umcg.genetica.math.stats

Class QuantileNormalization

java.lang.Object

umcg.genetica.math.stats.QuantileNormalization

public class QuantileNormalization
extends Object

Author:

Harm Jan & Marc Jan Bonder

Constructor Summary

Constructors

Constructor and Description
QuantileNormalization()

Method Summary

Modifier and Type	Method and Description
static void	QuantileNormAdressingNaValuesAfterInitialQN(DoubleMatrixDataset<String,String> dataset, boolean retainNA, boolean useRow, boolean keepZero) Quantile normalize data where missing values are allowed Do QN on all non-missing values, after which the missing values are replaced and a second QN is performed.
static void	QuantileNormAdressingNaValuesBeforeQN(double[][] rawData, boolean useMedian, boolean retainNA) Quantile normalize data where missing values are allowed.
static void	quantilenormalize(double[][] rawData) Quantile normalize a double[][] double[probes][sample]

Methods inherited from class java.lang.Object

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

Constructor Detail

QuantileNormalization

public QuantileNormalization()

Method Detail

quantilenormalize

public static void quantilenormalize(double[][] rawData)

Quantile normalize a double[][] double[probes][sample]

Parameters:

rawData - matrix containing expression/methylation data

QuantileNormAdressingNaValuesBeforeQN

public static void QuantileNormAdressingNaValuesBeforeQN(double[][] rawData,
                                                         boolean useMedian,
                                                         boolean retainNA)

Quantile normalize data where missing values are allowed. the missing values are replaced with the median or mean of the other probes in the sample If chosen keep median value instead of the NA value. NB: For now NA is defined as "-999"!!

Parameters:

rawData - matrix containing expression/methylation data

useMean - use mean for guessing the NA value if false use median

retainNA - retain the NA values, put NA values back after normalization

QuantileNormAdressingNaValuesAfterInitialQN

public static void QuantileNormAdressingNaValuesAfterInitialQN(DoubleMatrixDataset<String,String> dataset,
                                                               boolean retainNA,
                                                               boolean useRow,
                                                               boolean keepZero)

Quantile normalize data where missing values are allowed Do QN on all non-missing values, after which the missing values are replaced and a second QN is performed.

Parameters:

dataset -

retainNA - retain the NA values, put NA values back after normalization

useRow - use row to guess the median expression value, instead of column