void RooFitMacro()
{
gROOT->Reset();
gSystem->Load("libRooFit");
gROOT->LoadMacro("RooCMSShape.cc+");
//gStyle->SetOptStat(111111);
//gROOT->ProcessLine(".x ~/rootlogon.C");
TFile f("TagAndProbeResults_Marco.root");
TH1F * Z_mass = new TH1F("Z","Z" , 200, 0, 200);
//TH1D *results = new TH1D("results", "results", 200, 0, 200);
//results->Sumw2();
// We need to include NJetCut because there is depedence w.r.t. jet multiplicity
Z_mass = (TH1F*)f.Get("Zmass_Inclusive_NoElectronVeto_Numerator;1");
//Z_mass = (TH1F*)f.Get("Zmass_Inclusive_NoElectronVeto_Denominator;1");
//float bincontent = Z_mass->GetBinContent(60);
//cout << "60.bin content : " << bincontent << endl;
double hmin = 60;
double hmax = 120;
double r = .25;
double sl = 1. / ( 1. - r );
// Declare observable x
RooRealVar x("x","x",hmin,hmax) ;
RooDataHist dh("dh","dh",x,Import(*Z_mass)) ;
// Breit-Wigner Lineshape
// Parameters for Breit-Wigner Distribution
RooRealVar M("M_{Z^{0}}", "Z0 Resonance Mass", 91.18, 85.0, 95.0, "GeV/c^{2}");
RooRealVar gamma("#Gamma", "#Gamma", 2.4952, 2.0, 3.0, "GeV/c^{2}");
M.setConstant();
//gamma.setConstant();
RooBreitWigner bw("bw", "A Breit-Wigner Distribution", x, M, gamma);
// Mass resolution model - Crystal Ball Lineshape
// Parameters for Crystal Ball Lineshape
RooRealVar M_CB("#Delta m_{0}", "Bias", -2.0, -10.0, 10.0, "GeV/c^{2}");
RooRealVar sigma("#sigma_{CB}", "Width", 1.6, 0.1, 15.0, "GeV/c^{2}");
// RooRealVar sigma("#sigma_{CB}", "Width", 2.5, 0.1, 15.0, "GeV/c^{2}");
RooRealVar alpha("#alpha", "Cut", 1.5, 0.1, 15.0);
RooRealVar n("n", "Power", 1.8, 0.5, 5.0);
RooCBShape cb("resCBF", "A Crystal Ball Lineshape", x, M_CB, sigma, alpha, n);
// background p.d.f
RooRealVar cms_alpha("alpha", "alpha", 60, 50, 70);
RooRealVar cms_beta("beta", "beta", 0.01, 0.001, 0.05);
RooRealVar cms_gamma("gamma", "gamma", 0.05, 0.005, 0.5);
RooRealVar peak("peak", "peak", 60, 50, 70);
RooCMSShape bag("bag", "RooCMSShape", x, cms_alpha, cms_beta, cms_gamma, peak);
RooRealVar frac("frac", "f", 0.01, 0, 1.0);
RooFFTConvPdf signal("signal", "Convolution", x, bw, cb);
RooRealVar signal_yield("signal_yield", "signal_yield", 100, 0, Z_mass->Integral());
RooRealVar bag_yield("bag_yield", "bag_yield", 0, 0, Z_mass->Integral());
RooArgList shapes;
RooArgList yields;
shapes.add(signal);
shapes.add(bag);
yields.add(signal_yield);
yields.add(bag_yield);
RooRealVar fsig("fsig", "fsig", 0, 0, 1.);
//RooAddPdf model("model","model",shapes,yields);
// model = fsig x signal + (1-fsig) x bag
RooAddPdf model("model","model",shapes,fsig,kTRUE);
// Signal p.d.f.
//RooFFTConvPdf sum("sum", "Convolution", x, bw, cb);
//RooAddPdf sum("sum","sum",bw,cb,frac);
//RooAddPdf model("model","model",RooArgList(sum, bag),frac);
//RooFFTConvPdf model("model","model",x,sum,bag);
//RooFitResult* filters = sum.fitTo(dh,Range(0,200),"qr");
//filters->Print("v");
//RooFitResult* filters = model.fitTo(dh,Range(0,200),"qr");
RooFitResult* filters = model.fitTo(dh, RooFit::Minimizer("Minuit", "migradimproved"), RooFit::NumCPU(4), RooFit::Save(true), RooFit::Extended(false), RooFit::PrintLevel(-1));
//RooFitResult* filters = model.fitTo(dh, RooFit::Extended(true), RooFit::PrintLevel(-1));
//RooFitResult* filters = model.fitTo(dh, "mhe");
//RooFitResult* filters = cb.fitTo(dh,Range(0,200),"qr");
//RooFitResult* filters = bw.fitTo(dh,Range(0,100),"qr");
//RooFitResult* filters = bag.fitTo(dh,"qr");
TCanvas* canvas = new TCanvas("ZmassHisto","ZmassHisto",0, 0, 1000,700) ;
canvas->cd() ; //gPad->SetLeftMargin(0.15);
//gPad->SetLogy();
RooPlot* frame = x.frame(Title("e #gamma invariant mass fit")) ;
dh.plotOn(frame,MarkerSize(0.5),Name("data_hist")); //this will show histogram data points on canvas
dh.statOn(frame,Layout(0.15,0.37,0.85),What("N")) ;
//dh.statOn(frame); //this will display hist stat on canvas
//sum.plotOn(frame,LineColor(4));//this will show fit overlay on canvas
//sum.paramOn(frame); //this will display the fit parameters on canvas
//bag.plotOn(frame,LineColor(2));//this will show fit overlay on canvas
//bag.paramOn(frame); //this will display the fit parameters on canvas
model.plotOn(frame,Components(bag),LineColor(2));//this will show fit overlay on canvas
model.plotOn(frame,Components(signal),LineColor(3));//this will show fit overlay on canvas
model.plotOn(frame,LineColor(kBlue),Name("main_curve"));//this will show fit overlay on canvas
// model.paramOn(frame); //this will display the fit parameters on canvas
model.paramOn(frame, Layout(0.6, 0.95, 0.92)); //this will display the fit parameters on canvas
//model.paramOn(frame, Layout(0.6, 0.99, 0.75));
// model.plotOn(frame,LineColor(4));//this will show fit overlay on canvas
// model.paramOn(frame); //this will display the fit parameters on canvas
RooHist* histogram = frame->getHist("data_hist");
RooCurve* curve = frame->getCurve("main_curve");
TH1D* hresidual = residualHist(histogram,curve);
hresidual->Sumw2();
canvas->Divide( 1, 2, .1, .1 );
TPad* padHisto = (TPad*) canvas->cd(1);
TPad* padResid = (TPad*) canvas->cd(2);
double small = 0.1;
padHisto->SetPad( 0., r , 1., 1. );
padHisto->SetBottomMargin( small );
padResid->SetPad( 0., 0., 1., r );
padResid->SetBottomMargin( 0.3 );
padResid->SetTopMargin ( small );
padHisto->cd();
//float fitvalue = frame.GetX()[60];
//Double_t nX = x.getVal();
//cout << "nX : " << nX << endl;
//results->SetBinContent(results->GetXaxis()->FindBin(60), nX);
//float fitvalue = results->GetBinContent(59);
//cout << "59.5daki fit value : " << fitvalue << endl;
//cb.plotOn(frame,LineColor(2));//this will show fit overlay on canvas
//cb.paramOn(frame); //this will display the fit parameters on canvas
//bw.plotOn(frame,LineColor(4));//this will show fit overlay on canvas
//bw.paramOn(frame); //this will display the fit parameters on canvas
//gPad->SetLogy();
cout << "chisquare : " << frame->chiSquare() << endl ;
//cout << "Total Number of events: " << Z_mass->Integral() << endl;
//cout << "Number of signal events: " << fsig.getVal() * Z_mass->Integral() << endl;
//cout << "Number of background events: " << (1 - fsig.getVal()) * Z_mass->Integral() << endl;
//Draw all frames on a canvas
//TPaveLabel *label1 = new TPaveLabel(1,500,15,700,"Chisquare:");
frame->GetXaxis()->SetTitle("Invariant mass w/ NoElectronVeto photon (in GeV/c^{2})");
//frame->GetXaxis()->SetTitle("Invariant mass w/ TIGHT photon (in GeV/c^{2})");
//frame->GetXaxis()->SetTitleOffset(1.2);
frame->Draw();
//float binsize = Z_mass->GetBinWidth(1);
//char Bsize[50];
//sprintf(Bsize,"Events per %2.2f",binsize);
//frame->GetYaxis()->SetTitle(Bsize);
//frame->GetYaxis()->SetTitleOffset(1.2);
//results->Sumw2();
padResid->cd();
hresidual->Draw();
Lines( hresidual );
hresidual->Draw( "SAME" );
//frame->Draw();
//results->Draw();
//canvas->Update();
//title->Draw("same");
//hresidual->Draw();
canvas->SaveAs("pdf_TagAndProbe/ResidualNumerator_binned_negligible_errorfit.pdf");
//canvas->SaveAs("pdf_TagAndProbe/ResidualDenominator_binned_negligible_errorfit.pdf");
}
