void FitDijetMass_Data() {
TFile *inf = new TFile("MassResults_ak7calo.root");
TH1F *hCorMassDen = (TH1F*) inf->Get("DiJetMass");
hCorMassDen->SetXTitle("Corrected Dijet Mass (GeV)");
hCorMassDen->SetYTitle("Events/GeV");
hCorMassDen->GetYaxis()->SetTitleOffset(1.5);
hCorMassDen->SetMarkerStyle(20);
hCorMassDen->GetXaxis()->SetRangeUser(120.,900.);
gROOT->ProcessLine(".L tdrstyle.C");
setTDRStyle();
tdrStyle->SetErrorX(0.5);
tdrStyle->SetPadRightMargin(0.08);
tdrStyle->SetLegendBorderSize(0);
gStyle->SetOptFit(1111);
tdrStyle->SetOptStat(0);
TCanvas* c2 = new TCanvas("c2","DijetMass", 500, 500);
/////// perform 4 parameters fit
TF1 *func = new TF1("func", "[0]*((1-x/7000.+[3]*(x/7000)^2)^[1])/(x^[2])",
100., 1000.);
func->SetParameter(0, 1.0e+08);
func->SetParameter(1, -1.23);
func->SetParameter(2, 4.13);
func->SetParameter(3, 1.0);
func->SetLineColor(4);
func->SetLineWidth(3);
TVirtualFitter::SetMaxIterations( 10000 );
TVirtualFitter *fitter;
TMatrixDSym* cov_matrix;
int fitStatus = hCorMassDen->Fit("func","LLI","",130.0, 800.0); // QCD fit
TH1F *hFitUncertainty = hCorMassDen->Clone("hFitUncertainty");
hFitUncertainty->SetLineColor(5);
hFitUncertainty->SetFillColor(5);
hFitUncertainty->SetMarkerColor(5);
if (fitStatus == 0) {
fitter = TVirtualFitter::GetFitter();
double* m_elements = fitter->GetCovarianceMatrix();
cov_matrix = new TMatrixDSym( func->GetNumberFreeParameters(),m_elements);
cov_matrix->Print();
double x, y, e;
for(int i=0;i<hFitUncertainty->GetNbinsX();i++)
{
x = hFitUncertainty->GetBinCenter(i+1);
y = func->Eval(x);
e = QCDFitUncertainty( func, *cov_matrix, x);
hFitUncertainty->SetBinContent(i+1,y);
hFitUncertainty->SetBinError(i+1,e);
}
}
hCorMassDen->Draw("ep");
gPad->Update();
TPaveStats *st = (TPaveStats*)hCorMassDen->FindObject("stats");
st->SetName("stats1");
st->SetX1NDC(0.3); //new x start position
st->SetX2NDC(0.6); //new x end position
st->SetTextColor(4);
hCorMassDen->GetListOfFunctions()->Add(st);
/////// perform 2 parameters fit
TF1 *func2 = new TF1("func2", "[0]*(1-x/7000.)/(x^[1])", 100., 1000.);
func2->SetParameter(0, 10000.);
func2->SetParameter(1, 5.0);
func2->SetLineWidth(3);
fitStatus = hCorMassDen->Fit("func2","LLI","",130.0, 800.0); // QCD fit
TH1F *hFitUncertainty2 = hCorMassDen->Clone("hFitUncertainty2");
hFitUncertainty2->SetLineColor(kGray);
hFitUncertainty2->SetFillColor(kGray);
hFitUncertainty2->SetMarkerColor(kGray);
if (fitStatus == 0) {
fitter = TVirtualFitter::GetFitter();
double* m_elements = fitter->GetCovarianceMatrix();
cov_matrix = new TMatrixDSym( func2->GetNumberFreeParameters(),m_elements);
cov_matrix->Print();
double x, y, e;
for(int i=0;i<hFitUncertainty2->GetNbinsX();i++)
{
x = hFitUncertainty2->GetBinCenter(i+1);
y = func2->Eval(x);
e = QCDFitUncertainty( func2, *cov_matrix, x);
hFitUncertainty2->SetBinContent(i+1,y);
hFitUncertainty2->SetBinError(i+1,e);
}
}
hFitUncertainty->Draw("E3 same");
hCorMassDen->Draw("ep sames");
hFitUncertainty2->Draw("E3 same");
hCorMassDen->Draw("ep sames");
func2->Draw("same");
c2->SetLogy(1);
/*
TH1F *hCorMass = hCorMassDen->Clone("hCorMass");
for(int i=0; i<hCorMass->GetNbinsX(); i++){
hCorMass->SetBinContent(i+1, hCorMassDen->GetBinContent(i+1) * hCorMassDen->GetBinWidth(i+1));
hCorMass->SetBinError(i+1, hCorMassDen->GetBinError(i+1) * hCorMassDen->GetBinWidth(i+1));
}
// Our observable is the invariant mass
RooRealVar invMass("invMass", "Corrected dijet mass",
100., 1000.0, "GeV");
RooDataHist data( "data", "", invMass, hCorMass);
//////////////////////////////////////////////
// make QCD model
RooRealVar p0("p0", "# events", 600.0, 0.0, 10000000000.);
RooRealVar p1("p1","p1", 3.975, -10., 10.) ;
RooRealVar p2("p2","p2", 5.302, 4., 8.) ;
RooRealVar p3("p3","p3", -1.51, -100., 100.) ;
// // define QCD line shape
RooGenericPdf qcdModel("qcdModel", "pow(1-@0/7000.+@3*(@0/7000.)*(@0/7000.),@1)*pow(@0/7000.,-@2)",
RooArgList(invMass,p1,p2,p3));
// full model
RooAddPdf model("model","qcd",RooArgList(qcdModel), RooArgList(p0));
//plot sig candidates, full model, and individual componenets
// __ _ _
// / _(_) |_
// | |_| | __|
// | _| | |_
// |_| |_|\__|
// Important: fit integrating f(x) over ranges defined by X errors, rather
// than taking point at center of bin
RooFitResult* fit = model.fitTo(data, Minos(kFALSE), Extended(kTRUE),
SumW2Error(kFALSE),Save(kTRUE), Range(130.,800.),
Integrate(kTRUE) );
// to perform chi^2 minimization fit instead
// RooFitResult* fit = model.chi2FitTo(data, Extended(kTRUE),
// Save(),Range(50.,526.),Integrate(kTRUE) );
fit->Print();
//plot data
TCanvas* cdataNull = new TCanvas("cdataNull","fit to dijet mass",500,500);
RooPlot* frame1 = invMass.frame() ;
data.plotOn(frame1, DataError(RooAbsData::SumW2) ) ;
model.plotOn(frame1, LineColor(kBlue)) ;
model.plotOn(frame1, VisualizeError(*fit, 1),FillColor(kYellow)) ;
data.plotOn(frame1, DataError(RooAbsData::SumW2) ) ;
model.plotOn(frame1, LineColor(kBlue)) ;
model.paramOn(frame1, Layout(0.4, 0.85, 0.92));
TPaveText* dataPave = (TPaveText*) frame1->findObject("model_paramBox");
dataPave->SetY1(0.77);
gPad->SetLogy();
frame1->GetYaxis()->SetNoExponent();
frame1->GetYaxis()->SetRangeUser(5E-2,5E+4);
frame1->GetYaxis()->SetTitle("Events / bin");
frame1->GetYaxis()->SetTitleOffset(1.35);
frame1->SetTitle("fit to data with QCD lineshape");
frame1->Draw() ;
// S h o w r e s i d u a l a n d p u l l d i s t s
// -------------------------------------------------------
//// Construct a histogram with the residuals of the data w.r.t. the curve
RooHist* hresid = frame1->residHist() ;
// Create a new frame to draw the residual distribution and add the distribution to the frame
RooPlot* frame2 = invMass.frame(Title("Residual Distribution")) ;
frame2->addPlotable(hresid,"P") ;
///// Construct a histogram with the pulls of the data w.r.t the curve
RooHist* hpull = frame1->pullHist() ;
//// Create a new frame to draw the pull distribution and add the distribution to the frame
RooPlot* frame3 = invMass.frame(Title("Pull Distribution")) ;
frame3->addPlotable(hpull,"P") ;
TCanvas* cResidual = new TCanvas("cResidual","Residual Distribution",1000,500);
cResidual->Divide(2) ;
cResidual->cd(1) ; gPad->SetLeftMargin(0.15) ; frame1->GetYaxis()->SetTitleOffset(1.6) ; frame2->Draw() ;
cResidual->cd(2) ; gPad->SetLeftMargin(0.15) ; frame1->GetYaxis()->SetTitleOffset(1.6) ; frame3->Draw() ;
*/
}
