Helper class for weighted linear regression. More...

Collaboration diagram for ModifiedSincSmoother::LinearRegression:

Public Member Functions
void	clear ()
	Reset accumulator for new regression.

void	addPointW (double x, double y, double weight)
	Add a weighted data point to the regression.

void	calculate ()
	Compute least-squares line parameters.

double	getOffset ()
	Get regression intercept (y-value at x=0).

double	getSlope ()
	Get regression slope (dy/dx).

Public Attributes
double	sum_weights = 0
	Sum of weights.

double	sum_x = 0
	Sum of weighted x values.

double	sum_y = 0
	Sum of weighted y values

double	sum_xy = 0
	Sum of weighted x*y products.

double	sum_x2 = 0
	Sum of weighted x² values.

double	offset = NAN
	Regression intercept (cached)

double	slope = NAN
	Regression slope (cached)

bool	calculated = false
	Whether regression has been computed.

Detailed Description

Helper class for weighted linear regression.

Accumulates weighted (x,y) points and computes least-squares line fit on demand. Used for boundary extrapolation.

Member Function Documentation

◆ addPointW()

void addPointW	(	double	x,
		double	y,
		double	weight
	)

Add a weighted data point to the regression.

Parameters

x	independent variable value
y	dependent variable value
weight	point weight (≥ 0)

◆ calculate()

void calculate ( )

Compute least-squares line parameters.

Solves normal equations for weighted linear regression. Sets calculated = true and caches offset/slope.

Note: Handles singular cases by setting slope = 0; Automatically called by getOffset()/getSlope() if needed

◆ clear()

void clear ( )

Reset accumulator for new regression.

◆ getOffset()

double getOffset ( )

Get regression intercept (y-value at x=0).

Returns: regression offset parameter

Note: Triggers calculate() if not already computed

◆ getSlope()

double getSlope ( )

Get regression slope (dy/dx).

Returns: regression slope parameter

Note: Triggers calculate() if not already computed

Member Data Documentation

◆ calculated

bool calculated = false

Whether regression has been computed.

◆ offset

double offset = NAN

Regression intercept (cached)

◆ slope

double slope = NAN

Regression slope (cached)

◆ sum_weights

double sum_weights = 0

Sum of weights.

◆ sum_x

double sum_x = 0

Sum of weighted x values.

◆ sum_x2

double sum_x2 = 0

Sum of weighted x² values.

◆ sum_xy

double sum_xy = 0

Sum of weighted x*y products.

◆ sum_y

double sum_y = 0

Sum of weighted y values

Public Member Functions

Public Attributes