Classifier (LBJava 1.2.22 API)

java.lang.Object
- edu.illinois.cs.cogcomp.lbjava.classify.Classifier

All Implemented Interfaces:

Serializable, Cloneable

Direct Known Subclasses:

FeatureVectorReturner, LabelVectorReturner, Learner, ParameterizedConstraint, ValueComparer
```
public abstract class Classifier
extends Object
implements Cloneable, Serializable
```
Objects of this class represent functions that make some multi-valued decision about an object.

Author:

Nick Rizzolo

See Also:

Serialized Form

Field Summary

Fields
Modifier and Type	Field and Description
`String`	`containingPackage` The name of the package containing this classifier.
`String`	`name` The name of the classifier usually becomes the identifier of produced features.

Constructor Summary

Constructors
Modifier Constructor and Description

protected Classifier()
Does nothing.

protected Classifier(String n)
Initializing constructor.

Constructors
Modifier	Constructor and Description
`protected`	`Classifier()` Does nothing.
`protected`	`Classifier(String n)` Initializing constructor.

Method Summary

All Methods Static Methods Instance Methods Abstract Methods Concrete Methods
Modifier and Type	Method and Description
`String[]`	`allowableValues()` Returns the array of allowable values that a feature returned by this classifier may take.
`abstract FeatureVector`	`classify(Object o)` This method makes one or more decisions about a single object, returning those decisions as `Feature`s in a vector.
`FeatureVector[]`	`classify(Object[] o)` Use this method to make a batch of classification decisions about several objects.
`Object`	`clone()` This method returns a shallow clone.
`String`	`discreteValue(Object o)` Returns the value of the discrete feature that would be returned by this classifier.
`String[]`	`discreteValueArray(Object o)` Returns the values of the discrete array of features that would be returned by this classifier.
`Feature`	`featureValue(Object o)` Returns the classification of the given example object as a single feature instead of a `FeatureVector`.
`LinkedList`	`getCompositeChildren()` If this classifier is a composite generator, this method will be overridden such that it returns all the classifiers it calls on in a list.
`String`	`getInputType()` Returns a string describing the input type of this classifier.
`String`	`getOutputType()` Returns a string describing the output feature type of this classifier.
`double`	`realValue(Object o)` Returns the value of the real feature that would be returned by this classifier.
`double[]`	`realValueArray(Object o)` Returns the values of the real array of features that would be returned by this classifier.
`static double`	`test(Classifier subject, Classifier oracle, Object[] o)` Measures the performance of a classifier as compared with the values produced by an oracle.
`String`	`toString()` Simply returns the name of the classifier.
`short`	`valueIndexOf(String value)` Locates the specified discrete feature value in the array of allowable values defined for this classifier.

Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

- Field Detail
  - containingPackage
```
public String containingPackage
```
    The name of the package containing this classifier.
  - name
```
public String name
```
    The name of the classifier usually becomes the identifier of produced features.
- Constructor Detail
  - Classifier
```
protected Classifier()
```
    Does nothing.
  - Classifier
```
protected Classifier(String n)
```
    Initializing constructor.
    
    Parameters:
    
    n - The name of the classifier, which can be fully qualified.
- Method Detail
  - test
```
public static double test(Classifier subject,
                          Classifier oracle,
                          Object[] o)
```
    Measures the performance of a classifier as compared with the values produced by an oracle.
    
    Parameters:
    
    subject - The classifier to test.
    
    oracle - A classifier that produces the correct classifications.
    
    o - The set of objects to test the subject on.
    
    Returns:
    
    The accuracy of the subject classifier.
  - classify
```
public abstract FeatureVector classify(Object o)
```
    This method makes one or more decisions about a single object, returning those decisions as Features in a vector.
    
    Parameters:
    
    o - The object to make decisions about.
    
    Returns:
    
    A vector of Features about the input object.
  - classify
```
public FeatureVector[] classify(Object[] o)
```
    Use this method to make a batch of classification decisions about several objects. This function is implemented in the most naive way (simply calling classify(Object) repeatedly) and should be overridden if there is a more efficient implementation.
    
    Parameters:
    
    o - The objects to make decisions about.
    
    Returns:
    
    An array of feature vectors, one per input object.
  - getInputType
```
public String getInputType()
```
    Returns a string describing the input type of this classifier. The type name must be fully specified (i.e. including its package name). For example, the default return value of this method is:
```
 "java.lang.Object"
 
```
    This method should be overridden by derived classes.
    Returns:
    
    A string representation of the expected input type of this classifier.
  - getOutputType
```
public String getOutputType()
```
    Returns a string describing the output feature type of this classifier. It should either contain the basic type (discrete or real) and square brackets or a percent sign if the classifier returns an array or is a generator respectively, or simply mixed%. In the case that the basic type is discrete, the curly braces containing a list of allowable values should be omitted, as this list is provided by the allowableValues() method. The default return value of this method is:
```
 "discrete"
 
```
    This method should be overridden by derived classes.
    Returns:
    
    A string representation of the output feature type of this classifier.
  - allowableValues
```
public String[] allowableValues()
```
    Returns the array of allowable values that a feature returned by this classifier may take. If the array has length 0, it means either that the feature has discrete type and allowable values were not specified or that the feature has real or mixed type. The default return value of this method is a 0 length array.
    This method should be overridden by derived classes.
    
    Returns:
    
    The allowable values that a feature returned by this classifier may take.
  - valueIndexOf
```
public short valueIndexOf(String value)
```
    Locates the specified discrete feature value in the array of allowable values defined for this classifier.
    
    Parameters:
    
    value - The value to locate.
    
    Returns:
    
    The index of the specified value, or -1 if it wasn't found.
  - featureValue
```
public Feature featureValue(Object o)
```
    Returns the classification of the given example object as a single feature instead of a FeatureVector. By default, this method is implemented to simply throw an UnsupportedOperationException since some classifiers return zero or multiple features at once.
    
    Parameters:
    
    o - The object to classify.
    
    Returns:
    
    The classification of o as a feature.
  - discreteValue
```
public String discreteValue(Object o)
```
    Returns the value of the discrete feature that would be returned by this classifier. This method should only be called when overridden by a classifier returning a single discrete feature.
    
    Parameters:
    
    o - The object to classify.
    
    Returns:
    
    The value of the feature produced for the input object.
  - realValue
```
public double realValue(Object o)
```
    Returns the value of the real feature that would be returned by this classifier. This method should only be called when overridden by a classifier returning a single real feature.
    
    Parameters:
    
    o - The object to classify.
    
    Returns:
    
    The value of the feature produced for the input object.
  - discreteValueArray
```
public String[] discreteValueArray(Object o)
```
    Returns the values of the discrete array of features that would be returned by this classifier. This method should only be called when overridden by a classifier returning an array of discrete features.
    
    Parameters:
    
    o - The object to classify.
    
    Returns:
    
    The values of the array of features produced for the input object.
  - realValueArray
```
public double[] realValueArray(Object o)
```
    Returns the values of the real array of features that would be returned by this classifier. This method should only be called when overridden by a classifier returning an array of real features.
    
    Parameters:
    
    o - The object to classify.
    
    Returns:
    
    The value of the array of features produced for the input object.
  - getCompositeChildren
```
public LinkedList getCompositeChildren()
```
    If this classifier is a composite generator, this method will be overridden such that it returns all the classifiers it calls on in a list.
    
    Returns:
    
    All the classifiers that take part in this composite classifier, or null if this classifier is not a composite classifier.
  - toString
```
public String toString()
```
    Simply returns the name of the classifier.
    
    Overrides:
    
    toString in class Object
    
    Returns:
    
    The name of the classifier.
  - clone
```
public Object clone()
```
    This method returns a shallow clone.
    
    Overrides:
    
    clone in class Object
    
    Returns:
    
    A shallow clone.

Class Classifier

Field Summary

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

containingPackage

name

Constructor Detail

Classifier

Classifier

Method Detail

test

classify

classify

getInputType

getOutputType

allowableValues

valueIndexOf

featureValue

discreteValue

realValue

discreteValueArray

realValueArray

getCompositeChildren

toString

clone