LeastSqFit Class Reference
[Statistics]

Constructs a linear least squares fit to a straight line from input data. More...

#include <rw/lsqfit.h>

List of all members.

Public Member Functions
	LeastSqFit (const RWMathVec< double > &x, const RWMathVec< double > &y)
	LeastSqFit (const RWMathVec< double > &x, const RWMathVec< double > &y, const RWMathVec< double > &sigmay)
double	correlationCoeff () const
double	intercept () const
double	interceptStandardDev () const
double	slope () const
double	slopeStandardDev () const
double	xPosition (double y) const
double	yPosition (double x) const
Friends
std::ostream &	operator<< (std::ostream &s, const LeastSqFit &f)

---

Detailed Description

The class LeastSqFit constructs a linear least squares fit to a straight line from input data, with or without weighting factors. This class is of precision double. Once a LeastSqFit is constructed, various methods are provided to interrogate it.

Synopsis

 #include <rw/lsqfit.h>
 #include <rw/math/mathvec.h>
 RWMathVec<double> x, y;
 LeastSqFit lsf(x, y);

---

Constructor & Destructor Documentation

LeastSqFit::LeastSqFit	(	const RWMathVec< double > &	x,
		const RWMathVec< double > &	y
	)

Constructs a linear least squares fit to the data in vectors x and y, without weights. Similar indices of the vectors x(i) and y(i) are assumed to represent (x,y) pairs. The vectors need not be the same length; trailing unmatched points are ignored.

LeastSqFit::LeastSqFit	(	const RWMathVec< double > &	x,
		const RWMathVec< double > &	y,
		const RWMathVec< double > &	sigmay
	)

Constructs a linear least squares fit to the data in vectors x and y, using a set of known standard deviations for the y values. Similar indices of the vectors x(i) and y(i) are assumed to represent (x,y) pairs. These values are weighted according to the y standard deviations given as . The vectors need not be the same length; trailing unmatched points are ignored.

---

Member Function Documentation

double LeastSqFit::correlationCoeff

(

)

const [inline]

Returns the sample linear correlation coefficient of the fit.

double LeastSqFit::intercept

(

)

const [inline]

Returns the intercept of the line.

double LeastSqFit::interceptStandardDev

(

)

const [inline]

Returns the standard deviation of the calculated intercept. If the standard deviations of the data is given, the standard deviation for the intercept is calculated appropriately. In the unweighted case, the standard deviation is estimated by factoring in the variance of the data from the calculated line.

double LeastSqFit::slope

(

)

const [inline]

Returns the slope of the line.

double LeastSqFit::slopeStandardDev

(

)

const [inline]

Returns the standard deviation of the calculated slope. If the standard deviations of the data are given, the standard deviation for the slope is calculated appropriately. In the unweighted case, the standard deviation is estimated by factoring in the variance of the data from the calculated line.

double LeastSqFit::xPosition

(

double

y

)

const [inline]

Using the linear fit, returns the value x corresponding to y.

double LeastSqFit::yPosition

(

double

x

)

const [inline]

Using the linear fit, returns the value y corresponding to x.

---

Friends And Related Function Documentation

std::ostream& operator<<	(	std::ostream &	s,
		const LeastSqFit &	f
	)			[friend]

Outputs a summary of the linear fit f on std::ostream s.