gov.nist.microanalysis.Utility
Class LinearRegression

java.lang.Object
  gov.nist.microanalysis.Utility.LinearRegression

public class LinearRegression
extends java.lang.Object

Implements a class for computing the linear regression (the best fit line in a least-squares sense.)

Constructor Summary

LinearRegression()
Constructs an object implementing the linear regression for best fitting a line to data points (in a least-squares sense.)

Method Summary

void	addData(double[] x, double[] y) addData - Add two equal lengthed arrays of data points to the current accumulation.
void	addDatum(double x, double y) addDatum - Add a single data point
void	clear() clear - Clear the current accumulation and start a new fit.
double	computeX(double y) computeX - Computes the x such that y(x) = x*getSlope() + getIntercept();
double	computeY(double x) computeY - Computes Y(x) = x*getSlope() + getIntercept();
double	getIntercept() getIntercept - Returns the current best estimate of the y-intercept.
double	getR() getR - A measure of the goodness of fit to a linear model
double	getRSquared() getR - The standard measure of the goodness of fit to a linear model.
double	getSlope() getSlope - Returns the current best estimate of the slope.
void	removeDatum(double x, double y) removeDatum - Removes a single data point from the set of data to be fit.
void	setData(double[] x, double[] y) setData - Clear the current accumulation and add two equal lengthed arrays of data points.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LinearRegression

public LinearRegression()

Constructs an object implementing the linear regression for best fitting a line to data points (in a least-squares sense.)

Method Detail

clear

public void clear()

clear - Clear the current accumulation and start a new fit.

setData

public void setData(double[] x,
                    double[] y)

setData - Clear the current accumulation and add two equal lengthed arrays of data points.

Parameters:

x -

y -

addData

public void addData(double[] x,
                    double[] y)

addData - Add two equal lengthed arrays of data points to the current accumulation.

Parameters:

x -

y -

addDatum

public void addDatum(double x,
                     double y)

addDatum - Add a single data point

Parameters:

x -

y -

removeDatum

public void removeDatum(double x,
                        double y)

removeDatum - Removes a single data point from the set of data to be fit. The values x & y should have been previously added using addDatum(x,y) or as part of an setData() or addData() command otherwise the meaning of the operations performed by removeDatum(...) is not defined as expected. removeDatum(...) does not check to see that the pair x,y were previously added.

Parameters:

x -

y -

getSlope

public double getSlope()

getSlope - Returns the current best estimate of the slope. y(x) = x*getSlope() + getIntercept();

Returns:

double

getIntercept

public double getIntercept()

getIntercept - Returns the current best estimate of the y-intercept. y(x) = x*getSlope() + getIntercept();

Returns:

double

computeY

public double computeY(double x)

computeY - Computes Y(x) = x*getSlope() + getIntercept();

Parameters:

x -

Returns:

double

computeX

public double computeX(double y)

computeX - Computes the x such that y(x) = x*getSlope() + getIntercept();

Parameters:

y -

Returns:

double

getR

public double getR()

getR - A measure of the goodness of fit to a linear model

Returns:

double

getRSquared

public double getRSquared()

getR - The standard measure of the goodness of fit to a linear model. 0 -> does not fit. 1 -> fits a linear model perfectly.

Returns:

double