ConditionalGaussGenerator Class Reference

Class implementing n-dimensional random number generator from Gaussian distribution where the first component of the vector is generated first (using some external source) and the remaining components are generated based on the value of the first component. More...

#include <ConditionalGaussGenerator.h>

Collaboration diagram for ConditionalGaussGenerator:

Public Member Functions
	ConditionalGaussGenerator ()
	dummy constructor without arguments
	ConditionalGaussGenerator (const Eigen::VectorXd &mu, const Eigen::MatrixXd &covMat)
	constructor which takes vector of central values and covariance matrix as input
Eigen::VectorXd	generate (double x0) const
	generate random vector based on the provided first component x0 More...
double	getX0spread () const
	get the spread of first component which can be used by external generator
Eigen::VectorXd	getMu () const
	get the vector including central values of the distribution
Eigen::MatrixXd	getCovMat () const
	get the covariance matrix describing n-dimensional Gaussian distribution

Private Attributes
Eigen::VectorXd	m_mu
	central values of the distribution
Eigen::MatrixXd	m_covMat
	covariance matrix of the distribution
Eigen::MatrixXd	m_vBaseMat
	transformation matrix between eigen-system of m_covMat and nominal system
std::vector< Eigen::VectorXd >	m_cOrt
	array of vectors describing free degrees of freedom of random generator
Eigen::VectorXd	m_v0norm
	vector which scales with dx0

Detailed Description

Class implementing n-dimensional random number generator from Gaussian distribution where the first component of the vector is generated first (using some external source) and the remaining components are generated based on the value of the first component.

The obtained numbers have identical statistical behaviour like when all components are generated simultaneously

Definition at line 25 of file ConditionalGaussGenerator.h.

Member Function Documentation

generate()

VectorXd generate

(

double

x0

)

const

generate random vector based on the provided first component x0

Parameters

x0	the first component of the random vector

Returns

the vector of the random numbers, it starts with x0

Definition at line 99 of file ConditionalGaussGenerator.cc.

{
  double dx0 = x0 - m_mu[0];
 
  // in case of zero cov matrix
  if (m_vBaseMat.cols() == 0)
    return m_mu;
 
  VectorXd x = m_v0norm * dx0;
 
  for (const VectorXd& co : m_cOrt) {
    double rr = gRandom->Gaus();
    x += rr * co;
  }
 
  VectorXd vec = m_vBaseMat * x;
 
  return (m_mu + vec);
}

---

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

• framework/utilities/include/ConditionalGaussGenerator.h
• framework/utilities/src/ConditionalGaussGenerator.cc