utilities
Class Distribution

java.lang.Object
  |
  +--java.util.AbstractCollection
        |
        +--java.util.AbstractList
              |
              +--java.util.Vector
                    |
                    +--utilities.Distribution

All Implemented Interfaces:

java.lang.Cloneable, java.util.Collection, java.util.List, java.io.Serializable

public class Distribution

extends java.util.Vector

See Also:

Serialized Form

Constructor Summary

Distribution()
Makes an empty distribution

Distribution(double Value, double Distribution)
Makes a distribution with one entry

Distribution(float Value, float Distribution)
Makes a distribution with one entry

Distribution(java.lang.Float Value, java.lang.Float Distribution)
Makes a distribution with one entry

Distribution(java.lang.String str)
Parses a string of the form "5 0.2 10 0.5 20 0.3" into a distribution object.

Method Summary

Distribution	appendDistribution(Distribution d) Appends a distribution to another one, no normalization
void	appendTerm(double v, double d) Appends a new term onto the distribution, no normalization
void	appendTerm(float v, float d) Appends a new term onto the distribution, no normalization
void	appendTerm(java.lang.Float v, java.lang.Float d) Appends a new term onto the distribution, no normalization
Distribution	applyDensity(float density) Apply a density weight to a Distribution.
Distribution	applyOffset(float offset) Apply a offset to the value of a Distribution.
Distribution	applyPower(float power) Apply a power to the value of a Distribution.
float	calculateAvg() Compute the average of a given Distribution
float	calculateDeviation() Standart deviation : Sum(f X^2)/ Sum(f) - X_bar^2 Easy computation....
float	calculateMax() calculateMax() method computes the maximal value from a Distribution instance.
float	calculateMin() calculateMin() method computes the minimal value, you can get from a distribution instance.
float	calculateMostLikely() Compute the most likely outcome of a given Distribution
float	calculateTotalDensity() Calculate the total density represented by the distribution.
void	cluster(int pointsNumber) Clusters the distribution by combining terms to reach at most a given number of points
void	compact() Compacts the distribution by combining like valued terms
double	compare(Distribution b) Statistic described in our AAAI paper to compare two distributions.
static Distribution	computeDifferenceJointDistribution(Distribution d1, Distribution d2)
static Distribution	computeDifferenceJointDistribution(java.util.Vector v) Computes the overall joint probability for a vector of distributions.
static Distribution	computeJointDistribution(Distribution d1, Distribution d2)
static Distribution	computeJointDistribution(java.util.Vector v) Computes the overall joint probability for a vector of distributions.
static Distribution	computeMaxJointDistribution(Distribution d1, Distribution d2)
static Distribution	computeMaxJointDistribution(java.util.Vector v) Computes the overall joint probability for a vector of distributions.
static Distribution	computeMinJointDistribution(Distribution d1, Distribution d2)
static Distribution	computeMinJointDistribution(java.util.Vector v) Computes the overall joint probability for a vector of distributions.
boolean	containsValue(float v, float t) Determines if the value is contained in the Distribution
boolean	equals(Distribution d) This algorithm assumes no equal values (e.g.
float	findClosestValue(float v) Searches for the closest match to the supplied value in the distribution.
float	findValue(float v) Searches for the supplied value in the distribution
float[]	getArray() getArray method.
java.util.Hashtable	getHashtable() getHashtable, converts the Distribution is an hashtable where the pair is for as
void	normalize() Normalizes the probabilities to 1
java.lang.String	output() Stringify in a TTaemsy way
void	sort() Should be removed if nobody uses it.
void	sortByDistribution() Sort the distribution by frequency.
void	sortByValue() Sort the distribution by value.

Methods inherited from class java.util.Vector

add, add, addAll, addAll, addElement, capacity, clear, clone, contains, containsAll, copyInto, elementAt, elements, ensureCapacity, equals, firstElement, get, hashCode, indexOf, indexOf, insertElementAt, isEmpty, lastElement, lastIndexOf, lastIndexOf, remove, remove, removeAll, removeAllElements, removeElement, removeElementAt, retainAll, set, setElementAt, setSize, size, subList, toArray, toArray, toString, trimToSize

Methods inherited from class java.util.AbstractList

iterator, listIterator, listIterator

Methods inherited from class java.lang.Object

getClass, notify, notifyAll, wait, wait, wait

Methods inherited from interface java.util.List

iterator, listIterator, listIterator

Constructor Detail

Distribution

public Distribution()

Makes an empty distribution

Distribution

public Distribution(java.lang.Float Value,
                    java.lang.Float Distribution)

Makes a distribution with one entry

Distribution

public Distribution(float Value,
                    float Distribution)

Makes a distribution with one entry

Distribution

public Distribution(double Value,
                    double Distribution)

Makes a distribution with one entry

Distribution

public Distribution(java.lang.String str)

Parses a string of the form "5 0.2 10 0.5 20 0.3" into a distribution object. Will also trip leading and trailing parentheses if present.

This will crash if you feed it an incorrectly formatted string.

Parameters:

str - The string representation of the distribution

Method Detail

getArray

public float[] getArray()

getArray method. This method is used by the simulator to get the distribution in a float array. This method is called in the simulator.ExecuteEvent constructor.

Returns:

float [] array

See Also:

ExecuteEvent#ExecuteEvent(int, int, Agent, String, Distribution, Distribution, Distribution)

getHashtable

public java.util.Hashtable getHashtable()

getHashtable, converts the Distribution is an hashtable where the pair is for as

Returns:

a hashtable

findValue

public float findValue(float v)

Searches for the supplied value in the distribution

Parameters:

v - The value to look for

Returns:

The value's probability (0 if not found)

findClosestValue

public float findClosestValue(float v)

Searches for the closest match to the supplied value in the distribution. In case of a tie, the element with the greater probability is chosen.

Parameters:

v - The value to look for

Returns:

The value

containsValue

public boolean containsValue(float v,
                             float t)

Determines if the value is contained in the Distribution

Parameters:

v - The value to look for

t - The threshold (+/-) around each value to be considered a match

Returns:

True, if the value is contained

calculateMin

public float calculateMin()

calculateMin() method computes the minimal value, you can get from a distribution instance. Useful for the diagnostic component.

calculateMax

public float calculateMax()

calculateMax() method computes the maximal value from a Distribution instance.

calculateAvg

public float calculateAvg()

Compute the average of a given Distribution

calculateMostLikely

public float calculateMostLikely()

Compute the most likely outcome of a given Distribution

calculateDeviation

public float calculateDeviation()

Standart deviation : Sum(f X^2)/ Sum(f) - X_bar^2 Easy computation....

calculateTotalDensity

public float calculateTotalDensity()

Calculate the total density represented by the distribution.

Returns:

the total density Example:
[0.0, 0.1, 1.0, 0.2, 10.0, 0.3] = total density 0.6

applyDensity

public Distribution applyDensity(float density)

Apply a density weight to a Distribution. Changes are made to a clone, which is returned, the original is left unchanged.

Parameters:

density - Example:
Dc = [0.0, 1.0] Dd = [1.0, 0.5] Dq = [10.0, 0.5]
Applying a density of 0.5 give this result:
Dc = [0.0, 0.5] Dd = [1.0, 0.25] Dq = [10.0, 0.25]

applyPower

public Distribution applyPower(float power)

Apply a power to the value of a Distribution. Changes are made to a clone, which is returned, the original is left unchanged.

Parameters:

power - Example:
Dc = [0.0, 1.0] Dd = [1.0, 0.5] Dq = [10.0, 0.5]
Applying a power of 0.5 give this result:
Dc = [0.0, 1.0] Dd = [0.5, 0.5] Dq = [5.0, 0.5]

applyOffset

public Distribution applyOffset(float offset)

Apply a offset to the value of a Distribution. Changes are made to a clone, which is returned, the original is left unchanged.

Parameters:

offset - Example:
Dc = [0.0, 1.0] Dd = [1.0, 0.5] Dq = [10.0, 0.5]
Applying a power of 2.5 give this result:
Dc = [2.5, 1.0] Dd = [3.5, 0.5] Dq = [12.5, 0.5]

appendTerm

public void appendTerm(java.lang.Float v,
                       java.lang.Float d)

Appends a new term onto the distribution, no normalization

appendTerm

public void appendTerm(float v,
                       float d)

Appends a new term onto the distribution, no normalization

appendTerm

public void appendTerm(double v,
                       double d)

Appends a new term onto the distribution, no normalization

appendDistribution

public Distribution appendDistribution(Distribution d)

Appends a distribution to another one, no normalization

computeJointDistribution

public static Distribution computeJointDistribution(java.util.Vector v)

Computes the overall joint probability for a vector of distributions. Values are summed and compacted.

computeJointDistribution

public static Distribution computeJointDistribution(Distribution d1,
                                                    Distribution d2)

computeDifferenceJointDistribution

public static Distribution computeDifferenceJointDistribution(java.util.Vector v)

Computes the overall joint probability for a vector of distributions. Values are differenced and compacted.

computeDifferenceJointDistribution

public static Distribution computeDifferenceJointDistribution(Distribution d1,
                                                              Distribution d2)

computeMaxJointDistribution

public static Distribution computeMaxJointDistribution(java.util.Vector v)

Computes the overall joint probability for a vector of distributions. Values are maxed and compacted.

computeMaxJointDistribution

public static Distribution computeMaxJointDistribution(Distribution d1,
                                                       Distribution d2)

computeMinJointDistribution

public static Distribution computeMinJointDistribution(java.util.Vector v)

Computes the overall joint probability for a vector of distributions. Values are maxed and compacted.

computeMinJointDistribution

public static Distribution computeMinJointDistribution(Distribution d1,
                                                       Distribution d2)

compact

public void compact()

Compacts the distribution by combining like valued terms

normalize

public void normalize()

Normalizes the probabilities to 1

cluster

public void cluster(int pointsNumber)

Clusters the distribution by combining terms to reach at most a given number of points

sort

public void sort()

Should be removed if nobody uses it. Was used in compare().

sortByValue

public void sortByValue()

Sort the distribution by value. I'm not sure anyone will use this but I need it to compare 2 Distributions

sortByDistribution

public void sortByDistribution()

Sort the distribution by frequency. I'm not sure anyone will use this either but I need it for the cluster() method.

equals

public boolean equals(Distribution d)

This algorithm assumes no equal values (e.g. they're compacted)

compare

public double compare(Distribution b)

Statistic described in our AAAI paper to compare two distributions. It's based on the difference of area between the cumulative probability plots of the distributions.

Returns:

a value significatif depending of the errors between two Distributions

output

public java.lang.String output()

Stringify in a TTaemsy way