Package weka.classifiers.evaluation

Class RegressionAnalysis

java.lang.Object

weka.classifiers.evaluation.RegressionAnalysis

public class RegressionAnalysis
extends Object

Analyzes linear regression model by using the Student's t-test on each coefficient. Also calculates R^2 value and F-test value. More information: http://en.wikipedia.org/wiki/Student's_t-test http://en.wikipedia.org/wiki/Linear_regression http://en.wikipedia.org/wiki/Ordinary_least_squares

Version:

$Revision: $

Author:

Chris Meyer: cmeyer@udel.edu University of Delaware, Newark, DE, USA CISC 612: Design extension implementation

Constructor Summary

Constructors

Constructor	Description
RegressionAnalysis()

Method Summary

Modifier and Type	Method	Description
static double	calculateAdjRSquared(double rsq, int n, int k)	Returns the adjusted R-squared value for a linear regression model.
static double	calculateFStat(double rsq, int n, int k)	Returns the F-statistic for a linear regression model.
static double	calculateRSquared(Instances data, double ssr)	Returns the R-squared value for a linear regression model, where sum of squared residuals is already calculated.
static double	calculateSSR(Instances data, Attribute chosen, double slope, double intercept)	Returns the sum of squared residuals of the simple linear regression model: y = a + bx.
static double[]	calculateStdErrorOfCoef(Instances data, boolean[] selected, double ssr, int n, int k)	Returns an array of the standard errors of the coefficients in a multiple linear regression.
static double[]	calculateStdErrorOfCoef(Instances data, Attribute chosen, double slope, double intercept, int df)	Returns the standard errors of slope and intercept for a simple linear regression model: y = a + bx.
static double[]	calculateTStats(double[] coef, double[] stderror, int k)	Returns an array of the t-statistic of each coefficient in a multiple linear regression model.
String	getRevision()	Returns the revision string.

Methods inherited from class java.lang.Object

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

Constructor Details

RegressionAnalysis

public RegressionAnalysis()

Method Details

calculateSSR

public static double calculateSSR(Instances data,
 Attribute chosen,
 double slope,
 double intercept)
                           throws Exception

Returns the sum of squared residuals of the simple linear regression model: y = a + bx.

Parameters:

data - (the data set)

chosen - (chosen x-attribute)

slope - (slope determined by simple linear regression model)

intercept - (intercept determined by simple linear regression model)

Returns:

sum of squared residuals

Throws:

Exception - if there is a missing class value in data

calculateRSquared

public static double calculateRSquared(Instances data,
 double ssr)
                                throws Exception

Returns the R-squared value for a linear regression model, where sum of squared residuals is already calculated. This works for either a simple or a multiple linear regression model.

Parameters:

data - (the data set)

ssr - (sum of squared residuals)

Returns:

R^2 value

Throws:

Exception - if there is a missing class value in data

calculateAdjRSquared

public static double calculateAdjRSquared(double rsq,
 int n,
 int k)

Returns the adjusted R-squared value for a linear regression model. This works for either a simple or a multiple linear regression model.

Parameters:

rsq - (the model's R-squared value)

n - (the number of instances in the data)

k - (the number of coefficients in the model: k>=2)

Returns:

the adjusted R squared value

calculateFStat

public static double calculateFStat(double rsq,
 int n,
 int k)

Returns the F-statistic for a linear regression model.

Parameters:

rsq - (the model's R-squared value)

n - (the number of instances in the data)

k - (the number of coefficients in the model: k>=2)

Returns:

F-statistic

calculateStdErrorOfCoef

public static double[] calculateStdErrorOfCoef(Instances data,
 Attribute chosen,
 double slope,
 double intercept,
 int df)
                                        throws Exception

Returns the standard errors of slope and intercept for a simple linear regression model: y = a + bx. The first element is the standard error of slope, the second element is standard error of intercept.

Parameters:

data - (the data set)

chosen - (chosen x-attribute)

slope - (slope determined by simple linear regression model)

intercept - (intercept determined by simple linear regression model)

df - (number of instances - 2)

Returns:

array of standard errors of slope and intercept

Throws:

Exception - if there is a missing class value in data

calculateStdErrorOfCoef

public static double[] calculateStdErrorOfCoef(Instances data,
 boolean[] selected,
 double ssr,
 int n,
 int k)
                                        throws Exception

Returns an array of the standard errors of the coefficients in a multiple linear regression. The last element in the array is the standard error of the constant coefficient. The standard error array is used to calculate the t-statistics.

Parameters:

data - (the data set

selected - (flags indicating variables used in the regression)

ssr - (sum of squared residuals)

n - (number of instances)

k - (number of coefficients; includes constant)

Returns:

array of standard errors of coefficients

Throws:

Exception - if there is a missing class value in data

calculateTStats

public static double[] calculateTStats(double[] coef,
 double[] stderror,
 int k)

Returns an array of the t-statistic of each coefficient in a multiple linear regression model.

Parameters:

coef - (array holding the value of each coefficient)

stderror - (array holding each coefficient's standard error)

k - (number of coefficients, includes constant)

Returns:

array of t-statistics of coefficients

getRevision

public String getRevision()

Returns the revision string.

Returns:

the revision