de.lmu.ifi.dbs.elki.distance
Class CorrelationDistance<D extends CorrelationDistance<D>>

java.lang.Object
  de.lmu.ifi.dbs.elki.distance.AbstractDistance<D>
      de.lmu.ifi.dbs.elki.distance.CorrelationDistance<D>

Type Parameters:

D - distance type

All Implemented Interfaces:

Distance<D>, Externalizable, Serializable, Comparable<D>

Direct Known Subclasses:

PCACorrelationDistance, PreferenceVectorBasedCorrelationDistance

public abstract class CorrelationDistance<D extends CorrelationDistance<D>>
extends AbstractDistance<D>

The correlation distance is a special Distance that indicates the dissimilarity between correlation connected objects. The correlation distance between two points is a pair consisting of the correlation dimension of two points and the euclidean distance between the two points.

Author:

Elke Achtert

See Also:

Serialized Form

Field Summary

static Pattern	CORRELATION_DISTANCE_PATTERN The pattern used for correlation distances
protected int	correlationValue The correlation dimension.
protected double	euclideanValue The euclidean distance.
static String	SEPARATOR The component separator used by correlation distances.
private static long	serialVersionUID Generated SerialVersionUID.

Fields inherited from class de.lmu.ifi.dbs.elki.distance.AbstractDistance

DOUBLE_PATTERN, INFINITY_PATTERN, INTEGER_PATTERN

Constructor Summary

CorrelationDistance()
Empty constructor for serialization purposes.

CorrelationDistance(int correlationValue, double euclideanValue)
Constructs a new CorrelationDistance object consisting of the specified correlation value and euclidean value.

Method Summary

int	compareTo(D other) Compares this CorrelationDistance with the given CorrelationDistance wrt the represented correlation values.
boolean	equals(Object obj) Returns true if this == o has the value true or o is not null and o is of the same class as this instance and this.compareTo(o) is 0, false otherwise.
int	externalizableSize() Returns the number of Bytes this distance uses if it is written to an external file.
int	getCorrelationValue() Returns the correlation dimension between the objects.
double	getEuclideanValue() Returns the euclidean distance between the objects.
int	hashCode() Any extending class should implement a proper hashCode method.
void	readExternal(ObjectInput in) Reads the correlation value and the euclidean value of this CorrelationDistance from the specified stream.
String	toString() Returns a string representation of this CorrelationDistance.
void	writeExternal(ObjectOutput out) Writes the correlation value and the euclidean value of this CorrelationDistance to the specified stream.

Methods inherited from class de.lmu.ifi.dbs.elki.distance.AbstractDistance

getPattern, isInfiniteDistance, isNullDistance, isUndefinedDistance, requiredInputPattern, testInputPattern

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.lmu.ifi.dbs.elki.distance.Distance

infiniteDistance, minus, nullDistance, parseString, plus, undefinedDistance

Field Detail

SEPARATOR

public static final String SEPARATOR

The component separator used by correlation distances. Note: Do NOT use regular expression syntax characters!

See Also:

Constant Field Values

CORRELATION_DISTANCE_PATTERN

public static final Pattern CORRELATION_DISTANCE_PATTERN

The pattern used for correlation distances

serialVersionUID

private static final long serialVersionUID

Generated SerialVersionUID.

See Also:

Constant Field Values

correlationValue

protected int correlationValue

The correlation dimension.

euclideanValue

protected double euclideanValue

The euclidean distance.

Constructor Detail

CorrelationDistance

public CorrelationDistance()

Empty constructor for serialization purposes.

CorrelationDistance

public CorrelationDistance(int correlationValue,
                           double euclideanValue)

Constructs a new CorrelationDistance object consisting of the specified correlation value and euclidean value.

Parameters:

correlationValue - the correlation dimension to be represented by the CorrelationDistance

euclideanValue - the euclidean distance to be represented by the CorrelationDistance

Method Detail

toString

public String toString()

Returns a string representation of this CorrelationDistance.

Specified by:

toString in interface Distance<D extends CorrelationDistance<D>>

Overrides:

toString in class Object

Returns:

the correlation value and the euclidean value separated by blank

compareTo

public int compareTo(D other)

Compares this CorrelationDistance with the given CorrelationDistance wrt the represented correlation values. If both values are considered to be equal, the euclidean values are compared. Subclasses may need to overwrite this method if necessary.

Returns:

the value of Integer.compareTo(Integer) this.correlationValue.compareTo(other.correlationValue)} if it is a non zero value, the value of Double.compare(this.euclideanValue, other.euclideanValue) otherwise

hashCode

public int hashCode()

Description copied from class: AbstractDistance

Any extending class should implement a proper hashCode method.

Specified by:

hashCode in class AbstractDistance<D extends CorrelationDistance<D>>

equals

public boolean equals(Object obj)

Description copied from class: AbstractDistance

Returns true if this == o has the value true or o is not null and o is of the same class as this instance and this.compareTo(o) is 0, false otherwise.

Overrides:

equals in class AbstractDistance<D extends CorrelationDistance<D>>

getCorrelationValue

public int getCorrelationValue()

Returns the correlation dimension between the objects.

Returns:

the correlation dimension

getEuclideanValue

public double getEuclideanValue()

Returns the euclidean distance between the objects.

Returns:

the euclidean distance

writeExternal

public void writeExternal(ObjectOutput out)
                   throws IOException

Writes the correlation value and the euclidean value of this CorrelationDistance to the specified stream.

Throws:

IOException

readExternal

public void readExternal(ObjectInput in)
                  throws IOException

Reads the correlation value and the euclidean value of this CorrelationDistance from the specified stream.

Throws:

IOException

externalizableSize

public int externalizableSize()

Returns the number of Bytes this distance uses if it is written to an external file.

Returns:

12 (4 Byte for an integer, 8 Byte for a double value)