nl.liacs.timeseries.main

Class LinearRegression

java.lang.Object

nl.liacs.timeseries.main.LinearRegression

All Implemented Interfaces:

java.lang.Runnable

public class LinearRegression
extends java.lang.Object
implements java.lang.Runnable

LinearRegression constructs can be used to calculate the slope, intercept, Pearson's r and mean squared error for two equal length input columns (usually from the same Table).

Author:

Marvin Meeng Last updated 11.07.2012

See Also:

Table, PlotWindow.PlotType, PlotWindow

Field Summary

Fields

Modifier and Type	Field and Description
static int	INTERCEPT_INDEX Index of intercept in result of toArray().
static int	MSE_INDEX Index of mean squared error (mse) in result of toArray().
static int	R_INDEX Index of Pearson's r in result of toArray().
static int	SLOPE_INDEX Index of slope in result of toArray().

Constructor Summary

Constructors

Constructor and Description
LinearRegression(double[] theXColumn, double[] theYColumn, boolean calculateMSE) Creates a LinearRegression.

Method Summary

Methods

Modifier and Type	Method and Description
boolean	getCalculateMSE() Returns whether the mean squared error (mse) will be calculated.
double	getIntercept() Return the intercept for the LinearRegression.
double	getMSE() Return the mean squared error (mse) for the LinearRegression.
double	getR() Return the correlation as calculated by Pearson's r for the LinearRegression.
double	getSlope() Return the slope for the LinearRegression.
boolean	isCalculated() Return whether the result for this LinearRegression has been calculated.
void	run()
LinearRegression	setCalculateMSE(boolean theSetting) Sets whether the mean squared error will be calculated.
double[]	toArray() Returns an array containing the slope, intercept, Pearson's r and mean squared error that are calculated for the LinearRegression.
java.lang.String	toString() Returns the String representation of the LinearRegression.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

SLOPE_INDEX

public static final int SLOPE_INDEX

Index of slope in result of toArray().

See Also:

Constant Field Values

INTERCEPT_INDEX

public static final int INTERCEPT_INDEX

Index of intercept in result of toArray().

See Also:

Constant Field Values

R_INDEX

public static final int R_INDEX

Index of Pearson's r in result of toArray().

See Also:

Constant Field Values

MSE_INDEX

public static final int MSE_INDEX

Index of mean squared error (mse) in result of toArray().

See Also:

Constant Field Values

Constructor Detail

LinearRegression

public LinearRegression(double[] theXColumn,
                double[] theYColumn,
                boolean calculateMSE)

Creates a LinearRegression.

Parameters:

theXColumn - The x-column.

theYColumn - The y-column.

calculateMSE - boolean that controls whether the mean squared error will be calculated.

Method Detail

setCalculateMSE

public LinearRegression setCalculateMSE(boolean theSetting)

Sets whether the mean squared error will be calculated.

Note the default is to calculate the mean squared error.

Parameters:

theSetting - The setting.

Returns:

The LinearRegression the method was called on.

getCalculateMSE

public boolean getCalculateMSE()

Returns whether the mean squared error (mse) will be calculated.

Returns:

true iff the mse will be calculated, false otherwise.

isCalculated

public boolean isCalculated()

Return whether the result for this LinearRegression has been calculated.

This means the run() method has been called on this LinearRegression, and the calculation has finished.

Returns:

true iff the calculation has finished, false otherwise.

See Also:

run()

getSlope

public double getSlope()

Return the slope for the LinearRegression.

Note returns NaN if the slope is not calculated yet, and when an error occurred during the calculation.

Returns:

A double that represents the slope.

See Also:

isCalculated()

getIntercept

public double getIntercept()

Return the intercept for the LinearRegression.

Note returns NaN if the intercept is not calculated yet, and when an error occurred during the calculation.

Returns:

A double that represents the intercept .

See Also:

isCalculated()

getR

public double getR()

Return the correlation as calculated by Pearson's r for the LinearRegression.

Note returns -1.0 if the r is not calculated yet, and NaN when an error occurred during the calculation.

Returns:

A double that represents the r.

See Also:

isCalculated()

getMSE

public double getMSE()

Return the mean squared error (mse) for the LinearRegression.

Note returns -1.0 if the mse is not calculated yet, and NaN when an error occurred during the calculation.

Returns:

A double that represents the mse.

See Also:

isCalculated()

toArray

public double[] toArray()

Returns an array containing the slope, intercept, Pearson's r and mean squared error that are calculated for the LinearRegression.

Note that the index of each of the different values is given by the corresponding public static final int *_INDEX member variable of this class.

Returns:

A double[] containing all calculated values.

See Also:

isCalculated()

toString

public java.lang.String toString()

Returns the String representation of the LinearRegression.

This is a formula of the form: y = a*x + b (r = ?, mse = ?).

Overrides:

toString in class java.lang.Object

run

public void run()

Specified by:

run in interface java.lang.Runnable