pad.SetFillStyle(0) pad.Draw() pad.cd(0) frame2.SetMinimum(-5) frame2.SetMaximum(5) frame2.GetYaxis().SetNdivisions(505) frame2.GetXaxis().SetTitleOffset(0.9) frame2.GetYaxis().SetTitleOffset(0.8) frame2.GetYaxis().SetTickLength(0.06) frame2.GetYaxis().SetTitleSize(0.05) frame2.GetYaxis().SetLabelSize(0.03) frame2.GetYaxis().SetTitle('(Data-Fit)/Error') frame2.GetXaxis().SetTitle('m_{jj} (GeV)') frame2.Draw() parsBkg = background.getParameters(roohistBkg) if bkgConst: parsBkg.setAttribAll('Constant', True) if histpdfBkg: # ----------------------------------------- # hist pdf bkg canBhist_name = 'can_Mjj' + str(mass) canBhist = TCanvas(canSname, canSname, 900, 600) gPad.SetLogy() roohistBkg = RooDataHist('roohist', 'roohist', RooArgList(x), hBkg) roohistBkg.Print() background = RooHistPdf('background', 'background', RooArgSet(x), roohistBkg)
# decorrelated background parameters for Bayesian limits if args.decoBkg: signal_norm.setConstant() res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) res.Print() ## temp workspace for the PDF diagonalizer w_tmp = RooWorkspace("w_tmp") deco = PdfDiagonalizer("deco", w_tmp, res) # here diagonalizing only the shape parameters since the overall normalization is already decorrelated background_deco = deco.diagonalize(background) print "##################### workspace for decorrelation" w_tmp.Print("v") print "##################### original parameters" background.getParameters(rooDataHist).Print("v") print "##################### decorrelated parameters" # needed if want to evaluate limits without background systematics if args.fixBkg: w_tmp.var("deco_eig1").setConstant() w_tmp.var("deco_eig2").setConstant() if not args.fixP3: w_tmp.var("deco_eig3").setConstant() background_deco.getParameters(rooDataHist).Print("v") print "##################### original pdf" background.Print() print "##################### decorrelated pdf" background_deco.Print() # release signal normalization signal_norm.setConstant(kFALSE) # set the background normalization range to +/- 5 sigma bkg_val = background_norm.getVal()
pad.cd(0) frame2.SetMinimum(-5) frame2.SetMaximum(5) frame2.GetYaxis().SetNdivisions(505) frame2.GetXaxis().SetTitleOffset(0.9) frame2.GetYaxis().SetTitleOffset(0.8) frame2.GetYaxis().SetTickLength(0.06) frame2.GetYaxis().SetTitleSize(0.05) frame2.GetYaxis().SetLabelSize(0.03) frame2.GetYaxis().SetTitle('(Data-Fit)/Error') frame2.GetXaxis().SetTitle('m_{jj} (GeV)') if useSub: frame2.GetXaxis().SetRangeUser(900,2500) frame2.Draw(); parsBkg = background.getParameters(roohistBkg) parsBkg.setAttribAll('Constant', True) if fitSig and fitDat: # ----------------------------------------- # write everything to a workspace to make a datacard dcFN = 'RS'+str(mass)+'_datacard.txt' wsFN = 'RS'+str(mass)+'_workspace.root' if useSub: dcFN = 'RS'+str(mass)+'_sub_datacard.txt' wsFN = 'RS'+str(mass)+'_sub_workspace.root' nObs = roohistBkg.sumEntries();
def main(): # usage description usage = "Example: ./scripts/createDatacards.py --inputData inputs/rawhistV7_Run2015D_scoutingPFHT_UNBLINDED_649_838_JEC_HLTplusV7_Mjj_cor_smooth.root --dataHistname mjj_mjjcor_gev --inputSig inputs/ResonanceShapes_gg_13TeV_Scouting_Spring15.root -f gg -o datacards -l 1866 --lumiUnc 0.027 --massrange 1000 1500 50 --runFit --p1 5 --p2 7 --p3 0.4 --massMin 838 --massMax 2037 --fitStrategy 2" # input parameters parser = ArgumentParser(description='Script that creates combine datacards and corresponding RooFit workspaces',epilog=usage) parser.add_argument("--inputData", dest="inputData", required=True, help="Input data spectrum", metavar="INPUT_DATA") parser.add_argument("--dataHistname", dest="dataHistname", required=True, help="Data histogram name", metavar="DATA_HISTNAME") parser.add_argument("--inputSig", dest="inputSig", required=True, help="Input signal shapes", metavar="INPUT_SIGNAL") parser.add_argument("-f", "--final_state", dest="final_state", required=True, help="Final state (e.g. qq, qg, gg)", metavar="FINAL_STATE") parser.add_argument("-f2", "--type", dest="atype", required=True, help="Type (e.g. hG, lG, hR, lR)") parser.add_argument("-o", "--output_path", dest="output_path", required=True, help="Output path where datacards and workspaces will be stored", metavar="OUTPUT_PATH") parser.add_argument("-l", "--lumi", dest="lumi", required=True, default=1000., type=float, help="Integrated luminosity in pb-1 (default: %(default).1f)", metavar="LUMI") parser.add_argument("--massMin", dest="massMin", default=500, type=int, help="Lower bound of the mass range used for fitting (default: %(default)s)", metavar="MASS_MIN") parser.add_argument("--massMax", dest="massMax", default=1200, type=int, help="Upper bound of the mass range used for fitting (default: %(default)s)", metavar="MASS_MAX") parser.add_argument("--p1", dest="p1", default=5.0000e-03, type=float, help="Fit function p1 parameter (default: %(default)e)", metavar="P1") parser.add_argument("--p2", dest="p2", default=9.1000e+00, type=float, help="Fit function p2 parameter (default: %(default)e)", metavar="P2") parser.add_argument("--p3", dest="p3", default=5.0000e-01, type=float, help="Fit function p3 parameter (default: %(default)e)", metavar="P3") parser.add_argument("--lumiUnc", dest="lumiUnc", required=True, type=float, help="Relative uncertainty in the integrated luminosity", metavar="LUMI_UNC") parser.add_argument("--jesUnc", dest="jesUnc", type=float, help="Relative uncertainty in the jet energy scale", metavar="JES_UNC") parser.add_argument("--jerUnc", dest="jerUnc", type=float, help="Relative uncertainty in the jet energy resolution", metavar="JER_UNC") parser.add_argument("--sqrtS", dest="sqrtS", default=13000., type=float, help="Collision center-of-mass energy (default: %(default).1f)", metavar="SQRTS") parser.add_argument("--fixP3", dest="fixP3", default=False, action="store_true", help="Fix the fit function p3 parameter") parser.add_argument("--runFit", dest="runFit", default=False, action="store_true", help="Run the fit") parser.add_argument("--fitBonly", dest="fitBonly", default=False, action="store_true", help="Run B-only fit") parser.add_argument("--fixBkg", dest="fixBkg", default=False, action="store_true", help="Fix all background parameters") parser.add_argument("--decoBkg", dest="decoBkg", default=False, action="store_true", help="Decorrelate background parameters") parser.add_argument("--fitStrategy", dest="fitStrategy", type=int, default=1, help="Fit strategy (default: %(default).1f)") parser.add_argument("--debug", dest="debug", default=False, action="store_true", help="Debug printout") parser.add_argument("--postfix", dest="postfix", default='', help="Postfix for the output file names (default: %(default)s)") parser.add_argument("--pyes", dest="pyes", default=False, action="store_true", help="Make files for plots") parser.add_argument("--jyes", dest="jyes", default=False, action="store_true", help="Make files for JES/JER plots") parser.add_argument("--pdir", dest="pdir", default='testarea', help="Name a directory for the plots (default: %(default)s)") parser.add_argument("--chi2", dest="chi2", default=False, action="store_true", help="Compute chi squared") parser.add_argument("--widefit", dest="widefit", default=False, action="store_true", help="Fit with wide bin hist") mass_group = parser.add_mutually_exclusive_group(required=True) mass_group.add_argument("--mass", type=int, nargs = '*', default = 1000, help="Mass can be specified as a single value or a whitespace separated list (default: %(default)i)" ) mass_group.add_argument("--massrange", type=int, nargs = 3, help="Define a range of masses to be produced. Format: min max step", metavar = ('MIN', 'MAX', 'STEP') ) mass_group.add_argument("--masslist", help = "List containing mass information" ) args = parser.parse_args() if args.atype == 'hG': fstr = "bbhGGBB" in2 = 'bcorrbin/binmodh.root' elif args.atype == 'hR': fstr = "bbhRS" in2 = 'bcorrbin/binmodh.root' elif args.atype == 'lG': fstr = "bblGGBB" in2 = 'bcorrbin/binmodl.root' else: fstr = "bblRS" in2 = 'bcorrbin/binmodl.root' # check if the output directory exists if not os.path.isdir( os.path.join(os.getcwd(),args.output_path) ): os.mkdir( os.path.join(os.getcwd(),args.output_path) ) # mass points for which resonance shapes will be produced masses = [] if args.massrange != None: MIN, MAX, STEP = args.massrange masses = range(MIN, MAX+STEP, STEP) elif args.masslist != None: # A mass list was provided print "Will create mass list according to", args.masslist masslist = __import__(args.masslist.replace(".py","")) masses = masslist.masses else: masses = args.mass # sort masses masses.sort() # import ROOT stuff from ROOT import gStyle, TFile, TH1F, TH1D, TGraph, kTRUE, kFALSE, TCanvas, TLegend, TPad, TLine from ROOT import RooHist, RooRealVar, RooDataHist, RooArgList, RooArgSet, RooAddPdf, RooFit, RooGenericPdf, RooWorkspace, RooMsgService, RooHistPdf if not args.debug: RooMsgService.instance().setSilentMode(kTRUE) RooMsgService.instance().setStreamStatus(0,kFALSE) RooMsgService.instance().setStreamStatus(1,kFALSE) # input data file inputData = TFile(args.inputData) # input data histogram hData = inputData.Get(args.dataHistname) inData2 = TFile(in2) hData2 = inData2.Get('h_data') # input sig file inputSig = TFile(args.inputSig) sqrtS = args.sqrtS # mass variable mjj = RooRealVar('mjj','mjj',float(args.massMin),float(args.massMax)) # integrated luminosity and signal cross section lumi = args.lumi signalCrossSection = 1. # set to 1. so that the limit on r can be interpreted as a limit on the signal cross section for mass in masses: print ">> Creating datacard and workspace for %s resonance with m = %i GeV..."%(args.final_state, int(mass)) # get signal shape hSig = inputSig.Get( "h_" + args.final_state + "_" + str(int(mass)) ) # normalize signal shape to the expected event yield (works even if input shapes are not normalized to unity) hSig.Scale(signalCrossSection*lumi/hSig.Integral()) # divide by a number that provides roughly an r value of 1-10 rooSigHist = RooDataHist('rooSigHist','rooSigHist',RooArgList(mjj),hSig) print 'Signal acceptance:', (rooSigHist.sumEntries()/hSig.Integral()) signal = RooHistPdf('signal','signal',RooArgSet(mjj),rooSigHist) signal_norm = RooRealVar('signal_norm','signal_norm',0,-1e+05,1e+05) if args.fitBonly: signal_norm.setConstant() p1 = RooRealVar('p1','p1',args.p1,0.,100.) p2 = RooRealVar('p2','p2',args.p2,0.,60.) p3 = RooRealVar('p3','p3',args.p3,-10.,10.) p4 = RooRealVar('p4','p4',5.6,-50.,50.) p5 = RooRealVar('p5','p5',10.,-50.,50.) p6 = RooRealVar('p6','p6',.016,-50.,50.) p7 = RooRealVar('p7','p7',8.,-50.,50.) p8 = RooRealVar('p8','p8',.22,-50.,50.) p9 = RooRealVar('p9','p9',14.1,-50.,50.) p10 = RooRealVar('p10','p10',8.,-50.,50.) p11 = RooRealVar('p11','p11',4.8,-50.,50.) p12 = RooRealVar('p12','p12',7.,-50.,50.) p13 = RooRealVar('p13','p13',7.,-50.,50.) p14 = RooRealVar('p14','p14',7.,-50.,50.) p15 = RooRealVar('p15','p15',1.,-50.,50.) p16 = RooRealVar('p16','p16',9.,-50.,50.) p17 = RooRealVar('p17','p17',0.6,-50.,50.) if args.fixP3: p3.setConstant() background = RooGenericPdf('background','(pow(1-@0/%.1f,@1)/pow(@0/%.1f,@2+@3*log(@0/%.1f)))'%(sqrtS,sqrtS,sqrtS),RooArgList(mjj,p1,p2,p3)) dataInt = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm = RooRealVar('background_norm','background_norm',dataInt,0.,1e+08) background2 = RooGenericPdf('background2','(pow(@0/%.1f,-@1)*pow(1-@0/%.1f,@2))'%(sqrtS,sqrtS),RooArgList(mjj,p4,p5)) dataInt2 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm2 = RooRealVar('background_norm2','background_norm2',dataInt2,0.,1e+08) background3 = RooGenericPdf('background3','(1/pow(@1+@0/%.1f,@2))'%(sqrtS),RooArgList(mjj,p6,p7)) dataInt3 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm3 = RooRealVar('background_norm3','background_norm3',dataInt3,0.,1e+08) background4 = RooGenericPdf('background4','(1/pow(@1+@2*@0/%.1f+pow(@0/%.1f,2),@3))'%(sqrtS,sqrtS),RooArgList(mjj,p8,p9,p10)) dataInt4 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm4 = RooRealVar('background_norm4','background_norm4',dataInt4,0.,1e+08) background5 = RooGenericPdf('background5','(pow(@0/%.1f,-@1)*pow(1-pow(@0/%.1f,1/3),@2))'%(sqrtS,sqrtS),RooArgList(mjj,p11,p12)) dataInt5 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm5 = RooRealVar('background_norm5','background_norm5',dataInt5,0.,1e+08) background6 = RooGenericPdf('background6','(pow(@0/%.1f,2)+@1*@0/%.1f+@2)'%(sqrtS,sqrtS),RooArgList(mjj,p13,p14)) dataInt6 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm6 = RooRealVar('background_norm6','background_norm6',dataInt6,0.,1e+08) background7 = RooGenericPdf('background7','((-1+@1*@0/%.1f)*pow(@0/%.1f,@2+@3*log(@0/%.1f)))'%(sqrtS,sqrtS,sqrtS),RooArgList(mjj,p15,p16,p17)) dataInt7 = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)),hData.GetXaxis().FindBin(float(args.massMax))) background_norm7 = RooRealVar('background_norm7','background_norm7',dataInt7,0.,1e+08) # S+B model model = RooAddPdf("model","s+b",RooArgList(background,signal),RooArgList(background_norm,signal_norm)) model2 = RooAddPdf("model2","s+b2",RooArgList(background2,signal),RooArgList(background_norm2,signal_norm)) model3 = RooAddPdf("model3","s+b3",RooArgList(background3,signal),RooArgList(background_norm3,signal_norm)) model4 = RooAddPdf("model4","s+b4",RooArgList(background4,signal),RooArgList(background_norm4,signal_norm)) model5 = RooAddPdf("model5","s+b5",RooArgList(background5,signal),RooArgList(background_norm5,signal_norm)) model6 = RooAddPdf("model6","s+b6",RooArgList(background6,signal),RooArgList(background_norm6,signal_norm)) model7 = RooAddPdf("model7","s+b7",RooArgList(background7,signal),RooArgList(background_norm7,signal_norm)) rooDataHist = RooDataHist('rooDatahist','rooDathist',RooArgList(mjj),hData) if args.runFit: mframe = mjj.frame() rooDataHist.plotOn(mframe, ROOT.RooFit.Name("setonedata"), ROOT.RooFit.Invisible()) res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model.plotOn(mframe, ROOT.RooFit.Name("model1"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) res2 = model2.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model2.plotOn(mframe, ROOT.RooFit.Name("model2"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kOrange)) res3 = model3.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model3.plotOn(mframe, ROOT.RooFit.Name("model3"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) res4 = model4.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model4.plotOn(mframe, ROOT.RooFit.Name("model4"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) res5 = model5.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model5.plotOn(mframe, ROOT.RooFit.Name("model5"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kViolet)) res6 = model6.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) # model6.plotOn(mframe, ROOT.RooFit.Name("model6"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kPink)) res7 = model7.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) # model7.plotOn(mframe, ROOT.RooFit.Name("model7"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kAzure)) rooDataHist2 = RooDataHist('rooDatahist2','rooDathist2',RooArgList(mjj),hData2) rooDataHist2.plotOn(mframe, ROOT.RooFit.Name("data")) canvas = TCanvas("cdouble", "cdouble", 800, 1000) gStyle.SetOptStat(0); gStyle.SetOptTitle(0); top = TPad("top", "top", 0., 0.5, 1., 1.) top.SetBottomMargin(0.03) top.Draw() top.SetLogy() bottom = TPad("bottom", "bottom", 0., 0., 1., 0.5) bottom.SetTopMargin(0.02) bottom.SetBottomMargin(0.2) bottom.Draw() top.cd() frame_top = TH1D("frame_top", "frame_top", 100, 526, 1500) frame_top.GetXaxis().SetTitleSize(0) frame_top.GetXaxis().SetLabelSize(0) frame_top.GetYaxis().SetLabelSize(0.04) frame_top.GetYaxis().SetTitleSize(0.04) frame_top.GetYaxis().SetTitle("Events") frame_top.SetMaximum(1000.) frame_top.SetMinimum(0.1) frame_top.Draw("axis") mframe.Draw("p e1 same") bottom.cd() frame_bottom = TH1D("frame_bottom", "frame_bottom", 100, 526, 1500) frame_bottom.GetXaxis().SetTitle("m_{jj} [GeV]") frame_bottom.GetYaxis().SetTitle("Pull") frame_bottom.GetXaxis().SetLabelSize(0.04) frame_bottom.GetXaxis().SetTitleSize(0.06) frame_bottom.GetXaxis().SetLabelOffset(0.01) frame_bottom.GetXaxis().SetTitleOffset(1.1) frame_bottom.GetYaxis().SetLabelSize(0.04) frame_bottom.GetYaxis().SetTitleSize(0.04) frame_bottom.GetYaxis().SetTitleOffset(0.85) frame_bottom.SetMaximum(4.) frame_bottom.SetMinimum(-3.) frame_bottom.Draw("axis") zero = TLine(526., 0., 1500., 0.) zero.SetLineColor(ROOT.EColor.kBlack) zero.SetLineStyle(1) zero.SetLineWidth(2) zero.Draw("same") # Ratio histogram with no errors (not so well defined, since this isn't a well-defined efficiency) newHist = mframe.getHist("data") curve = mframe.getObject(1) hresid = newHist.makePullHist(curve,kTRUE) resframe = mjj.frame() mframe.SetAxisRange(526.,1500.) resframe.addPlotable(hresid,"B X") resframe.Draw("same") canvas.cd() canvas.SaveAs("testdouble.pdf") if args.pyes: c = TCanvas("c","c",800,800) mframe.SetAxisRange(300.,1300.) c.SetLogy() # mframe.SetMaximum(10) # mframe.SetMinimum(1) mframe.Draw() fitname = args.pdir+'/5funcfit_m'+str(mass)+fstr+'.pdf' c.SaveAs(fitname) cpull = TCanvas("cpull","cpull",800,800) pulls = mframe.pullHist("data","model3") pulls.Draw("ABX") pullname = args.pdir+'/pull_m'+str(mass)+fstr+'.pdf' cpull.SaveAs(pullname) cpull2 = TCanvas("cpull2","cpull2",800,800) pulls2 = mframe.pullHist("setonedata","model1") pulls2.Draw("ABX") pull2name = args.pdir+'/pull2_m'+str(mass)+fstr+'.pdf' cpull2.SaveAs(pull2name) if args.widefit: mframew = mjj.frame() rooDataHist2.plotOn(mframew, ROOT.RooFit.Name("data")) res6 = model.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model.plotOn(mframew, ROOT.RooFit.Name("model1"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) res7 = model2.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model2.plotOn(mframew, ROOT.RooFit.Name("model2"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kOrange)) res8 = model3.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model3.plotOn(mframew, ROOT.RooFit.Name("model3"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) res9 = model4.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model4.plotOn(mframew, ROOT.RooFit.Name("model4"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) res10 = model5.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model5.plotOn(mframew, ROOT.RooFit.Name("model5"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kViolet)) if args.pyes: c = TCanvas("c","c",800,800) mframew.SetAxisRange(300.,1300.) c.SetLogy() # mframew.SetMaximum(10) # mframew.SetMinimum(1) mframew.Draw() fitname = args.pdir+'/5funcfittowide_m'+str(mass)+fstr+'.pdf' c.SaveAs(fitname) cpull = TCanvas("cpull","cpull",800,800) pulls = mframew.pullHist("data","model1") pulls.Draw("ABX") pullname = args.pdir+'/pullwidefit_m'+str(mass)+fstr+'.pdf' cpull.SaveAs(pullname) if args.chi2: fullInt = model.createIntegral(RooArgSet(mjj)) norm = dataInt/fullInt.getVal() chi1 = 0. fullInt2 = model2.createIntegral(RooArgSet(mjj)) norm2 = dataInt2/fullInt2.getVal() chi2 = 0. fullInt3 = model3.createIntegral(RooArgSet(mjj)) norm3 = dataInt3/fullInt3.getVal() chi3 = 0. fullInt4 = model4.createIntegral(RooArgSet(mjj)) norm4 = dataInt4/fullInt4.getVal() chi4 = 0. fullInt5 = model5.createIntegral(RooArgSet(mjj)) norm5 = dataInt5/fullInt5.getVal() chi5 = 0. for i in range(args.massMin, args.massMax): new = 0 new2 = 0 new3 = 0 new4 = 0 new5 = 0 height = hData.GetBinContent(i) xLow = hData.GetXaxis().GetBinLowEdge(i) xUp = hData.GetXaxis().GetBinLowEdge(i+1) obs = height*(xUp-xLow) mjj.setRange("intrange",xLow,xUp) integ = model.createIntegral(RooArgSet(mjj),ROOT.RooFit.NormSet(RooArgSet(mjj)),ROOT.RooFit.Range("intrange")) exp = integ.getVal()*norm new = pow(exp-obs,2)/exp chi1 = chi1 + new integ2 = model2.createIntegral(RooArgSet(mjj),ROOT.RooFit.NormSet(RooArgSet(mjj)),ROOT.RooFit.Range("intrange")) exp2 = integ2.getVal()*norm2 new2 = pow(exp2-obs,2)/exp2 chi2 = chi2 + new2 integ3 = model3.createIntegral(RooArgSet(mjj),ROOT.RooFit.NormSet(RooArgSet(mjj)),ROOT.RooFit.Range("intrange")) exp3 = integ3.getVal()*norm3 new3 = pow(exp3-obs,2)/exp3 chi3 = chi3 + new3 integ4 = model4.createIntegral(RooArgSet(mjj),ROOT.RooFit.NormSet(RooArgSet(mjj)),ROOT.RooFit.Range("intrange")) exp4 = integ4.getVal()*norm4 if exp4 != 0: new4 = pow(exp4-obs,2)/exp4 else: new4 = 0 chi4 = chi4 + new4 integ5 = model5.createIntegral(RooArgSet(mjj),ROOT.RooFit.NormSet(RooArgSet(mjj)),ROOT.RooFit.Range("intrange")) exp5 = integ5.getVal()*norm5 new5 = pow(exp5-obs,2)/exp5 chi5 = chi5 + new5 print "chi1 %d "%(chi1) print "chi2 %d "%(chi2) print "chi3 %d "%(chi3) print "chi4 %d "%(chi4) print "chi5 %d "%(chi5) if not args.decoBkg: print " " res.Print() # res2.Print() # res3.Print() # res4.Print() # res5.Print() # res6.Print() # res7.Print() # decorrelated background parameters for Bayesian limits if args.decoBkg: signal_norm.setConstant() res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) res.Print() ## temp workspace for the PDF diagonalizer w_tmp = RooWorkspace("w_tmp") deco = PdfDiagonalizer("deco",w_tmp,res) # here diagonalizing only the shape parameters since the overall normalization is already decorrelated background_deco = deco.diagonalize(background) print "##################### workspace for decorrelation" w_tmp.Print("v") print "##################### original parameters" background.getParameters(rooDataHist).Print("v") print "##################### decorrelated parameters" # needed if want to evaluate limits without background systematics if args.fixBkg: w_tmp.var("deco_eig1").setConstant() w_tmp.var("deco_eig2").setConstant() if not args.fixP3: w_tmp.var("deco_eig3").setConstant() background_deco.getParameters(rooDataHist).Print("v") print "##################### original pdf" background.Print() print "##################### decorrelated pdf" background_deco.Print() # release signal normalization signal_norm.setConstant(kFALSE) # set the background normalization range to +/- 5 sigma bkg_val = background_norm.getVal() bkg_error = background_norm.getError() background_norm.setMin(bkg_val-5*bkg_error) background_norm.setMax(bkg_val+5*bkg_error) background_norm.Print() # change background PDF names background.SetName("background_old") background_deco.SetName("background") # needed if want to evaluate limits without background systematics if args.fixBkg: background_norm.setConstant() p1.setConstant() p2.setConstant() p3.setConstant() # ----------------------------------------- # dictionaries holding systematic variations of the signal shape hSig_Syst = {} hSig_Syst_DataHist = {} sigCDF = TGraph(hSig.GetNbinsX()+1) # JES and JER uncertainties if args.jesUnc != None or args.jerUnc != None: sigCDF.SetPoint(0,0.,0.) integral = 0. for i in range(1, hSig.GetNbinsX()+1): x = hSig.GetXaxis().GetBinLowEdge(i+1) integral = integral + hSig.GetBinContent(i) sigCDF.SetPoint(i,x,integral) if args.jesUnc != None: hSig_Syst['JESUp'] = copy.deepcopy(hSig) hSig_Syst['JESDown'] = copy.deepcopy(hSig) if args.jerUnc != None: hSig_Syst['JERUp'] = copy.deepcopy(hSig) hSig_Syst['JERDown'] = copy.deepcopy(hSig) # reset signal histograms for key in hSig_Syst.keys(): hSig_Syst[key].Reset() hSig_Syst[key].SetName(hSig_Syst[key].GetName() + '_' + key) # produce JES signal shapes if args.jesUnc != None: for i in range(1, hSig.GetNbinsX()+1): xLow = hSig.GetXaxis().GetBinLowEdge(i) xUp = hSig.GetXaxis().GetBinLowEdge(i+1) jes = 1. - args.jesUnc xLowPrime = jes*xLow xUpPrime = jes*xUp hSig_Syst['JESUp'].SetBinContent(i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) jes = 1. + args.jesUnc xLowPrime = jes*xLow xUpPrime = jes*xUp hSig_Syst['JESDown'].SetBinContent(i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) hSig_Syst_DataHist['JESUp'] = RooDataHist('hSig_JESUp','hSig_JESUp',RooArgList(mjj),hSig_Syst['JESUp']) hSig_Syst_DataHist['JESDown'] = RooDataHist('hSig_JESDown','hSig_JESDown',RooArgList(mjj),hSig_Syst['JESDown']) if args.jyes: c2 = TCanvas("c2","c2",800,800) mframe2 = mjj.frame(ROOT.RooFit.Title("JES One Sigma Shifts")) mframe2.SetAxisRange(525.,1200.) hSig_Syst_DataHist['JESUp'].plotOn(mframe2, ROOT.RooFit.Name("JESUP"),ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kRed), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) hSig_Syst_DataHist['JESDown'].plotOn(mframe2,ROOT.RooFit.Name("JESDOWN"),ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kBlue), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) rooSigHist.plotOn(mframe2, ROOT.RooFit.DataError(2),ROOT.RooFit.Name("SIG"),ROOT.RooFit.DrawOption("L"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kGreen), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) mframe2.Draw() mframe2.GetXaxis().SetTitle("Dijet Mass (GeV)") leg = TLegend(0.7,0.8,0.9,0.9) leg.AddEntry(mframe2.findObject("SIG"),"Signal Model","l") leg.AddEntry(mframe2.findObject("JESUP"),"+1 Sigma","l") leg.AddEntry(mframe2.findObject("JESDOWN"),"-1 Sigma","l") leg.Draw() jesname = args.pdir+'/jes_m'+str(mass)+fstr+'.pdf' c2.SaveAs(jesname) # produce JER signal shapes if args.jesUnc != None: for i in range(1, hSig.GetNbinsX()+1): xLow = hSig.GetXaxis().GetBinLowEdge(i) xUp = hSig.GetXaxis().GetBinLowEdge(i+1) jer = 1. - args.jerUnc xLowPrime = jer*(xLow-float(mass))+float(mass) xUpPrime = jer*(xUp-float(mass))+float(mass) hSig_Syst['JERUp'].SetBinContent(i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) jer = 1. + args.jerUnc xLowPrime = jer*(xLow-float(mass))+float(mass) xUpPrime = jer*(xUp-float(mass))+float(mass) hSig_Syst['JERDown'].SetBinContent(i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) hSig_Syst_DataHist['JERUp'] = RooDataHist('hSig_JERUp','hSig_JERUp',RooArgList(mjj),hSig_Syst['JERUp']) hSig_Syst_DataHist['JERDown'] = RooDataHist('hSig_JERDown','hSig_JERDown',RooArgList(mjj),hSig_Syst['JERDown']) if args.jyes: c3 = TCanvas("c3","c3",800,800) mframe3 = mjj.frame(ROOT.RooFit.Title("JER One Sigma Shifts")) mframe3.SetAxisRange(525.,1200.) hSig_Syst_DataHist['JERUp'].plotOn(mframe3,ROOT.RooFit.Name("JERUP"),ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kRed), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) hSig_Syst_DataHist['JERDown'].plotOn(mframe3,ROOT.RooFit.Name("JERDOWN"),ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kBlue), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) rooSigHist.plotOn(mframe3,ROOT.RooFit.DrawOption("L"),ROOT.RooFit.Name("SIG"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kGreen), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) mframe3.Draw() mframe3.GetXaxis().SetTitle("Dijet Mass (GeV)") leg = TLegend(0.7,0.8,0.9,0.9) leg.AddEntry(mframe3.findObject("SIG"),"Signal Model","l") leg.AddEntry(mframe3.findObject("JERUP"),"+1 Sigma","l") leg.AddEntry(mframe3.findObject("JERDOWN"),"-1 Sigma","l") leg.Draw() jername = args.pdir+'/jer_m'+str(mass)+fstr+'.pdf' c3.SaveAs(jername) # ----------------------------------------- # create a datacard and corresponding workspace postfix = (('_' + args.postfix) if args.postfix != '' else '') dcName = 'datacard_' + args.final_state + '_m' + str(mass) + postfix + '.txt' wsName = 'workspace_' + args.final_state + '_m' + str(mass) + postfix + '.root' w = RooWorkspace('w','workspace') getattr(w,'import')(rooSigHist,RooFit.Rename("signal")) if args.jesUnc != None: getattr(w,'import')(hSig_Syst_DataHist['JESUp'],RooFit.Rename("signal__JESUp")) getattr(w,'import')(hSig_Syst_DataHist['JESDown'],RooFit.Rename("signal__JESDown")) if args.jerUnc != None: getattr(w,'import')(hSig_Syst_DataHist['JERUp'],RooFit.Rename("signal__JERUp")) getattr(w,'import')(hSig_Syst_DataHist['JERDown'],RooFit.Rename("signal__JERDown")) if args.decoBkg: getattr(w,'import')(background_deco,ROOT.RooCmdArg()) else: getattr(w,'import')(background,ROOT.RooCmdArg(),RooFit.Rename("background")) #if use different fits for shape uncertainties #getattr(w,'import')(,ROOT.RooCmdArg(),RooFit.Rename("background__bkgUp")) #getattr(w,'import')(,ROOT.RooCmdArg(),RooFit.Rename("background__bkgDown")) getattr(w,'import')(background_norm,ROOT.RooCmdArg()) getattr(w,'import')(rooDataHist,RooFit.Rename("data_obs")) w.Print() w.writeToFile(os.path.join(args.output_path,wsName)) beffUnc = 0.3 boffUnc = 0.06 datacard = open(os.path.join(args.output_path,dcName),'w') datacard.write('imax 1\n') datacard.write('jmax 1\n') datacard.write('kmax *\n') datacard.write('---------------\n') if args.jesUnc != None or args.jerUnc != None: datacard.write('shapes * * '+wsName+' w:$PROCESS w:$PROCESS__$SYSTEMATIC\n') else: datacard.write('shapes * * '+wsName+' w:$PROCESS\n') datacard.write('---------------\n') datacard.write('bin 1\n') datacard.write('observation -1\n') datacard.write('------------------------------\n') datacard.write('bin 1 1\n') datacard.write('process signal background\n') datacard.write('process 0 1\n') datacard.write('rate -1 1\n') datacard.write('------------------------------\n') datacard.write('lumi lnN %f -\n'%(1.+args.lumiUnc)) datacard.write('beff lnN %f -\n'%(1.+beffUnc)) datacard.write('boff lnN %f -\n'%(1.+boffUnc)) datacard.write('bkg lnN - 1.03\n') if args.jesUnc != None: datacard.write('JES shape 1 -\n') if args.jerUnc != None: datacard.write('JER shape 1 -\n') # flat parameters --- flat prior datacard.write('background_norm flatParam\n') if args.decoBkg: datacard.write('deco_eig1 flatParam\n') datacard.write('deco_eig2 flatParam\n') if not args.fixP3: datacard.write('deco_eig3 flatParam\n') else: datacard.write('p1 flatParam\n') datacard.write('p2 flatParam\n') if not args.fixP3: datacard.write('p3 flatParam\n') datacard.close() print '>> Datacards and workspaces created and stored in %s/'%( os.path.join(os.getcwd(),args.output_path) )
def main(): # usage description usage = "Example: ./scripts/createDatacards.py --inputData inputs/rawhistV7_Run2015D_scoutingPFHT_UNBLINDED_649_838_JEC_HLTplusV7_Mjj_cor_smooth.root --dataHistname mjj_mjjcor_gev --inputSig inputs/ResonanceShapes_gg_13TeV_Scouting_Spring15.root -f gg -o datacards -l 1866 --lumiUnc 0.027 --massrange 1000 1500 50 --runFit --p1 5 --p2 7 --p3 0.4 --massMin 838 --massMax 2037 --fitStrategy 2" # input parameters parser = ArgumentParser( description= 'Script that creates combine datacards and corresponding RooFit workspaces', epilog=usage) parser.add_argument("--inputData", dest="inputData", required=True, help="Input data spectrum", metavar="INPUT_DATA") parser.add_argument("--dataHistname", dest="dataHistname", required=True, help="Data histogram name", metavar="DATA_HISTNAME") parser.add_argument("--inputSig", dest="inputSig", required=True, help="Input signal shapes", metavar="INPUT_SIGNAL") parser.add_argument("-f", "--final_state", dest="final_state", required=True, help="Final state (e.g. qq, qg, gg)", metavar="FINAL_STATE") parser.add_argument("-f2", "--type", dest="atype", required=True, help="Type (e.g. hG, lG, hR, lR)") parser.add_argument( "-o", "--output_path", dest="output_path", required=True, help="Output path where datacards and workspaces will be stored", metavar="OUTPUT_PATH") parser.add_argument( "-l", "--lumi", dest="lumi", required=True, default=1000., type=float, help="Integrated luminosity in pb-1 (default: %(default).1f)", metavar="LUMI") parser.add_argument( "--massMin", dest="massMin", default=500, type=int, help= "Lower bound of the mass range used for fitting (default: %(default)s)", metavar="MASS_MIN") parser.add_argument( "--massMax", dest="massMax", default=1200, type=int, help= "Upper bound of the mass range used for fitting (default: %(default)s)", metavar="MASS_MAX") parser.add_argument( "--p1", dest="p1", default=5.0000e-03, type=float, help="Fit function p1 parameter (default: %(default)e)", metavar="P1") parser.add_argument( "--p2", dest="p2", default=9.1000e+00, type=float, help="Fit function p2 parameter (default: %(default)e)", metavar="P2") parser.add_argument( "--p3", dest="p3", default=5.0000e-01, type=float, help="Fit function p3 parameter (default: %(default)e)", metavar="P3") parser.add_argument( "--lumiUnc", dest="lumiUnc", required=True, type=float, help="Relative uncertainty in the integrated luminosity", metavar="LUMI_UNC") parser.add_argument("--jesUnc", dest="jesUnc", type=float, help="Relative uncertainty in the jet energy scale", metavar="JES_UNC") parser.add_argument( "--jerUnc", dest="jerUnc", type=float, help="Relative uncertainty in the jet energy resolution", metavar="JER_UNC") parser.add_argument( "--sqrtS", dest="sqrtS", default=13000., type=float, help="Collision center-of-mass energy (default: %(default).1f)", metavar="SQRTS") parser.add_argument("--fixP3", dest="fixP3", default=False, action="store_true", help="Fix the fit function p3 parameter") parser.add_argument("--runFit", dest="runFit", default=False, action="store_true", help="Run the fit") parser.add_argument("--fitBonly", dest="fitBonly", default=False, action="store_true", help="Run B-only fit") parser.add_argument("--fixBkg", dest="fixBkg", default=False, action="store_true", help="Fix all background parameters") parser.add_argument("--decoBkg", dest="decoBkg", default=False, action="store_true", help="Decorrelate background parameters") parser.add_argument("--fitStrategy", dest="fitStrategy", type=int, default=1, help="Fit strategy (default: %(default).1f)") parser.add_argument("--debug", dest="debug", default=False, action="store_true", help="Debug printout") parser.add_argument( "--postfix", dest="postfix", default='', help="Postfix for the output file names (default: %(default)s)") parser.add_argument("--pyes", dest="pyes", default=False, action="store_true", help="Make files for plots") parser.add_argument("--jyes", dest="jyes", default=False, action="store_true", help="Make files for JES/JER plots") parser.add_argument( "--pdir", dest="pdir", default='testarea', help="Name a directory for the plots (default: %(default)s)") parser.add_argument("--chi2", dest="chi2", default=False, action="store_true", help="Compute chi squared") parser.add_argument("--widefit", dest="widefit", default=False, action="store_true", help="Fit with wide bin hist") mass_group = parser.add_mutually_exclusive_group(required=True) mass_group.add_argument( "--mass", type=int, nargs='*', default=1000, help= "Mass can be specified as a single value or a whitespace separated list (default: %(default)i)" ) mass_group.add_argument( "--massrange", type=int, nargs=3, help="Define a range of masses to be produced. Format: min max step", metavar=('MIN', 'MAX', 'STEP')) mass_group.add_argument("--masslist", help="List containing mass information") args = parser.parse_args() if args.atype == 'hG': fstr = "bbhGGBB" in2 = 'bcorrbin/binmodh.root' elif args.atype == 'hR': fstr = "bbhRS" in2 = 'bcorrbin/binmodh.root' elif args.atype == 'lG': fstr = "bblGGBB" in2 = 'bcorrbin/binmodl.root' else: fstr = "bblRS" in2 = 'bcorrbin/binmodl.root' # check if the output directory exists if not os.path.isdir(os.path.join(os.getcwd(), args.output_path)): os.mkdir(os.path.join(os.getcwd(), args.output_path)) # mass points for which resonance shapes will be produced masses = [] if args.massrange != None: MIN, MAX, STEP = args.massrange masses = range(MIN, MAX + STEP, STEP) elif args.masslist != None: # A mass list was provided print "Will create mass list according to", args.masslist masslist = __import__(args.masslist.replace(".py", "")) masses = masslist.masses else: masses = args.mass # sort masses masses.sort() # import ROOT stuff from ROOT import gStyle, TFile, TH1F, TH1D, TGraph, kTRUE, kFALSE, TCanvas, TLegend, TPad, TLine from ROOT import RooHist, RooRealVar, RooDataHist, RooArgList, RooArgSet, RooAddPdf, RooFit, RooGenericPdf, RooWorkspace, RooMsgService, RooHistPdf, RooExtendPdf if not args.debug: RooMsgService.instance().setSilentMode(kTRUE) RooMsgService.instance().setStreamStatus(0, kFALSE) RooMsgService.instance().setStreamStatus(1, kFALSE) # input data file inputData = TFile(args.inputData) # input data histogram hData = inputData.Get(args.dataHistname) inData2 = TFile(in2) hData2 = inData2.Get('h_data') # input sig file inputSig = TFile(args.inputSig) sqrtS = args.sqrtS # mass variable mjj = RooRealVar('mjj', 'mjj', float(args.massMin), float(args.massMax)) # integrated luminosity and signal cross section lumi = args.lumi signalCrossSection = 1. # set to 1. so that the limit on r can be interpreted as a limit on the signal cross section for mass in masses: print ">> Creating datacard and workspace for %s resonance with m = %i GeV..." % ( args.final_state, int(mass)) # get signal shape hSig = inputSig.Get("h_" + args.final_state + "_" + str(int(mass))) # normalize signal shape to the expected event yield (works even if input shapes are not normalized to unity) hSig.Scale( signalCrossSection * lumi / hSig.Integral() ) # divide by a number that provides roughly an r value of 1-10 rooSigHist = RooDataHist('rooSigHist', 'rooSigHist', RooArgList(mjj), hSig) print 'Signal acceptance:', (rooSigHist.sumEntries() / hSig.Integral()) signal = RooHistPdf('signal', 'signal', RooArgSet(mjj), rooSigHist) signal_norm = RooRealVar('signal_norm', 'signal_norm', 0, -1e+05, 1e+05) signal_norm2 = RooRealVar('signal_norm2', 'signal_norm2', 0, -1e+05, 1e+05) signal_norm3 = RooRealVar('signal_norm3', 'signal_norm3', 0, -1e+05, 1e+05) signal_norm4 = RooRealVar('signal_norm4', 'signal_norm4', 0, -1e+05, 1e+05) signal_norm5 = RooRealVar('signal_norm5', 'signal_norm5', 0, -1e+05, 1e+05) if args.fitBonly: signal_norm.setConstant() signal_norm2.setConstant() signal_norm3.setConstant() signal_norm4.setConstant() signal_norm5.setConstant() p1 = RooRealVar('p1', 'p1', args.p1, 0., 100.) p2 = RooRealVar('p2', 'p2', args.p2, 0., 60.) p3 = RooRealVar('p3', 'p3', args.p3, -10., 10.) p4 = RooRealVar('p4', 'p4', 5.6, -50., 50.) p5 = RooRealVar('p5', 'p5', 10., -50., 50.) p6 = RooRealVar('p6', 'p6', .016, -50., 50.) p7 = RooRealVar('p7', 'p7', 8., -50., 50.) p8 = RooRealVar('p8', 'p8', .22, -50., 50.) p9 = RooRealVar('p9', 'p9', 14.1, -50., 50.) p10 = RooRealVar('p10', 'p10', 8., -50., 50.) p11 = RooRealVar('p11', 'p11', 4.8, -50., 50.) p12 = RooRealVar('p12', 'p12', 7., -50., 50.) p13 = RooRealVar('p13', 'p13', 7., -50., 50.) p14 = RooRealVar('p14', 'p14', 7., -50., 50.) p15 = RooRealVar('p15', 'p15', 1., -50., 50.) p16 = RooRealVar('p16', 'p16', 9., -50., 50.) p17 = RooRealVar('p17', 'p17', 0.6, -50., 50.) if args.fixP3: p3.setConstant() background = RooGenericPdf( 'background', '(pow(1-@0/%.1f,@1)/pow(@0/%.1f,@2+@3*log(@0/%.1f)))' % (sqrtS, sqrtS, sqrtS), RooArgList(mjj, p1, p2, p3)) dataInt = hData.Integral(hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background_norm = RooRealVar('background_norm', 'background_norm', dataInt, 0., 1e+08) background2 = RooGenericPdf( 'background2', '(pow(@0/%.1f,-@1)*pow(1-@0/%.1f,@2))' % (sqrtS, sqrtS), RooArgList(mjj, p4, p5)) dataInt2 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background2_norm = RooRealVar('background2_norm', 'background2_norm', dataInt2, 0., 1e+08) background3 = RooGenericPdf('background3', '(1/pow(@1+@0/%.1f,@2))' % (sqrtS), RooArgList(mjj, p6, p7)) dataInt3 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background3_norm = RooRealVar('background3_norm', 'background3_norm', dataInt3, 0., 1e+08) background4 = RooGenericPdf( 'background4', '(1/pow(@1+@2*@0/%.1f+pow(@0/%.1f,2),@3))' % (sqrtS, sqrtS), RooArgList(mjj, p8, p9, p10)) dataInt4 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background4_norm = RooRealVar('background4_norm', 'background4_norm', dataInt4, 0., 1e+08) background5 = RooGenericPdf( 'background5', '(pow(@0/%.1f,-@1)*pow(1-pow(@0/%.1f,1/3),@2))' % (sqrtS, sqrtS), RooArgList(mjj, p11, p12)) dataInt5 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background5_norm = RooRealVar('background5_norm', 'background5_norm', dataInt5, 0., 1e+08) background6 = RooGenericPdf( 'background6', '(pow(@0/%.1f,2)+@1*@0/%.1f+@2)' % (sqrtS, sqrtS), RooArgList(mjj, p13, p14)) dataInt6 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background_norm6 = RooRealVar('background_norm6', 'background_norm6', dataInt6, 0., 1e+08) background7 = RooGenericPdf( 'background7', '((-1+@1*@0/%.1f)*pow(@0/%.1f,@2+@3*log(@0/%.1f)))' % (sqrtS, sqrtS, sqrtS), RooArgList(mjj, p15, p16, p17)) dataInt7 = hData.Integral( hData.GetXaxis().FindBin(float(args.massMin)), hData.GetXaxis().FindBin(float(args.massMax))) background_norm7 = RooRealVar('background_norm7', 'background_norm7', dataInt7, 0., 1e+08) #Extend PDFs exts = RooExtendPdf('extsignal', 'Extended Signal Pdf', signal, signal_norm) extb = RooExtendPdf('extbackground', 'Extended Background Pdf', background, background_norm) exts2 = RooExtendPdf('extsignal2', 'Extended Signal Pdf2', signal, signal_norm2) extb2 = RooExtendPdf('extbackground2', 'Extended Background Pdf2', background2, background2_norm) exts3 = RooExtendPdf('extsignal3', 'Extended Signal Pdf3', signal, signal_norm3) extb3 = RooExtendPdf('extbackground3', 'Extended Background Pdf3', background3, background3_norm) exts4 = RooExtendPdf('extsignal4', 'Extended Signal Pdf4', signal, signal_norm4) extb4 = RooExtendPdf('extbackground4', 'Extended Background Pdf4', background4, background4_norm) exts5 = RooExtendPdf('extsignal5', 'Extended Signal Pdf5', signal, signal_norm5) extb5 = RooExtendPdf('extbackground5', 'Extended Background Pdf5', background5, background5_norm) # S+B model model = RooAddPdf("model", "s+b", RooArgList(extb, exts)) model2 = RooAddPdf("model2", "s+b2", RooArgList(extb2, exts2)) model3 = RooAddPdf("model3", "s+b3", RooArgList(extb3, exts3)) model4 = RooAddPdf("model4", "s+b4", RooArgList(extb4, exts4)) model5 = RooAddPdf("model5", "s+b5", RooArgList(extb5, exts5)) #model6 = RooAddPdf("model6","s+b6",RooArgList(background6,signal),RooArgList(background_norm6,signal_norm)) #model7 = RooAddPdf("model7","s+b7",RooArgList(background7,signal),RooArgList(background_norm7,signal_norm)) rooDataHist = RooDataHist('rooDatahist', 'rooDathist', RooArgList(mjj), hData) if args.runFit: mframe = mjj.frame() rooDataHist.plotOn(mframe, ROOT.RooFit.Name("setonedata")) res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Extended(kTRUE), RooFit.Strategy(args.fitStrategy)) model.plotOn(mframe, ROOT.RooFit.Name("model1"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) res2 = model2.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Extended(kTRUE), RooFit.Strategy(args.fitStrategy)) # model2.plotOn(mframe, ROOT.RooFit.Name("model2"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kOrange)) res3 = model3.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Extended(kTRUE), RooFit.Strategy(args.fitStrategy)) # model3.plotOn(mframe, ROOT.RooFit.Name("model3"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) res4 = model4.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Extended(kTRUE), RooFit.Strategy(args.fitStrategy)) # model4.plotOn(mframe, ROOT.RooFit.Name("model4"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) res5 = model5.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Extended(kTRUE), RooFit.Strategy(args.fitStrategy)) # model5.plotOn(mframe, ROOT.RooFit.Name("model5"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kViolet)) # res6 = model6.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) # model6.plotOn(mframe, ROOT.RooFit.Name("model6"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kPink)) # res7 = model7.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) # model7.plotOn(mframe, ROOT.RooFit.Name("model7"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kAzure)) rooDataHist2 = RooDataHist('rooDatahist2', 'rooDathist2', RooArgList(mjj), hData2) # rooDataHist2.plotOn(mframe, ROOT.RooFit.Name("data")) if args.pyes: c = TCanvas("c", "c", 800, 800) mframe.SetAxisRange(300., 1300.) c.SetLogy() # mframe.SetMaximum(10) # mframe.SetMinimum(1) mframe.Draw() fitname = args.pdir + '/5funcfit_m' + str(mass) + fstr + '.pdf' c.SaveAs(fitname) # cpull = TCanvas("cpull","cpull",800,800) # pulls = mframe.pullHist("data","model1") # pulls.Draw("ABX") # pullname = args.pdir+'/pull_m'+str(mass)+fstr+'.pdf' # cpull.SaveAs(pullname) # cpull2 = TCanvas("cpull2","cpull2",800,800) # pulls2 = mframe.pullHist("setonedata","model1") # pulls2.Draw("ABX") # pull2name = args.pdir+'/pull2_m'+str(mass)+fstr+'.pdf' # cpull2.SaveAs(pull2name) if args.widefit: mframew = mjj.frame() rooDataHist2.plotOn(mframew, ROOT.RooFit.Name("data")) res6 = model.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model.plotOn(mframew, ROOT.RooFit.Name("model1"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) res7 = model2.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model2.plotOn(mframew, ROOT.RooFit.Name("model2"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kOrange)) res8 = model3.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model3.plotOn(mframew, ROOT.RooFit.Name("model3"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) res9 = model4.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model4.plotOn(mframew, ROOT.RooFit.Name("model4"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) res10 = model5.fitTo(rooDataHist2, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) model5.plotOn(mframew, ROOT.RooFit.Name("model5"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.LineWidth(1), ROOT.RooFit.LineColor(ROOT.EColor.kViolet)) if args.pyes: c = TCanvas("c", "c", 800, 800) mframew.SetAxisRange(300., 1300.) c.SetLogy() # mframew.SetMaximum(10) # mframew.SetMinimum(1) mframew.Draw() fitname = args.pdir + '/5funcfittowide_m' + str( mass) + fstr + '.pdf' c.SaveAs(fitname) cpull = TCanvas("cpull", "cpull", 800, 800) pulls = mframew.pullHist("data", "model1") pulls.Draw("ABX") pullname = args.pdir + '/pullwidefit_m' + str( mass) + fstr + '.pdf' cpull.SaveAs(pullname) if args.chi2: fullInt = model.createIntegral(RooArgSet(mjj)) norm = dataInt / fullInt.getVal() chi1 = 0. fullInt2 = model2.createIntegral(RooArgSet(mjj)) norm2 = dataInt2 / fullInt2.getVal() chi2 = 0. fullInt3 = model3.createIntegral(RooArgSet(mjj)) norm3 = dataInt3 / fullInt3.getVal() chi3 = 0. fullInt4 = model4.createIntegral(RooArgSet(mjj)) norm4 = dataInt4 / fullInt4.getVal() chi4 = 0. fullInt5 = model5.createIntegral(RooArgSet(mjj)) norm5 = dataInt5 / fullInt5.getVal() chi5 = 0. for i in range(args.massMin, args.massMax): new = 0 new2 = 0 new3 = 0 new4 = 0 new5 = 0 height = hData.GetBinContent(i) xLow = hData.GetXaxis().GetBinLowEdge(i) xUp = hData.GetXaxis().GetBinLowEdge(i + 1) obs = height * (xUp - xLow) mjj.setRange("intrange", xLow, xUp) integ = model.createIntegral( RooArgSet(mjj), ROOT.RooFit.NormSet(RooArgSet(mjj)), ROOT.RooFit.Range("intrange")) exp = integ.getVal() * norm new = pow(exp - obs, 2) / exp chi1 = chi1 + new integ2 = model2.createIntegral( RooArgSet(mjj), ROOT.RooFit.NormSet(RooArgSet(mjj)), ROOT.RooFit.Range("intrange")) exp2 = integ2.getVal() * norm2 new2 = pow(exp2 - obs, 2) / exp2 chi2 = chi2 + new2 integ3 = model3.createIntegral( RooArgSet(mjj), ROOT.RooFit.NormSet(RooArgSet(mjj)), ROOT.RooFit.Range("intrange")) exp3 = integ3.getVal() * norm3 new3 = pow(exp3 - obs, 2) / exp3 chi3 = chi3 + new3 integ4 = model4.createIntegral( RooArgSet(mjj), ROOT.RooFit.NormSet(RooArgSet(mjj)), ROOT.RooFit.Range("intrange")) exp4 = integ4.getVal() * norm4 if exp4 != 0: new4 = pow(exp4 - obs, 2) / exp4 else: new4 = 0 chi4 = chi4 + new4 integ5 = model5.createIntegral( RooArgSet(mjj), ROOT.RooFit.NormSet(RooArgSet(mjj)), ROOT.RooFit.Range("intrange")) exp5 = integ5.getVal() * norm5 new5 = pow(exp5 - obs, 2) / exp5 chi5 = chi5 + new5 print "chi1 %d " % (chi1) print "chi2 %d " % (chi2) print "chi3 %d " % (chi3) print "chi4 %d " % (chi4) print "chi5 %d " % (chi5) if not args.decoBkg: print " " res.Print() res2.Print() res3.Print() res4.Print() res5.Print() # res6.Print() # res7.Print() # decorrelated background parameters for Bayesian limits if args.decoBkg: signal_norm.setConstant() res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) res.Print() ## temp workspace for the PDF diagonalizer w_tmp = RooWorkspace("w_tmp") deco = PdfDiagonalizer("deco", w_tmp, res) # here diagonalizing only the shape parameters since the overall normalization is already decorrelated background_deco = deco.diagonalize(background) print "##################### workspace for decorrelation" w_tmp.Print("v") print "##################### original parameters" background.getParameters(rooDataHist).Print("v") print "##################### decorrelated parameters" # needed if want to evaluate limits without background systematics if args.fixBkg: w_tmp.var("deco_eig1").setConstant() w_tmp.var("deco_eig2").setConstant() if not args.fixP3: w_tmp.var("deco_eig3").setConstant() background_deco.getParameters(rooDataHist).Print("v") print "##################### original pdf" background.Print() print "##################### decorrelated pdf" background_deco.Print() # release signal normalization signal_norm.setConstant(kFALSE) # set the background normalization range to +/- 5 sigma bkg_val = background_norm.getVal() bkg_error = background_norm.getError() background_norm.setMin(bkg_val - 5 * bkg_error) background_norm.setMax(bkg_val + 5 * bkg_error) background_norm.Print() # change background PDF names background.SetName("background_old") background_deco.SetName("background") # needed if want to evaluate limits without background systematics if args.fixBkg: background_norm.setConstant() p1.setConstant() p2.setConstant() p3.setConstant() # ----------------------------------------- # dictionaries holding systematic variations of the signal shape hSig_Syst = {} hSig_Syst_DataHist = {} sigCDF = TGraph(hSig.GetNbinsX() + 1) # JES and JER uncertainties if args.jesUnc != None or args.jerUnc != None: sigCDF.SetPoint(0, 0., 0.) integral = 0. for i in range(1, hSig.GetNbinsX() + 1): x = hSig.GetXaxis().GetBinLowEdge(i + 1) integral = integral + hSig.GetBinContent(i) sigCDF.SetPoint(i, x, integral) if args.jesUnc != None: hSig_Syst['JESUp'] = copy.deepcopy(hSig) hSig_Syst['JESDown'] = copy.deepcopy(hSig) if args.jerUnc != None: hSig_Syst['JERUp'] = copy.deepcopy(hSig) hSig_Syst['JERDown'] = copy.deepcopy(hSig) # reset signal histograms for key in hSig_Syst.keys(): hSig_Syst[key].Reset() hSig_Syst[key].SetName(hSig_Syst[key].GetName() + '_' + key) # produce JES signal shapes if args.jesUnc != None: for i in range(1, hSig.GetNbinsX() + 1): xLow = hSig.GetXaxis().GetBinLowEdge(i) xUp = hSig.GetXaxis().GetBinLowEdge(i + 1) jes = 1. - args.jesUnc xLowPrime = jes * xLow xUpPrime = jes * xUp hSig_Syst['JESUp'].SetBinContent( i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) jes = 1. + args.jesUnc xLowPrime = jes * xLow xUpPrime = jes * xUp hSig_Syst['JESDown'].SetBinContent( i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) hSig_Syst_DataHist['JESUp'] = RooDataHist('hSig_JESUp', 'hSig_JESUp', RooArgList(mjj), hSig_Syst['JESUp']) hSig_Syst_DataHist['JESDown'] = RooDataHist( 'hSig_JESDown', 'hSig_JESDown', RooArgList(mjj), hSig_Syst['JESDown']) if args.jyes: c2 = TCanvas("c2", "c2", 800, 800) mframe2 = mjj.frame(ROOT.RooFit.Title("JES One Sigma Shifts")) mframe2.SetAxisRange(args.massMin, args.massMax) hSig_Syst_DataHist['JESUp'].plotOn( mframe2, ROOT.RooFit.Name("JESUP"), ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kRed), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) hSig_Syst_DataHist['JESDown'].plotOn( mframe2, ROOT.RooFit.Name("JESDOWN"), ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kBlue), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) rooSigHist.plotOn(mframe2, ROOT.RooFit.DataError(2), ROOT.RooFit.Name("SIG"), ROOT.RooFit.DrawOption("L"), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kGreen), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) mframe2.Draw() mframe2.GetXaxis().SetTitle("Dijet Mass (GeV)") leg = TLegend(0.7, 0.8, 0.9, 0.9) leg.SetFillColor(0) leg.AddEntry(mframe2.findObject("SIG"), "Signal Model", "l") leg.AddEntry(mframe2.findObject("JESUP"), "+1 Sigma", "l") leg.AddEntry(mframe2.findObject("JESDOWN"), "-1 Sigma", "l") leg.Draw() jesname = args.pdir + '/jes_m' + str(mass) + fstr + '.pdf' c2.SaveAs(jesname) # produce JER signal shapes if args.jesUnc != None: for i in range(1, hSig.GetNbinsX() + 1): xLow = hSig.GetXaxis().GetBinLowEdge(i) xUp = hSig.GetXaxis().GetBinLowEdge(i + 1) jer = 1. - args.jerUnc xLowPrime = jer * (xLow - float(mass)) + float(mass) xUpPrime = jer * (xUp - float(mass)) + float(mass) hSig_Syst['JERUp'].SetBinContent( i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) jer = 1. + args.jerUnc xLowPrime = jer * (xLow - float(mass)) + float(mass) xUpPrime = jer * (xUp - float(mass)) + float(mass) hSig_Syst['JERDown'].SetBinContent( i, sigCDF.Eval(xUpPrime) - sigCDF.Eval(xLowPrime)) hSig_Syst_DataHist['JERUp'] = RooDataHist('hSig_JERUp', 'hSig_JERUp', RooArgList(mjj), hSig_Syst['JERUp']) hSig_Syst_DataHist['JERDown'] = RooDataHist( 'hSig_JERDown', 'hSig_JERDown', RooArgList(mjj), hSig_Syst['JERDown']) if args.jyes: c3 = TCanvas("c3", "c3", 800, 800) mframe3 = mjj.frame(ROOT.RooFit.Title("JER One Sigma Shifts")) mframe3.SetAxisRange(args.massMin, args.massMax) hSig_Syst_DataHist['JERUp'].plotOn( mframe3, ROOT.RooFit.Name("JERUP"), ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kRed), ROOT.RooFit.LineColor(ROOT.EColor.kRed)) hSig_Syst_DataHist['JERDown'].plotOn( mframe3, ROOT.RooFit.Name("JERDOWN"), ROOT.RooFit.DrawOption("L"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kBlue), ROOT.RooFit.LineColor(ROOT.EColor.kBlue)) rooSigHist.plotOn(mframe3, ROOT.RooFit.DrawOption("L"), ROOT.RooFit.Name("SIG"), ROOT.RooFit.DataError(2), ROOT.RooFit.LineStyle(1), ROOT.RooFit.MarkerColor(ROOT.EColor.kGreen), ROOT.RooFit.LineColor(ROOT.EColor.kGreen)) mframe3.Draw() mframe3.GetXaxis().SetTitle("Dijet Mass (GeV)") leg = TLegend(0.7, 0.8, 0.9, 0.9) leg.SetFillColor(0) leg.AddEntry(mframe3.findObject("SIG"), "Signal Model", "l") leg.AddEntry(mframe3.findObject("JERUP"), "+1 Sigma", "l") leg.AddEntry(mframe3.findObject("JERDOWN"), "-1 Sigma", "l") leg.Draw() jername = args.pdir + '/jer_m' + str(mass) + fstr + '.pdf' c3.SaveAs(jername) # ----------------------------------------- # create a datacard and corresponding workspace postfix = (('_' + args.postfix) if args.postfix != '' else '') dcName = 'datacard_' + args.final_state + '_m' + str( mass) + postfix + '.txt' wsName = 'workspace_' + args.final_state + '_m' + str( mass) + postfix + '.root' w = RooWorkspace('w', 'workspace') getattr(w, 'import')(rooSigHist, RooFit.Rename("signal")) if args.jesUnc != None: getattr(w, 'import')(hSig_Syst_DataHist['JESUp'], RooFit.Rename("signal__JESUp")) getattr(w, 'import')(hSig_Syst_DataHist['JESDown'], RooFit.Rename("signal__JESDown")) if args.jerUnc != None: getattr(w, 'import')(hSig_Syst_DataHist['JERUp'], RooFit.Rename("signal__JERUp")) getattr(w, 'import')(hSig_Syst_DataHist['JERDown'], RooFit.Rename("signal__JERDown")) if args.decoBkg: getattr(w, 'import')(background_deco, ROOT.RooCmdArg()) else: getattr(w, 'import')(background, ROOT.RooCmdArg(), RooFit.Rename("background")) getattr(w, 'import')(background2, ROOT.RooCmdArg(), RooFit.Rename("background2")) getattr(w, 'import')(background3, ROOT.RooCmdArg(), RooFit.Rename("background3")) getattr(w, 'import')(background4, ROOT.RooCmdArg(), RooFit.Rename("background4")) getattr(w, 'import')(background5, ROOT.RooCmdArg(), RooFit.Rename("background5")) getattr(w, 'import')(background_norm, ROOT.RooCmdArg(), RooFit.Rename("background_norm")) getattr(w, 'import')(background2_norm, ROOT.RooCmdArg(), RooFit.Rename("background2_norm")) getattr(w, 'import')(background3_norm, ROOT.RooCmdArg(), RooFit.Rename("background3_norm")) getattr(w, 'import')(background4_norm, ROOT.RooCmdArg(), RooFit.Rename("background4_norm")) getattr(w, 'import')(background5_norm, ROOT.RooCmdArg(), RooFit.Rename("background5_norm")) getattr(w, 'import')(res) getattr(w, 'import')(res2) getattr(w, 'import')(res3) getattr(w, 'import')(res4) getattr(w, 'import')(res5) getattr(w, 'import')(background_norm, ROOT.RooCmdArg()) getattr(w, 'import')(signal_norm, ROOT.RooCmdArg()) getattr(w, 'import')(rooDataHist, RooFit.Rename("data_obs")) w.Print() w.writeToFile(os.path.join(args.output_path, wsName)) beffUnc = 0.3 boffUnc = 0.06 datacard = open(os.path.join(args.output_path, dcName), 'w') datacard.write('imax 1\n') datacard.write('jmax 1\n') datacard.write('kmax *\n') datacard.write('---------------\n') if args.jesUnc != None or args.jerUnc != None: datacard.write('shapes * * ' + wsName + ' w:$PROCESS w:$PROCESS__$SYSTEMATIC\n') else: datacard.write('shapes * * ' + wsName + ' w:$PROCESS\n') datacard.write('---------------\n') datacard.write('bin 1\n') datacard.write('observation -1\n') datacard.write('------------------------------\n') datacard.write('bin 1 1\n') datacard.write('process signal background\n') datacard.write('process 0 1\n') datacard.write('rate -1 1\n') datacard.write('------------------------------\n') datacard.write('lumi lnN %f -\n' % (1. + args.lumiUnc)) datacard.write('beff lnN %f -\n' % (1. + beffUnc)) datacard.write('boff lnN %f -\n' % (1. + boffUnc)) datacard.write('bkg lnN - 1.03\n') if args.jesUnc != None: datacard.write('JES shape 1 -\n') if args.jerUnc != None: datacard.write('JER shape 1 -\n') # flat parameters --- flat prior datacard.write('background_norm flatParam\n') if args.decoBkg: datacard.write('deco_eig1 flatParam\n') datacard.write('deco_eig2 flatParam\n') if not args.fixP3: datacard.write('deco_eig3 flatParam\n') else: datacard.write('p1 flatParam\n') datacard.write('p2 flatParam\n') if not args.fixP3: datacard.write('p3 flatParam\n') datacard.close() print '>> Datacards and workspaces created and stored in %s/' % ( os.path.join(os.getcwd(), args.output_path))
if not args.decoBkg: res.Print() # decorrelated background parameters for Bayesian limits if args.decoBkg: signal_norm.setConstant() res = model.fitTo(rooDataHist, RooFit.Save(kTRUE), RooFit.Strategy(args.fitStrategy)) res.Print() ## temp workspace for the PDF diagonalizer w_tmp = RooWorkspace("w_tmp") deco = PdfDiagonalizer("deco",w_tmp,res) # here diagonalizing only the shape parameters since the overall normalization is already decorrelated background_deco = deco.diagonalize(background) print "##################### workspace for decorrelation" w_tmp.Print("v") print "##################### original parameters" background.getParameters(rooDataHist).Print("v") print "##################### decorrelated parameters" # needed if want to evaluate limits without background systematics if args.fixBkg: w_tmp.var("deco_eig1").setConstant() w_tmp.var("deco_eig2").setConstant() if not args.fixP3: w_tmp.var("deco_eig3").setConstant() background_deco.getParameters(rooDataHist).Print("v") print "##################### original pdf" background.Print() print "##################### decorrelated pdf" background_deco.Print() # release signal normalization signal_norm.setConstant(kFALSE) # set the background normalization range to +/- 5 sigma bkg_val = background_norm.getVal()