class FakeBNormalizationManager:
'''
Base class for QCD measurement normalization from which specialized algorithm classes inherit
'''
def __init__(self,
binLabels,
resultDirName,
moduleInfoString,
verbose=False):
self._verbose = verbose
self._templates = {}
self._binLabels = binLabels
self._sources = {}
self._commentLines = []
self._BinLabelMap = {}
self._TF = {} # Transfer Factor (TF)
self._TF_Error = {}
self._TF_Up = {}
self._TF_Down = {}
self._dqmKeys = OrderedDict()
self._myPath = os.path.join(resultDirName, "normalisationPlots")
if not isinstance(binLabels, list):
raise Exception("Error: binLabels needs to be a list of strings")
self.Verbose("__init__")
# No optimisation mode
if moduleInfoString == "":
moduleInfoString = "Default"
if not os.path.exists(self._myPath):
self.Print("Creating new directory %s" % (self._myPath), True)
os.mkdir(self._myPath)
self._plotDirName = os.path.join(resultDirName, "normalisationPlots",
moduleInfoString)
# If already exists, Delete an entire directory tree
if os.path.exists(self._plotDirName):
msg = "Removing directory tree %s" % (self._plotDirName)
self.Verbose(
ShellStyles.NoteStyle() + msg + ShellStyles.NormalStyle(),
True)
shutil.rmtree(self._plotDirName)
msg = "Creating directory %s" % (self._plotDirName)
self.Verbose(
ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
False)
os.mkdir(self._plotDirName)
return
def Print(self, msg, printHeader=False):
fName = __file__.split("/")[-1]
if printHeader == True:
print "=== ", fName + ": class " + self.__class__.__name__
print "\t", msg
else:
print "\t", msg
return
def Verbose(self, msg, printHeader=True, verbose=False):
if not self._verbose:
return
self.Print(msg, printHeader)
return
def GetQCDNormalization(self, binLabel):
if binLabel in self._TF.keys():
return self._TF[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! " %
(binLabel))
def GetTransferFactor(self, binLabel):
return self.GetQCDNormalization(binLabel)
def GetQCDNormalizationError(self, binLabel):
if binLabel in self._TF_Error.keys():
return self._TF_Error[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! " %
(binLabel))
def CalculateTransferFactor(self,
binLabel,
hFakeB_Baseline,
hFakeB_Inverted,
verbose=False):
'''
Calculates the combined normalization and, if specified,
varies it up or down by factor (1+variation)
TF = Transfer Factor
SR = Signal Region
CR = Control Region
VR = Verification Region
'''
self.verbose = verbose
# Obtain counts for QCD and EWK fakes
lines = []
# NOTES: Add EWKGenuineB TF, Add Data TF, add QCD TF, Add EWK TF, add MCONLY TFs
nSR_Error = ROOT.Double(0.0)
nCR_Error = ROOT.Double(0.0)
# nTotalError = ROOT.TMath.Sqrt(nSRerror**2 + nCRError**2)
nSR = hFakeB_Baseline.IntegralAndError(1,
hFakeB_Baseline.GetNbinsX() + 1,
nSR_Error)
nCR = hFakeB_Inverted.IntegralAndError(1,
hFakeB_Inverted.GetNbinsX() + 1,
nCR_Error)
# nTotal = nSR + nCR
# Calculate Transfer Factor (TF) from Control Region (R) to Signal Region (SR): R = N_CR1/ N_CR2
TF = None
TF_Up = None
TF_Down = None
TF_Error = None
if 1: ## nTotal > 0.0:
TF = nSR / nCR
TF_Error = errorPropagation.errorPropagationForDivision(
nSR, nSR_Error, nCR, nCR_Error)
TF_Up = TF + TF_Error
if TF_Up > 1.0:
TF_Up = 1.0
TF_Down = TF - TF_Error
if TF_Down < 0.0:
TF_Down = 0.0
lines.append("TF (bin=%s) = N_CR1 / N_CR2 = %f / %f = %f +- %f" %
(binLabel, nSR, nCR, TF, TF_Error))
# Calculate the combined normalization factor (f_fakes = w*f_QCD + (1-w)*f_EWKfakes)
fakeRate = None
fakeRateError = None
fakeRateUp = None
fakeRateDown = None
if TF != None:
# fakeRate = w*self._TF[binLabel] + (1.0-w)*self._ewkNormalization[binLabel]
# fakeRateUp = wUp*self._TF[binLabel] + (1.0-wUp)*self._ewkNormalization[binLabel]
# fakeRateDown = wDown*self._TF[binLabel] + (1.0-wDown)*self._ewkNormalization[binLabel]
# fakeRateErrorPart1 = errorPropagation.errorPropagationForProduct(w, wError, self._TF[binLabel], self._TFError[binLabel])
# fakeRateErrorPart2 = errorPropagation.errorPropagationForProduct(w, wError, self._ewkNormalization[binLabel], self._ewkNormalizationError[binLabel])
# fakeRateError = ROOT.TMath.Sqrt(fakeRateErrorPart1**2 + fakeRateErrorPart2**2)
# Replace bin label with histo title (has exact binning info)
self._BinLabelMap[binLabel] = hFakeB_Inverted.GetTitle()
self._TF[binLabel] = TF
self._TF_Error[binLabel] = TF_Error
self._TF_Up[binLabel] = TF_Up
self._TF_Down[binLabel] = TF_Down
# self._combinedFakesNormalizationError[binLabel] = fakeRateError
# self._combinedFakesNormalizationUp[binLabel] = fakeRateUp
# self._combinedFakesNormalizationDown[binLabel] = fakeRateDown
# Store all information for later used (write to file)
self._commentLines.extend(lines)
# Print output and store comments
if 0:
for i, line in enumerate(lines, 1):
Print(line, i == 1)
return
def writeNormFactorFile(self, filename, opts):
'''
Save the fit results for QCD and EWK.
The results will are stored in a python file starting with name:
"QCDInvertedNormalizationFactors_" + moduleInfoString
The script also summarizes warnings and errors encountered:
- Green means deviation from normal is 0-3 %,
- Yellow means deviation of 3-10 %, and
- Red means deviation of >10 % (i.e. something is clearly wrong).
If necessary, do adjustments to stabilize the fits to get rid of the errors/warnings.
The first things to work with are:
a) Make sure enough events are in the histograms used
b) Adjust fit parameters and/or fit functions and re-fit results
Move on only once you are pleased with the normalisation coefficients
'''
s = ""
s += "# Generated on %s\n" % datetime.datetime.now().ctime()
s += "# by %s\n" % os.path.basename(sys.argv[0])
s += "\n"
s += "import sys\n"
s += "\n"
s += "def QCDInvertedNormalizationSafetyCheck(era, searchMode, optimizationMode):\n"
s += " validForEra = \"%s\"\n" % opts.dataEra
s += " validForSearchMode = \"%s\"\n" % opts.searchMode
s += " validForOptMode = \"%s\"\n" % opts.optMode
s += " if not era == validForEra:\n"
s += " raise Exception(\"Error: inconsistent era, normalisation factors valid for\",validForEra,\"but trying to use with\",era)\n"
s += " if not searchMode == validForSearchMode:\n"
s += " raise Exception(\"Error: inconsistent search mode, normalisation factors valid for\",validForSearchMode,\"but trying to use with\",searchMode)\n"
s += " if not optimizationMode == validForOptMode:\n"
s += " raise Exception(\"Error: inconsistent optimization mode, normalisation factors valid for\",validForOptMode,\"but trying to use with\",optimizationMode)\n"
s += " return"
s += "\n"
s += "QCDNormalization = {\n"
for k in self._TF:
if 0:
print "key = %s, value = %s" % (k, self._TF[k])
s += ' "%s": %f,\n' % (k, self._TF[k])
s += "}\n"
s += "QCDNormalizationError = {\n"
for k in self._TF_Error:
s += ' "%s": %f,\n' % (k, self._TF_Error[k])
s += "}\n"
s += "QCDNormalizationErrorUp = {\n"
for k in self._TF_Up:
s += ' "%s": %f,\n' % (k, self._TF_Up[k])
s += "}\n"
s += "QCDNormalizationErrorDown = {\n"
for k in self._TF_Down:
s += ' "%s": %f,\n' % (k, self._TF_Down[k])
s += "}\n"
self.Verbose("Writing results in file %s" % filename, True)
fOUT = open(filename, "w")
fOUT.write(s)
fOUT.write("'''\n")
for l in self._commentLines:
fOUT.write(l + "\n")
fOUT.write("'''\n")
fOUT.close()
msg = "Results written in file %s" % (
ShellStyles.SuccessStyle() + filename + ShellStyles.NormalStyle())
self.Print(msg, True)
# Create the transfer factors plot (for each bin of FakeB measurement)
self._generateCoefficientPlot()
# self._generateDQMplot()
return
def _generateCoefficientPlot(self):
'''
This probably is needed in the case the measurement is done in
bins of a correlated quantity (e.g. pT in the case of inverted tau isolation
'''
def makeGraph(markerStyle, color, binList, valueDict, upDict,
downDict):
g = ROOT.TGraphAsymmErrors(len(binList))
for i in range(len(binList)):
g.SetPoint(i, i + 0.5, valueDict[binList[i]])
g.SetPointEYhigh(i, upDict[binList[i]])
g.SetPointEYlow(i, downDict[binList[i]])
g.SetMarkerSize(1.2)
g.SetMarkerStyle(markerStyle)
g.SetLineColor(color)
g.SetLineWidth(3)
g.SetMarkerColor(color)
return g
# Obtain bin list in right order
keyList = []
keys = self._TF.keys()
keys.sort()
# for k in keys:
# if "lt" in k:
# keyList.append(k)
# for k in keys:
# if "eq" in k:
# keyList.append(k)
# for k in keys:
# if "gt" in k:
# keyList.append(k)
# For-loop: All Fake-b measurement bins
for k in keys:
keyList.append(k)
#if "Inclusive" in keys:
# keyList.append("Inclusive")
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(True)
style.setGridY(True)
# Create graphs
gFakeB = makeGraph(ROOT.kFullCircle, ROOT.kRed, keyList, self._TF,
self._TF_Error, self._TF_Error)
# Make plot
hFrame = ROOT.TH1F("frame", "frame", len(keyList), 0, len(keyList))
# Change bin labels to text
for i, binLabel in enumerate(keyList, 1):
# for i in range(len(keyList)):
binLabelText = self.getFormattedBinLabelString(binLabel)
#hFrame.GetXaxis().SetBinLabel(i+1, binLabelText)
hFrame.GetXaxis().SetBinLabel(i, binLabelText)
# Set axes names
hFrame.GetYaxis().SetTitle("transfer factor ")
# hFrame.GetYaxis().SetTitle("transfer factors (R_{i})")
# hFrame.GetXaxis().SetTitle("Fake-b bin")
# Customise axes
hFrame.SetMinimum(0.6e-1)
hFrame.SetMaximum(2e0)
if len(self._BinLabelMap) > 12:
lSize = 8
elif len(self._BinLabelMap) > 8:
lSize = 12
else:
lSize = 16
hFrame.GetXaxis().SetLabelSize(lSize) # 20
hFrame.GetXaxis().LabelsOption("d")
# Label Style options
# "a" sort by alphabetic order
# ">" sort by decreasing values
# "<" sort by increasing values
# "h" draw labels horizonthal
# "v" draw labels vertical
# "u" draw labels up (end of label right adjusted)
# "d" draw labels down (start of label left adjusted)
# Create canvas
c = ROOT.TCanvas()
c.SetLogy(True)
c.SetGridx()
c.SetGridy()
hFrame.Draw()
gFakeB.Draw("p same")
histograms.addStandardTexts(cmsTextPosition="outframe")
# Create the legend & draw it
l = ROOT.TLegend(0.65, 0.80, 0.90, 0.90)
l.SetFillStyle(-1)
l.SetBorderSize(0)
# l.AddEntry(gFakeB, "Fake-#it{b} #pm Stat.", "LP")
l.AddEntry(gFakeB, "Value #pm Stat.", "LP")
l.SetTextSize(0.035)
l.Draw()
# Store ROOT ignore level to normal before changing it
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
# Save the plot
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName +
"/QCDNormalisationCoefficients.%s" % item)
# Reset the ROOT ignore level to normal
ROOT.gErrorIgnoreLevel = backup
# Inform user
msg = "Transfer-factors written in %s " % (ShellStyles.SuccessStyle() +
self._plotDirName +
ShellStyles.NormalStyle())
self.Print(msg, False)
return
def getFormattedBinLabelString(self, binLabel):
'''
Dirty trick to get what I want
'''
if binLabel not in self._BinLabelMap:
# for k in self._BinLabelMap:
# print k
raise Exception("Got unexpected bin label \"%s\"!" % binLabel)
newLabel = self._BinLabelMap[binLabel]
newLabel = newLabel.replace("abs(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace(":", ",")
newLabel = newLabel.replace("TetrajetBjet", "") #"b^{ldg} ")
newLabel = newLabel.replace("Pt", "p_{T} ")
newLabel = newLabel.replace("Eta", "#eta ")
if "inclusive" in binLabel.lower():
newLabel = "Inclusive"
return newLabel
def _generateDQMplot(self):
'''
Create a DQM style plot
'''
# Check the uncertainties on the normalization factors
for k in self._dqmKeys.keys():
self._addDqmEntry(k, "norm.coeff.uncert::QCD", self._TFError[k],
0.03, 0.10)
self._addDqmEntry(k, "norm.coeff.uncert::fake",
self._ewkNormalizationError[k], 0.03, 0.10)
value = abs(self._combinedFakesNormalizationUp[k] -
self._combinedFakesNormalization[k])
value = max(
value,
abs(self._combinedFakesNormalizationDown[k] -
self._combinedFakesNormalizationUp[k]))
self._addDqmEntry(k, "norm.coeff.uncert::combined", value, 0.03,
0.10)
# Construct the DQM histogram
h = ROOT.TH2F("QCD DQM", "QCD DQM",
len(self._dqmKeys[self._dqmKeys.keys()[0]].keys()), 0,
len(self._dqmKeys[self._dqmKeys.keys()[0]].keys()),
len(self._dqmKeys.keys()), 0, len(self._dqmKeys.keys()))
h.GetXaxis().SetLabelSize(15)
h.GetYaxis().SetLabelSize(15)
h.SetMinimum(0)
h.SetMaximum(3)
#h.GetXaxis().LabelsOption("v")
nWarnings = 0
nErrors = 0
for i in range(h.GetNbinsX()):
for j in range(h.GetNbinsY()):
ykey = self._dqmKeys.keys()[j]
xkey = self._dqmKeys[ykey].keys()[i]
h.SetBinContent(i + 1, j + 1, self._dqmKeys[ykey][xkey])
h.GetYaxis().SetBinLabel(j + 1, ykey)
h.GetXaxis().SetBinLabel(i + 1, xkey)
if self._dqmKeys[ykey][xkey] > 2:
nErrors += 1
elif self._dqmKeys[ykey][xkey] > 1:
nWarnings += 1
palette = array.array("i", [ROOT.kGreen + 1, ROOT.kYellow, ROOT.kRed])
ROOT.gStyle.SetPalette(3, palette)
c = ROOT.TCanvas()
c.SetBottomMargin(0.2)
c.SetLeftMargin(0.2)
c.SetRightMargin(0.2)
h.Draw("colz")
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName + "/QCDNormalisationDQM.%s" % item)
ROOT.gErrorIgnoreLevel = backup
ROOT.gStyle.SetPalette(1)
print "Obtained %d warnings and %d errors for the normalization" % (
nWarnings, nErrors)
if nWarnings > 0 or nErrors > 0:
print "Please have a look at %s/QCDNormalisationDQM.png to see the origin of the warning(s) and error(s)" % self._plotDirName
return
def _addDqmEntry(self, binLabel, name, value, okTolerance, warnTolerance):
if not binLabel in self._dqmKeys.keys():
self._dqmKeys[binLabel] = OrderedDict()
result = 2.5
if abs(value) < okTolerance:
result = 0.5
elif abs(value) < warnTolerance:
result = 1.5
self._dqmKeys[binLabel][name] = result
return
def _getSanityCheckTextForFractions(self,
dataTemplate,
binLabel,
saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
NOTE: Should one divide the fractions with dataTemplate.getFittedParameters()[0] ?
Right now not because the correction is so small.
'''
self.Verbose("_getSanityCheckTextForFractions()", True)
# Get variables
label = "QCD"
fraction = dataTemplate.getFittedParameters()[1]
fractionError = dataTemplate.getFittedParameterErrors()[1]
nBaseline = self._templates["%s_Baseline" %
label].getNeventsFromHisto(False)
nCalculated = fraction * dataTemplate.getNeventsFromHisto(False)
if nCalculated > 0:
ratio = nBaseline / nCalculated
else:
ratio = 0
lines = []
lines.append("Fitted %s fraction: %f +- %f" %
(label, fraction, fractionError))
lines.append(
"Sanity check: ratio = %.3f: baseline = %.1f vs. fitted = %.1f" %
(ratio, nBaseline, nCalculated))
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
def _checkOverallNormalization(self,
template,
binLabel,
saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
'''
self.Verbose("_checkOverallNormalization()")
# Calculatotions
value = template.getFittedParameters()[0]
error = template.getFittedParameterErrors()[0]
# Definitions
lines = []
lines.append(
"The fitted overall normalization factor for purity is: (should be 1.0)"
)
lines.append("NormFactor = %f +/- %f" % (value, error))
self._addDqmEntry(binLabel, "OverallNormalization(par0)", value - 1.0,
0.03, 0.10)
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
## Helper method to be called from parent class when calculating norm.coefficients
def _getResultOutput(self, binLabel):
lines = []
lines.append(" Normalization factor (QCD): %f +- %f" %
(self._TF[binLabel], self._TFError[binLabel]))
lines.append(" Normalization factor (EWK fake taus): %f +- %f" %
(self._ewkNormalization[binLabel],
self._ewkNormalizationError[binLabel]))
lines.append(" Combined norm. factor: %f +- %f" %
(self._combinedFakesNormalization[binLabel],
self._combinedFakesNormalizationError[binLabel]))
# Store all information for later used (write to file)
self._commentLines.extend(lines)
return lines
class SystTopBDTNormalizationManager:
'''
Base class for QCD measurement normalization from which specialized algorithm classes inherit
'''
def __init__(self, resultDirName, moduleInfoString, verbose=False):
self._verbose = verbose
self._templates = {}
self._sources = {}
self._commentLines = []
self._BinLabelMap = {}
self._TF = {} # Transfer Factor (TF)
self._TF_Error = {}
self._TF_Up = {}
self._TF_Down = {}
self._dqmKeys = OrderedDict()
self._myPath = os.path.join(resultDirName, "normalisationPlots")
self.Verbose("__init__")
# No optimisation mode
if moduleInfoString == "":
moduleInfoString = "Default"
if not os.path.exists(self._myPath):
self.Print("Creating new directory %s" % (self._myPath), True )
os.mkdir(self._myPath)
self._plotDirName = os.path.join(resultDirName, "normalisationPlots", moduleInfoString)
# If already exists, Delete an entire directory tree
if os.path.exists(self._plotDirName):
msg = "Removing directory tree %s" % (self._plotDirName)
self.Verbose(ShellStyles.NoteStyle() + msg + ShellStyles.NormalStyle(), True)
shutil.rmtree(self._plotDirName)
msg = "Creating directory %s" % (self._plotDirName)
self.Verbose(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), False)
os.mkdir(self._plotDirName)
return
def Print(self, msg, printHeader=False):
fName = __file__.split("/")[-1]
if printHeader==True:
print "=== ", fName + ": class " + self.__class__.__name__
print "\t", msg
else:
print "\t", msg
return
def Verbose(self, msg, printHeader=True, verbose=False):
if not self._verbose:
return
self.Print(msg, printHeader)
return
def GetQCDNormalization(self, binLabel):
if binLabel in self._TF.keys():
return self._TF[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! "% (binLabel) )
def GetTransferFactor(self, binLabel):
return self.GetQCDNormalization(binLabel)
def GetQCDNormalizationError(self, binLabel):
if binLabel in self._TF_Error.keys():
return self._TF_Error[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! "% (binLabel) )
def CalculateTransferFactor(self, binLabel, hFakeB_Baseline, hFakeB_Inverted, verbose=False):
'''
Calculates the combined normalization and, if specified,
varies it up or down by factor (1+variation)
TF = Transfer Factor
SR = Signal Region
CR = Control Region
VR = Verification Region
'''
self.verbose = verbose
print "======= Calculate TransferFactor "
# Obtain counts for QCD and EWK fakes
lines = []
# NOTES: Add EWKGenuineB TF, Add Data TF, add QCD TF, Add EWK TF, add MCONLY TFs
nSR_Error = ROOT.Double(0.0)
nCR_Error = ROOT.Double(0.0)
# nTotalError = ROOT.TMath.Sqrt(nSRerror**2 + nCRError**2)
nSR = hFakeB_Baseline.IntegralAndError(1, hFakeB_Baseline.GetNbinsX()+1, nSR_Error)
nCR = hFakeB_Inverted.IntegralAndError(1, hFakeB_Inverted.GetNbinsX()+1, nCR_Error)
# nTotal = nSR + nCR
# Calculate Transfer Factor (TF) from Control Region (R) to Signal Region (SR): R = N_CR1/ N_CR2
TF = None
TF_Up = None
TF_Down = None
TF_Error = None
if 1: ## nTotal > 0.0:
TF = nSR / nCR
TF_Error = errorPropagation.errorPropagationForDivision(nSR, nSR_Error, nCR, nCR_Error)
TF_Up = TF + TF_Error
if TF_Up > 1.0:
TF_Up = 1.0
TF_Down = TF - TF_Error
if TF_Down < 0.0:
TF_Down = 0.0
lines.append("TF (bin=%s) = N_CR1 / N_CR2 = %f / %f = %f +- %f" % (binLabel, nSR, nCR, TF, TF_Error) )
# Calculate the combined normalization factor (f_fakes = w*f_QCD + (1-w)*f_EWKfakes)
fakeRate = None
fakeRateError = None
fakeRateUp = None
fakeRateDown = None
if TF != None:
# fakeRate = w*self._TF[binLabel] + (1.0-w)*self._ewkNormalization[binLabel]
# fakeRateUp = wUp*self._TF[binLabel] + (1.0-wUp)*self._ewkNormalization[binLabel]
# fakeRateDown = wDown*self._TF[binLabel] + (1.0-wDown)*self._ewkNormalization[binLabel]
# fakeRateErrorPart1 = errorPropagation.errorPropagationForProduct(w, wError, self._TF[binLabel], self._TFError[binLabel])
# fakeRateErrorPart2 = errorPropagation.errorPropagationForProduct(w, wError, self._ewkNormalization[binLabel], self._ewkNormalizationError[binLabel])
# fakeRateError = ROOT.TMath.Sqrt(fakeRateErrorPart1**2 + fakeRateErrorPart2**2)
# Replace bin label with histo title (has exact binning info)
self._BinLabelMap[binLabel] = hFakeB_Inverted.GetTitle()
self._TF[binLabel ] = TF
self._TF_Error[binLabel] = TF_Error
self._TF_Up[binLabel] = TF_Up
self._TF_Down[binLabel] = TF_Down
# self._combinedFakesNormalizationError[binLabel] = fakeRateError
# self._combinedFakesNormalizationUp[binLabel] = fakeRateUp
# self._combinedFakesNormalizationDown[binLabel] = fakeRateDown
# Store all information for later used (write to file)
self._commentLines.extend(lines)
# Print output and store comments
if 1:
for i, line in enumerate(lines, 1):
Print(line, i==1)
return
def writeNormFactorFile(self, filename, opts):
'''
Save the fit results for QCD and EWK.
The results will are stored in a python file starting with name:
"QCDInvertedNormalizationFactors_" + moduleInfoString
The script also summarizes warnings and errors encountered:
- Green means deviation from normal is 0-3 %,
- Yellow means deviation of 3-10 %, and
- Red means deviation of >10 % (i.e. something is clearly wrong).
If necessary, do adjustments to stabilize the fits to get rid of the errors/warnings.
The first things to work with are:
a) Make sure enough events are in the histograms used
b) Adjust fit parameters and/or fit functions and re-fit results
Move on only once you are pleased with the normalisation coefficients
'''
s = ""
s += "# Generated on %s\n"% datetime.datetime.now().ctime()
s += "# by %s\n" % os.path.basename(sys.argv[0])
s += "\n"
s += "import sys\n"
s += "\n"
s += "def QCDInvertedNormalizationSafetyCheck(era, searchMode, optimizationMode):\n"
s += " validForEra = \"%s\"\n" % opts.dataEra
s += " validForSearchMode = \"%s\"\n" % opts.searchMode
s += " validForOptMode = \"%s\"\n" % opts.optMode
s += " if not era == validForEra:\n"
s += " raise Exception(\"Error: inconsistent era, normalisation factors valid for\",validForEra,\"but trying to use with\",era)\n"
s += " if not searchMode == validForSearchMode:\n"
s += " raise Exception(\"Error: inconsistent search mode, normalisation factors valid for\",validForSearchMode,\"but trying to use with\",searchMode)\n"
s += " if not optimizationMode == validForOptMode:\n"
s += " raise Exception(\"Error: inconsistent optimization mode, normalisation factors valid for\",validForOptMode,\"but trying to use with\",optimizationMode)\n"
s += " return"
s += "\n"
s += "QCDNormalization = {\n"
for k in self._TF:
if 0:
print "key = %s, value = %s" % (k, self._TF[k])
s += ' "%s": %f,\n'%(k, self._TF[k])
s += "}\n"
s += "QCDNormalizationError = {\n"
for k in self._TF_Error:
s += ' "%s": %f,\n'%(k, self._TF_Error[k])
s += "}\n"
s += "QCDNormalizationErrorUp = {\n"
for k in self._TF_Up:
s += ' "%s": %f,\n'%(k, self._TF_Up[k])
s += "}\n"
s += "QCDNormalizationErrorDown = {\n"
for k in self._TF_Down:
s += ' "%s": %f,\n'%(k, self._TF_Down[k])
s += "}\n"
self.Verbose("Writing results in file %s" % filename, True)
fOUT = open(filename,"w")
fOUT.write(s)
fOUT.write("'''\n")
for l in self._commentLines:
fOUT.write(l + "\n")
fOUT.write("'''\n")
fOUT.close()
msg = "Results written in file %s" % (ShellStyles.SuccessStyle() + filename + ShellStyles.NormalStyle())
self.Print(msg, True)
# Create the transfer factors plot (for each bin of FakeB measurement)
self._generateCoefficientPlot()
# self._generateDQMplot()
return
def _generateCoefficientPlot(self):
'''
This probably is needed in the case the measurement is done in
bins of a correlated quantity (e.g. pT in the case of inverted tau isolation
'''
def makeGraph(markerStyle, color, binList, valueDict, upDict, downDict):
g = ROOT.TGraphAsymmErrors(len(binList))
for i in range(len(binList)):
g.SetPoint(i, i+0.5, valueDict[binList[i]])
g.SetPointEYhigh(i, upDict[binList[i]])
g.SetPointEYlow(i, downDict[binList[i]])
g.SetMarkerSize(1.2)
g.SetMarkerStyle(markerStyle)
g.SetLineColor(color)
g.SetLineWidth(3)
g.SetMarkerColor(color)
return g
# Obtain bin list in right order
keyList = []
keys = self._TF.keys()
keys.sort()
# for k in keys:
# if "lt" in k:
# keyList.append(k)
# for k in keys:
# if "eq" in k:
# keyList.append(k)
# for k in keys:
# if "gt" in k:
# keyList.append(k)
# For-loop: All Fake-b measurement bins
for k in keys:
keyList.append(k)
#if "Inclusive" in keys:
# keyList.append("Inclusive")
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(True)
style.setGridY(True)
# Create graphs
gFakeB = makeGraph(ROOT.kFullCircle, ROOT.kRed, keyList, self._TF, self._TF_Error, self._TF_Error)
# Make plot
hFrame = ROOT.TH1F("frame","frame", len(keyList), 0, len(keyList))
# Change bin labels to text
for i, binLabel in enumerate(keyList, 1):
# for i in range(len(keyList)):
binLabelText = self.getFormattedBinLabelString(binLabel)
#hFrame.GetXaxis().SetBinLabel(i+1, binLabelText)
hFrame.GetXaxis().SetBinLabel(i, binLabelText)
# Set axes names
hFrame.GetYaxis().SetTitle("transfer factor ")
# hFrame.GetYaxis().SetTitle("transfer factors (R_{i})")
# hFrame.GetXaxis().SetTitle("Fake-b bin")
# Customise axes
hFrame.SetMinimum(0.6e-1)
hFrame.SetMaximum(2e0)
if len(self._BinLabelMap) > 12:
lSize = 8
elif len(self._BinLabelMap) > 8:
lSize = 12
else:
lSize = 16
hFrame.GetXaxis().SetLabelSize(lSize) # 20
hFrame.GetXaxis().LabelsOption("d")
# Label Style options
# "a" sort by alphabetic order
# ">" sort by decreasing values
# "<" sort by increasing values
# "h" draw labels horizonthal
# "v" draw labels vertical
# "u" draw labels up (end of label right adjusted)
# "d" draw labels down (start of label left adjusted)
# Create canvas
c = ROOT.TCanvas()
c.SetLogy(True)
c.SetGridx()
c.SetGridy()
hFrame.Draw()
gFakeB.Draw("p same")
histograms.addStandardTexts(cmsTextPosition="outframe")
# Create the legend & draw it
l = ROOT.TLegend(0.65, 0.80, 0.90, 0.90)
l.SetFillStyle(-1)
l.SetBorderSize(0)
# l.AddEntry(gFakeB, "Fake-#it{b} #pm Stat.", "LP")
l.AddEntry(gFakeB, "Value #pm Stat.", "LP")
l.SetTextSize(0.035)
l.Draw()
# Store ROOT ignore level to normal before changing it
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
# Save the plot
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName + "/QCDNormalisationCoefficients.%s" % item)
# Reset the ROOT ignore level to normal
ROOT.gErrorIgnoreLevel = backup
# Inform user
msg = "Transfer-factors written in %s " % (ShellStyles.SuccessStyle() + self._plotDirName + ShellStyles.NormalStyle())
self.Print(msg, False)
return
def getFormattedBinLabelString(self, binLabel):
'''
Dirty trick to get what I want
'''
if binLabel not in self._BinLabelMap:
# for k in self._BinLabelMap:
# print k
raise Exception("Got unexpected bin label \"%s\"!" % binLabel)
newLabel = self._BinLabelMap[binLabel]
newLabel = newLabel.replace("abs(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace(":", ",")
newLabel = newLabel.replace("TetrajetBjet", "") #"b^{ldg} ")
newLabel = newLabel.replace("Pt", "p_{T} ")
newLabel = newLabel.replace("Eta", "#eta ")
if "inclusive" in binLabel.lower():
newLabel = "Inclusive"
return newLabel
def _generateDQMplot(self):
'''
Create a DQM style plot
'''
# Check the uncertainties on the normalization factors
for k in self._dqmKeys.keys():
self._addDqmEntry(k, "norm.coeff.uncert::QCD" , self._TFError[k], 0.03, 0.10)
self._addDqmEntry(k, "norm.coeff.uncert::fake", self._ewkNormalizationError[k], 0.03, 0.10)
value = abs(self._combinedFakesNormalizationUp[k]-self._combinedFakesNormalization[k])
value = max(value, abs(self._combinedFakesNormalizationDown[k]-self._combinedFakesNormalizationUp[k]))
self._addDqmEntry(k, "norm.coeff.uncert::combined", value, 0.03, 0.10)
# Construct the DQM histogram
h = ROOT.TH2F("QCD DQM", "QCD DQM",
len(self._dqmKeys[self._dqmKeys.keys()[0]].keys()), 0, len(self._dqmKeys[self._dqmKeys.keys()[0]].keys()),
len(self._dqmKeys.keys()), 0, len(self._dqmKeys.keys()))
h.GetXaxis().SetLabelSize(15)
h.GetYaxis().SetLabelSize(15)
h.SetMinimum(0)
h.SetMaximum(3)
#h.GetXaxis().LabelsOption("v")
nWarnings = 0
nErrors = 0
for i in range(h.GetNbinsX()):
for j in range(h.GetNbinsY()):
ykey = self._dqmKeys.keys()[j]
xkey = self._dqmKeys[ykey].keys()[i]
h.SetBinContent(i+1, j+1, self._dqmKeys[ykey][xkey])
h.GetYaxis().SetBinLabel(j+1, ykey)
h.GetXaxis().SetBinLabel(i+1, xkey)
if self._dqmKeys[ykey][xkey] > 2:
nErrors += 1
elif self._dqmKeys[ykey][xkey] > 1:
nWarnings += 1
palette = array.array("i", [ROOT.kGreen+1, ROOT.kYellow, ROOT.kRed])
ROOT.gStyle.SetPalette(3, palette)
c = ROOT.TCanvas()
c.SetBottomMargin(0.2)
c.SetLeftMargin(0.2)
c.SetRightMargin(0.2)
h.Draw("colz")
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName+"/QCDNormalisationDQM.%s"%item)
ROOT.gErrorIgnoreLevel = backup
ROOT.gStyle.SetPalette(1)
print "Obtained %d warnings and %d errors for the normalization"%(nWarnings, nErrors)
if nWarnings > 0 or nErrors > 0:
print "Please have a look at %s/QCDNormalisationDQM.png to see the origin of the warning(s) and error(s)"%self._plotDirName
return
def _addDqmEntry(self, binLabel, name, value, okTolerance, warnTolerance):
if not binLabel in self._dqmKeys.keys():
self._dqmKeys[binLabel] = OrderedDict()
result = 2.5
if abs(value) < okTolerance:
result = 0.5
elif abs(value) < warnTolerance:
result = 1.5
self._dqmKeys[binLabel][name] = result
return
def _getSanityCheckTextForFractions(self, dataTemplate, binLabel, saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
NOTE: Should one divide the fractions with dataTemplate.getFittedParameters()[0] ?
Right now not because the correction is so small.
'''
self.Verbose("_getSanityCheckTextForFractions()", True)
# Get variables
label = "QCD"
fraction = dataTemplate.getFittedParameters()[1]
fractionError = dataTemplate.getFittedParameterErrors()[1]
nBaseline = self._templates["%s_Baseline" % label].getNeventsFromHisto(False)
nCalculated = fraction * dataTemplate.getNeventsFromHisto(False)
if nCalculated > 0:
ratio = nBaseline / nCalculated
else:
ratio = 0
lines = []
lines.append("Fitted %s fraction: %f +- %f" % (label, fraction, fractionError))
lines.append("Sanity check: ratio = %.3f: baseline = %.1f vs. fitted = %.1f" % (ratio, nBaseline, nCalculated))
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
def _checkOverallNormalization(self, template, binLabel, saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
'''
self.Verbose("_checkOverallNormalization()")
# Calculatotions
value = template.getFittedParameters()[0]
error = template.getFittedParameterErrors()[0]
# Definitions
lines = []
lines.append("The fitted overall normalization factor for purity is: (should be 1.0)")
lines.append("NormFactor = %f +/- %f" % (value, error))
self._addDqmEntry(binLabel, "OverallNormalization(par0)", value-1.0, 0.03, 0.10)
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
## Helper method to be called from parent class when calculating norm.coefficients
def _getResultOutput(self, binLabel):
lines = []
lines.append(" Normalization factor (QCD): %f +- %f"%(self._TF[binLabel], self._TFError[binLabel]))
lines.append(" Normalization factor (EWK fake taus): %f +- %f"%(self._ewkNormalization[binLabel], self._ewkNormalizationError[binLabel]))
lines.append(" Combined norm. factor: %f +- %f"%(self._combinedFakesNormalization[binLabel], self._combinedFakesNormalizationError[binLabel]))
# Store all information for later used (write to file)
self._commentLines.extend(lines)
return lines
class FakeBNormalizationManager:
'''
Base class for QCD measurement normalization from which specialized algorithm classes inherit
'''
def __init__(self, binLabels, resultDirName, moduleInfoString, verbose=False):
self._verbose = verbose
self._templates = {}
self._binLabels = binLabels
self._sources = {}
self._commentLines = []
self._NEvtsCR1 = {}
self._NEvtsCR1_Error = {}
self._NEvtsCR2 = {}
self._NEvtsCR2_Error = {}
self._NEvtsCR3 = {}
self._NEvtsCR3_Error = {}
self._NEvtsCR4 = {}
self._NEvtsCR4_Error = {}
self._TF = {} # Transfer Factor (TF)
self._TF_Error = {}
self._TF_Up = {}
self._TF_Down = {}
self._TF_Up2x = {}
self._TF_Down2x = {}
self._TF_Up3x = {}
self._TF_Down3x = {}
self._dqmKeys = OrderedDict()
self._myPath = os.path.join(resultDirName, "normalisationPlots")
self._BinLabelMap = {}
self._FakeBNormalization = {} # for the time being same as TF
self._FakeBNormalizationError = {} # for the time being same as TF_Error
self._FakeBNormalizationUp = {} # for the time being same as TF_Up
self._FakeBNormalizationUp2x = {}
self._FakeBNormalizationUp3x = {}
self._FakeBNormalizationDown = {} # for the time being same as TF_Down
self._FakeBNormalizationDown2x = {}
self._FakeBNormalizationDown3x = {}
if not isinstance(binLabels, list):
raise Exception("Error: binLabels needs to be a list of strings")
self.Verbose("__init__")
# No optimisation mode
if moduleInfoString == "":
moduleInfoString = "Default"
if not os.path.exists(self._myPath):
self.Print("Creating new directory %s" % (self._myPath), True )
os.mkdir(self._myPath)
self._plotDirName = os.path.join(resultDirName, "normalisationPlots", moduleInfoString)
# If already exists, Delete an entire directory tree
if os.path.exists(self._plotDirName):
msg = "Removing directory tree %s" % (self._plotDirName)
self.Verbose(ShellStyles.NoteStyle() + msg + ShellStyles.NormalStyle(), True)
shutil.rmtree(self._plotDirName)
msg = "Creating directory %s" % (self._plotDirName)
self.Verbose(ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(), False)
os.mkdir(self._plotDirName)
return
def Print(self, msg, printHeader=False):
fName = __file__.split("/")[-1]
if printHeader==True:
# print "=== ", fName + ": class " + self.__class__.__name__
print "=== ", fName
print "\t", msg
else:
print "\t", msg
return
def Verbose(self, msg, printHeader=True, verbose=False):
if not self._verbose:
return
self.Print(msg, printHeader)
return
def GetQCDNormalization(self, binLabel):
if binLabel in self._TF.keys():
return self._TF[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! "% (binLabel) )
def GetTransferFactor(self, binLabel):
return self.GetQCDNormalization(binLabel)
def GetQCDNormalizationError(self, binLabel):
if binLabel in self._TF_Error.keys():
return self._TF_Error[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! "% (binLabel) )
def CalculateTransferFactor(self, binLabel, hFakeB_CR1, hFakeB_CR2, hFakeB_CR3, hFakeB_CR4, verbose=False):
'''
Calculates the combined normalization and, if specified,
varies it up or down by factor (1+variation)
TF = Transfer Factor
SR = Signal Region
CR = Control Region
VR = Verification Region
'''
self.verbose = verbose
# Obtain counts for QCD and EWK fakes
lines = []
# NOTES: Add EWKGenuineB TF, Add Data TF, add QCD TF, Add EWK TF, add MCONLY TFs
nCR1_Error = ROOT.Double(0.0)
nCR2_Error = ROOT.Double(0.0)
nCR3_Error = ROOT.Double(0.0)
nCR4_Error = ROOT.Double(0.0)
# Get Events in all CRs and their associated errors
nCR1 = hFakeB_CR1.IntegralAndError(1, hFakeB_CR1.GetNbinsX()+1, nCR1_Error)
nCR2 = hFakeB_CR2.IntegralAndError(1, hFakeB_CR2.GetNbinsX()+1, nCR2_Error)
nCR3 = hFakeB_CR3.IntegralAndError(1, hFakeB_CR3.GetNbinsX()+1, nCR3_Error)
nCR4 = hFakeB_CR4.IntegralAndError(1, hFakeB_CR4.GetNbinsX()+1, nCR4_Error)
# Calculate Transfer Factor (TF) from Control Region (R) to Signal Region (SR): R = N_CR1/ N_CR2
TF = None
TF_Up = None
TF_Up2x = None
TF_Up3x = None
TF_Down = None
TF_Down2x = None
TF_Down3x = None
TF_Error = None
TF = (nCR1 / nCR2)
TF_Error = errorPropagation.errorPropagationForDivision(nCR1, nCR1_Error, nCR2, nCR2_Error)
# Up variations
TF_Up = TF + TF_Error
TF_Up2x = TF + 2*TF_Error
TF_Up3x = TF + 3*TF_Error
if TF_Up > 1.0:
Print("Forcing TF_Up (=%.3f) to be equal to 1!" % ( TF_Up), True) # added 23 Oct 2018
TF_Up = 1.0
if TF_Up2x > 1.0:
Print("Forcing TF_Up2x (=%.3f) to be equal to 1!" % ( TF_Up2x), True) # added 23 Oct 2018
TF_Up2x = 1.0
if TF_Up3x > 1.0:
Print("Forcing TF_Up3x (=%.3f) to be equal to 1!" % ( TF_Up3x), True) # added 23 Oct 2018
TF_Up3x = 1.0
# Down variations
TF_Down = TF - TF_Error
TF_Down2x = TF - 2*TF_Error
TF_Down3x = TF - 3*TF_Error
if TF_Down < 0.0:
Print("Forcing TF_Down (=%.3f) to be equal to 0" % (TF_Down), True) # added 23 Oct 2018
TF_Down = 0.0
if TF_Down2x < 0.0:
Print("Forcing TF_Down2x (=%.3f) to be equal to 0" % (TF_Down2x), True) # added 23 Oct 2018
TF_Down2x = 0.0
if TF_Down3x < 0.0:
Print("Forcing TF_Down3x (=%.3f) to be equal to 0" % (TF_Down3x), True) # added 23 Oct 2018
TF_Down3x = 0.0
lines.append("TF (bin=%s) = N_CR1 / N_CR2 = %f / %f = %f +- %f" % (binLabel, nCR1, nCR2, TF, TF_Error) )
# Calculate the transfer factors (R_{i}) where i is index of bin the Fake-b measurement is made in (pT and/or eta of ldg b-jet)
if TF != None:
# Replace bin label with histo title (has exact binning info)
self._BinLabelMap[binLabel] = self.getNiceBinLabel(hFakeB_CR2.GetTitle())
self._NEvtsCR1[binLabel] = nCR1
self._NEvtsCR1_Error[binLabel] = nCR1_Error
self._NEvtsCR2[binLabel] = nCR2
self._NEvtsCR2_Error[binLabel] = nCR2_Error
self._NEvtsCR3[binLabel] = nCR3
self._NEvtsCR3_Error[binLabel] = nCR3_Error
self._NEvtsCR4[binLabel] = nCR4
self._NEvtsCR4_Error[binLabel] = nCR4_Error
self._TF[binLabel ] = TF
self._TF_Error[binLabel] = TF_Error
self._TF_Up[binLabel] = TF_Up
self._TF_Up2x[binLabel] = TF_Up2x
self._TF_Up3x[binLabel] = TF_Up3x
self._TF_Down[binLabel] = TF_Down
self._TF_Down2x[binLabel] = TF_Down2x
self._TF_Down3x[binLabel] = TF_Down3x
self._FakeBNormalization[binLabel] = TF # TF
self._FakeBNormalizationError[binLabel] = TF_Error # Error(TF)
self._FakeBNormalizationUp[binLabel] = TF_Up # TF + Error
self._FakeBNormalizationUp2x[binLabel] = TF_Up2x # TF + 2*Error
self._FakeBNormalizationUp3x[binLabel] = TF_Up3x # TF + 3*Error
self._FakeBNormalizationDown[binLabel] = TF_Down # TF - Error
self._FakeBNormalizationDown2x[binLabel] = TF_Down2x # TF - 2*Error
self._FakeBNormalizationDown3x[binLabel] = TF_Down3x # TF - 3*Error
# Store all information for later used (write to file)
self._commentLines.extend(lines)
# Print output and store comments
if 0:
for i, line in enumerate(lines, 1):
Print(line, i==1)
return
def getNiceBinLabel(self, binLabel):
newLabel = binLabel.replace("abs", "")
if 1:
newLabel = newLabel.replace("TetrajetBjet", " ")
newLabel = newLabel.replace("TetrajetBJet", " ")
else: #more room
newLabel = newLabel.replace("TetrajetBjet", "b-jet ")
newLabel = newLabel.replace("TetrajetBJet", "b-jet ")
newLabel = newLabel.replace("_", " ")
newLabel = newLabel.replace("(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("Eta", "#eta")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace("CRtwo", "")
return newLabel
def writeTransferFactorsToFile(self, filename, opts):
'''
Save the fit results for QCD and EWK.
The results will are stored in a python file starting with name:
"QCDInvertedNormalizationFactors_" + moduleInfoString
The script also summarizes warnings and errors encountered:
- Green means deviation from normal is 0-3 %,
- Yellow means deviation of 3-10 %, and
- Red means deviation of >10 % (i.e. something is clearly wrong).
If necessary, do adjustments to stabilize the fits to get rid of the errors/warnings.
The first things to work with are:
a) Make sure enough events are in the histograms used
b) Adjust fit parameters and/or fit functions and re-fit results
Move on only once you are pleased with the normalisation coefficients
'''
s = ""
s += "# Generated on %s\n"% datetime.datetime.now().ctime()
s += "# by %s\n" % os.path.basename(sys.argv[0])
s += "\n"
s += "import sys\n"
s += "\n"
s += "def FakeBNormalisationSafetyCheck(era, searchMode, optimizationMode):\n"
s += " validForEra = \"%s\"\n" % opts.dataEra
s += " validForSearchMode = \"%s\"\n" % opts.searchMode
s += " validForOptMode = \"%s\"\n" % opts.optMode
s += " if not era == validForEra:\n"
s += " raise Exception(\"Error: inconsistent era, normalisation factors valid for\",validForEra,\"but trying to use with\",era)\n"
s += " if not searchMode == validForSearchMode:\n"
s += " raise Exception(\"Error: inconsistent search mode, normalisation factors valid for\",validForSearchMode,\"but trying to use with\",searchMode)\n"
s += " if not optimizationMode == validForOptMode:\n"
s += " raise Exception(\"Error: inconsistent optimization mode, normalisation factors valid for\",validForOptMode,\"but trying to use with\",optimizationMode)\n"
s += " return"
s += "\n"
s += "\n"
# First write the transfer factor (for each Fake-b measurement bin)
s += "FakeBNormalisation_Value = {\n"
for binLabel in self._TF:
s += ' "%s": %f,\n' % (binLabel, self._TF[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factor error (for each Fake-b measurement bin)
s += "FakeBNormalisation_Error = {\n"
for binLabel in self._TF_Error:
s += ' "%s": %f,\n' % (binLabel, self._TF_Error[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors + error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorUp = {\n"
for binLabel in self._TF_Up:
s += ' "%s": %f,\n' % (binLabel, self._TF_Up[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors + 2*error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorUp2x = {\n"
for binLabel in self._TF_Up2x:
s += ' "%s": %f,\n' % (binLabel, self._TF_Up2x[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors + 3*error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorUp3x = {\n"
for binLabel in self._TF_Up3x:
s += ' "%s": %f,\n' % (binLabel, self._TF_Up3x[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors - error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorDown = {\n"
for binLabel in self._TF_Down:
s += ' "%s": %f,\n' % (binLabel, self._TF_Down[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors - 2*error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorDown2x = {\n"
for binLabel in self._TF_Down2x:
s += ' "%s": %f,\n' % (binLabel, self._TF_Down2x[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors - 3*error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorDown3x = {\n"
for binLabel in self._TF_Down3x:
s += ' "%s": %f,\n' % (binLabel, self._TF_Down3x[binLabel])
s += "}\n"
s += "\n"
# Then write bin label map
s += "BinLabelMap = {\n"
for binLabel in self._BinLabelMap:
s += ' "%s": \"%s\",\n' % (binLabel, self._BinLabelMap[binLabel])
s += "}\n"
s += "\n"
self.Verbose("Writing results in file %s" % filename, True)
fOUT = open(filename,"w")
fOUT.write(s)
fOUT.write("'''\n")
for l in self._commentLines:
fOUT.write(l + "\n")
fOUT.write("'''\n")
fOUT.close()
msg = "Results written in file %s" % (ShellStyles.SuccessStyle() + filename + ShellStyles.NormalStyle())
self.Print(msg, True)
# Create the transfer factors plot (for each bin of FakeB measurement)
self._generateTransferFactorsPlot()
self._generateDQMPlot()
return
def _generateTransferFactorsPlot(self):
'''
The resulting plot will contain the transfer factors (y-axis) for a given measurement bin (x-axis)
This is needed in the case the Fake-b measurement is done in
bins of a correlated quantity (e.g. eta of leading b-jet in Fake-b for HToTB, and tau-pT in the
case of inverted tau isolatio for HToTau)
'''
def makeGraph(markerStyle, color, binList, valueDict, upDict, downDict):
g = ROOT.TGraphAsymmErrors(len(binList))
for i in range(len(binList)):
g.SetPoint(i, i+0.5, valueDict[binList[i]])
g.SetPointEYhigh(i, upDict[binList[i]])
g.SetPointEYlow(i, downDict[binList[i]])
g.SetMarkerSize(1.2)
g.SetMarkerStyle(markerStyle)
g.SetLineColor(color)
g.SetLineWidth(3)
g.SetMarkerColor(color)
return g
# Obtain bin list in right order
keyList = []
keys = self._TF.keys()
keys.sort()
# for k in keys:
# if "lt" in k:
# keyList.append(k)
# for k in keys:
# if "eq" in k:
# keyList.append(k)
# for k in keys:
# if "gt" in k:
# keyList.append(k)
# For-loop: All Fake-b measurement bins
for k in keys:
keyList.append(k)
#if "Inclusive" in keys:
# keyList.append("Inclusive")
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Create graphs
gFakeB = makeGraph(ROOT.kFullCircle, ROOT.kAzure, keyList, self._TF, self._TF_Error, self._TF_Error)
# Make plot
hFrame = ROOT.TH1F("frame","frame", len(keyList), 0, len(keyList))
# Change bin labels to text
for i, binLabel in enumerate(keyList, 1):
binLabelText = self.getFormattedBinLabelString(binLabel)
hFrame.GetXaxis().SetBinLabel(i, binLabelText)
# Set axes names
hFrame.GetYaxis().SetTitle("transfer factor") #R_{i}
# hFrame.GetXaxis().SetTitle("Fake-b bin")
# Customise axes
logy = False
if logy:
hFrame.SetMinimum(0.6e-1)
hFrame.SetMaximum(2e0)
else:
hFrame.SetMinimum(0.0)
hFrame.SetMaximum(1.0)
if len(self._BinLabelMap) > 12:
lSize = 8
elif len(self._BinLabelMap) > 8:
lSize = 12
else:
lSize = 16
hFrame.GetXaxis().SetLabelSize(lSize) # 20
hFrame.GetXaxis().LabelsOption("d")
# Label Style options
# "a" sort by alphabetic order
# ">" sort by decreasing values
# "<" sort by increasing values
# "h" draw labels horizonthal
# "v" draw labels vertical
# "u" draw labels up (end of label right adjusted)
# "d" draw labels down (start of label left adjusted)
# Create canvas
c = ROOT.TCanvas()
c.SetLogy(logy)
c.SetGridx(False)
c.SetGridy(False)
hFrame.Draw()
gFakeB.Draw("p same")
histograms.addStandardTexts(cmsTextPosition="outframe")
# Create the legend & draw it
l = ROOT.TLegend(0.65, 0.80, 0.90, 0.90)
l.SetFillStyle(-1)
l.SetBorderSize(0)
l.AddEntry(gFakeB, "Value #pm stat.", "LP") # "Fake-#it{b} #pm Stat.", "LP"
l.SetTextSize(0.035)
if 0:
l.Draw()
# Store ROOT ignore level to normal before changing it
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
# Save the plot
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName + "/FakeBNormalisationCoefficients.%s" % item)
# Reset the ROOT ignore level to normal
ROOT.gErrorIgnoreLevel = backup
# Inform user
msg = "Plot saved under %s" % (ShellStyles.SuccessStyle() + self._plotDirName + "/" + ShellStyles.NormalStyle())
self.Print(msg, True)
return
def getFormattedBinLabelString(self, binLabel):
'''
Dirty trick to get what I want
'''
if binLabel not in self._BinLabelMap:
raise Exception("Got unexpected bin label \"%s\"!" % binLabel)
newLabel = self._BinLabelMap[binLabel]
newLabel = newLabel.replace("abs(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace(":", ",")
newLabel = newLabel.replace("TetrajetBJet", "") #"b^{ldg} ")
newLabel = newLabel.replace("Pt", "p_{T} ")
newLabel = newLabel.replace("Eta", "#eta ")
if "inclusive" in binLabel.lower():
newLabel = "Inclusive"
return newLabel
def _generateDQMPlot(self):
'''
Create a Data Quality Monitor (DQM) style plot
to easily check the error for each transfer factor
and whether it is within an acceptable relative error
'''
# Define error warning/tolerance on relative errors
okay = 1.0/(len(self._BinLabelMap.keys()))
warn = 0.5*okay
NEvts = []
# Check the uncertainties on the normalization factors
for k in self._BinLabelMap:
relErrorUp = abs(self._TF_Up[k])/(self._TF[k])
relErrorDown = abs(self._TF_Down[k])/(self._TF[k])
relError = self._TF_Error[k]/self._TF[k]
if 0:
print "bin = %s , relErrorUp = %s, relErrorDown = %s " % (k, relErrorUp, relErrorDown)
# Add DQM entries
NCR1 = 0
NCR2 = 0
NCR3 = 0
NCR4 = 0
for j in self._NEvtsCR1:
NCR1 += self._NEvtsCR1[j]
NEvts.append(self._NEvtsCR1[j])
for j in self._NEvtsCR2:
NCR2 += self._NEvtsCR2[j]
NEvts.append(self._NEvtsCR2[j])
for j in self._NEvtsCR3:
NCR3 += self._NEvtsCR3[j]
NEvts.append(self._NEvtsCR3[j])
for j in self._NEvtsCR4:
NCR4 += self._NEvtsCR4[j]
NEvts.append(self._NEvtsCR4[j])
if 0:
print "NCR1[%s] = %0.1f, NCR2[%s] = %0.1f, k = %s" % (k, self._NEvtsCR1[k], k, self._NEvtsCR2[k], k)
print "NCR1 = %s, NCR2 = %s, k = %s" % (NCR1, NCR2, k)
print "error/NCR1[%s] = %0.2f, error/NCR2[%s] = %0.2f" % (k, self._NEvtsCR1_Error[k]/self._NEvtsCR1[k], k, self._NEvtsCR2_Error[k]/self._NEvtsCR2[k])
# Add DQM plot entries
self._addDqmEntry(self._BinLabelMap[k], "N_{CR1}", self._NEvtsCR1[k], NCR1*okay, NCR1*warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR2}", self._NEvtsCR2[k], NCR2*okay, NCR2*warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR3}", self._NEvtsCR3[k], NCR3*okay, NCR3*warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR4}", self._NEvtsCR4[k], NCR4*okay, NCR4*warn)
# self._addDqmEntry(self._BinLabelMap[k], "#frac{#sigma_{CR1}}{N_{CR1}}", self._NEvtsCR1_Error[k]/NCR1, 0.05, 0.15)
# self._addDqmEntry(self._BinLabelMap[k], "#frac{#sigma_{CR2}}{N_{CR2}}", self._NEvtsCR2_Error[k]/NCR2, 0.05, 0.15)
# Construct the DQM histogram
nBinsX = len(self._dqmKeys[self._dqmKeys.keys()[0]].keys())
nBinsY = len(self._dqmKeys.keys())
h = ROOT.TH2F("FakeB DQM", "FakeB DQM", nBinsX, 0, nBinsX, nBinsY, 0, nBinsY)
# Customise axes
h.GetXaxis().SetLabelSize(15)
h.GetYaxis().SetLabelSize(10)
# Set Min and Max of z-axis
if 0: # red, yellow, green for DQM
h.SetMinimum(0)
h.SetMaximum(3)
else: # pure entries instead of red, yellow, green
h.SetMinimum(min(NEvts)*0.25)
h.SetMaximum(round(max(NCR1, NCR2, NCR3, NCR4)))
h.SetContour(10)
#h.SetContour(3)
if 0:
h.GetXaxis().LabelsOption("v")
h.GetYaxis().LabelsOption("v")
nWarnings = 0
nErrors = 0
# For-loop: All x-axis bins
for i in range(h.GetNbinsX()):
# For-loop: All y-axis bins
for j in range(h.GetNbinsY()):
ykey = self._dqmKeys.keys()[j]
xkey = self._dqmKeys[ykey].keys()[i]
# Set the bin content
h.SetBinContent(i+1, j+1, self._dqmKeys[ykey][xkey])
h.GetXaxis().SetBinLabel(i+1, xkey)
h.GetYaxis().SetBinLabel(j+1, ykey)
if self._dqmKeys[ykey][xkey] > 2:
nErrors += 1
elif self._dqmKeys[ykey][xkey] > 1:
nWarnings += 1
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(False)
style.setGridY(False)
style.setWide(True, 0.15)
# Set the colour styling (red, yellow, green)
if 0:
palette = array.array("i", [ROOT.kGreen+1, ROOT.kYellow, ROOT.kRed])
ROOT.gStyle.SetPalette(3, palette)
else:
# https://root.cern.ch/doc/master/classTColor.html
ROOT.gStyle.SetPalette(ROOT.kLightTemperature)
# ROOT.gStyle.SetPalette(ROOT.kColorPrintableOnGrey)
#tdrstyle.setRainBowPalette()
#tdrstyle.setDeepSeaPalette()
# Create canvas
c = ROOT.TCanvas()
c.SetLogx(False)
c.SetLogy(False)
c.SetLogz(True)
c.SetGridx()
c.SetGridy()
h.Draw("COLZ") #"COLZ TEXT"
# Add CMS text and text with colour keys
histograms.addStandardTexts(cmsTextPosition="outframe")
if 0:
histograms.addText(0.55, 0.80, "green < %.0f %%" % (okay*100), size=20)
histograms.addText(0.55, 0.84, "yellow < %.0f %%" % (warn*100), size=20)
histograms.addText(0.55, 0.88, "red > %.0f %%" % (warn*100), size=20)
# Save the canvas to a file
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
plotName = os.path.join(self._plotDirName, "FakeBNormalisationDQM")
# For-loop: Save formats
for ext in ["png", "C", "pdf"]:
saveName ="%s.%s" % (plotName, ext)
c.Print(saveName)
ROOT.gErrorIgnoreLevel = backup
ROOT.gStyle.SetPalette(1)
msg = "Obtained %d warnings and %d errors for the normalisation" % (nWarnings, nErrors)
self.Verbose(msg)
if nWarnings > 0:
msg = "DQM has %d warnings and %d errors! Please have a look at %s.png." % (nWarnings, nErrors, os.path.basename(plotName))
self.Verbose(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(), True)
#if nWarnings > 0 or nErrors > 0:
if nErrors > 0:
msg = "DQM has %d warnings and %d errors! Please have a look at %s.png." % (nWarnings, nErrors, os.path.basename(plotName))
self.Verbose(ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(), True)
return
def _addDqmEntry(self, binLabel, name, value, okTolerance, warnTolerance):
# Define colour codes
red = 2.5
yellow = 1.5
green = 0.5
if not binLabel in self._dqmKeys.keys():
self._dqmKeys[binLabel] = OrderedDict()
result = red
if abs(value) > okTolerance:
result = green
elif abs(value) > warnTolerance:
result = yellow
else:
pass
#self._dqmKeys[binLabel][name] = result
self._dqmKeys[binLabel][name] = value
return
def _getSanityCheckTextForFractions(self, dataTemplate, binLabel, saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
NOTE: Should one divide the fractions with dataTemplate.getFittedParameters()[0] ?
Right now not because the correction is so small.
'''
self.Verbose("_getSanityCheckTextForFractions()", True)
# Get variables
label = "QCD"
fraction = dataTemplate.getFittedParameters()[1]
fractionError = dataTemplate.getFittedParameterErrors()[1]
nBaseline = self._templates["%s_Baseline" % label].getNeventsFromHisto(False)
nCalculated = fraction * dataTemplate.getNeventsFromHisto(False)
if nCalculated > 0:
ratio = nBaseline / nCalculated
else:
ratio = 0
lines = []
lines.append("Fitted %s fraction: %f +- %f" % (label, fraction, fractionError))
lines.append("Sanity check: ratio = %.3f: baseline = %.1f vs. fitted = %.1f" % (ratio, nBaseline, nCalculated))
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
def _checkOverallNormalization(self, template, binLabel, saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
'''
self.Verbose("_checkOverallNormalization()")
# Calculatotions
value = template.getFittedParameters()[0]
error = template.getFittedParameterErrors()[0]
# Definitions
lines = []
lines.append("The fitted overall normalization factor for purity is: (should be 1.0)")
lines.append("NormFactor = %f +/- %f" % (value, error))
self._addDqmEntry(binLabel, "OverallNormalization(par0)", value-1.0, 0.03, 0.10)
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
## Helper method to be called from parent class when calculating norm.coefficients
def _getResultOutput(self, binLabel):
lines = []
lines.append("Transfer Factor (bin=%s): %f +- %f" % (binLabel, self._TF[binLabel], self._TFError[binLabel]) )
lines.append("Normalisation (bin=%s) : %f +- %f" % (binLabel, self._FakeBNormalization[binLabel], self._FakeBNormalizationError[binLabel]))
# Store all information for later used (write to file)
self._commentLines.extend(lines)
return lines
class FakeBNormalizationManager:
'''
Base class for QCD measurement normalization from which specialized algorithm classes inherit
'''
def __init__(self,
binLabels,
resultDirName,
moduleInfoString,
verbose=False):
self._verbose = verbose
self._templates = {}
self._binLabels = binLabels
self._sources = {}
self._commentLines = []
self._NEvtsCR1 = {}
self._NEvtsCR1_Error = {}
self._NEvtsCR2 = {}
self._NEvtsCR2_Error = {}
self._NEvtsCR3 = {}
self._NEvtsCR3_Error = {}
self._NEvtsCR4 = {}
self._NEvtsCR4_Error = {}
self._TF = {} # Transfer Factor (TF)
self._TF_Error = {}
self._TF_Up = {}
self._TF_Down = {}
self._dqmKeys = OrderedDict()
self._myPath = os.path.join(resultDirName, "normalisationPlots")
self._BinLabelMap = {}
self._FakeBNormalization = {} # for the time being same as TF
self._FakeBNormalizationError = {
} # for the time being same as TF_Error
self._FakeBNormalizationUp = {} # for the time being same as TF_Up
self._FakeBNormalizationDown = {} # for the time being same as TF_Down
if not isinstance(binLabels, list):
raise Exception("Error: binLabels needs to be a list of strings")
self.Verbose("__init__")
# No optimisation mode
if moduleInfoString == "":
moduleInfoString = "Default"
if not os.path.exists(self._myPath):
self.Print("Creating new directory %s" % (self._myPath), True)
os.mkdir(self._myPath)
self._plotDirName = os.path.join(resultDirName, "normalisationPlots",
moduleInfoString)
# If already exists, Delete an entire directory tree
if os.path.exists(self._plotDirName):
msg = "Removing directory tree %s" % (self._plotDirName)
self.Verbose(
ShellStyles.NoteStyle() + msg + ShellStyles.NormalStyle(),
True)
shutil.rmtree(self._plotDirName)
msg = "Creating directory %s" % (self._plotDirName)
self.Verbose(
ShellStyles.SuccessStyle() + msg + ShellStyles.NormalStyle(),
False)
os.mkdir(self._plotDirName)
return
def Print(self, msg, printHeader=False):
fName = __file__.split("/")[-1]
if printHeader == True:
# print "=== ", fName + ": class " + self.__class__.__name__
print "=== ", fName
print "\t", msg
else:
print "\t", msg
return
def Verbose(self, msg, printHeader=True, verbose=False):
if not self._verbose:
return
self.Print(msg, printHeader)
return
def GetQCDNormalization(self, binLabel):
if binLabel in self._TF.keys():
return self._TF[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! " %
(binLabel))
def GetTransferFactor(self, binLabel):
return self.GetQCDNormalization(binLabel)
def GetQCDNormalizationError(self, binLabel):
if binLabel in self._TF_Error.keys():
return self._TF_Error[binLabel]
else:
raise Exception("Error: _TF dictionary has no key \"%s\"! " %
(binLabel))
def CalculateTransferFactor(self,
binLabel,
hFakeB_CR1,
hFakeB_CR2,
hFakeB_CR3,
hFakeB_CR4,
verbose=False):
'''
Calculates the combined normalization and, if specified,
varies it up or down by factor (1+variation)
TF = Transfer Factor
SR = Signal Region
CR = Control Region
VR = Verification Region
'''
self.verbose = verbose
# Obtain counts for QCD and EWK fakes
lines = []
# NOTES: Add EWKGenuineB TF, Add Data TF, add QCD TF, Add EWK TF, add MCONLY TFs
nCR1_Error = ROOT.Double(0.0)
nCR2_Error = ROOT.Double(0.0)
nCR3_Error = ROOT.Double(0.0)
nCR4_Error = ROOT.Double(0.0)
# Get Events in all CRs and their associated errors
nCR1 = hFakeB_CR1.IntegralAndError(1,
hFakeB_CR1.GetNbinsX() + 1,
nCR1_Error)
nCR2 = hFakeB_CR2.IntegralAndError(1,
hFakeB_CR2.GetNbinsX() + 1,
nCR2_Error)
nCR3 = hFakeB_CR3.IntegralAndError(1,
hFakeB_CR3.GetNbinsX() + 1,
nCR3_Error)
nCR4 = hFakeB_CR4.IntegralAndError(1,
hFakeB_CR4.GetNbinsX() + 1,
nCR4_Error)
# Calculate Transfer Factor (TF) from Control Region (R) to Signal Region (SR): R = N_CR1/ N_CR2
TF = None
TF_Up = None
TF_Down = None
TF_Error = None
TF = (nCR1 / nCR2)
TF_Error = errorPropagation.errorPropagationForDivision(
nCR1, nCR1_Error, nCR2, nCR2_Error)
TF_Up = TF + TF_Error
if TF_Up > 1.0:
TF_Up = 1.0
TF_Down = TF - TF_Error
if TF_Down < 0.0:
TF_Down = 0.0
lines.append("TF (bin=%s) = N_CR1 / N_CR2 = %f / %f = %f +- %f" %
(binLabel, nCR1, nCR2, TF, TF_Error))
# Calculate the transfer factors (R_{i}) where i is index of bin the Fake-b measurement is made in (pT and/or eta of ldg b-jet)
if TF != None:
# Replace bin label with histo title (has exact binning info)
self._BinLabelMap[binLabel] = self.getNiceBinLabel(
hFakeB_CR2.GetTitle())
self._NEvtsCR1[binLabel] = nCR1
self._NEvtsCR1_Error[binLabel] = nCR1_Error
self._NEvtsCR2[binLabel] = nCR2
self._NEvtsCR2_Error[binLabel] = nCR2_Error
self._NEvtsCR3[binLabel] = nCR3
self._NEvtsCR3_Error[binLabel] = nCR3_Error
self._NEvtsCR4[binLabel] = nCR4
self._NEvtsCR4_Error[binLabel] = nCR4_Error
self._TF[binLabel] = TF
self._TF_Error[binLabel] = TF_Error
self._TF_Up[binLabel] = TF_Up
self._TF_Down[binLabel] = TF_Down
self._FakeBNormalization[binLabel] = TF # TF
self._FakeBNormalizationError[binLabel] = TF_Error # Error(TF)
self._FakeBNormalizationUp[binLabel] = TF_Up # TF + Error
self._FakeBNormalizationDown[binLabel] = TF_Down # TF - Error
# Store all information for later used (write to file)
self._commentLines.extend(lines)
# Print output and store comments
if 0:
for i, line in enumerate(lines, 1):
Print(line, i == 1)
return
def getNiceBinLabel(self, binLabel):
newLabel = binLabel.replace("abs", "")
if 1:
newLabel = newLabel.replace("TetrajetBjet", " ")
newLabel = newLabel.replace("TetrajetBJet", " ")
else: #more room
newLabel = newLabel.replace("TetrajetBjet", "b-jet ")
newLabel = newLabel.replace("TetrajetBJet", "b-jet ")
newLabel = newLabel.replace("_", " ")
newLabel = newLabel.replace("(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("Eta", "#eta")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace("CRtwo", "")
return newLabel
def writeTransferFactorsToFile(self, filename, opts):
'''
Save the fit results for QCD and EWK.
The results will are stored in a python file starting with name:
"QCDInvertedNormalizationFactors_" + moduleInfoString
The script also summarizes warnings and errors encountered:
- Green means deviation from normal is 0-3 %,
- Yellow means deviation of 3-10 %, and
- Red means deviation of >10 % (i.e. something is clearly wrong).
If necessary, do adjustments to stabilize the fits to get rid of the errors/warnings.
The first things to work with are:
a) Make sure enough events are in the histograms used
b) Adjust fit parameters and/or fit functions and re-fit results
Move on only once you are pleased with the normalisation coefficients
'''
s = ""
s += "# Generated on %s\n" % datetime.datetime.now().ctime()
s += "# by %s\n" % os.path.basename(sys.argv[0])
s += "\n"
s += "import sys\n"
s += "\n"
s += "def FakeBNormalisationSafetyCheck(era, searchMode, optimizationMode):\n"
s += " validForEra = \"%s\"\n" % opts.dataEra
s += " validForSearchMode = \"%s\"\n" % opts.searchMode
s += " validForOptMode = \"%s\"\n" % opts.optMode
s += " if not era == validForEra:\n"
s += " raise Exception(\"Error: inconsistent era, normalisation factors valid for\",validForEra,\"but trying to use with\",era)\n"
s += " if not searchMode == validForSearchMode:\n"
s += " raise Exception(\"Error: inconsistent search mode, normalisation factors valid for\",validForSearchMode,\"but trying to use with\",searchMode)\n"
s += " if not optimizationMode == validForOptMode:\n"
s += " raise Exception(\"Error: inconsistent optimization mode, normalisation factors valid for\",validForOptMode,\"but trying to use with\",optimizationMode)\n"
s += " return"
s += "\n"
s += "\n"
# First write the transfer factor (for each Fake-b measurement bin)
s += "FakeBNormalisation_Value = {\n"
for binLabel in self._TF:
s += ' "%s": %f,\n' % (binLabel, self._TF[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factor error (for each Fake-b measurement bin)
s += "FakeBNormalisation_Error = {\n"
for binLabel in self._TF_Error:
s += ' "%s": %f,\n' % (binLabel, self._TF_Error[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors + error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorUp = {\n"
for binLabel in self._TF_Up:
s += ' "%s": %f,\n' % (binLabel, self._TF_Up[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors - error (for each Fake-b measurement bin)
s += "FakeBNormalisation_ErrorDown = {\n"
for binLabel in self._TF_Down:
s += ' "%s": %f,\n' % (binLabel, self._TF_Down[binLabel])
s += "}\n"
s += "\n"
# Then write the transfer factors - error (for each Fake-b measurement bin)
s += "BinLabelMap = {\n"
for binLabel in self._BinLabelMap:
s += ' "%s": \"%s\",\n' % (binLabel,
self._BinLabelMap[binLabel])
s += "}\n"
s += "\n"
self.Verbose("Writing results in file %s" % filename, True)
fOUT = open(filename, "w")
fOUT.write(s)
fOUT.write("'''\n")
for l in self._commentLines:
fOUT.write(l + "\n")
fOUT.write("'''\n")
fOUT.close()
msg = "Results written in file %s" % (
ShellStyles.SuccessStyle() + filename + ShellStyles.NormalStyle())
self.Print(msg, True)
# Create the transfer factors plot (for each bin of FakeB measurement)
self._generateTransferFactorsPlot()
self._generateDQMPlot()
return
def _generateTransferFactorsPlot(self):
'''
The resulting plot will contain the transfer factors (y-axis) for a given measurement bin (x-axis)
This is needed in the case the Fake-b measurement is done in
bins of a correlated quantity (e.g. eta of leading b-jet in Fake-b for HToTB, and tau-pT in the
case of inverted tau isolatio for HToTau)
'''
def makeGraph(markerStyle, color, binList, valueDict, upDict,
downDict):
g = ROOT.TGraphAsymmErrors(len(binList))
for i in range(len(binList)):
g.SetPoint(i, i + 0.5, valueDict[binList[i]])
g.SetPointEYhigh(i, upDict[binList[i]])
g.SetPointEYlow(i, downDict[binList[i]])
g.SetMarkerSize(1.2)
g.SetMarkerStyle(markerStyle)
g.SetLineColor(color)
g.SetLineWidth(3)
g.SetMarkerColor(color)
return g
# Obtain bin list in right order
keyList = []
keys = self._TF.keys()
keys.sort()
# for k in keys:
# if "lt" in k:
# keyList.append(k)
# for k in keys:
# if "eq" in k:
# keyList.append(k)
# for k in keys:
# if "gt" in k:
# keyList.append(k)
# For-loop: All Fake-b measurement bins
for k in keys:
keyList.append(k)
#if "Inclusive" in keys:
# keyList.append("Inclusive")
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setGridX(False)
style.setGridY(False)
style.setOptStat(False)
# Create graphs
gFakeB = makeGraph(ROOT.kFullCircle, ROOT.kAzure, keyList, self._TF,
self._TF_Error, self._TF_Error)
# Make plot
hFrame = ROOT.TH1F("frame", "frame", len(keyList), 0, len(keyList))
# Change bin labels to text
for i, binLabel in enumerate(keyList, 1):
binLabelText = self.getFormattedBinLabelString(binLabel)
hFrame.GetXaxis().SetBinLabel(i, binLabelText)
# Set axes names
hFrame.GetYaxis().SetTitle("transfer factor") #R_{i}
# hFrame.GetXaxis().SetTitle("Fake-b bin")
# Customise axes
logy = False
if logy:
hFrame.SetMinimum(0.6e-1)
hFrame.SetMaximum(2e0)
else:
hFrame.SetMinimum(0.0)
hFrame.SetMaximum(1.0)
if len(self._BinLabelMap) > 12:
lSize = 8
elif len(self._BinLabelMap) > 8:
lSize = 12
else:
lSize = 16
hFrame.GetXaxis().SetLabelSize(lSize) # 20
hFrame.GetXaxis().LabelsOption("d")
# Label Style options
# "a" sort by alphabetic order
# ">" sort by decreasing values
# "<" sort by increasing values
# "h" draw labels horizonthal
# "v" draw labels vertical
# "u" draw labels up (end of label right adjusted)
# "d" draw labels down (start of label left adjusted)
# Create canvas
c = ROOT.TCanvas()
c.SetLogy(logy)
c.SetGridx(False)
c.SetGridy(False)
hFrame.Draw()
gFakeB.Draw("p same")
histograms.addStandardTexts(cmsTextPosition="outframe")
# Create the legend & draw it
l = ROOT.TLegend(0.65, 0.80, 0.90, 0.90)
l.SetFillStyle(-1)
l.SetBorderSize(0)
l.AddEntry(gFakeB, "Value #pm stat.",
"LP") # "Fake-#it{b} #pm Stat.", "LP"
l.SetTextSize(0.035)
if 0:
l.Draw()
# Store ROOT ignore level to normal before changing it
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
# Save the plot
for item in ["png", "C", "pdf"]:
c.Print(self._plotDirName +
"/FakeBNormalisationCoefficients.%s" % item)
# Reset the ROOT ignore level to normal
ROOT.gErrorIgnoreLevel = backup
# Inform user
msg = "Plot saved under %s" % (ShellStyles.SuccessStyle() +
self._plotDirName + "/" +
ShellStyles.NormalStyle())
self.Print(msg, True)
return
def getFormattedBinLabelString(self, binLabel):
'''
Dirty trick to get what I want
'''
if binLabel not in self._BinLabelMap:
raise Exception("Got unexpected bin label \"%s\"!" % binLabel)
newLabel = self._BinLabelMap[binLabel]
newLabel = newLabel.replace("abs(", "|")
newLabel = newLabel.replace(")", "|")
newLabel = newLabel.replace("..", "-")
newLabel = newLabel.replace(":", ",")
newLabel = newLabel.replace("TetrajetBJet", "") #"b^{ldg} ")
newLabel = newLabel.replace("Pt", "p_{T} ")
newLabel = newLabel.replace("Eta", "#eta ")
if "inclusive" in binLabel.lower():
newLabel = "Inclusive"
return newLabel
def _generateDQMPlot(self):
'''
Create a Data Quality Monitor (DQM) style plot
to easily check the error for each transfer factor
and whether it is within an acceptable relative error
'''
# Define error warning/tolerance on relative errors
okay = 1.0 / (len(self._BinLabelMap.keys()))
warn = 0.5 * okay
NEvts = []
# Check the uncertainties on the normalization factors
for k in self._BinLabelMap:
relErrorUp = abs(self._TF_Up[k]) / (self._TF[k])
relErrorDown = abs(self._TF_Down[k]) / (self._TF[k])
relError = self._TF_Error[k] / self._TF[k]
if 0:
print "bin = %s , relErrorUp = %s, relErrorDown = %s " % (
k, relErrorUp, relErrorDown)
# Add DQM entries
NCR1 = 0
NCR2 = 0
NCR3 = 0
NCR4 = 0
for j in self._NEvtsCR1:
NCR1 += self._NEvtsCR1[j]
NEvts.append(self._NEvtsCR1[j])
for j in self._NEvtsCR2:
NCR2 += self._NEvtsCR2[j]
NEvts.append(self._NEvtsCR2[j])
for j in self._NEvtsCR3:
NCR3 += self._NEvtsCR3[j]
NEvts.append(self._NEvtsCR3[j])
for j in self._NEvtsCR4:
NCR4 += self._NEvtsCR4[j]
NEvts.append(self._NEvtsCR4[j])
if 0:
print "NCR1[%s] = %0.1f, NCR2[%s] = %0.1f, k = %s" % (
k, self._NEvtsCR1[k], k, self._NEvtsCR2[k], k)
print "NCR1 = %s, NCR2 = %s, k = %s" % (NCR1, NCR2, k)
print "error/NCR1[%s] = %0.2f, error/NCR2[%s] = %0.2f" % (
k, self._NEvtsCR1_Error[k] / self._NEvtsCR1[k], k,
self._NEvtsCR2_Error[k] / self._NEvtsCR2[k])
# Add DQM plot entries
self._addDqmEntry(self._BinLabelMap[k], "N_{CR1}",
self._NEvtsCR1[k], NCR1 * okay, NCR1 * warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR2}",
self._NEvtsCR2[k], NCR2 * okay, NCR2 * warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR3}",
self._NEvtsCR3[k], NCR3 * okay, NCR3 * warn)
self._addDqmEntry(self._BinLabelMap[k], "N_{CR4}",
self._NEvtsCR4[k], NCR4 * okay, NCR4 * warn)
# self._addDqmEntry(self._BinLabelMap[k], "#frac{#sigma_{CR1}}{N_{CR1}}", self._NEvtsCR1_Error[k]/NCR1, 0.05, 0.15)
# self._addDqmEntry(self._BinLabelMap[k], "#frac{#sigma_{CR2}}{N_{CR2}}", self._NEvtsCR2_Error[k]/NCR2, 0.05, 0.15)
# Construct the DQM histogram
nBinsX = len(self._dqmKeys[self._dqmKeys.keys()[0]].keys())
nBinsY = len(self._dqmKeys.keys())
h = ROOT.TH2F("FakeB DQM", "FakeB DQM", nBinsX, 0, nBinsX, nBinsY, 0,
nBinsY)
# Customise axes
h.GetXaxis().SetLabelSize(15)
h.GetYaxis().SetLabelSize(10)
# Set Min and Max of z-axis
if 0: # red, yellow, green for DQM
h.SetMinimum(0)
h.SetMaximum(3)
else: # pure entries instead of red, yellow, green
h.SetMinimum(min(NEvts) * 0.25)
h.SetMaximum(round(max(NCR1, NCR2, NCR3, NCR4)))
h.SetContour(10)
#h.SetContour(3)
if 0:
h.GetXaxis().LabelsOption("v")
h.GetYaxis().LabelsOption("v")
nWarnings = 0
nErrors = 0
# For-loop: All x-axis bins
for i in range(h.GetNbinsX()):
# For-loop: All y-axis bins
for j in range(h.GetNbinsY()):
ykey = self._dqmKeys.keys()[j]
xkey = self._dqmKeys[ykey].keys()[i]
# Set the bin content
h.SetBinContent(i + 1, j + 1, self._dqmKeys[ykey][xkey])
h.GetXaxis().SetBinLabel(i + 1, xkey)
h.GetYaxis().SetBinLabel(j + 1, ykey)
if self._dqmKeys[ykey][xkey] > 2:
nErrors += 1
elif self._dqmKeys[ykey][xkey] > 1:
nWarnings += 1
# Apply TDR style
style = tdrstyle.TDRStyle()
style.setOptStat(False)
style.setGridX(False)
style.setGridY(False)
style.setWide(True, 0.15)
# Set the colour styling (red, yellow, green)
if 0:
palette = array.array("i",
[ROOT.kGreen + 1, ROOT.kYellow, ROOT.kRed])
ROOT.gStyle.SetPalette(3, palette)
else:
# https://root.cern.ch/doc/master/classTColor.html
ROOT.gStyle.SetPalette(ROOT.kLightTemperature)
# ROOT.gStyle.SetPalette(ROOT.kColorPrintableOnGrey)
#tdrstyle.setRainBowPalette()
#tdrstyle.setDeepSeaPalette()
# Create canvas
c = ROOT.TCanvas()
c.SetLogx(False)
c.SetLogy(False)
c.SetLogz(True)
c.SetGridx()
c.SetGridy()
h.Draw("COLZ") #"COLZ TEXT"
# Add CMS text and text with colour keys
histograms.addStandardTexts(cmsTextPosition="outframe")
if 0:
histograms.addText(0.55,
0.80,
"green < %.0f %%" % (okay * 100),
size=20)
histograms.addText(0.55,
0.84,
"yellow < %.0f %%" % (warn * 100),
size=20)
histograms.addText(0.55,
0.88,
"red > %.0f %%" % (warn * 100),
size=20)
# Save the canvas to a file
backup = ROOT.gErrorIgnoreLevel
ROOT.gErrorIgnoreLevel = ROOT.kWarning
plotName = os.path.join(self._plotDirName, "FakeBNormalisationDQM")
# For-loop: Save formats
for ext in ["png", "C", "pdf"]:
saveName = "%s.%s" % (plotName, ext)
c.Print(saveName)
ROOT.gErrorIgnoreLevel = backup
ROOT.gStyle.SetPalette(1)
msg = "Obtained %d warnings and %d errors for the normalisation" % (
nWarnings, nErrors)
self.Verbose(msg)
if nWarnings > 0:
msg = "DQM has %d warnings and %d errors! Please have a look at %s.png." % (
nWarnings, nErrors, os.path.basename(plotName))
self.Verbose(
ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(),
True)
#if nWarnings > 0 or nErrors > 0:
if nErrors > 0:
msg = "DQM has %d warnings and %d errors! Please have a look at %s.png." % (
nWarnings, nErrors, os.path.basename(plotName))
self.Verbose(
ShellStyles.ErrorStyle() + msg + ShellStyles.NormalStyle(),
True)
return
def _addDqmEntry(self, binLabel, name, value, okTolerance, warnTolerance):
# Define colour codes
red = 2.5
yellow = 1.5
green = 0.5
if not binLabel in self._dqmKeys.keys():
self._dqmKeys[binLabel] = OrderedDict()
result = red
if abs(value) > okTolerance:
result = green
elif abs(value) > warnTolerance:
result = yellow
else:
pass
#self._dqmKeys[binLabel][name] = result
self._dqmKeys[binLabel][name] = value #iro
return
def _getSanityCheckTextForFractions(self,
dataTemplate,
binLabel,
saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
NOTE: Should one divide the fractions with dataTemplate.getFittedParameters()[0] ?
Right now not because the correction is so small.
'''
self.Verbose("_getSanityCheckTextForFractions()", True)
# Get variables
label = "QCD"
fraction = dataTemplate.getFittedParameters()[1]
fractionError = dataTemplate.getFittedParameterErrors()[1]
nBaseline = self._templates["%s_Baseline" %
label].getNeventsFromHisto(False)
nCalculated = fraction * dataTemplate.getNeventsFromHisto(False)
if nCalculated > 0:
ratio = nBaseline / nCalculated
else:
ratio = 0
lines = []
lines.append("Fitted %s fraction: %f +- %f" %
(label, fraction, fractionError))
lines.append(
"Sanity check: ratio = %.3f: baseline = %.1f vs. fitted = %.1f" %
(ratio, nBaseline, nCalculated))
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
def _checkOverallNormalization(self,
template,
binLabel,
saveToComments=False):
'''
Helper method to be called from parent class when calculating norm.coefficients
'''
self.Verbose("_checkOverallNormalization()")
# Calculatotions
value = template.getFittedParameters()[0]
error = template.getFittedParameterErrors()[0]
# Definitions
lines = []
lines.append(
"The fitted overall normalization factor for purity is: (should be 1.0)"
)
lines.append("NormFactor = %f +/- %f" % (value, error))
self._addDqmEntry(binLabel, "OverallNormalization(par0)", value - 1.0,
0.03, 0.10)
# Store all information for later used (write to file)
if saveToComments:
self._commentLines.extend(lines)
return lines
## Helper method to be called from parent class when calculating norm.coefficients
def _getResultOutput(self, binLabel):
lines = []
lines.append("Transfer Factor (bin=%s): %f +- %f" %
(binLabel, self._TF[binLabel], self._TFError[binLabel]))
lines.append("Normalisation (bin=%s) : %f +- %f" %
(binLabel, self._FakeBNormalization[binLabel],
self._FakeBNormalizationError[binLabel]))
# Store all information for later used (write to file)
self._commentLines.extend(lines)
return lines