Gambit::Cholesky Class Reference

#include <cholesky.hpp>

Collaboration diagram for Gambit::Cholesky:

Public Member Functions
	Cholesky ()
	Cholesky (const int num)
bool	EnterMat (std::vector< std::vector< double >> &a)
void	ElMult (std::vector< double > &y) const
std::vector< double >	invElMult (const std::vector< double > &y) const
	x = L^-1 y where L is the lower-diagonal Cholesky matrix More...
double	Square (const std::vector< double > &y, const std::vector< double > &y0)
double	DetSqrt ()

Private Attributes
std::vector< std::vector< double > >	el

Detailed Description

Definition at line 26 of file cholesky.hpp.

Constructor & Destructor Documentation

Cholesky() [1/2]

Gambit::Cholesky::Cholesky ( )

inline

Definition at line 32 of file cholesky.hpp.

{}

Cholesky() [2/2]

Gambit::Cholesky::Cholesky ( const int  num )

inline

Definition at line 34 of file cholesky.hpp.

: el(num, std::vector<double>(num)) {}

Member Function Documentation

DetSqrt()

double Gambit::Cholesky::DetSqrt ( )

inline

Definition at line 127 of file cholesky.hpp.

Referenced by Gambit::Priors::Cauchy::operator()(), Gambit::Priors::Gaussian::operator()(), and Gambit::Priors::LogNormal::operator()().

        {
            double temp = 1.0;
            int num = el.size();
            for (int i = 0; i < num; i++)
                temp *= el[i][i];
            return temp;
        }

Here is the caller graph for this function:

ElMult()

void Gambit::Cholesky::ElMult ( std::vector< double > &  y ) const

inline

Definition at line 69 of file cholesky.hpp.

References b.

Referenced by Gambit::Priors::Cauchy::transform(), Gambit::Priors::Gaussian::transform(), and Gambit::Priors::LogNormal::transform().

        {
            int i, j;
            int num = el.size();
            std::vector<double> b(num);
            for(i = 0; i < num; i++)
            {
                b[i] = 0.0;
                for (j = 0; j <= i; j++)
                {
                    b[i] += el[i][j]*y[j];
                }
            }
            
            y = b;
        }

Here is the caller graph for this function:

EnterMat()

bool Gambit::Cholesky::EnterMat ( std::vector< std::vector< double >> &  a )

inline

Definition at line 36 of file cholesky.hpp.

Referenced by Gambit::Priors::Cauchy::Cauchy(), Gambit::Priors::Gaussian::Gaussian(), Gambit::Priors::LogNormal::LogNormal(), and objective_plugin().

        {
            el = a;
            int num = el.size();
            double sum = 0;
            int i, j, k;
            
            for (i = 0; i < num; i++)
            {
                for (j = i; j < num; j++)
                {
                    for(sum = el[i][j], k = i - 1; k >= 0; k--)
                        sum -= el[i][k]*el[j][k];
                    if(i == j)
                    {
                        if(sum <= 0.0)
                        {
                                return false;
                        }
                        el[i][i] = std::sqrt(sum);
                    }
                    else
                        el[j][i] = sum/el[i][i];
                }
            }
            
            for (i = 0; i < num; i++)
                for (j = 0; j < i; j++)
                    el[j][i] = 0.0;
                    
            return true;
        }

Here is the caller graph for this function:

invElMult()

std::vector<double> Gambit::Cholesky::invElMult ( const std::vector< double > &  y ) const

inline

x = L^-1 y where L is the lower-diagonal Cholesky matrix

Found by forward substituion since L is lower-diagonal.

Definition at line 91 of file cholesky.hpp.

References generate_raster_scan_settings::n.

Referenced by Gambit::Priors::Cauchy::inverse_transform(), Gambit::Priors::Gaussian::inverse_transform(), and Gambit::Priors::LogNormal::inverse_transform().

        {
            std::vector<double> x(y.size(), 0.);
            for (int i = 0, n = x.size(); i < n; i++)
            {
                double sum_ = 0;
                for (int j = 0; j < i; j++)
                {
                    sum_ += el[i][j] * x[j];
                }
                x[i] = (y[i] - sum_) / el[i][i];
            }
            return x;
        }

Here is the caller graph for this function:

Square()

double Gambit::Cholesky::Square ( const std::vector< double > &  y, const std::vector< double > &  y0  )

inline

Definition at line 106 of file cholesky.hpp.

Referenced by objective_plugin(), Gambit::Priors::Cauchy::operator()(), Gambit::Priors::Gaussian::operator()(), and Gambit::Priors::LogNormal::operator()().

        {
            int i, j, num = y.size();
            double sum;
            std::vector<double> x(num);
            
            for (i = 0; i < num; i++)
            {
                for (sum = (y[i]-y0[i]), j=0; j < i; j++)
                    sum -= el[i][j]*x[j];
                
                x[i]=sum/el[i][i];
            }
            
            sum = 0.0;
            for (i = 0; i < num; i++)
                sum += x[i]*x[i];
            
            return sum;
        }

Here is the caller graph for this function:

Member Data Documentation

el

std::vector<std::vector<double> > Gambit::Cholesky::el

private

Definition at line 29 of file cholesky.hpp.

---

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

• ScannerBit/include/gambit/ScannerBit/cholesky.hpp