SSJ  3.3.1
Stochastic Simulation in Java
Public Member Functions | List of all members
NI1 Class Reference

Extends the class NortaInitDisc and implements the algorithm NI1. More...

Inheritance diagram for NI1:
[legend]
Collaboration diagram for NI1:
[legend]

Public Member Functions

 NI1 (double rX, DiscreteDistributionInt dist1, DiscreteDistributionInt dist2, double tr, double tolerance)
 Constructor with the target rank correlation rX, the two discrete marginals dist1 and dist2, the parameter for truncation tr (see the constructor of class NortaInitDisc ) and the specific parameter \(\epsilon=\) tolerance defined above for the algorithm NI1.
 
double computeCorr ()
 Computes and returns the correlation \(\rho_Z\) using the algorithm NI1. More...
 
String toString ()
 
- Public Member Functions inherited from NortaInitDisc
 NortaInitDisc (double rX, DiscreteDistributionInt dist1, DiscreteDistributionInt dist2, double tr)
 Constructor with the target rank correlation rX, the two discrete marginals dist1 and dist2 and the parameter for the truncation tr. More...
 
abstract double computeCorr ()
 Computes and returns the correlation \(\rho_Z\). More...
 
void computeParams ()
 Computes the following inputs of each marginal distribution: More...
 
double integ (double r)
 Computes the function. More...
 
double deriv (double r)
 Computes the derivative of \(g_r\), given by. More...
 
String toString ()
 

Additional Inherited Members

- Protected Attributes inherited from NortaInitDisc
double rX
 
DiscreteDistributionInt dist1
 
DiscreteDistributionInt dist2
 
double tr
 
double mu1
 
- Package Attributes inherited from NortaInitDisc
double mu2
 
double sd1
 
double sd2
 
int m2
 

Detailed Description

Extends the class NortaInitDisc and implements the algorithm NI1.

It uses an algorithm based on Brent method for root-finding, which combines root-bracketing, bisection and inverse quadratic interpolation. It calls the method integ to compute the function \(g_r\) given in ( gr_M ). The search should be done in the interval \([-1,0]\) if \(r_X\in[-1,0]\), or \([0,1]\) if \(r_X\in[0,1]\). At each iteration, the algorithm halves the interval length and uses an accuracy \(\epsilon\) to find the root \(\rho_Z\) of equation ( fr ).

Member Function Documentation

◆ computeCorr()

double computeCorr ( )

Computes and returns the correlation \(\rho_Z\) using the algorithm NI1.

a, b, c, fa, fb, fc, and tolerance correspond to a, b, c, f(a) f(b), f(c) and epsilon, respectively, in the paper (paragraph "Method NI1" of section 3).


The documentation for this class was generated from the following file: