uk.ac.starlink.ttools.plot
Class XYStats

java.lang.Object
  uk.ac.starlink.ttools.plot.XYStats

public class XYStats
extends Object

Calculates X-Y correlation statistics.

Since:

15 Jul 2004

Author:

Mark Taylor (Starlink)

Constructor Summary

XYStats(boolean xLog, boolean yLog)
Constructs a new correlation statistics calculator.

Method Summary

void	addPoint(double x, double y) Submits a data point for calculations.
double	getCorrelation() Returns the product moment correlation coefficient.
double[]	getLinearCoefficients() Returns the polynomial coefficients of a linear regression line for the submitted data.
double[]	linearRegressionLine() Calculates the linear regression line for the submitted points.

Methods inherited from class java.lang.Object

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

Constructor Detail

XYStats

public XYStats(boolean xLog,
               boolean yLog)

Constructs a new correlation statistics calculator.

Parameters:

xLog - whether the X coordinates are being plotted on a logarithmic scale

yLog - whether the Y coordinates are being plotted on a logarithmic scale

Method Detail

addPoint

public void addPoint(double x,
                     double y)

Submits a data point for calculations.

Parameters:

x - X coordinate

y - Y coordinate

getLinearCoefficients

public double[] getLinearCoefficients()

Returns the polynomial coefficients of a linear regression line for the submitted data.

Returns:

2-element array: (intercept, gradient)

getCorrelation

public double getCorrelation()

Returns the product moment correlation coefficient.

Returns:

correlation coefficient

linearRegressionLine

public double[] linearRegressionLine()

Calculates the linear regression line for the submitted points. The endpoints of the line correspond to the lowest and highest X values submitted.

Returns:

4-element array giving the endpoints (low-X, low-Y, high-X, high-Y) of a linear regression line for these data