using namespace RooFit ;

void ex5()
{
  // Make model

  // A workspace is a container class for RooFit data modeling objects
  // (variables, functions, probability density functions etc...)
  // Here we make an empty workspace

  RooWorkspace w("w") ;

  // Now we use the 'factory' to fill the workspace with the definition
  // of three variables (x,mean and sigma) and the definition of a Gaussian
  // probability density function 'g' in terms of (x,mean,sigma)s

  w.factory("Gaussian::g(x[-10,10],mean[-10,10],sigma[3,0.1,10])") ;

  // Generate an unbinned toy dataset 
  // of 10000 events in observable x from pdf g

  RooDataSet* data = w->pdf("g")->generate(*w->var("x"),10000) ;

  // Fit model g to the data using an unbinned ML fit

  RooFitResult* r = w->pdf("g")->fitTo(*data,Save()) ;
  
  // Make plot
  // First make a plot frame in x, then plot the data on that frame,
  // and finally overlay the fitted pdf over the data
  
  RooPlot* frame = w.var("x")->frame(100) ;
  data->plotOn(frame) ;
  w.pdf("g")->plotOn(frame) ;
  frame->Draw() ;  

//   // BLOCK 1
//   cout << "the correlation matrix" << endl ;
//   r->correlationMatrix().Print() ;
//   cout << "the covariance matrix" << endl ;
//   r->covarianceMatrix().Print() ;
  
//   // BLOCK 2
//   w.pdf("g")->plotOn(frame,VisualizeError(*r),MoveToBack()) ;
//   frame->Draw() ;

//   // BLOCK 3
//   w.pdf("g")->plotOn(frame,VisualizeError(*r,*w.var("mean")),FillColor(kBlue),Name("mean_only")) ;
//   frame->Draw() ;


}