magpie.data

Class MultiPropertyEntry

java.lang.Object

magpie.data.BaseEntry

magpie.data.MultiPropertyEntry

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, java.lang.Comparable, java.util.Comparator

Direct Known Subclasses:

CompositionEntry, ElementEntry

public class MultiPropertyEntry
extends BaseEntry

Entry that can have multiple properties. Consider these properties as potential class variables. This class simply allows a user to designate which property to use as the class variable, and seamlessly stores predictions of these properties internally.

Users of entries that extend this class can designate a property as the class variable using setTargetProperty(int). Often, this is performed through various commands from the MultiPropertyDataset. You may notice that names of properties are not stored in this class, and they are actually stored in MultiPropertyDataset. This is to reminder users that they should access properties through the Dataset.

If the "target" property is set to "-1", this class functions as a normal BaseEntry.

See Also:

MultiPropertyDataset, Serialized Form

Constructor Summary

Constructors

Constructor and Description
MultiPropertyEntry()
MultiPropertyEntry(double[] attributes) Create a new entry, given attributes

Method Summary

Modifier and Type	Method and Description
void	addProperty() Add a new property to this entry.
void	addProperty(double measuredValue) Add a new property and define measured property.
void	addProperty(double measuredValue, double predictedValue) Add a new property to this entry, set measured and predicted values
void	clearPropertyData() Remove all information about properties
MultiPropertyEntry	clone() Generates a clone of an entry.
void	deleteMeasuredClass() Delete measured class variable.
void	deletePredictedClass() Delete predicted class variable.
double[]	getClassProbilities() Get the probability of an entry existing in each class.
double	getMeasuredClass() Get the measured class variable
double[]	getMeasuredProperties() Get copy all measured properties.
double	getMeasuredProperty(int index) Get a single measured property.
double	getPredictedClass() Get the predicted class variable
double	getPredictedProperty(int index) Get the predicted properties for an instance.
double[]	getPropertyClassProbabilties(int index) Get the predicted class probabilities for a certain property.
int	getTargetProperty() Look up which property is being used as the class variable.
boolean	hasClassProbabilities() Whether this entry has predicted probabilities
boolean	hasMeasuredProperty(int index) Whether a specific property of this entry has been measured
boolean	hasMeasurement()
boolean	hasPredictedProperty(int index) Whether an specific property has any predicted value
boolean	hasPrediction()
boolean	hasPropertyClassProbabilities(int index) Check whether a property has predicted class probabilities
int	NProperties()
void	setClassProbabilities(double[] probabilites) Set the predicted probability of a entry existing in each class
void	setMeasuredClass(double x) Set the measured class variable
void	setMeasuredProperties(double[] newValues) Define the values of all measured properties of this entry.
void	setMeasuredProperty(int index, double newValue) Set the measured value of a specific property of this entry.
void	setNProperties(int N) Define the number of properties this entry can support.
void	setPredictedClass(double x) Set the predicted class variable
void	setPredictedProperty(int index, double newValue) Directly set the predicted value of a property.
void	setPredictedProperty(int index, double[] newValues) Directly set the probabilities of an entry existing in one of several classes.
void	setTargetProperty(int index) Define which property to use as the class variable.
org.json.JSONObject	toJSON() Print entry as a JSON object
boolean	usingPropertyAsClass()

Methods inherited from class magpie.data.BaseEntry

addAttribute, addAttributes, clearAttributes, compare, compareTo, equals, getAttribute, getAttributes, hashCode, NAttributes, reduceMemoryFootprint, setAttribute, setAttributes, toHTMLString, toString

Methods inherited from class java.lang.Object

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

Methods inherited from interface java.util.Comparator

comparing, comparing, comparingDouble, comparingInt, comparingLong, naturalOrder, nullsFirst, nullsLast, reversed, reverseOrder, thenComparing, thenComparing, thenComparing, thenComparingDouble, thenComparingInt, thenComparingLong

Constructor Detail

MultiPropertyEntry

public MultiPropertyEntry()

MultiPropertyEntry

public MultiPropertyEntry(double[] attributes)

Create a new entry, given attributes

Parameters:

attributes - Attribute values

Method Detail

clone

public MultiPropertyEntry clone()

Description copied from class: BaseEntry

Generates a clone of an entry. It creates a new list to store the attributes. So, the attribute data is preserved, but you can change the list as desired.

Overrides:

clone in class BaseEntry

Returns:

Clone

clearPropertyData

public void clearPropertyData()

Remove all information about properties

setNProperties

public void setNProperties(int N)

Define the number of properties this entry can support.

Parameters:

N - Number of properties this entry should be able to store

addProperty

public void addProperty()

Add a new property to this entry.

See Also:

MultiPropertyDataset.addProperty(java.lang.String, java.lang.String[])

addProperty

public void addProperty(double measuredValue)

Add a new property and define measured property.

Parameters:

measuredValue - Measured value of new property

See Also:

MultiPropertyDataset.addProperty(java.lang.String, java.lang.String[])

addProperty

public void addProperty(double measuredValue,
                        double predictedValue)

Add a new property to this entry, set measured and predicted values

Parameters:

measuredValue - Measured value of this property. Set to Double.NaN if unknown.

predictedValue - Predicted value of this property. Set to Double.NaN if unknown.

See Also:

MultiPropertyDataset.addProperty(java.lang.String, java.lang.String[]);

getMeasuredClass

public double getMeasuredClass()

Description copied from class: BaseEntry

Get the measured class variable

Overrides:

getMeasuredClass in class BaseEntry

Returns:

Measured class, if known. NaN otherwise

setMeasuredClass

public void setMeasuredClass(double x)

Description copied from class: BaseEntry

Set the measured class variable

Overrides:

setMeasuredClass in class BaseEntry

Parameters:

x - Measured class

getMeasuredProperties

public double[] getMeasuredProperties()

Get copy all measured properties.

Returns:

Clone of internal array of measured properties

setMeasuredProperties

public void setMeasuredProperties(double[] newValues)

Define the values of all measured properties of this entry. Simply use Double.NaN for all properties that have not been measured.

This command will automatically reset the predicted properties, so use with caution.

Parameters:

newValues - Values of all measured properties

getMeasuredProperty

public double getMeasuredProperty(int index)

Get a single measured property. If not measured, will return NaN.

Parameters:

index - Index of measured property to retrieve

Returns:

Value of that property

getPredictedClass

public double getPredictedClass()

Description copied from class: BaseEntry

Get the predicted class variable

Overrides:

getPredictedClass in class BaseEntry

Returns:

Predicted class, if known. NaN otherwise

setPredictedClass

public void setPredictedClass(double x)

Description copied from class: BaseEntry

Set the predicted class variable

Overrides:

setPredictedClass in class BaseEntry

Parameters:

x - Predicted class

getClassProbilities

public double[] getClassProbilities()

Description copied from class: BaseEntry

Get the probability of an entry existing in each class. Returns null if no class probabilities have been stored

Overrides:

getClassProbilities in class BaseEntry

Returns:

Class probabilities

getPredictedProperty

public double getPredictedProperty(int index)

Get the predicted properties for an instance. If the property has not been predicted, returns Double.NaN.

Parameters:

index - Index of property to retrieve

Returns:

Predicted property

getPropertyClassProbabilties

public double[] getPropertyClassProbabilties(int index)

Get the predicted class probabilities for a certain property. If the property has not been predicted, returns null.

Parameters:

index - Index of property to retrieve

Returns:

Predicted class probabilities for that property

getTargetProperty

public int getTargetProperty()

Look up which property is being used as the class variable. Returns -1 if the class variable is something else.

Returns:

Target property

setTargetProperty

public void setTargetProperty(int index)

Define which property to use as the class variable. Set to -1 when using some other variable.

Parameters:

index - Index of property to use

usingPropertyAsClass

public boolean usingPropertyAsClass()

Returns:

Whether a property is being used as the class variable.

hasMeasuredProperty

public boolean hasMeasuredProperty(int index)

Whether a specific property of this entry has been measured

Parameters:

index - Index of property to test

Returns:

Whether it has been measured

hasPredictedProperty

public boolean hasPredictedProperty(int index)

Whether an specific property has any predicted value

Parameters:

index - Index of property to test

Returns:

Whether that property has been predicted

hasMeasurement

public boolean hasMeasurement()

Overrides:

hasMeasurement in class BaseEntry

Returns:

Whether a measured class has been set for this entry

hasPrediction

public boolean hasPrediction()

Overrides:

hasPrediction in class BaseEntry

Returns:

Whether a predicted class has been set for this entry

hasClassProbabilities

public boolean hasClassProbabilities()

Description copied from class: BaseEntry

Whether this entry has predicted probabilities

Overrides:

hasClassProbabilities in class BaseEntry

Returns:

Whether it has probabilities

hasPropertyClassProbabilities

public boolean hasPropertyClassProbabilities(int index)

Check whether a property has predicted class probabilities

Parameters:

index - Index of property

Returns:

Whether it has class probabilities

deleteMeasuredClass

public void deleteMeasuredClass()

Description copied from class: BaseEntry

Delete measured class variable.

Overrides:

deleteMeasuredClass in class BaseEntry

deletePredictedClass

public void deletePredictedClass()

Description copied from class: BaseEntry

Delete predicted class variable.

Overrides:

deletePredictedClass in class BaseEntry

setClassProbabilities

public void setClassProbabilities(double[] probabilites)

Description copied from class: BaseEntry

Set the predicted probability of a entry existing in each class

Overrides:

setClassProbabilities in class BaseEntry

Parameters:

probabilites - Probability of entry being in each class

setMeasuredProperty

public void setMeasuredProperty(int index,
                                double newValue)

Set the measured value of a specific property of this entry. Use Double.NaN if it has not been measured (though I am not sure why you would do this).

Parameters:

index - Index of property to set

newValue - New measured value

setPredictedProperty

public void setPredictedProperty(int index,
                                 double newValue)

Directly set the predicted value of a property.

Parameters:

index - Index of property

newValue - Its new value.

setPredictedProperty

public void setPredictedProperty(int index,
                                 double[] newValues)

Directly set the probabilities of an entry existing in one of several classes.

Parameters:

index - Index of property

newValues - Class probabilities. If null, marks this class as unset

NProperties

public int NProperties()

Returns:

Number of properties defined for this entry

toJSON

public org.json.JSONObject toJSON()

Description copied from class: BaseEntry

Print entry as a JSON object

Overrides:

toJSON in class BaseEntry

Returns:

JSON object