Class LogisticCore

java.lang.Object

solvers.queueingnet.montecarloLogistic.solver.LogisticCore

Direct Known Subclasses:

LogisticCoreAdd, LogisticCoreMult

public abstract class LogisticCore
extends Object

Implements the core Logistic Sampling algorithm by calling the required stationary point calculations, Hessian calculations, and the correct integration scheme for solution of the model, to calculate the NC. See Ong 2018. Subclassed by the additive and multiplicative core, which are responsible for choosing the correct logistic functions for use by the core algorithm

Field Summary

Fields

Modifier and Type	Field	Description
protected cern.colt.matrix.linalg.Algebra	A
protected cern.colt.matrix.DoubleMatrix2D	cov
protected cern.colt.matrix.DoubleMatrix1D	delays
protected cern.colt.matrix.DoubleMatrix2D	demands
protected double	epsilon
protected LogisticFunctionBase	F
protected boolean	has_delay
protected boolean	has_multi
protected cern.colt.matrix.DoubleMatrix2D	hess
protected boolean	invalid_model
protected int	K
protected MathContext	MC
protected MCIntegrator	MCI
protected int	N
protected cern.colt.matrix.DoubleMatrix1D	pop
protected int	R
protected cern.colt.matrix.DoubleMatrix1D	serv
protected double	sigma
protected cern.colt.matrix.DoubleMatrix1D	x_stat
protected boolean	zero_population

Constructor Summary

Constructors

Constructor	Description
LogisticCore(double[][] demands, int[] serv, double[] delays, int[] population, double epsilon, int max_samples, MathContext mc, int nthreads)

Method Summary

Modifier and Type	Method	Description
BigDecimal	calculateNC()	Calculate the NC, and multiply by the appropriate coefficients.
double[][]	getCovariance()
double[][]	getHessian()	Get Hessian
double[]	getStationaryPoint()
double[]	getStationaryPointSimplex()
protected abstract LogisticFunctionBase	initialize_logistic_function()
boolean	is_invalid_model()

Methods inherited from class Object

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

Field Details

demands

protected cern.colt.matrix.DoubleMatrix2D demands

delays

protected cern.colt.matrix.DoubleMatrix1D delays

has_delay

protected boolean has_delay

has_multi

protected boolean has_multi

sigma

protected double sigma

pop

protected cern.colt.matrix.DoubleMatrix1D pop

serv

protected cern.colt.matrix.DoubleMatrix1D serv

N

protected int N

K

protected int K

R

protected int R

epsilon

protected double epsilon

x_stat

protected cern.colt.matrix.DoubleMatrix1D x_stat

hess

protected cern.colt.matrix.DoubleMatrix2D hess

cov

protected cern.colt.matrix.DoubleMatrix2D cov

invalid_model

protected boolean invalid_model

zero_population

protected boolean zero_population

A

protected cern.colt.matrix.linalg.Algebra A

MCI

protected MCIntegrator MCI

F

protected LogisticFunctionBase F

MC

protected MathContext MC

Constructor Details

LogisticCore

public LogisticCore(double[][] demands,
 int[] serv,
 double[] delays,
 int[] population,
 double epsilon,
 int max_samples,
 MathContext mc,
 int nthreads)
             throws InternalErrorException

Throws:

InternalErrorException

Method Details

initialize_logistic_function

protected abstract LogisticFunctionBase initialize_logistic_function()
                                                              throws InternalErrorException

Throws:

InternalErrorException

calculateNC

public BigDecimal calculateNC()
                       throws InternalErrorException

Calculate the NC, and multiply by the appropriate coefficients. Details see Casale 2017 and Ong 2018

Returns:

The Normalising constant in BigDecimal form

Throws:

InternalErrorException

getHessian

public double[][] getHessian()

Get Hessian

Returns:

getCovariance

public double[][] getCovariance()

getStationaryPoint

public double[] getStationaryPoint()

getStationaryPointSimplex

public double[] getStationaryPointSimplex()

is_invalid_model

public boolean is_invalid_model()