Package rocca

Class RoccaRFModel

java.lang.Object

rocca.RoccaRFModel

All Implemented Interfaces:

Serializable

public class RoccaRFModel
extends Object
implements Serializable

Class to hold information about the random forest classifier. This class was created to facilitate multiple stages of classification. The AbstractClassifier roccaClassifierModel contains the random forest model itself. The Instances trainedDataset contains the list of species as well as the list of attributes. The RoccaRFModel modelList array is the same length as the species list. If a particular species is a 'trigger' for another stage of classification (eg. 'Small dolphin' which triggers a second classifer based on Common or Striped dolphin) the next-stage RoccaRFModel is stored here. If the species is not a trigger, the array position should be loaded with a null.

Author:

Mike

See Also:

• Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
static final String	AMBIG

Constructor Summary

Constructors

Constructor	Description
RoccaRFModel(weka.classifiers.AbstractClassifier roccaClassifierModel, weka.core.Instances trainedDataset, RoccaRFModel[] modelList)	Constructor

Method Summary

Modifier and Type	Method	Description
void	classifyContour(RoccaClassifier rc, RoccaContourDataBlock rcdb)
String	classifySighting(weka.core.DenseInstance sightingParams, int sightingThreshold)	Classifies a sighting based on the Instance passed
void	generateSpList(ArrayList<String> spList)	Generates a list of species in this model.
RoccaRFModel[]	getModelList()
weka.classifiers.AbstractClassifier	getRoccaClassifierModel()
weka.core.Instances	getTrainedDataset()
void	setModelList(RoccaRFModel[] modelList)
void	setRoccaClassifierModel(weka.classifiers.AbstractClassifier roccaClassifierModel)
void	setTrainedDataset(weka.core.Instances trainedDataset)

Methods inherited from class java.lang.Object

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

Field Details

AMBIG

public static final String AMBIG

See Also:

• Constant Field Values

Constructor Details

RoccaRFModel

public RoccaRFModel(weka.classifiers.AbstractClassifier roccaClassifierModel,
 weka.core.Instances trainedDataset,
 RoccaRFModel[] modelList)

Constructor

Parameters:

roccaClassifierModel -

trainedDataset -

modelList -

Method Details

classifyContour

public void classifyContour(RoccaClassifier rc,
 RoccaContourDataBlock rcdb)

classifySighting

public String classifySighting(weka.core.DenseInstance sightingParams,
 int sightingThreshold)

Classifies a sighting based on the Instance passed

Parameters:

sightingParams - DenseInstance containing the parameters

Returns:

The species with the highest tree votes

generateSpList

public void generateSpList(ArrayList<String> spList)

Generates a list of species in this model. Also, if it finds a next-stage model it will call the same method in that one in order to include those species as well.

Parameters:

spList - original species list

getRoccaClassifierModel

public weka.classifiers.AbstractClassifier getRoccaClassifierModel()

Returns:

the roccaClassifierModel

setRoccaClassifierModel

public void setRoccaClassifierModel(weka.classifiers.AbstractClassifier roccaClassifierModel)

Parameters:

roccaClassifierModel - the roccaClassifierModel to set

getTrainedDataset

public weka.core.Instances getTrainedDataset()

Returns:

the trainedDataset

setTrainedDataset

public void setTrainedDataset(weka.core.Instances trainedDataset)

Parameters:

trainedDataset - the trainedDataset to set

getModelList

public RoccaRFModel[] getModelList()

Returns:

the modelList

setModelList

public void setModelList(RoccaRFModel[] modelList)

Parameters:

modelList - the modelList to set