import static java.lang.System.out;

import java.util.*;

import javastat.*;

import javastat.regression.lm.*;

import static javastat.util.Argument.*;

import static javastat.util.Output.*;

import javastat.util.*;

/**

*

* <p>Example: class LinearRegression.</p>

* <p>Data Source: Anderson, D. R., Sweeney, D. J. and Williams, T. A. (2001).

* Contemporary Business Statistics with Microsoft Excel.

* South-Western, p. 468, p.530.

</p>

*/

public class LinearRegressionExample

{

public static void main(String arg[])

{

double[] pizzaResponse = {58, 105, 88, 118, 117, 137, 157, 169, 149,

202};

double[] pizzaCovariate = {2, 6, 8, 8, 12, 16, 20, 20, 22, 26};

double[] hellerResponse = {102, 100, 120, 77, 46, 93, 26, 69, 65, 85};

double[] hellerCovariate1 = {120, 140, 190, 130, 155, 175, 125, 145,

180, 150};

double[] hellerCovariate2 = {100, 110, 90, 150, 210, 150, 250, 270, 300,

250};

DataManager dm = new DataManager();

LinearRegression myClass1 = new LinearRegression(pizzaResponse,

pizzaCovariate);

double[] coefficients = myClass1.coefficients;

double[] fittedValues = myClass1.fittedValues;

double[] residuals = myClass1.residuals;

double[] testStatistic = myClass1.testStatistic;

double[] pValue = myClass1.pValue;

double[][] confidenceInterval = {myClass1.confidenceInterval[0],

myClass1.confidenceInterval[1]};

double rSquare = myClass1.rSquare;

double testFStatistic = myClass1.testFStatistic;

double fPValue = myClass1.fPValue;

LinearRegression myClass2 = new LinearRegression();

coefficients = myClass2.coefficients(hellerResponse, hellerCovariate1,

hellerCovariate2);

confidenceInterval = myClass2.confidenceInterval(0.05, hellerResponse,

hellerCovariate1, hellerCovariate2);

testStatistic = myClass2.testStatistic(hellerResponse, hellerCovariate1,

hellerCovariate2);

pValue = myClass2.pValue(hellerResponse, hellerCovariate1,

hellerCovariate2);

testFStatistic = myClass2.testFStatistic(hellerResponse,

hellerCovariate1,

hellerCovariate2);

fPValue = myClass2.fPValue(hellerResponse, hellerCovariate1,

hellerCovariate2);

confidenceInterval = myClass2.confidenceInterval(hellerResponse,

hellerCovariate1, hellerCovariate2);

Hashtable argument1 = new Hashtable();

argument1.put(ALPHA, 0.05);

StatisticalAnalysis myClass3 = new LinearRegression(argument1,

pizzaResponse, pizzaCovariate).statisticalAnalysis;

confidenceInterval =

(double[][]) myClass3.output.get(CONFIDENCE_INTERVAL);

testStatistic = (double[]) myClass3.output.get(TEST_STATISTIC);

pValue = (double[]) myClass3.output.get(PVALUE);

rSquare = (Double) myClass3.output.get(R_SQUARE);

testFStatistic = (Double) myClass3.output.get(F_STATISTIC);

fPValue = (Double) myClass3.output.get(F_PVALUE);

residuals = (double[]) myClass3.output.get(RESIDUALS);

fittedValues = (double[]) myClass3.output.get(FITTED_VALUES);

Hashtable argument2 = new Hashtable();

LinearRegression myClass4 = new LinearRegression(argument2, null);

coefficients = myClass4.coefficients(

argument2, hellerResponse, hellerCovariate1, hellerCovariate2);

argument2.put(ALPHA, 0.05);

confidenceInterval = myClass4.confidenceInterval(argument2,

hellerResponse, hellerCovariate1, hellerCovariate2);

testStatistic = myClass4.testStatistic(argument2, hellerResponse,

hellerCovariate1,

hellerCovariate2);

pValue = myClass4.pValue(argument2, hellerResponse,

hellerCovariate1, hellerCovariate2);

testFStatistic = myClass4.testFStatistic(argument2, hellerResponse,

hellerCovariate1,

hellerCovariate2);

fPValue = myClass4.fPValue(argument2, hellerResponse,

hellerCovariate1, hellerCovariate2);

}

}

Results:

The estimated coefficients based on non-null constructor = [60.0 , 5.0]

The t statistics based on non-null constructor = [6.503 , 8.617]

The p-values for the t statistics based on non-null constructor = [0.0 , 0.0]

The confidence interval for the intercept based on non-null constructor =

[41.917 , 78.083]

The confidence interval for the slope based on non-null constructor =

[3.863 , 6.137]

The r-square based on non-null constructor = 0.903

The f statistic based on non-null constructor = 74.248

The p-value for the F statistic based on non-null constructor = 0.0

{FITTED_VALUES=[D@e89b94, RESIDUALS=[D@13e205f,

RSS=1530.00, CONFIDENCE_INTERVAL=[[D@1bf73fa,

MSE=191.25, HAT_MATRIX=[[D@5740bb, R_SQUARE=0.9027, SSR=14200,

COEFFICIENT_VARIANCE=[[D@5ac072, F_PVALUE=2.549E-5,

PVALUE=[D@109a4c, TEST_STATISTIC=[D@201f9,

DEGREE_OF_FREEDOM=[D@1cf8583, F_STATISTIC=74.248,

COEFFICIENTS=[D@14693c7, SST=15730.0, SSE=1530.00}

The estimated coefficients based on null constructor = [66.518 , 0.414 , -0.27]

The t statistics based on null constructor = [1.588 , 1.59 , -3.335]

The p-values for the t statistics based on null constructor = [0.156 , 0.156 , 0.013]

The confidence interval for parameter 1 based on null constructor =

[-15.56 , 148.595]

The confidence interval for parameter 2 based on null constructor =

[-0.096 , 0.924]

The confidence interval for parameter 3 based on null constructor =

[-0.428 , -0.111]

The f statistic based on null constructor = 6.579

The p-value for the f statistic based on null constructor = 0.025

The predicted quantity sold = 93.718

{FITTED_VALUES=[D@901887, RESIDUALS=[D@3a6727, RSS=2457.311,

CONFIDENCE_INTERVAL=[[D@4a65e0, MSE=351.044,

HAT_MATRIX=[[D@665753, R_SQUARE=0.653, SSR=4618.789,

COEFFICIENT_VARIANCE=[[D@ef22f8, F_PVALUE=0.0247,

PVALUE=[D@1e0cf70, TEST_STATISTIC=[D@52fe85,

DEGREE_OF_FREEDOM=[D@c40c80, F_STATISTIC=6.579,

COEFFICIENTS=[D@110d81b, SST=7076.100, SSE=2457.311}