marytts.util.math

Class Regression

java.lang.Object

marytts.util.math.Regression

public class Regression
extends Object

Multiple linear regression For the case of k independent variables x_1, x_2, ... x_k the sample regression equation is given by y_i = b_0 + b_1*x_1i + b_2*x_2i + ... + b_k*x_ki + e_i y = X*b Ref: Walpole, Myers, Myers and Ye, "Probability and statistics for engineers and scientists", Prentice Hall, chapter 12, pg. 400, 2002.

Author:

marcela

Constructor Summary

Constructors

Constructor and Description
Regression()

Method Summary

Methods

Modifier and Type	Method and Description
double[]	getCoeffs()
double	getCorrelation()
double[]	getPredictedValues()
double[]	getResiduals()
double	getRMSE()
static void	main(String[] args)
double[]	multipleLinearRegression(double[] datay, double[][] datax, boolean interceptTerm)
double[]	multipleLinearRegression(double[] data, int rows, int cols, boolean interceptTerm)
void	multipleLinearRegression(Jama.Matrix datay, Jama.Matrix dataX, boolean interceptTerm)
void	multipleLinearRegression(String fileName, boolean interceptTerm)
void	multipleLinearRegression(String fileName, int indVariable, int[] c, String[] factors, boolean interceptTerm, int rowIni, int rowEnd) multipleLinearRegression providing index numbers for the columns in fileName, index 0 correspond to column 1
double[]	multipleLinearRegression(Vector<Double> data, int rows, int cols, boolean interceptTerm)
double[]	multipleLinearRegression(Vector<Double> vectory, Vector<Double> vectorx, int rows, int cols, boolean interceptTerm)
void	predictValues(String fileName, int indVariable, int[] c, boolean interceptTerm, int rowIni, int rowEnd)
void	printCoefficients()
void	printCoefficients(int[] indices, String[] factors)
void	setCoeffs(double[] coeffsVal)

Methods inherited from class java.lang.Object

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

Constructor Detail

Regression

public Regression()

Method Detail

getCoeffs

public double[] getCoeffs()

setCoeffs

public void setCoeffs(double[] coeffsVal)

getResiduals

public double[] getResiduals()

getPredictedValues

public double[] getPredictedValues()

getCorrelation

public double getCorrelation()

getRMSE

public double getRMSE()

multipleLinearRegression

public double[] multipleLinearRegression(double[] data,
                                int rows,
                                int cols,
                                boolean interceptTerm)

Parameters:

data - dependent and independent variables data={{y1, x11, x12, ... x1k}, {y2, x21, x22, ... x2k}, ... {yn, xn1, xn2, ... xnk}}

interceptTerm - number of the column that will be used as dependent variable → vector y by default the first column is y

rows - number of rows

cols - number of cols including the dependent variable

Returns:

coefficients or null if problems found

multipleLinearRegression

public double[] multipleLinearRegression(double[] datay,
                                double[][] datax,
                                boolean interceptTerm)

multipleLinearRegression

public double[] multipleLinearRegression(Vector<Double> vectory,
                                Vector<Double> vectorx,
                                int rows,
                                int cols,
                                boolean interceptTerm)

multipleLinearRegression

public double[] multipleLinearRegression(Vector<Double> data,
                                int rows,
                                int cols,
                                boolean interceptTerm)

Parameters:

data - Vector contains dependent variable first and then independent variables

rows - rows

cols - cols

interceptTerm - intercept term

Returns:

coeffs

multipleLinearRegression

public void multipleLinearRegression(Jama.Matrix datay,
                            Jama.Matrix dataX,
                            boolean interceptTerm)

printCoefficients

public void printCoefficients(int[] indices,
                     String[] factors)

printCoefficients

public void printCoefficients()

multipleLinearRegression

public void multipleLinearRegression(String fileName,
                            boolean interceptTerm)

multipleLinearRegression

public void multipleLinearRegression(String fileName,
                            int indVariable,
                            int[] c,
                            String[] factors,
                            boolean interceptTerm,
                            int rowIni,
                            int rowEnd)

multipleLinearRegression providing index numbers for the columns in fileName, index 0 correspond to column 1

Parameters:

fileName - file name

indVariable - column number (index) of the independent variable

c - int[] column numbers array (indices) of dependent variables

factors - factors

rowIni - and rowEnd should be given from 0 - maxData-1

rowEnd - row end

interceptTerm - intercept term

predictValues

public void predictValues(String fileName,
                 int indVariable,
                 int[] c,
                 boolean interceptTerm,
                 int rowIni,
                 int rowEnd)

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception