papaya
Class Descriptive.Pooled

java.lang.Object
  papaya.Descriptive.Pooled

Enclosing class:

Descriptive

public static class Descriptive.Pooled
extends Object

Class for computing the pooled mean and variance of data sequences

Method Summary

static float	mean(float[] means, int[] sizes) Returns the pooled mean of a k data sequences.
static float	mean(float mean1, int size1, float mean2, int size2) Returns the pooled mean of two data sequences.
static float	var(float[] variances, int[] sizes, boolean unbiased) Returns the pooled variance of k data sequences.
static float	var(float variance1, int size1, float variance2, int size2, boolean unbiased) Returns the pooled variance of two data sequences.

Methods inherited from class java.lang.Object

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

Method Detail

mean

public static float mean(float mean1,
                         int size1,
                         float mean2,
                         int size2)

Returns the pooled mean of two data sequences. That is (size1 * mean1 + size2 * mean2) / (size1 + size2).

Parameters:

size1 - the number of elements in data sequence 1.

mean1 - the mean of data sequence 1.

size2 - the number of elements in data sequence 2.

mean2 - the mean of data sequence 2.

mean

public static float mean(float[] means,
                         int[] sizes)

Returns the pooled mean of a k data sequences. That is (size1 * mean1 + size2 * mean1 + ... + sizek * meank) / (size1 + size2 + ... + sizek).

var

public static float var(float variance1,
                        int size1,
                        float variance2,
                        int size2,
                        boolean unbiased)

Returns the pooled variance of two data sequences. The unbiased estimate of the pooled variance is computed as

 unbiasedPooledVar = Sum( (size_i -1)*variance_i ) / Sum( (size_i-1) ),
 

while the biased pooled variance is computed using

 biasedPooledVar = Sum( (size_i)*variance_i ) / Sum(size_i),
 
 For more than 2 datasets, use var(float[],int[],boolean).


Parameters:
size1 - the number of elements in data sequence 1.
variance1 - the variance of data sequence 1.
size2 - the number of elements in data sequence 2.
variance2 - the variance of data sequence 2.
unbiased - set to true to return the unbiased pooled variance,
 false to return the biased version.

var

public static float var(float[] variances,
                        int[] sizes,
                        boolean unbiased)

Returns the pooled variance of k data sequences. The unbiased estimate of the pooled variance is computed as

 unbiasedPooledVar = Sum( (sizes[i] -1)*variances[i] ) / Sum( (sizes[i]-1) ),
 

while the biased pooled variance is computed using

 biasedPooledVar = Sum( (sizes[i])*variances[i] ) / Sum(sizes[i]),