jphase
Class ErlangCoxianVar

java.lang.Object
  jphase.AbstractContPhaseVar
      jphase.ErlangCoxianVar

All Implemented Interfaces:

JMarkovElement, ContPhaseVar, PhaseVar

public class ErlangCoxianVar
extends AbstractContPhaseVar
implements PhaseVar

Phase-Type representation of an ErlangCoxian distribution as defined by Osogami and Harchol in "Closed form solutions for mapping general distributions to quasi-minimal PH distributions", 2005.

Constructor Summary

ErlangCoxianVar()
Constructor of an Erlang Coxian variable in dense representation.

ErlangCoxianVar(int n)
Constructor of a ErlangCoxian variable in dense representation

ErlangCoxianVar(int n, double p, double lambdaY, double lambdaX1, double lambdaX2, double px)
Constructor of a ErlangCoxian variable in dense representation

Method Summary

PhaseVar	convoExpo(double lambda) Creates a Dense Continuous Phase Variable that represents the convolution of the original ErlangCoxian distribution and an exponential phase with rate lambda
ContPhaseVar	copy() Creates a deep copy of the original Phase-Type Variable
java.lang.String	description() This method return a complete verbal describtion of this element.
double	getLambdaX1()
double	getLambdaX2()
double	getLambdaY()
no.uib.cipr.matrix.Matrix	getMatrix() Returns the transition matrix of the Phase-Type Distribution
int	getN()
double	getP()
double	getPx()
no.uib.cipr.matrix.Vector	getVector() Returns the initial probability mass vector
PhaseVar	mixtureExpo(double lambda, double p) Creates a Dense Continuous Phase Variable that represents the mixture of the original ErlangCoxian distribution (p) and an exponential phase with rate lambda (1-p)
ContPhaseVar	newVar(int n) Creates a new variable of the same class of the original Continuous Phase-Type Variable
void	setLambdaX1(double lambdaX1)
void	setLambdaX2(double lambdaX2)
void	setLambdaY(double lambdaY)
void	setMatrix(no.uib.cipr.matrix.Matrix A) Rate Matrix = A
void	setN(int n)
void	setP(double p)
void	setPx(double px)
void	setVector(no.uib.cipr.matrix.Vector alpha) Initial Probability vector = alpha

Methods inherited from class jphase.AbstractContPhaseVar

cdf, cdf, CV, eqResidualTime, expectedValue, getMat0, getMat0Array, getMatrixArray, getNumPhases, getVec0, getVectorArray, label, lossFunction1, lossFunction2, max, max, median, min, min, mix, mix, moment, pdf, pdf, prob, quantil, residualTime, residualVar, stdDeviation, sum, sum, sumGeom, sumPH, sumPH, survival, survival, times, toString, variance, waitingQ

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface jphase.PhaseVar

cdf, cdf, CV, expectedValue, getMat0, getMat0Array, getMatrixArray, getNumPhases, getVec0, getVectorArray, lossFunction1, lossFunction2, median, moment, prob, quantil, stdDeviation, survival, survival, variance

Methods inherited from interface jmarkov.basic.JMarkovElement

equals, label, toString

Constructor Detail

ErlangCoxianVar

public ErlangCoxianVar()

Constructor of an Erlang Coxian variable in dense representation. As default it has just one phase in the Erlang that is taken with probability one. This phase and those of the Coxian distribution has rate 1 per time unit.

ErlangCoxianVar

public ErlangCoxianVar(int n,
                       double p,
                       double lambdaY,
                       double lambdaX1,
                       double lambdaX2,
                       double px)

Constructor of a ErlangCoxian variable in dense representation

Parameters:

n - total number of phases (Erlang degree: n-2)

p - probability of having a positive elapse time in the distribution. 1-p: mass probability at zero

lambdaY - rate of the Erlang distribution

lambdaX1 - rate of the first stage of the Coxian distribution

lambdaX2 - rate of the second stage of the Coxian distribution

px - probability of going from the first to the second stage in the Coxian distribution. 1-p: probability of absorption at the first stage of the Coxian distribution

ErlangCoxianVar

public ErlangCoxianVar(int n)

Constructor of a ErlangCoxian variable in dense representation

Parameters:

n - total number of phases (Erlang degree: n-2) absorption at the first stage of the Coxian distribution

Method Detail

getN

public int getN()

Returns:

Total number of phases

setN

public void setN(int n)

Parameters:

n - Total number of phases

getP

public double getP()

Returns:

Probability of having a positive elapse time in the distribution. 1-p: mass probability at zero

setP

public void setP(double p)

Parameters:

p - Probability of having a positive elapse time in the distribution. 1-p: mass probability at zero

getLambdaY

public double getLambdaY()

Returns:

Erlang distribution rate

setLambdaY

public void setLambdaY(double lambdaY)

Parameters:

lambdaY - Erlang distribution rate

getLambdaX1

public double getLambdaX1()

Returns:

Rate of the first stage at the Coxian distribution

setLambdaX1

public void setLambdaX1(double lambdaX1)

Parameters:

lambdaX1 - Rate of the first stage at the Coxian distribution

getLambdaX2

public double getLambdaX2()

Returns:

Rate of the secons stage at the Coxian distribution

setLambdaX2

public void setLambdaX2(double lambdaX2)

Parameters:

lambdaX2 - Rate of the second stage at the Coxian distribution

getPx

public double getPx()

Returns:

Probability of going from the first to the second stage in the Coxian distribution. 1-p: probability of absorption at the first stage if the Coxian distribution

setPx

public void setPx(double px)

Parameters:

px - Probability of going from the first to the second stage in the Coxian distribution. 1-p: probability of absorption at the first stage if the Coxian distribution

getMatrix

public no.uib.cipr.matrix.Matrix getMatrix()

Description copied from interface: PhaseVar

Returns the transition matrix of the Phase-Type Distribution

Specified by:

getMatrix in interface PhaseVar

Returns:

Transition matrix for transient states of the Phase-Type Distribution

setMatrix

public void setMatrix(no.uib.cipr.matrix.Matrix A)

Description copied from interface: PhaseVar

Rate Matrix = A

Specified by:

setMatrix in interface PhaseVar

Parameters:

A - Transition matrix for transient states of the Phase-Type Distribution

getVector

public no.uib.cipr.matrix.Vector getVector()

Description copied from interface: PhaseVar

Returns the initial probability mass vector

Specified by:

getVector in interface PhaseVar

Returns:

Initial probability mass vector

setVector

public void setVector(no.uib.cipr.matrix.Vector alpha)

Description copied from interface: PhaseVar

Initial Probability vector = alpha

Specified by:

setVector in interface PhaseVar

Parameters:

alpha - Initial probability mass vector

copy

public ContPhaseVar copy()

Description copied from interface: ContPhaseVar

Creates a deep copy of the original Phase-Type Variable

Specified by:

copy in interface ContPhaseVar

Specified by:

copy in interface PhaseVar

Returns:

A deep copy of the original Phase-Type Variable

newVar

public ContPhaseVar newVar(int n)

Description copied from interface: ContPhaseVar

Creates a new variable of the same class of the original Continuous Phase-Type Variable

Specified by:

newVar in interface ContPhaseVar

Parameters:

n - number of Phases of the new Variable

Returns:

A new variable of the same class of the original Continuous Phase-Type Variable

mixtureExpo

public PhaseVar mixtureExpo(double lambda,
                            double p)

Creates a Dense Continuous Phase Variable that represents the mixture of the original ErlangCoxian distribution (p) and an exponential phase with rate lambda (1-p)

Parameters:

lambda - rate of the exponential phase to be included in the mixture

p - probability mass of the ErlangCoxian distribution in the mixture

Returns:

Dense Continuous Phase Variable that represents the mixture of the original ErlangCoxian distribution (p) and an exponential phase with rate lambda (1-p)

convoExpo

public PhaseVar convoExpo(double lambda)

Creates a Dense Continuous Phase Variable that represents the convolution of the original ErlangCoxian distribution and an exponential phase with rate lambda

Parameters:

lambda - rate of the exponential phase to be included in the convolution

Returns:

Dense Continuous Phase Variable that represents the convolution of the original ErlangCoxian distribution and an exponential phase with rate lambda

description

public java.lang.String description()

Description copied from interface: JMarkovElement

This method return a complete verbal describtion of this element. This description may contain multiple text rows.

Specified by:

description in interface JMarkovElement

Overrides:

description in class AbstractContPhaseVar

Returns:

A String describing this element.

See Also:

JMarkovElement.label()