de.jstacs.classifier
Class MeasureParameters

java.lang.Object
  de.jstacs.parameters.Parameter
      de.jstacs.parameters.CollectionParameter
          de.jstacs.parameters.MultiSelectionCollectionParameter
              de.jstacs.classifier.MeasureParameters

All Implemented Interfaces:

GalaxyConvertible, Rangeable, RangeIterator, Storable, Cloneable

public final class MeasureParameters
extends MultiSelectionCollectionParameter

This class holds the parameters for the evaluate-methods of a classifier.

Author:

Jan Grau, Jens Keilwagen

See Also:

AbstractClassifier.evaluate(MeasureParameters, boolean, de.jstacs.data.Sample...), AbstractClassifier.evaluateAll(MeasureParameters, boolean, de.jstacs.data.Sample...)

Nested Class Summary

static class

MeasureParameters.Measure This enum defines all measures that are currently implemented in Jstacs.

Nested classes/interfaces inherited from class de.jstacs.parameters.CollectionParameter

CollectionParameter.InconsistentCollectionException

Field Summary

Fields inherited from class de.jstacs.parameters.CollectionParameter

errorMessage, parameters, userSelected

Fields inherited from class de.jstacs.parameters.Parameter

neededReference, neededReferenceId, parent

Constructor Summary

MeasureParameters(boolean evaluateAll)
Creates a new empty instance of MeasureParameters.

MeasureParameters(boolean evaluateAll, double sp, double snForFPR, double snForPPV)
Creates a new instance of MeasureParameters.

MeasureParameters(boolean evaluateAll, double sp, double snForFPR, double snForPPV, double minSpec, double maxSpec, int steps, RangeParameter.Scale scale)
Creates a new instance of MeasureParameters.

MeasureParameters(StringBuffer representation)
The standard constructor for the interface Storable.

Method Summary

void	deselectAll() Deselects all measures, i.e. calls setSelected(Measure, boolean) for all MeasureParameters.Measures.
LinkedList<Result>	getAnnotation() Returns the selected parameters and their values as a LinkedList of Results.
boolean	isSelected(MeasureParameters.Measure sel) Indicates if the option sel is selected.
void	setSelected(MeasureParameters.Measure sel, boolean b) Selects or deselects an option sel depending on the new selection b.

Methods inherited from class de.jstacs.parameters.MultiSelectionCollectionParameter

checkValue, clone, fromXML, getNumberOfNexts, getNumberOfValues, getSelected, getValue, getValueFor, getValueFor, getValues, hasDefaultOrIsSet, isAtomic, isRanged, isSelected, isSelected, isSet, next, reset, resetToFirst, setDefault, setSelected, setSelected, setValue, simplify, toXML, valuesToString

Methods inherited from class de.jstacs.parameters.CollectionParameter

appendCollection, createParameterSet, equals, extractCollection, fromGalaxy, getComment, getDatatype, getDefault, getErrorMessage, getName, getParametersInCollection, getRangedInstance, hasDefault, isRangeable, isRequired, isUserSelected, setRangeable, toGalaxy, toString

Methods inherited from class de.jstacs.parameters.Parameter

getId, getNeededReference, getNeededReferenceId, getParent, setNeededReference, setParent

Methods inherited from class java.lang.Object

finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

MeasureParameters

public MeasureParameters(boolean evaluateAll)
                  throws ParameterException

Creates a new empty instance of MeasureParameters.

Parameters:

evaluateAll - indicates if the instance contains the parameters for the method AbstractClassifier.evaluateAll(MeasureParameters, boolean, de.jstacs.data.Sample...) or only the parameters for the method AbstractClassifier.evaluate(MeasureParameters, boolean, de.jstacs.data.Sample...)

Throws:

ParameterException - if something went wrong while constructing the parameters

See Also:

MeasureParameters.Measure, MultiSelectionCollectionParameter.MultiSelectionCollectionParameter(DataType, String[], Object[], String[], String, String, boolean)

MeasureParameters

public MeasureParameters(boolean evaluateAll,
                         double sp,
                         double snForFPR,
                         double snForPPV)
                  throws ParameterException

Creates a new instance of MeasureParameters. All measures are switched on and the parameters of specificity and sensitivity are set to the given values.

Parameters:

evaluateAll - indicates if the instance contains the parameters for the method AbstractClassifier.evaluateAll(MeasureParameters, boolean, de.jstacs.data.Sample...) or only the parameters for the method AbstractClassifier.evaluate(MeasureParameters, boolean, de.jstacs.data.Sample...)

sp - the (fixed) specificity

snForFPR - the (fixed) sensitivity for the FPR

snForPPV - the (fixed) sensitivity for the PPV

Throws:

ParameterException - if something went wrong while constructing the parameters

See Also:

MeasureParameters.Measure, MultiSelectionCollectionParameter.MultiSelectionCollectionParameter(DataType, String[], Object[], String[], String, String, boolean)

MeasureParameters

public MeasureParameters(boolean evaluateAll,
                         double sp,
                         double snForFPR,
                         double snForPPV,
                         double minSpec,
                         double maxSpec,
                         int steps,
                         RangeParameter.Scale scale)
                  throws ParameterException

Creates a new instance of MeasureParameters. All measures are switched on and the parameters of specificity and sensitivity are set to the given values.

Parameters:

evaluateAll - indicates if the instance contains the parameters for the method AbstractClassifier.evaluateAll(MeasureParameters, boolean, de.jstacs.data.Sample...) or only the parameters for the method AbstractClassifier.evaluate(MeasureParameters, boolean, de.jstacs.data.Sample...)

sp - the (fixed) specificity

snForFPR - the (fixed) sensitivity for the FPR

snForPPV - the (fixed) sensitivity for the PPV

minSpec - only used for MeasureParameters.Measure.PartialROCCurve: the minimal specificity

maxSpec - only used for MeasureParameters.Measure.PartialROCCurve: the maximal specificity

steps - only used for MeasureParameters.Measure.PartialROCCurve: the number of steps

scale - only used for MeasureParameters.Measure.PartialROCCurve: see RangeParameter

Throws:

ParameterException - if something went wrong while constructing the parameters

See Also:

RangeParameter.Scale, MeasureParameters.Measure, RangeParameter, MultiSelectionCollectionParameter.MultiSelectionCollectionParameter(DataType, String[], Object[], String[], String, String, boolean)

MeasureParameters

public MeasureParameters(StringBuffer representation)
                  throws NonParsableException

The standard constructor for the interface Storable. Creates a new instance of MeasureParameters out of its XML representation.

Parameters:

representation - the XML representation as StringBuffer

Throws:

NonParsableException - if the MeasureParameters could not be reconstructed out of the XML representation (the StringBuffer representation could not be parsed)

See Also:

MultiSelectionCollectionParameter.MultiSelectionCollectionParameter(StringBuffer), Storable

Method Detail

setSelected

public void setSelected(MeasureParameters.Measure sel,
                        boolean b)

Selects or deselects an option sel depending on the new selection b.

Parameters:

sel - the option

b - the new selection

See Also:

MultiSelectionCollectionParameter.setSelected(String, boolean)

isSelected

public boolean isSelected(MeasureParameters.Measure sel)

Indicates if the option sel is selected.

Parameters:

sel - the option

Returns:

true if the option sel is selected, false otherwise

See Also:

MultiSelectionCollectionParameter.isSelected(String)

deselectAll

public void deselectAll()

Deselects all measures, i.e. calls setSelected(Measure, boolean) for all MeasureParameters.Measures.

See Also:

setSelected(Measure, boolean)

getAnnotation

public LinkedList<Result> getAnnotation()

Returns the selected parameters and their values as a LinkedList of Results.

Returns:

a LinkedList of all selected parameters and their values