counter.TableFormatLaTeX(cellFormat))
else:
cellFormat = counter.CellFormatText(valueOnly=opts.valueOnly, valueFormat=opts.format)
formatFunc = lambda table: table.format(counter.TableFormatText(cellFormat))
csvSplitter = counter.TableSplitter([" +- ", " +", " -"])
# Optional: Format as comma-separated-variables (csv), presubambly for exporting to a spreadsheet
if opts.csv:
formatFunc = lambda table: table.format(counter.TableFormatText(cellFormat, columnSeparator=","), csvSplitter)
# Optional: Convert to (relative) efficienies
if opts.mode == "eff":
if opts.weighted:
Print("Cannot operate in \"eff\" mode while using weighted counters (opts.mode=\'%s\', opts.weighted=%s)" % (opts.mode, opts.weighted) )
return 1
else:
Print("Converting to efficiencies (opts.mode=%s)" % (opts.mode) )
cellFormat = counter.CellFormatText(valueFormat="%.4f", valueOnly=opts.valueOnly)
if opts.latex:
formatFunc = lambda table: counter.counterEfficiency(table).format(counter.TableFormatLaTeX(cellFormat))
else:
formatFunc = lambda table: counter.counterEfficiency(table).format(counter.TableFormatText(cellFormat))
quantity = "Cut efficiencies"
# Optional: Format as comma-separated-variables (csv), presubambly for exporting to a spreadsheet
if opts.csv:
formatFunc = lambda table: counter.counterEfficiency(table).format(counter.TableFormatText(cellFormat, columnSeparator=","), csvSplitter)
# Optional: Print only this sub-counters
if opts.subCounter is not None:
msg = "Subcounter %s %s: " % (opts.subCounter, quantity)
Print(msg, True)
print formatFunc(eventCounter.getSubCounterTable(opts.subCounter))
print
return 0
# Print the main counters
hLine = "="*10
msg = " Main counter %s: " % quantity
print "\n" + hLine + msg + hLine
print formatFunc(eventCounter.getMainCounterTable())
print
# Optional: Print sub-counters (only if --mainCounterOnly is not called)
if not opts.mainCounterOnly:
names = eventCounter.getSubCounterNames()
names.sort()
for name in names:
hLine = "="*10
msg = " Subcounter %s %s: " % (name, quantity)
print "\n" + hLine + msg + hLine
print formatFunc(eventCounter.getSubCounterTable(name) )
return 0
