public class CorrelationCoefficient extends AbstractNumericalTwoClassPerformanceMeasure
PerformanceMeasure
using Pearson or Spearman correlation between prediction scores and
weighted class labels. The weighted class labels may be logit-transformed before computing correlaton.Modifier and Type | Class and Description |
---|---|
static class |
CorrelationCoefficient.Method
The type of correlation used.
|
ParameterSet.ParameterList
errorMessage, parameters, parent
Constructor and Description |
---|
CorrelationCoefficient()
Creates a new
CorrelationCoefficient using Spearman correlation and the raw
weighted labels. |
CorrelationCoefficient(CorrelationCoefficient.Method method,
boolean logit)
Creates a new
CorrelationCoefficient using the suppled type of correlation
and, optionally, logit transformation of weighted labels. |
Modifier and Type | Method and Description |
---|---|
NumericalResultSet |
compute(double[] sortedScoresClass0,
double[] weightsClass0,
double[] sortedScoresClass1,
double[] weightsClass1)
This method allows to compute the performance measure of given sorted score ratios.
|
String |
getName()
The method returns the name of the performance measure.
|
compute, compute, compute
getAllowedNumberOfClasses
findSplitIndex, findThreshold, getCollectionOfAllMeasures, getWeight, simpleWeights
clone, fromGalaxy, fromXML, getAllParameterNames, getComment, getComment, getErrorMessage, getIndex, getName, getName, getNumberOfParameters, getParameterAt, getParameterForName, getParent, hasDefaultOrIsSet, initParameterList, initParameterList, isAtomic, isComparable, parametersLoaded, reset, setParent, toGalaxy, toXML
equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
getAllowedNumberOfClasses
public CorrelationCoefficient()
CorrelationCoefficient
using Spearman correlation and the raw
weighted labels.public CorrelationCoefficient(CorrelationCoefficient.Method method, boolean logit)
CorrelationCoefficient
using the suppled type of correlation
and, optionally, logit transformation of weighted labels.method
- the type of correlationlogit
- if weights should be logit transformed.public String getName()
PerformanceMeasure
getName
in interface PerformanceMeasure
getName
in class AbstractPerformanceMeasure
public NumericalResultSet compute(double[] sortedScoresClass0, double[] weightsClass0, double[] sortedScoresClass1, double[] weightsClass1)
NumericalPerformanceMeasure
compute
in interface NumericalPerformanceMeasure
compute
in interface PerformanceMeasure
compute
in class AbstractNumericalTwoClassPerformanceMeasure
sortedScoresClass0
- the sorted score ratios of class 0weightsClass0
- the weights of the sequences of class 0 sorted along with the scores sortedScoresClass0
sortedScoresClass1
- the sorted score ratios of class 1weightsClass1
- the weights of the sequences of class 1 sorted along with the scores sortedScoresClass1
ToolBox.sortAlongWith(double[], double[][])