stira::common::FitCurve Class Reference

class to fit a parametric function to a set of data points For the moment, only linear, quadratic, exponential and polynomial fit are available The polynomial fit uses OpenCV's matrix operations More...

#include <FitCurve.h>

Public Member Functions
	FitCurve (std::vector< Point< int > > inputPoints)
	constructor More...
	FitCurve (std::vector< Point< double > > inputPoints)
	constructor More...
void	SetPoints (std::vector< Point< double > > inputPoints)
	sets the input data points (coordinates in double precision float format ) More...
void	SetPoints (std::vector< Point< int > > inputPoints)
	sets the input data points (coordinates in integer format ) More...
std::vector< double >	FitLinear ()
	fits a straight line to a set of 2D points (x, y = a * x + b) Source: http://mathworld.wolfram.com/LeastSquaresFitting.html More...
std::vector< double >	FitQuadratic ()
	fits a straight line to a set of 2D points (x, y = a * x^2 + b * x + c) Source: http://mathforum.org/library/drmath/view/72047.html More...
std::vector< double >	FitExponential ()
	fits an exponential to a set of 2D points ( x, y = a * exp( b * x ) ) Source http://mathworld.wolfram.com/LeastSquaresFittingExponential.html More...
std::vector< double >	GetAllAbsoluteErrors ()
	Gets a vector with all errors (in absolute value between the data points and the modeled points.
double	GetMeanAbsoluteError ()
	Gets a vector with the average absolute error.
std::vector< double >	PolynomialFit (int polynomialDegree)
	Fitting of a general polynomial of a given degree n to a set of 2D points (x, y = a * x^n + b * x^(n-1) + ... ) Math of the least square fitting: http://mathworld.wolfram.com/LeastSquaresFittingPolynomial.html Documentation of the used cv::Mat object: see http://opencv.willowgarage.com/documentation/cpp/basic_structures.html. More...
std::vector< double >	GaussianFitLeastSquares ()
	Fitting of a guassian profile to a set of 2D points (x, y = G(x) ) Reference: http://terpconnect.umd.edu/~toh/spectrum/CurveFitting.html. More...
bool	WriteToDisk (std::string fileName)
	writes source data points and points from the fitted model to disk More...

Detailed Description

class to fit a parametric function to a set of data points For the moment, only linear, quadratic, exponential and polynomial fit are available The polynomial fit uses OpenCV's matrix operations

Constructor & Destructor Documentation

stira::common::FitCurve::FitCurve

(

std::vector< Point< int > >

inputPoints

)

constructor

Parameters

inputPoints

the points to which the data model has to be fitted

stira::common::FitCurve::FitCurve

(

std::vector< Point< double > >

inputPoints

)

constructor

Parameters

inputPoints

the points to which the data model has to be fitted

Member Function Documentation

vector< double > stira::common::FitCurve::FitExponential

(

)

fits an exponential to a set of 2D points ( x, y = a * exp( b * x ) ) Source http://mathworld.wolfram.com/LeastSquaresFittingExponential.html

Returns

vector with output parameters ( a, b )

vector< double > stira::common::FitCurve::FitLinear

(

)

fits a straight line to a set of 2D points (x, y = a * x + b) Source: http://mathworld.wolfram.com/LeastSquaresFitting.html

Returns

vector with output parameters ( a, b )

Referenced by stira::common::QuantileNormalDistribution::AnalyseNormalizedHistogram().

vector< double > stira::common::FitCurve::FitQuadratic

(

)

fits a straight line to a set of 2D points (x, y = a * x^2 + b * x + c) Source: http://mathforum.org/library/drmath/view/72047.html

Returns

vector with output parameters ( a, b, c )

vector< double > stira::common::FitCurve::GaussianFitLeastSquares

(

)

Fitting of a guassian profile to a set of 2D points (x, y = G(x) ) Reference: http://terpconnect.umd.edu/~toh/spectrum/CurveFitting.html.

Returns

vector with in the following order ( peakHeight, peakPosition, sigma)

vector< double > stira::common::FitCurve::PolynomialFit

(

int

polynomialDegree

)

Fitting of a general polynomial of a given degree n to a set of 2D points (x, y = a * x^n + b * x^(n-1) + ... ) Math of the least square fitting: http://mathworld.wolfram.com/LeastSquaresFittingPolynomial.html Documentation of the used cv::Mat object: see http://opencv.willowgarage.com/documentation/cpp/basic_structures.html.

Parameters

polynomialDegree

the degree of the polynomial

Returns

vector with output parameters ( a, b, ... )

void stira::common::FitCurve::SetPoints

(

std::vector< Point< double > >

inputPoints

)

sets the input data points (coordinates in double precision float format )

Parameters

inputPoints

the input data points

void stira::common::FitCurve::SetPoints

(

std::vector< Point< int > >

inputPoints

)

sets the input data points (coordinates in integer format )

Parameters

inputPoints

the input data points

bool stira::common::FitCurve::WriteToDisk

(

std::string

fileName

)

writes source data points and points from the fitted model to disk

Parameters

fileName

name of file to write to

---

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

• common/common/FitCurve.h
• common/common/FitCurve.cpp