RandomNumber Class Reference

Parallel data structure for manipulating random numbers. More...

#include <random_number.h>

List of all members.

Public Member Functions

RandomNumber ()

RandomNumber (int num_samples)

RandomNumber (int num_samples, int num_ghosts)

~RandomNumber ()

void set_size (int num_samples)

Allocates space for a random number even distributed across the cores.

void set_ghost_size (int num_samples, int num_ghosts)

Allocates space for a ghost padded random number even distributed across the cores.

void make_uniform (int seed) const

Gives the random number a uniform distribution.

void make_normal (double mu, double sigma, int seed) const

Gives the random number a normal distribution.

double mean ()

Returns the sample mean and stores it for future use.

double variance ()

Returns the unbiased sample variance and stores it for future use.

double min ()

Returns the sample minimum and stores it for future use.

double max ()

Returns the sample maximum and stores it for future use.

void create_frequency_table (int num_bins, bool normalized)

Uses an STL vector array to store the frequency table.

void print_frequency_table (std::string file_name, int num_bins, Distribution &dist) const

double get_quantile (int alpha)

void view (void) const

Write the random number to the screen.

void print (std::string file_name) const

Public Attributes

std::vector< double > * frequency_table

ddata::Vector samples

Private Attributes

int num_samples

double * sample_min

double * sample_max

double * sample_mean

double * sample_variance

Detailed Description

Parallel data structure for manipulating random numbers.

Constructor & Destructor Documentation

RandomNumber::RandomNumber ( )

RandomNumber::RandomNumber ( int num_samples )

RandomNumber::RandomNumber	(	int	num_samples,
		int	num_ghosts
	)

RandomNumber::~RandomNumber ( )

Member Function Documentation

void RandomNumber::set_size ( int num_samples_ )

Allocates space for a random number even distributed across the cores.

Parameters:

num_samples

: The number of samples.

void RandomNumber::set_ghost_size	(	int	num_samples_,
		int	num_ghosts
	)

Allocates space for a ghost padded random number even distributed across the cores.

Parameters:

	num_samples	: The number of samples.
	num_ghosts	: The number of ghost points.

void RandomNumber::make_uniform ( int seed ) const

Gives the random number a uniform distribution.

Parameters:

seed

: True seeds the generator, false does not.

void RandomNumber::make_normal	(	double	mu,
		double	sigma,
		int	seed
	)			const

Gives the random number a normal distribution.

Parameters:

	mu	: The mean.
	sigma	: The square root of the variance.
	seed	: True seeds the generator, false does not.

This implementation uses the Box-Muller method. Given two independent random numbers $U_1~U(0,1)$

and

, the transformed variables $N_1=\sqrt{R}\sin{V}$ and $N_2=\sqrt{R}\cos{V}$ are independent with a distribution of $N(0,1)$

. Where
$R = -2\log{U_1}$
$V = 2 \pi U_2$

double RandomNumber::mean ( )

Returns the sample mean and stores it for future use.

double RandomNumber::variance ( )

Returns the unbiased sample variance and stores it for future use.

double RandomNumber::min ( )

Returns the sample minimum and stores it for future use.

double RandomNumber::max ( )

Returns the sample maximum and stores it for future use.

void RandomNumber::create_frequency_table	(	int	num_bins,
		bool	normalized
	)

Uses an STL vector array to store the frequency table.

Parameters:

	num_bins	: The number of bins.
	normalized	: If true the area under the frequency curve is 1.0. If false then the actual frequencies are computed.

The observed frequencies are stored in frequency_table[1] while the ranges are stored in frequency_table[0]. The frequency table only covers the sample min and sample max.
frequency_table[0][i]= $X_i$

frequency_table[1][i]= $F_i=\sum_j^n I(X_i \le x_j < X_{i+1})$

void RandomNumber::print_frequency_table	(	std::string	file_name,
		int	num_bins,
		Distribution &	dist
	)			const

double RandomNumber::get_quantile ( int alpha )

void RandomNumber::view ( void ) const

Write the random number to the screen.

void RandomNumber::print ( std::string file_name ) const

Member Data Documentation

int RandomNumber::num_samples [private]

double* RandomNumber::sample_min [private]

double* RandomNumber::sample_max [private]

double* RandomNumber::sample_mean [private]

double* RandomNumber::sample_variance [private]

std::vector<double>* RandomNumber::frequency_table

ddata::Vector RandomNumber::samples

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

include/random_number.h
src/Statistics/random_number.cpp


Public Member Functions
	RandomNumber ()
	RandomNumber (int num_samples)
	RandomNumber (int num_samples, int num_ghosts)
	~RandomNumber ()
void	set_size (int num_samples)
	Allocates space for a random number even distributed across the cores.
void	set_ghost_size (int num_samples, int num_ghosts)
	Allocates space for a ghost padded random number even distributed across the cores.
void	make_uniform (int seed) const
	Gives the random number a uniform distribution.
void	make_normal (double mu, double sigma, int seed) const
	Gives the random number a normal distribution.
double	mean ()
	Returns the sample mean and stores it for future use.
double	variance ()
	Returns the unbiased sample variance and stores it for future use.
double	min ()
	Returns the sample minimum and stores it for future use.
double	max ()
	Returns the sample maximum and stores it for future use.
void	create_frequency_table (int num_bins, bool normalized)
	Uses an STL vector array to store the frequency table.
void	print_frequency_table (std::string file_name, int num_bins, Distribution &dist) const
double	get_quantile (int alpha)
void	view (void) const
	Write the random number to the screen.
void	print (std::string file_name) const
Public Attributes
std::vector< double > *	frequency_table
ddata::Vector	samples
Private Attributes
int	num_samples
double *	sample_min
double *	sample_max
double *	sample_mean
double *	sample_variance