FittingUtils.hpp File Reference

Binned chi-squared photopeak fitting on a TF1 + Minuit2 backend. More...

#include "PlottingUtils.hpp"
#include <TCanvas.h>
#include <TF1.h>
#include <TFile.h>
#include <TFitResult.h>
#include <TH1.h>
#include <TMath.h>
#include <TPad.h>
#include <TROOT.h>
#include <TSystem.h>
#include <TTree.h>
#include <fstream>
#include <iomanip>

Include dependency graph for FittingUtils.hpp:

Go to the source code of this file.

Classes
struct	PeakFitResult
	Fitted parameters and errors for one peak. More...
struct	FitParameterDiagnostic
	One fitted parameter's value, error and proximity to its limits. More...
struct	FitResult
	Result of one fit: peaks, background, quality, and diagnostics. More...
class	FittingUtils
	Binned chi-squared photopeak fitting over a TH1. More...

Namespaces
namespace	FittingFunctions
	Raw model functions in TF1 form: f(Double_t *x, Double_t *par).

Functions
Bool_t	LaunchInteractiveFitEditor (TH1 *hist, TF1 *fit_func, Double_t range_low, Double_t range_high, Int_t num_peaks=1, const TString &info_label="")
Double_t	FittingFunctions::Gaussian (Double_t *x, Double_t *par)
	Gaussian peak.
Double_t	FittingFunctions::LinearBackground (Double_t *x, Double_t *par)
	Straight-line background.
Double_t	FittingFunctions::Step (Double_t *x, Double_t *par)
	Resolution-smeared step shelf below the peak.
Double_t	FittingFunctions::LowTail (Double_t *x, Double_t *par)
	Combined exponential and linear tail below the peak.
Double_t	FittingFunctions::HighTail (Double_t *x, Double_t *par)
	Exponential tail above the peak, from pileup.
Double_t	FittingFunctions::PeakFunction (Double_t *x, Double_t *par)
	One peak with every optional component, plus a linear background.
Double_t	FittingFunctions::DoublePeakFunction (Double_t *x, Double_t *par)
	Two peaks sharing one linear background.
Double_t	FittingFunctions::TriplePeakFunction (Double_t *x, Double_t *par)
	Three peaks sharing one linear background.

Detailed Description

Binned chi-squared photopeak fitting on a TF1 + Minuit2 backend.

The counterpart to RooFitUtils, which fits the same model by unbinned extended maximum likelihood. Both take the same constructor arguments, expose the same Set* component flags, and return the same FitResult, so a call site can swap one for the other. This backend is faster, because the model is evaluated pointwise and never normalised; the RooFit backend is the one to reach for when you need correct likelihood errors or simultaneous fits.

Saved interactive parameters use the .fits extension here and .roofits in the RooFit backend, so both can coexist for the same input.

Definition in file FittingUtils.hpp.

Function Documentation

LaunchInteractiveFitEditor()

Bool_t LaunchInteractiveFitEditor	(	TH1 *	hist,
		TF1 *	fit_func,
		Double_t	range_low,
		Double_t	range_high,
		Int_t	num_peaks = 1,
		const TString &	info_label = "" )

Definition at line 1037 of file InteractiveFitEditor.cpp.

Referenced by FittingUtils::FitDoublePeak(), FittingUtils::FitDoublePeak(), FittingUtils::FitSinglePeak(), and FittingUtils::FitTriplePeak().