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

Partial variance estimator. More...

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

Public Member Functions

 PartialVarianceEstimator (MonteCarloModelDoubleRQMC model, double approxMean, List< CoordinateSet > coordSets)
 
MonteCarloModelDoubleRQMC getModel ()
 
void setModel (MonteCarloModelDoubleRQMC model, double approxMean)
 Sets the model whose partial variances are to be estimated. More...
 
List< CoordinateSetgetCoordinateSets ()
 Returns the list of coordinate sets under consideration.
 
void setCoordinateSets (List< CoordinateSet > coordSets)
 Set the coordinate sets to consider to coordSets.
 
double getApproximateMean ()
 
void simulate (RandomStream stream)
 Simulates the estimator once. More...
 
double [] getPerformance ()
 Recovers and returns the realization of the performance measure, of type E.
 
int getDimension ()
 Returns the number of input dimensions.
 
String toString ()
 Returns a description of the partial variance estimator.
 

Protected Attributes

MonteCarloModelDoubleRQMC model
 
double approxMean
 
List< CoordinateSetcoordSets
 
CoordinateSet noCoordinate
 
CoordinateSet allCoordinates
 
double [] vars
 

Detailed Description

Partial variance estimator.

Estimates partial variances of a model with respect to multiple coordinate sets.

Reference: Monte Carlo estimators for small sensitivity indices I. M. Sobol' and E. E. Myshetskaya Monte Carlo Methods Appl. Vol. 13 No. 5-6 (2007), pp. 455-465

Member Function Documentation

◆ setModel()

void setModel ( MonteCarloModelDoubleRQMC  model,
double  approxMean 
)

Sets the model whose partial variances are to be estimated.

Best precision is achieved when the mean value of the model is close to approxMean.

◆ simulate()

void simulate ( RandomStream  stream)

Simulates the estimator once.

Returns vars such that:

  • vars.length is coordSets.size() + 2.
  • vars[nSets] contains the correction to the approximate mean.
  • vars[nSets+1] contains the square of the above correction.

Implements MonteCarloModel< E >.


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