: Class Mvndstpack

Overview

Package

Class

Use

Tree

Deprecated

Index

Help

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: INNER | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

com.kutsyy.util
Class Mvndstpack

java.lang.Object
  |
  +--com.kutsyy.util.Mvndstpack

public final class Mvndstpack
extends Object

This class implements multivariate normal integration.
Method is based on MCMC. class use its own random number generator,
and does not depends an any other classes
so no sinchronization is required for multithreading.

Description from original FORTRAN code:
A subroutine for computing multivariate normal probabilities.
This subroutine uses an algorithm given in the paper
"Numerical Computation of Multivariate Normal Probabilities", in
J. of Computational and Graphical Stat., 1(1992), pp. 141-149, by
Alan Genz
Department of Mathematics
Washington State University
Pullman, WA 99164-3113
Email : AlanGenz@wsu.edu
Parameters N INTEGER, the number of variables.
LOWER REAL, array of lower integration limits.
UPPER REAL, array of upper integration limits.
INFIN INTEGER, array of integration limits flags:
if INFIN(I) < 0, Ith limits are (-infinity, infinity);
if INFIN(I) = 0, Ith limits are (-infinity, UPPER(I)];
if INFIN(I) = 1, Ith limits are [LOWER(I), infinity);
if INFIN(I) = 2, Ith limits are [LOWER(I), UPPER(I)].
CORREL REAL, array of correlation coefficients; the correlation
coefficient in row I column J of the correlation matrix
should be stored in CORREL( J + ((I-2)*(I-1))/2 ), for J < I.
THe correlation matrix must be positive semidefinite.
MAXPTS INTEGER, maximum number of function values allowed. This
parameter can be used to limit the time. A sensible
strategy is to start with MAXPTS = 1000*N, and then
increase MAXPTS if ERROR is too large.
ABSEPS REAL absolute error tolerance.
RELEPS REAL relative error tolerance.
ERROR REAL estimated absolute error, with 99% confidence level.
VALUE REAL estimated value for the integral
INFORM INTEGER, termination status parameter:
if INFORM = 0, normal completion with ERROR < EPS;
if INFORM = 1, completion with ERROR >EPS and MAXPTS
function vaules used; increase MAXPTS to
decrease ERROR;
if INFORM = 2, N >100 or N < 1.

Author:: kutsyy

Field Summary

private double[] a
          Used internaly, see original FORTRAN code for details

private double abseps
          Used internaly, see original FORTRAN code for details

private double[] b
          Used internaly, see original FORTRAN code for details

private int[][] c
          Used internaly, see original FORTRAN code for details

private double[] correl
          Used internaly, see original FORTRAN code for details

private double[] cov
          Used internaly, see original FORTRAN code for details

private double[] d
          Used internaly, see original FORTRAN code for details

private double[] e
          Used internaly, see original FORTRAN code for details

double error
          Output : error of the estimator

private int hisum
          Used internaly, see original FORTRAN code for details

private int[] infi
          Used internaly, see original FORTRAN code for details

private int[] infin
          Used internaly, see original FORTRAN code for details

private int[] infis
          Used internaly, see original FORTRAN code for details

int inform
          Output: code of the computation (see description)

private double[] lower
          Used internaly, see original FORTRAN code for details

int maxpts
          Input: number of itteration

private int n
          Used internaly, see original FORTRAN code for details

private int[] N
          Used internaly, see original FORTRAN code for details

private int nl
          Used internaly, see original FORTRAN code for details

private int np
          Used internaly, see original FORTRAN code for details

private double olds
          Used internaly, see original FORTRAN code for details

private int[] p
          Used internaly, see original FORTRAN code for details

private double[] prime
          Used internaly, see original FORTRAN code for details

private double[] psqt
          Used internaly, see original FORTRAN code for details

double releps
          Input: relative error

private MersenneTwister rnd
          Random Number Generator (faster than Java's build in)

boolean running
          Description of the Field

private int sampls
          Used internaly, see original FORTRAN code for details

private double[] sd
          Used internaly, see original FORTRAN code for details

private double[] upper
          Used internaly, see original FORTRAN code for details

double value
          Output: value of the integral

private double varest
          Used internaly, see original FORTRAN code for details

Constructor Summary

Mvndstpack()
          Empty constructor

Mvndstpack(double[] Lower, double[] Upper, double[][] Cor, int Maxpts, double Abseps, double Releps)
          Constructor

Mvndstpack(double[] Lower, double[] Upper, double[] Correl, int Maxpts, double Abseps, double Releps)
          Constructor

Mvndstpack(double[] Lower, double[] Upper, int[] Infin, double[][] Cor, int Maxpts, double Abseps, double Releps)
          Constructor

Mvndstpack(double[] Lower, double[] Upper, int[] Infin, double[] Correl, int Maxpts, double Abseps, double Releps)
          Constructor

Method Summary

private void compute()
          Same as run

private void cor2to1(double[][] cor2)
          Insert the method's description here.

private void covsrt(double[] y, int[] infis)
          Used internaly, see original FORTRAN code for details

private void dkbvrc(int ndim, int minvls, int maxvls)
          Discription from FORTRAN code:
Automatic Multidimensional Integration Subroutine AUTHOR: Alan Genz Department of Mathematics Washington State University Pulman, WA 99164-3113 Email: AlanGenz@wsu.edu Last Change: 5/15/98 KRBVRC computes an approximation to the integral 1 1 1 I I ...

private void dkrcht(int s, double[] quasi)
          This subroutine generates a new quasi-random Richtmeyer vector.

private void dkrcht(int s, double[] quasi, int start)
          This subroutine generates a new quasi-random Richtmeyer vector.

private void dksmrc(int ndim, int klim, int prime, double[] vk, double[] x)
          Used internaly, see original FORTRAN code for details

private void dkswap(double[] x, int i, int j)
          Swap i and j element od x

private void dkswap(int[] x, int i, int j)
          Swap i and j element od x

private double functn(int ndim, double[] x)
          same as mvndfn(int, double[])

private double mvndfn(int n, double[] w)
          Insert the method's description here.

private double mvndnt(int[] infis, double[] d, double[] e)
          Used internaly, see original FORTRAN code for details

private void mvndst()
          Used internaly, see original FORTRAN code for details

void Mvndstpack(double[] Lower, double[] Upper, int[] Infin, double[][] Cov, int Maxpts, double Abseps, double Releps)
          main method

void Mvndstpack(double[] Lower, double[] Upper, int[] Infin, double[] Correl, int Maxpts, double Abseps, double Releps)
          main method

private void mvnlms(double a, double b, int infin, double[] lower, double[] upper)
          Used internaly, see original FORTRAN code for details

private void rcswp(int p, int q)
          Swaps rows and columns P and Q in situ, with P <= Q.

void run()
          To actualy re compute.

Methods inherited from class java.lang.Object

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

Field Detail

running

public boolean running

Description of the Field

error

public double error

Output : error of the estimator

inform

public int inform

Output: code of the computation (see description)

value

public double value

Output: value of the integral

releps

public double releps

Input: relative error

maxpts

public int maxpts

Input: number of itteration

lower

private double[] lower

Used internaly, see original FORTRAN code for details

sd

private double[] sd

Used internaly, see original FORTRAN code for details

upper

private double[] upper

Used internaly, see original FORTRAN code for details

infin

private int[] infin

Used internaly, see original FORTRAN code for details

correl

private double[] correl

Used internaly, see original FORTRAN code for details

abseps

private double abseps

Used internaly, see original FORTRAN code for details

nl

private int nl

Used internaly, see original FORTRAN code for details

cov

private double[] cov

Used internaly, see original FORTRAN code for details

a

private double[] a

Used internaly, see original FORTRAN code for details

b

private double[] b

Used internaly, see original FORTRAN code for details

infi

private int[] infi

Used internaly, see original FORTRAN code for details

n

private int n

Used internaly, see original FORTRAN code for details

c

private final int[][] c

Used internaly, see original FORTRAN code for details

hisum

private int hisum

Used internaly, see original FORTRAN code for details

np

private int np

Used internaly, see original FORTRAN code for details

olds

private double olds

Used internaly, see original FORTRAN code for details

p

private final int[] p

Used internaly, see original FORTRAN code for details

prime

private final double[] prime

Used internaly, see original FORTRAN code for details

psqt

private double[] psqt

Used internaly, see original FORTRAN code for details

sampls

private int sampls

Used internaly, see original FORTRAN code for details

varest

private double varest

Used internaly, see original FORTRAN code for details

infis

private int[] infis

Used internaly, see original FORTRAN code for details

e

private double[] e

Used internaly, see original FORTRAN code for details

d

private double[] d

Used internaly, see original FORTRAN code for details

rnd

private MersenneTwister rnd

Random Number Generator (faster than Java's build in)

N

private int[] N

Used internaly, see original FORTRAN code for details

Constructor Detail

Mvndstpack

public Mvndstpack(double[] Lower,
                  double[] Upper,
                  double[][] Cor,
                  int Maxpts,
                  double Abseps,
                  double Releps)

Constructor

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Cor - Covariance matrix; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror

Mvndstpack

public Mvndstpack(double[] Lower,
                  double[] Upper,
                  double[] Correl,
                  int Maxpts,
                  double Abseps,
                  double Releps)

Constructor

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror; Correl - Description of Parameter

Mvndstpack

public Mvndstpack(double[] Lower,
                  double[] Upper,
                  int[] Infin,
                  double[][] Cor,
                  int Maxpts,
                  double Abseps,
                  double Releps)

Constructor

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Cor - Covariance matrix; Infin - as above; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror

Mvndstpack

public Mvndstpack()

Empty constructor

Mvndstpack

public Mvndstpack(double[] Lower,
                  double[] Upper,
                  int[] Infin,
                  double[] Correl,
                  int Maxpts,
                  double Abseps,
                  double Releps)

Constructor

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Correl - Covariance matrix; Infin - as above; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror

Method Detail

Mvndstpack

public void Mvndstpack(double[] Lower,
                       double[] Upper,
                       int[] Infin,
                       double[][] Cov,
                       int Maxpts,
                       double Abseps,
                       double Releps)

main method

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Infin - as above; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror; Cov - Description of Parameter

Mvndstpack

public void Mvndstpack(double[] Lower,
                       double[] Upper,
                       int[] Infin,
                       double[] Correl,
                       int Maxpts,
                       double Abseps,
                       double Releps)

main method

Parameters:: Lower - Lower limit of integration; Upper - Upper limit of integration; Correl - Covariance matrix; Infin - as above; Maxpts - maximum number of itteration; Abseps - absolute error; Releps - relative arror

run

public void run()

To actualy re compute.

compute

private void compute()

Same as run

See Also:: run()

dkbvrc

private void dkbvrc(int ndim,
                    int minvls,
                    int maxvls)

Discription from FORTRAN code:
Automatic Multidimensional Integration Subroutine AUTHOR: Alan Genz Department of Mathematics Washington State University Pulman, WA 99164-3113 Email: AlanGenz@wsu.edu Last Change: 5/15/98 KRBVRC computes an approximation to the integral 1 1 1 I I ... I F(X) dx(NDIM)...dx(2)dx(1) 0 0 0 DKBVRC uses randomized Korobov rules for the first 20 variables. The primary references are "Randomization of Number Theoretic Methods for Multiple Integration" R. Cranley and T.N.L. Patterson, SIAM J Numer Anal, 13, pp. 904-14, and "Optimal Parameters for Multidimensional Integration", P. Keast, SIAM J Numer Anal, 10, pp.831-838. If there are more than 20 variables, the remaining variables are integrated using Richtmeyer rules. A reference is "Methods of Numerical Integration", P.J. Davis and P. Rabinowitz, Academic Press, 1984, pp. 482-483. Parameters Input parameters NDIM Number of variables, must exceed 1, but not exceed 40 MINVLS Integer minimum number of function evaluations allowed. MINVLS must not exceed MAXVLS. If MINVLS < 0 then the routine assumes a previous call has been made with the same integrand and continues that calculation. MAXVLS Integer maximum number of function evaluations allowed. FUNCTN EXTERNALly declared user defined function to be integrated. It must have parameters (NDIM,Z), where Z is a real array of dimension NDIM. ABSEPS Required absolute accuracy. RELEPS Required relative accuracy. Output parameters MINVLS Actual number of function evaluations used. ABSERR Estimated absolute accuracy of FINEST. FINEST Estimated value of integral. INFORM INFORM = 0 for normal exit, when ABSERR <= MAX(ABSEPS, RELEPS*ABS(FINEST)) and INTVLS <= MAXCLS. INFORM = 1 If MAXVLS was too small to obtain the required accuracy. In this case a value FINEST is returned with estimated absolute accuracy ABSERR. Creation date: (10/11/2000 2:43:21 AM)

Parameters:: ndim - int; minvls - int[]; maxvls - int

mvndst

private void mvndst()

Used internaly, see original FORTRAN code for details

covsrt

private void covsrt(double[] y,
                    int[] infis)

Used internaly, see original FORTRAN code for details

Parameters:: y - double[]; infis - int[]

dkrcht

private void dkrcht(int s,
                    double[] quasi)

This subroutine generates a new quasi-random Richtmeyer vector. A reference is "Methods of Numerical Integration", P.J. Davis and P. Rabinowitz, Academic Press, 1984, pp. 482-483. INPUTS: S - the number of dimensions; KRRCHT is initialized for each new S or S < 1. OUTPUTS: QUASI - a new quasi-random S-vector Creation date: (10/11/2000 11:20:54 AM)

Parameters:: s - double; quasi - double[]

dkrcht

private void dkrcht(int s,
                    double[] quasi,
                    int start)

Parameters:: s - double; quasi - double[]; start - Description of Parameter

dksmrc

private void dksmrc(int ndim,
                    int klim,
                    int prime,
                    double[] vk,
                    double[] x)

Used internaly, see original FORTRAN code for details

Parameters:: ndim - int; klim - int; prime - int; vk - double[]; x - double[]

dkswap

private void dkswap(double[] x,
                    int i,
                    int j)

Swap i and j element od x

Parameters:: x - double[]; i - int; j - int

dkswap

private void dkswap(int[] x,
                    int i,
                    int j)

Swap i and j element od x

Parameters:: x - int[]; i - int; j - int

functn

private double functn(int ndim,
                      double[] x)

same as mvndfn(int, double[])

Parameters:: ndim - int; x - double[]
Returns:: double
See Also:: mvndfn(int, double[])

mvndfn

private double mvndfn(int n,
                      double[] w)

Insert the method's description here. Creation date: (10/10/2000 10:50:14 PM)

Parameters:: n - int; w - double[]
Returns:: double

mvndnt

private double mvndnt(int[] infis,
                      double[] d,
                      double[] e)

Used internaly, see original FORTRAN code for details

Parameters:: infis - int[]; d - double[]; e - double[]
Returns:: double[]

mvnlms

private void mvnlms(double a,
                    double b,
                    int infin,
                    double[] lower,
                    double[] upper)

Used internaly, see original FORTRAN code for details

Parameters:: a - double; b - double; infin - int; lower - double[]; upper - double[]

rcswp

private void rcswp(int p,
                   int q)

Swaps rows and columns P and Q in situ, with P <= Q. Creation date: (10/12/2000 12:36:03 AM)

Parameters:: p - int; q - int

cor2to1

private void cor2to1(double[][] cor2)

Insert the method's description here. Creation date: (10/10/2000 6:47:53 PM) Transform 2 dimentional matric in to 1 dimentional

Parameters:: cor2 - Description of Parameter

Overview

Package

Class

Use

Tree

Deprecated

Index

Help

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: INNER | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

com.kutsyy.util Class Mvndstpack

running

error

inform

value

releps

maxpts

lower

sd

upper

infin

correl

abseps

nl

cov

a

b

infi

n

c

hisum

np

olds

p

prime

psqt

sampls

varest

infis

e

d

rnd

N

Mvndstpack

Mvndstpack

Mvndstpack

Mvndstpack

Mvndstpack

Mvndstpack

Mvndstpack

run

compute

dkbvrc

mvndst

covsrt

dkrcht

dkrcht

dksmrc

dkswap

dkswap

functn

mvndfn

mvndnt

mvnlms

rcswp

cor2to1

com.kutsyy.util
Class Mvndstpack