package jsat.math;
import jsat.linear.Vec;
/**
*
* @author Edward Raff
*/
public class SimpleLinearRegression
{
/**
* Performs a Simple Linear Regression on the data set, calculating the best fit a and b such that y = a + b * x
*
* @param yData the Y data set (to be predicted)
* @param xData the X data set (the predictor)
* @return an array containing the a and b, such that index 0 contains a and index 1 contains b
*/
static public double[] regres(Vec xData, Vec yData)
{
//find y = a + B *x
double[] toReturn = new double[2];
//B value
toReturn[1] = DescriptiveStatistics.sampleCorCoeff(xData, yData)*yData.standardDeviation()/xData.standardDeviation();
//a value
toReturn[0] = yData.mean() - toReturn[1]*xData.mean();
return toReturn;
}
}
Ads help maintain this website.