int main(){
RooRealVar zero("zero","",0,0,150);
//input parameters for the distribution
double alpha_double = 2.38;
double beta_double = 0.092;
double lambda_double = 2.21;
double theta_double = 1.12;
double c_frac = 0.5;
//RooRealVar fsig("fsig","signal fraction",0.3) ;
string epsname = "plots/gamma_pion20";
hit_info multi_shower("fullsim/example_pi20Gev.root");
//hit_info multi_shower("fullsim/example_ele_20GeV.root");
double beta_gflash_factor = 1.49;
double theta_gflash_factor = 16.42;
beta_double = beta_double/beta_gflash_factor;
theta_double = theta_double/theta_gflash_factor;
// --- Observable ---
RooRealVar depth("depth","depth [cm]",0,800);
depth.removeMax(); // set infinite range
// --- Weight ---
RooRealVar* weight = new RooRealVar("weight","1/E dE/dr", 0.0, 1.0);
// --- Gamma pdf ---
//if one of the Values is set to const you can not ask for it in RooFitModel, otherwise the program will crash
//RooRealVar alpha("alpha","alpha",alpha_double,0.0,4);
RooRealVar alpha("alpha","alpha",0.0,20.);
//alpha.removeMax(); // set infinite range
//RooRealVar beta("beta","beta",1./beta_double,0.0,100.);
RooRealVar beta("beta","beta",0.1,.1,3.5);
//beta.removeMax(); // set infinite range
RooRealVar mu("mu","mu",0); // no ranges means variable is fixed in fit
RooRealVar nu("nu","nu",0); // no ranges means variable is fixed in fit
//RooRealVar lambda("lambda","lambda",lambda_double,0.0,6.);
RooRealVar lambda("lambda","lambda",0.0,40.);
//lambda.removeMax(); // set infinite range
//RooRealVar theta("theta","theta",1./theta_double,0.0,100.) ;
RooRealVar theta("theta","theta",1.,47) ;
//theta.removeMax(); // set infinite range
RooGamma gamma_one("gamma_one","gamma_one pdf",depth,alpha,beta,mu) ;
RooGamma gamma_two("gamma_two","gamma_two pdf",depth,lambda,theta,nu) ;
RooRealVar fsig("fsig","signal fraction",c_frac,0.0,1.);
//RooRealVar fsig("fsig","signal fraction",c_frac);
RooAddPdf model("model","model",RooArgList(gamma_two,gamma_one),fsig) ;
RooDataSet mean("mean","mean",RooArgSet(depth,weight),"weight");
TH1F* mean_fithist = new TH1F("mean_fithist"," mean fit hist",50,0,150);
TH1F* log_alpha1 = new TH1F("log_alpha1"," log(#alpha_{1})",20,-1,3);
TH1F* log_beta1 = new TH1F("log_beta1"," log(#beta_{1})",20,-1,3);
TH2F* log_corr1 = new TH2F("log_corr1"," correlation #alpha_{1} #beta_{1}",20,-1,3,20,-1,3);
TH1F* log_alpha2 = new TH1F("log_alpha2"," log(#alpha_{2})",20,-1,3);
TH1F* log_beta2 = new TH1F("log_beta2"," log(#beta_{2})",20,-1,3);
TH2F* log_corr2 = new TH2F("log_corr2"," correlation #alpha_{2} #beta_{2}",20,-1,3,20,-1,3);
TH2F* mean_hist = new TH2F("mean_hist"," mean hist",100,0,150,100.,0.,.0015);
TCanvas * can = new TCanvas("can", "can", 600, 500);
can->cd();
set_style();
//can->Print(epsname+"[");
double c_tot =0;
unsigned int n_part =100;
for(unsigned int p = 0; p<n_part ; ++p){
TLegend * legend = new TLegend(0.5,0.65,.88,0.88);
legend->SetTextFont(72);
legend->SetTextSize(0.04);
legend->SetFillColor(kWhite);
legend->SetBorderSize(0);
one_shower shower = multi_shower.all_shower.at(p);
//DataSet I am looking at
RooDataSet full_sim_data("full_sim_data","full_sim_data",RooArgSet(depth,weight),"weight");
TH1F* datahisto = new TH1F("datahist","datahist",15,0,120);
double sumHCAL=0;
double deltar =-1;
for(unsigned int i = 0; i < shower.mytype.size(); ++i){
sumHCAL += shower.myenergy.at(i);
double x_val = shower.myz.at(i);
double y_val = shower.myy.at(i);
double z_val = shower.myz.at(i);
z_val = z_val - shower.interactPoint;
double distance=-1;
if(sqrt(x_val*x_val+y_val*y_val+z_val*z_val)/10 < 200) distance = sqrt(x_val*x_val+y_val*y_val+z_val*z_val)/10;
if(deltar < distance) deltar = z_val;
}
deltar=deltar/shower.mytype.size();
c_tot += shower.ePi0first_energy/n_part;
fsig.setVal(shower.ePi0first_energy);
fsig.setRange(shower.ePi0first_energy-0.05 > 0 ? shower.ePi0first_energy-0.05: 0,shower.ePi0first_energy+0.05<1?shower.ePi0first_energy+0.05:1);
//fsig.setRange(1-shower.ePi0first_energy-0.05 > 0 ? 1-shower.ePi0first_energy-0.05: 0,1-shower.ePi0first_energy+0.05<1 ? 1-shower.ePi0first_energy+0.05:1);
for(unsigned int i = 0; i < shower.mytype.size(); ++i){
double x_val = shower.myx.at(i);
double y_val = shower.myy.at(i);
double z_val = shower.myz.at(i);
// if(shower.mytype.at(i) !=2)continue;
z_val = z_val - shower.interactPoint;
//if(z_val<=0) continue;
depth.setVal(sqrt(x_val*x_val+y_val*y_val+z_val*z_val)/10);
//depth.setVal(z_val);
//cout <<depth.getVal()<<endl;
datahisto->Fill(depth.getVal(),shower.myenergy.at(i)/sumHCAL/deltar);
mean_fithist->Fill(depth.getVal(),shower.myenergy.at(i)/sumHCAL/deltar/n_part);
mean_hist->Fill(depth.getVal(),shower.myenergy.at(i)/sumHCAL/deltar);
full_sim_data.add(depth,shower.myenergy.at(i)/sumHCAL/deltar,0);
mean.add(depth,shower.myenergy.at(i)/sumHCAL/deltar,0);
}
RooDataHist roodatahist("roodatahist","roodatahist",depth,Import(*datahisto)) ;
//model.fitTo(roodatahist, Strategy(2) );
model.fitTo(full_sim_data);
RooDataHist* full_sim_binned = full_sim_data.binnedClone("full_sim_binned","full sim binned");
//prepare to plot fits
RooPlot * depthFrame = depth.frame(0,120,100);
depthFrame->SetTitle("");
depthFrame->SetTitleOffset(1.2,"Y");
depthFrame->SetLabelSize(0.02,"Y");
depthFrame->SetYTitle("1/E dE/dr");
//zero.plotOn(depthFrame,LineStyle(kDashed),LineColor(kBlack));
//full_sim_data.plotOn(depthFrame);
roodatahist.plotOn(depthFrame);
//full_sim_binned->plotOn(depthFrame);
model.plotOn(depthFrame,Components(gamma_one),LineStyle(kDashed),LineColor(kGreen),Name("gamma_one"));
model.plotOn(depthFrame,Components(gamma_two),LineStyle(kDashed),LineColor(kRed),Name("gamma_two"));
model.plotOn(depthFrame,Name("model"));
stringstream gamma_one_text;
gamma_one_text.precision(3);
gamma_one_text << "#Gamma_{1} ("<<alpha.getVal()<<","<<(1/beta.getVal()*beta_gflash_factor)<<")"<<endl;
stringstream gamma_two_text;
gamma_two_text.precision(3);
gamma_two_text << "#Gamma_{2} (" <<lambda.getVal()<<","<<(1/theta.getVal()*theta_gflash_factor)<< ")"<<endl;
stringstream gamma_text;
gamma_text.precision(2);
gamma_text << "#Gamma_{1/2} c=" <<fsig.getVal() <<"("<<shower.ePi0first_energy<<")"<<endl;
stringstream chi2;
chi2.precision(2);
chi2 << "#chi^{2} prob. = " << TMath::Prob(depthFrame->chiSquare(),14) <<endl;
legend->AddEntry(depthFrame->findObject("gamma_one"),gamma_one_text.str().c_str(),"L");
legend->AddEntry(depthFrame->findObject("gamma_two"),gamma_two_text.str().c_str(),"L");
legend->AddEntry(depthFrame->findObject("model"),gamma_text.str().c_str(),"L");
legend->AddEntry((TObject*)0, chi2.str().c_str(), "");
cout << "alpha, beta "<< alpha.getVal() << " " << 1/beta.getVal()*beta_gflash_factor << std::endl;
cout << "lambda, theta "<<lambda.getVal() << " " << 1/theta.getVal()*theta_gflash_factor << std::endl;
cout <<"c, chi2 "<< fsig.getVal() << " "<<depthFrame->chiSquare()<< std::endl;
cout <<"starting , interacting Points "<<shower.startingPoint<<" "<<shower.interactPoint <<endl;
cout<< "pi0first, pi0tot "<<shower.ePi0first_energy <<" "<<shower.ePi0tot_energy<<endl;
/*
log_alpha1->Fill(log10(alpha.getVal()));
log_alpha2->Fill(log10(lambda.getVal()));
//cout<<log(1/beta.getVal()*beta_gflash_factor)<<endl;
log_beta1 ->Fill(log10(1/beta.getVal()*beta_gflash_factor));
log_beta2 ->Fill(log10(1/theta.getVal()*theta_gflash_factor));
log_corr1 ->Fill(log10(alpha.getVal()),log10(1/beta.getVal()*beta_gflash_factor));
log_corr2 ->Fill(log10(lambda.getVal()),log10(1/theta.getVal()*theta_gflash_factor));
*/
if(fsig.getVal()>0.1)log_alpha1->Fill(log(alpha.getVal()));
if(fsig.getVal()<0.9)log_alpha2->Fill(log(lambda.getVal()));
//cout<<log(1/beta.getVal()*beta_gflash_factor)<<endl;
if(fsig.getVal()>0.1)log_beta1 ->Fill(log(1/beta.getVal()*beta_gflash_factor));
if(fsig.getVal()<0.9)log_beta2 ->Fill(log(1/theta.getVal()*theta_gflash_factor));
if(fsig.getVal()>0.1)log_corr1 ->Fill(log(alpha.getVal()),log(1/beta.getVal()*beta_gflash_factor));
if(fsig.getVal()<0.9)log_corr2 ->Fill(log(lambda.getVal()),log(1/theta.getVal()*theta_gflash_factor));
//prepare to dump all the plots in the ps file
// one data sample and its fit is drawn
/*
datahisto->Draw();
datahisto0->SetLineColor(kRed);
datahisto0->Draw("same");
datahisto1->SetLineColor(kGreen);
datahisto1->Draw("same");
datahisto2->SetLineColor(kBlue);
datahisto2->Draw("same");
can->Print(epsname);
*/
//if(depthFrame->chiSquare()>5) continue;
depthFrame->Draw("HIST p");
legend->Draw();
stringstream particle_number;
particle_number.precision(1);
particle_number << p <<endl;
can->Print((epsname+"_"+particle_number.str()+".pdf").c_str());
}
RooDataHist roomeandatahist("roomeandatahist","roomeandatahist",depth,Import(*mean_fithist)) ;
fsig.setVal(c_tot);
alpha.removeMax(); // set infinite range
beta.removeMax(); // set infinite range
lambda.removeMax(); // set infinite range
theta.removeMax(); // set infinite range
//model.fitTo(roomeandatahist);
//model.fitTo(mean,SumW2Error(kTRUE));
TLegend * mean_legend = new TLegend(0.5,0.65,.9,0.9);
mean_legend->SetTextFont(72);
mean_legend->SetTextSize(0.04);
mean_legend->SetFillColor(kWhite);
mean_legend->SetBorderSize(0);
stringstream mean_gamma_one_text;
mean_gamma_one_text.precision(3);
mean_gamma_one_text << "#Gamma_{1} ("<<alpha.getVal()<<","<<(1/beta.getVal()*beta_gflash_factor)<<")"<<endl;
stringstream mean_gamma_two_text;
mean_gamma_two_text.precision(3);
mean_gamma_two_text << "#Gamma_{2} (" <<lambda.getVal()<<","<<(1/theta.getVal()*theta_gflash_factor)<< ")"<<endl;
stringstream mean_gamma_text;
mean_gamma_text.precision(2);
mean_gamma_text << "#Gamma_{1/2} c=" <<fsig.getVal() <<"("<< c_tot<<")"<<endl;
stringstream mean_chi2;
mean_chi2.precision(2);
//mean_chi2 << "#chi^{2} " <<depthFrame->chiSquare() <<endl;
RooPlot * meanFrame = depth.frame(0,120,200);
//meanFrame->SetMaximum(1);
meanFrame->SetTitle("Mean Fit");
meanFrame->SetYTitle("1/E dE/dr");
model.plotOn(meanFrame,Components(gamma_one),LineStyle(kDashed),LineColor(kGreen),Name("gamma_one"));
model.plotOn(meanFrame,Components(gamma_two),LineStyle(kDashed),LineColor(kRed),Name("gamma_two"));
model.plotOn(meanFrame,Name("model"));
//mean.plotOn(meanFrame);
//roomeandatahist.plotOn(meanFrame);
mean_legend->AddEntry(meanFrame->findObject("gamma_one"),mean_gamma_one_text.str().c_str(),"L");
mean_legend->AddEntry(meanFrame->findObject("gamma_two"),mean_gamma_two_text.str().c_str(),"L");
mean_legend->AddEntry(meanFrame->findObject("model"),mean_gamma_text.str().c_str(),"L");
meanFrame->Draw();
mean_legend->Draw();
//datahisto->Draw();
//can->Print(epsname);
//mean_hist->Draw("box");
//can->Print(epsname);
can->Print((epsname+"_mean.pdf").c_str());
log_alpha1->Draw();
can->Print((epsname+"_alpha1_log.pdf").c_str());
log_beta1->Draw();
can->Print((epsname+"_beta1_log.pdf").c_str());
log_corr1->Draw("colz");
can->Print((epsname+"corr1_log.pdf").c_str());
log_alpha2->Draw();
can->Print((epsname+"apha2_log.pdf").c_str());
log_beta2->Draw();
can->Print((epsname+"beta2_log.pdf").c_str());
log_corr2->Draw("colz");
can->Print((epsname+"corr2_log.pdf").c_str());
///can->Print(epsname+"]");
return 0;
}
