Chi2FitFigureOfMerit Class Reference

Class implementing a \(\chi^2/ndf\) figure of merit. More...

#include <Chi2FitFigureOfMerit.h>

Inheritance diagram for Chi2FitFigureOfMerit:

Collaboration diagram for Chi2FitFigureOfMerit:

Public Member Functions
	Chi2FitFigureOfMerit ()
	~Chi2FitFigureOfMerit ()
void	addData (const std::vector< double > &x, double y, double ey)
void	clearData ()
double	evaluate (IModel *model) const
bool	isBetterThan (double scoreToTest, double referenceScore) const
Public Member Functions inherited from IFigureOfMerit
	IFigureOfMerit ()
virtual	~IFigureOfMerit ()
virtual bool	accept (IModel *model) const
virtual bool	isBetterThan (IModel *scoreToTest, IModel *referenceModel) const
void	setAcceptThreshold (double acceptThreshold)
double	getAcceptThreshold () const

Protected Attributes
std::vector< std::vector< double > >	m_x
	Stores the \(\vec{x_i}\) coordinates.
std::vector< double >	m_y
	Stores the \(y_i\) coordinates.
std::vector< double >	m_ey
	Stores the \(\sigma_{y_i}\) errors.
Protected Attributes inherited from IFigureOfMerit
double	m_acceptThreshold
	Stores the score threshold to accept a model as a final answer.

Additional Inherited Members
Protected Member Functions inherited from IFigureOfMerit
virtual bool	accept (double scrore) const

Detailed Description

Class implementing a \(\chi^2/ndf\) figure of merit.

The \(\chi^2/ndf\) is computed by comparing a model to a dataset:

• The data consists of \((\vec{x_i}, y_i)\) pairs with associated errors \(\sigma_{y_i}\) on \(y_i\).
• The model is a function of the form \(y = f(\vec{x})\).
• The figure of merit is then defined as:

  \[ \chi^2/ndf = \frac{1}{N}\sum_{i=0}^N \frac{(y_i - f(\vec{x_i}))^2}{\sigma_{y_i}^2} \]

Definition at line 21 of file Chi2FitFigureOfMerit.h.

Constructor & Destructor Documentation

Chi2FitFigureOfMerit()

Chi2FitFigureOfMerit::Chi2FitFigureOfMerit

(

)

Default Constructor

Definition at line 8 of file Chi2FitFigureOfMerit.cxx.

~Chi2FitFigureOfMerit()

Chi2FitFigureOfMerit::~Chi2FitFigureOfMerit

(

)

Destructor

Definition at line 14 of file Chi2FitFigureOfMerit.cxx.

Member Function Documentation

addData()

void Chi2FitFigureOfMerit::addData	(	const std::vector< double > &	x,
		double	y,
		double	ey
	)

Adds a data point

Parameters

x	\(\vec{x}\) coordinate.
y	\(y\) coordinate.
ey	\(\sigma_y\) error on \(y\) coordinate.

Definition at line 23 of file Chi2FitFigureOfMerit.cxx.

clearData()

void Chi2FitFigureOfMerit::clearData

(

)

Clear all data

Definition at line 30 of file Chi2FitFigureOfMerit.cxx.

evaluate()

double Chi2FitFigureOfMerit::evaluate ( IModel *  imodel ) const

virtual

Compute the score ( \(\chi^2/ndf\)) for a given model relative to the data points.

Parameters

imodel

Model to be evaluated.

Returns

\(\chi^2/ndf\) score.

The \(\chi^2/ndf\) score is then defined as:

\[ \chi^2/ndf = \frac{1}{N}\sum_{i=0}^N \frac{(y_i - f(\vec{x_i}))^2}{\sigma_{y_i}^2} \]

Implements IFigureOfMerit.

Definition at line 46 of file Chi2FitFigureOfMerit.cxx.

isBetterThan()

bool Chi2FitFigureOfMerit::isBetterThan ( double  scoreToTest, double  referenceScore  ) const

virtual

Compares two score values

Parameters

scoreToTest	Score value to be tested.
referenceScore	Score value to be compared to.

Returns

true if scoreToTest < referenceScore

Reimplemented from IFigureOfMerit.

Definition at line 76 of file Chi2FitFigureOfMerit.cxx.

---

The documentation for this class was generated from the following files:

• include/Chi2FitFigureOfMerit.h
• src/Chi2FitFigureOfMerit.cxx