TF1Convolution is not work for me

ROOT
1

Hello Experts,

TF1Convolution() and drawing it from a TF1 is not working for me. I’m tried following this example with my own TF1 function: Convolution of two TF1 , but it gives me a flat line on the canvas. Not really sure what I need to do and ChatGPT is not sure either. Any help would be much appreciated. Here is my code for reference:

#include <iostream>
#include <fstream>
#include <vector>
using namespace std;

#include "TFile.h"
#include "TMath.h"
#include "TH1.h"
#include "TF1.h"
#include "TPDF.h"
#include "TCanvas.h"
#include "TStyle.h"
#include "TLatex.h"
#include "TString.h"
#include "TLine.h"
#include "TPaveStats.h"
#include "TF1Convolution.h"

void landau (){

  //Float_t pi = TMath::Pi();

  Double_t low = 0;

  Double_t high = 70;
  /*
  //Parameters changed for fit.
  TF1 *Landau1 = new TF1("Landau1","[0]*TMath::Landau(x,[1],[2])",low,high);
  //[1] = location parameter, [2] = scaling parameter
  Landau1->SetParameters(649.5,22.95,7.534);
  Landau1->SetTitle("Landau vs. Gaus*Landau Comparison");
*/
  TF1 *GausLandau = new TF1("GausLandau","[0]*TMath::Gaus(x,[1],[2])*TMath::Landau(x,[3],[4])",low,high);
  GausLandau->SetParameters(649.5,11.29,7.856,22.95,7.534);
  //GausLandau->GetYaxis()->SetRangeUser(0,700);
  GausLandau->SetLineColor(kBlue);
  GausLandau->SetTitle("Landau vs. Gaussian*Landau Comparison");

  TF1 *Landau1 = new TF1("Landau1","[0]*TMath::Landau(x,[1],[2])",low,high);
  //[1] = location parameter, [2] = scaling parameter
  Landau1->SetParameters(649.5,22.95,7.534);
  Landau1->GetYaxis()->SetRangeUser(0,700);
  TF1 *Gaus = new TF1("Gaus","[0]*TMath::Gaus(x,[1],[2])",low,high);
  Gaus->SetParameters(649.5,11.29,7.856);
  Gaus->SetLineColor(kGreen);

  TF1Convolution *conv = new TF1Convolution(Gaus,Landau1,low,high);
  TF1 *f = new TF1("f", *conv, low, high, conv->GetNpar());
  //f->SetParameters(11.29,7.856,22.95,7.534);


  TCanvas *canvas = new TCanvas("c","c",1000,1000);

  //GausLandau->Draw();

  canvas->Update();

  TPaveStats *stats = new TPaveStats(50, 110, 70, 70, "stats");
  stats->SetFillColor(0); // Set background color to transparent
  stats->SetTextAlign(12); // Align text left
  stats->SetBorderSize(1);

  // Customize content of the stats box for different parameters
  stats->AddText(Form("#alpha: %.2f", GausLandau->GetParameter(0))); // Add amplitude
  stats->AddText(Form("#mu_{g}: %.2f", GausLandau->GetParameter(1))); // Add gaussian mean
  stats->AddText(Form("#sigma_{g}: %.2f", GausLandau->GetParameter(2))); // Add gaussian sigma
  stats->AddText(Form("#mu_{l}: %.2f", GausLandau->GetParameter(3))); // Add landau mean
  stats->AddText(Form("#sigma_{l}: %.2f", GausLandau->GetParameter(4))); // Add landau sigma

  canvas->cd();
  stats->Draw();

  // Update the canvas
  canvas->Modified();
  canvas->Update();

  Landau1->Draw();
  Gaus->Draw("SAME");
  f->Draw("SAME");


  TLegend *l = new TLegend(0.7,0.7,0.9,0.9);
  l->AddEntry(Landau1,"Landau","l");
  //l->AddEntry(GausLandau,"Gaussian*Landau","l");
  l->AddEntry(Gaus,"Gaussian","l");
  l->AddEntry(f,"convolution:GL","l");
  l->Draw();

  //canvas->SaveAs("GausLandau_Statbox.pdf");

  TCanvas *canvas2 = new TCanvas("can2","can2",1000,1000);
  f->Draw();

}

-Charlie Clark

2

Hi Charlie,

Thanks for the post.
Have you tried starting from the official example in ROOT?

Cheers,
D

3

Hi Danilo,

I started from the official example for TF1Convolution and it makes sense to me now what that function is really used for. I cannot plot it individually from the randomly generated function that the convolution is trying to fit. However, this gives me an idea for my own code. Thanks for the suggestion.

4

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.