LinearRegression Class Reference

Performs linear regression to fit a straight line to data points. More...

#include <gin_linearregression.h>

Public Member Functions
	LinearRegression ()
	Creates an empty linear regression object.
	LinearRegression (juce::Array< Point< double > > points)
	Creates a linear regression object and adds the given points.
void	addPoint (Point< double > pnt)
	Adds a data point to the regression.
void	calculate ()
	Calculates the regression coefficients from the added points.
void	clear ()
	Removes all data points and resets the regression.
bool	haveData () const
	Checks if enough data points have been added.
int	items () const
	Returns the number of data points.
double	getA () const
	Returns the intercept of the fitted line.
double	getB () const
	Returns the slope of the fitted line.
double	getCoefDeterm () const
	Returns the coefficient of determination (R²).
double	getCoefCorrel () const
	Returns the coefficient of correlation.
double	getStdErrorEst () const
	Returns the standard error of estimate.
double	estimateY (double x) const
	Estimates the y value for a given x using the fitted line.

Protected Attributes
int	n = 0
double	sumX = 0
double	sumY = 0
double	sumXsquared = 0
double	sumYsquared = 0
double	sumXY = 0
double	a = 0
double	b = 0
double	coefD = 0
double	coefC = 0
double	stdError = 0

Detailed Description

Performs linear regression to fit a straight line to data points.

This class fits a linear equation y = a + bx to a set of data points using the least squares method. It calculates the coefficients (a, b) that minimize the sum of squared differences between the data points and the line.

The class also computes:

• Coefficient of determination (R²): Measures goodness of fit (0 to 1)
• Coefficient of correlation: Measures strength of linear relationship (-1 to 1)
• Standard error of estimate: Measures typical prediction error

Example usage:

LinearRegression regression;
regression.addPoint (Point<double> (1.0, 2.3));
regression.addPoint (Point<double> (2.0, 4.1));
regression.addPoint (Point<double> (3.0, 5.9));
regression.calculate();
 
double a = regression.getA();  // Intercept
double b = regression.getB();  // Slope
double predicted = regression.estimateY (4.0);  // Predict y for x=4

See also

getA, getB, estimateY, getCoefDeterm

Constructor & Destructor Documentation

LinearRegression() [1/2]

LinearRegression::LinearRegression

(

)

Creates an empty linear regression object.

LinearRegression() [2/2]

LinearRegression::LinearRegression

(

juce::Array< Point< double > >

points

)

Creates a linear regression object and adds the given points.

Parameters

points

Initial set of data points

Member Function Documentation

addPoint()

void LinearRegression::addPoint

(

Point< double >

pnt

)

Adds a data point to the regression.

Parameters

pnt	The point to add

calculate()

void LinearRegression::calculate

(

)

Calculates the regression coefficients from the added points.

Must be called after adding points and before retrieving results.

clear()

void LinearRegression::clear

(

)

Removes all data points and resets the regression.

haveData()

bool LinearRegression::haveData

(

)

const

Checks if enough data points have been added.

Returns

true if more than 2 points have been added

References n.

items()

int LinearRegression::items

(

)

const

Returns the number of data points.

Returns

The number of points added

References n.

getA()

double LinearRegression::getA

(

)

const

Returns the intercept of the fitted line.

Returns

The 'a' coefficient in y = a + bx

References a.

getB()

double LinearRegression::getB

(

)

const

Returns the slope of the fitted line.

Returns

The 'b' coefficient in y = a + bx

References b.

getCoefDeterm()

double LinearRegression::getCoefDeterm

(

)

const

Returns the coefficient of determination (R²).

Returns

R² value (0 to 1), where 1 indicates perfect fit

References coefD.

getCoefCorrel()

double LinearRegression::getCoefCorrel

(

)

const

Returns the coefficient of correlation.

Returns

Correlation coefficient (-1 to 1)

References coefC.

getStdErrorEst()

double LinearRegression::getStdErrorEst

(

)

const

Returns the standard error of estimate.

Returns

The typical prediction error

References stdError.

estimateY()

double LinearRegression::estimateY

(

double

x

)

const

Estimates the y value for a given x using the fitted line.

Parameters

x	The x coordinate

Returns

The estimated y value

References a, and b.

Member Data Documentation

n

int LinearRegression::n = 0

protected

Referenced by haveData(), and items().

sumX

double LinearRegression::sumX = 0

protected

sumY

double LinearRegression::sumY = 0

protected

sumXsquared

double LinearRegression::sumXsquared = 0

protected

sumYsquared

double LinearRegression::sumYsquared = 0

protected

sumXY

double LinearRegression::sumXY = 0

protected

a

double LinearRegression::a = 0

protected

Referenced by estimateY(), and getA().

b

double LinearRegression::b = 0

protected

Referenced by estimateY(), and getB().

coefD

double LinearRegression::coefD = 0

protected

Referenced by getCoefDeterm().

coefC

double LinearRegression::coefC = 0

protected

Referenced by getCoefCorrel().

stdError

double LinearRegression::stdError = 0

protected

Referenced by getStdErrorEst().

---

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

• gin_linearregression.h