umcg.genetica.math.stats

Class ZScores

java.lang.Object

umcg.genetica.math.stats.ZScores

public class ZScores
extends Object

Author:

juha

Constructor Summary

Constructors

Constructor and Description
ZScores()

Method Summary

Modifier and Type	Method and Description
static double	betaToZ(double b, double se, double n)
static double	correlationToP(double correlation, int nrSamples, cern.jet.random.tdouble.StudentT tDist) Returns the two-sided p-value for a given correlation coefficient from Student's t distribution.
static double	correlationToZ(double correlation, int nrSamples, cern.jet.random.tdouble.StudentT tDist) Returns the Z-score for a given correlation coefficient from Student's t distribution.
static double	extrapolateZScore(int originalSampleSize, int newSampleSize, double originalZ)
static double	getWeightedZ(double[] zScores, int[] sampleSizes) Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.
static double	getWeightedZ(double[] zScores, int[] sampleSizes, double[] weights) Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.
static double	getWeightedZ(float[] zScores, int[] sampleSizes) Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.
static double	pToZ(double p) Returns the absolute Z-score for a given p-value using a normal distribution.
static double	zScoreToCorrelation(double obsZ, int nrSamples)
static double	zToP(double z) Returns the p-value corresponding to the given Z-score assuming a two tailed-test.

Methods inherited from class java.lang.Object

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

Constructor Detail

ZScores

public ZScores()

Method Detail

getWeightedZ

public static double getWeightedZ(float[] zScores,
                                  int[] sampleSizes)

Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.

Parameters:

zScores - Z-scores from individual tests

sampleSizes - sample sizes of these tests

Returns:

getWeightedZ

public static double getWeightedZ(double[] zScores,
                                  int[] sampleSizes)

Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.

Parameters:

zScores - Z-scores from individual tests

sampleSizes - sample sizes of these tests

Returns:

getWeightedZ

public static double getWeightedZ(double[] zScores,
                                  int[] sampleSizes,
                                  double[] weights)

Calculates a weighted Z-score according to Whitlock's paper: http://www.ncbi.nlm.nih.gov/pubmed/16135132 Square root of the sample size is used as the weight for each test.

Parameters:

zScores - Z-scores from individual tests

sampleSizes - sample sizes of these tests

Returns:

zToP

public static double zToP(double z)

Returns the p-value corresponding to the given Z-score assuming a two tailed-test.

Parameters:

z - A Z-score to convert

Returns:

The p-value corresponding to the Z-score

pToZ

public static double pToZ(double p)

Returns the absolute Z-score for a given p-value using a normal distribution.

Parameters:

p - p-value

Returns:

absolute Z-score

correlationToZ

public static double correlationToZ(double correlation,
                                    int nrSamples,
                                    cern.jet.random.tdouble.StudentT tDist)

Returns the Z-score for a given correlation coefficient from Student's t distribution.

Parameters:

correlation - correlation coefficient (r)

nrSamples - number of samples used

Returns:

Z-score

correlationToP

public static double correlationToP(double correlation,
                                    int nrSamples,
                                    cern.jet.random.tdouble.StudentT tDist)

Returns the two-sided p-value for a given correlation coefficient from Student's t distribution.

Parameters:

correlation - correlation coefficient (r)

nrSamples - number of samples used

Returns:

Z-score

zScoreToCorrelation

public static double zScoreToCorrelation(double obsZ,
                                         int nrSamples)

extrapolateZScore

public static double extrapolateZScore(int originalSampleSize,
                                       int newSampleSize,
                                       double originalZ)

betaToZ

public static double betaToZ(double b,
                             double se,
                             double n)