Package ubic.gemma.core.analysis.preprocess

Class OutlierDetectionServiceImpl

java.lang.Object

ubic.gemma.core.analysis.preprocess.OutlierDetectionServiceImpl

All Implemented Interfaces:

OutlierDetectionService

@Component
public class OutlierDetectionServiceImpl
extends Object
implements OutlierDetectionService

Methods to (attempt to) detect outliers in data sets.

Author:

paul

Constructor Summary

Constructors

Constructor	Description
OutlierDetectionServiceImpl()

Method Summary

Modifier and Type	Method	Description
Collection<OutlierDetails>	identifyOutliersByMedianCorrelation(DoubleMatrix<BioAssay,BioAssay> cormat)	Identify outliers by sorting by median, then looking for non-overlap of first quartile-second quartile range This is exposed for efficiency in geeq score calculation, use this#identifyOutliers(ExpressionExperiment, boolean, boolean) to have the correlation matrix computed correctly for you.
Collection<OutlierDetails>	identifyOutliersByMedianCorrelation(ExpressionExperiment ee)	Uses the identifyOutliersByMedianCorrelation(DoubleMatrix) method to identify outliers in the given experiment.

Methods inherited from class java.lang.Object

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

Constructor Detail

OutlierDetectionServiceImpl

public OutlierDetectionServiceImpl()

Method Detail

identifyOutliersByMedianCorrelation

public Collection<OutlierDetails> identifyOutliersByMedianCorrelation(ExpressionExperiment ee)
                                                               throws FilteringException

Description copied from interface: OutlierDetectionService

Uses the identifyOutliersByMedianCorrelation(DoubleMatrix) method to identify outliers in the given experiment.

Specified by:

identifyOutliersByMedianCorrelation in interface OutlierDetectionService

Parameters:

ee - The experiment to identify sample outliers in.

Returns:

the information about the identified outliers.

Throws:

FilteringException

identifyOutliersByMedianCorrelation

public Collection<OutlierDetails> identifyOutliersByMedianCorrelation(DoubleMatrix<BioAssay,BioAssay> cormat)

Description copied from interface: OutlierDetectionService

Identify outliers by sorting by median, then looking for non-overlap of first quartile-second quartile range This is exposed for efficiency in geeq score calculation, use this#identifyOutliers(ExpressionExperiment, boolean, boolean) to have the correlation matrix computed correctly for you.

Specified by:

identifyOutliersByMedianCorrelation in interface OutlierDetectionService

Parameters:

cormat - the correlation matrix to identify outliers in.

Returns:

the information about the identified outliers.