Package ai.libs.jaicore.basic

Class StatisticsUtil

java.lang.Object

ai.libs.jaicore.basic.StatisticsUtil

public class StatisticsUtil
extends java.lang.Object

Utils for computing some statistics from collections of doubles or arrays.

Method Summary

Modifier and Type	Method	Description
static boolean	mannWhitneyOneSidedSignificance(double[] sampleA, double[] sampleB)	Uses the MannWhitneyU test to carry out a single-sided significance test.
static boolean	mannWhitneyTwoSidedSignificance(double[] sampleA, double[] sampleB)	Uses the MannWhitneyU test to determine whether the distributions of the the two given samples is significantly different (two-sided test).
static boolean	mannWhitneyTwoSidedSignificance(java.util.Collection<java.lang.Double> sampleA, java.util.Collection<java.lang.Double> sampleB)	Uses the MannWhitneyU test to determine whether the distributions of the the two given samples is significantly different (two-sided test).
static double	mannWhitneyTwoSidedSignificanceP(double[] sampleA, double[] sampleB)	Computes the p-value according to the MannWhitneyU test for iid. samples A and B.
static double	max(java.util.Collection<? extends java.lang.Number> values)	Filters the maximum value of the given collection.
static double	mean(java.util.Collection<? extends java.lang.Number> values)	Computes the mean of the given collection.
static double	median(java.util.Collection<? extends java.lang.Number> values)	Computes the median of the given collection.
static double	min(java.util.Collection<? extends java.lang.Number> values)	Filters the minimum value of the given collection.
static double	standardDeviation(java.util.Collection<? extends java.lang.Number> values)	Computes the standard deviation of the given collection.
static double	sum(java.util.Collection<? extends java.lang.Number> values)	Computes the sum of the given collection.
static boolean	twoSampleTTestSignificance(double[] valuesA, double[] valuesB)	Carries out a two sample ttest to determine whether the distributions of the two given samples are significantly different.
static boolean	twoSampleTTestSignificance(double mean1, double variance1, double n1, double mean2, double variance2, double n2)	Carries out a two sample ttest when the sampled values have already been aggregated to mean, variance, and n, to determine whether the distributions of the two given samples are significantly different.
static boolean	twoSampleTTestSignificance(java.util.Collection<java.lang.Double> valuesA, java.util.Collection<java.lang.Double> valuesB)	Carries out a two sample ttest to determine whether the distributions of the two given samples are significantly different.
static double	variance(java.util.Collection<? extends java.lang.Number> values)	Computes the variance of the given collection.
static double	wilcoxonSignedRankSumTestP(double[] sampleA, double[] sampleB)	Computes the p-value according to the Wilcoxon signed rank test for related samples A and B.
static boolean	wilcoxonSignedRankSumTestSingleSided(double[] sampleA, double[] sampleB)	Uses the Wilcoxon Signed Rank test to carry out a single-sided significance test.
static boolean	wilcoxonSignedRankSumTestTwoSided(double[] sampleA, double[] sampleB)	Uses the Wilcoxon Signed Rank test to determine whether the difference of the distributions of the the two given samples is significantly different (two-sided test).

Methods inherited from class java.lang.Object

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

Method Detail

max

public static double max(java.util.Collection<? extends java.lang.Number> values)

Filters the maximum value of the given collection.

Parameters:

values - Values to take the maximum from.

Returns:

The maximum value of the provided collection.

min

public static double min(java.util.Collection<? extends java.lang.Number> values)

Filters the minimum value of the given collection.

Parameters:

values - Values to take the minimum from.

Returns:

The minimum value of the provided collection.

mean

public static double mean(java.util.Collection<? extends java.lang.Number> values)

Computes the mean of the given collection.

Parameters:

values - Values to take the mean of.

Returns:

The mean of the provided values.

median

public static double median(java.util.Collection<? extends java.lang.Number> values)

Computes the median of the given collection.

Parameters:

values - Values to take the mean of.

Returns:

The mean of the provided values.

sum

public static double sum(java.util.Collection<? extends java.lang.Number> values)

Computes the sum of the given collection.

Parameters:

values - Values to take the sum of.

Returns:

The sum of the provided values.

variance

public static double variance(java.util.Collection<? extends java.lang.Number> values)

Computes the variance of the given collection.

Parameters:

values - Values to compute the variance for.

Returns:

The variance of the provided values.

standardDeviation

public static double standardDeviation(java.util.Collection<? extends java.lang.Number> values)

Computes the standard deviation of the given collection.

Parameters:

values - Values to compute the standard deviation for.

Returns:

The standard deviation of the provided values.

wilcoxonSignedRankSumTestP

public static double wilcoxonSignedRankSumTestP(double[] sampleA,
                                                double[] sampleB)

Computes the p-value according to the Wilcoxon signed rank test for related samples A and B.

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

The p-value of the test for the given two samples.

wilcoxonSignedRankSumTestTwoSided

public static boolean wilcoxonSignedRankSumTestTwoSided(double[] sampleA,
                                                        double[] sampleB)

Uses the Wilcoxon Signed Rank test to determine whether the difference of the distributions of the the two given samples is significantly different (two-sided test).

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.05)

wilcoxonSignedRankSumTestSingleSided

public static boolean wilcoxonSignedRankSumTestSingleSided(double[] sampleA,
                                                           double[] sampleB)

Uses the Wilcoxon Signed Rank test to carry out a single-sided significance test.

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.01)

mannWhitneyTwoSidedSignificanceP

public static double mannWhitneyTwoSidedSignificanceP(double[] sampleA,
                                                      double[] sampleB)

Computes the p-value according to the MannWhitneyU test for iid. samples A and B.

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

The p-value of the test for the given two samples.

mannWhitneyTwoSidedSignificance

public static boolean mannWhitneyTwoSidedSignificance(double[] sampleA,
                                                      double[] sampleB)

Uses the MannWhitneyU test to determine whether the distributions of the the two given samples is significantly different (two-sided test).

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.05)

mannWhitneyTwoSidedSignificance

public static boolean mannWhitneyTwoSidedSignificance(java.util.Collection<java.lang.Double> sampleA,
                                                      java.util.Collection<java.lang.Double> sampleB)

Uses the MannWhitneyU test to determine whether the distributions of the the two given samples is significantly different (two-sided test).

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.05)

mannWhitneyOneSidedSignificance

public static boolean mannWhitneyOneSidedSignificance(double[] sampleA,
                                                      double[] sampleB)

Uses the MannWhitneyU test to carry out a single-sided significance test.

Parameters:

sampleA - The first sample.

sampleB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.01)

twoSampleTTestSignificance

public static boolean twoSampleTTestSignificance(java.util.Collection<java.lang.Double> valuesA,
                                                 java.util.Collection<java.lang.Double> valuesB)

Carries out a two sample ttest to determine whether the distributions of the two given samples are significantly different. Requires the distributions to be a normal distribution respectively.

Parameters:

valuesA - The first sample..

valuesB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.05)

twoSampleTTestSignificance

public static boolean twoSampleTTestSignificance(double[] valuesA,
                                                 double[] valuesB)

Carries out a two sample ttest to determine whether the distributions of the two given samples are significantly different. Requires the distributions to be a normal distribution respectively.

Parameters:

valuesA - The first sample..

valuesB - The second sample.

Returns:

True iff the difference is significant (p-value < 0.05)

twoSampleTTestSignificance

public static boolean twoSampleTTestSignificance(double mean1,
                                                 double variance1,
                                                 double n1,
                                                 double mean2,
                                                 double variance2,
                                                 double n2)

Carries out a two sample ttest when the sampled values have already been aggregated to mean, variance, and n, to determine whether the distributions of the two given samples are significantly different. Requires the distributions to be a normal distribution respectively.

Parameters:

mean1 - The mean of the first sample.

variance1 - The variance of the first sample.

n1 - The sample size of the first sample.

mean2 - The mean of the second sample.

variance2 - The variance of the second sample.

n2 - The sample size of the second sample.

Returns:

True iff the difference is significant (p-value < 0.05)