Statistics of two-bit quantization and artifact correction parameters. More...

#include <JenetAnderson98.h>

Inheritance diagram for JenetAnderson98:

Classes
class	EquationA5
class	Probability
	Interface to the measured distributions of Phi. More...

Public Member Functions
	JenetAnderson98 ()
	Default constructor.
void	set_Phi (double Phi)
	Set the fraction of samples in the low voltage state.
void	set_sigma_n (double sigma_n)
	Set the normalized, undigitized power.
double	get_hi () const
	Get the optimal high voltage output levels, JA98 Eq.40.
double	get_lo () const
	Get the optimal low voltage output levels, JA98 Eq.41.
double	get_A () const
	Get the slope of digitized vs undigitized correlation, JA98 Eq.43.
void	set_threshold (double t=get_optimal_spacing(2))
	Set the sampling threshold.
double	get_threshold () const
	Get the sampling threshold.
double	get_mean_Phi () const
	Get the expectation value of Phi, JA98 Eq.A2.
void	set_mean_Phi (double mean_Phi)
	Set the expectation value of Phi.
double	get_var_Phi () const
	Get the variance of Phi.
void	get_prob_Phi (unsigned L, std::vector< float > &prob_Phi)
	Get the theoretical probability distribution of Phi, JA98 Eq.A6.
double	A4 (double Phi)
	Given Phi, return the digitized power.
double	invert_A4 (double sigma_hat)
	Given the digitized power, return Phi.
double	A14 (double mean_Phi)
	Given <Phi>, return the digitized power.
double	invert_A14 (double sigma_hat)
	Given the digitized power, return <Phi>
void	set_measured_prob_Phi (const Probability *data)
	Set the measured probability distribution of Phi.
const Probability *	get_measured_prob_Phi () const
	Set the measured probability distribution of Phi.
Public Member Functions inherited from Reference::Able
	Able ()
	Default constructor.
	Able (const Able &)
	Copy constructor.
Able &	operator= (const Able &)
	Assignment operator.
virtual	~Able ()
	Destructor.
unsigned	get_reference_count () const
	Returns the number of references there are to this.
Public Member Functions inherited from Reference::HeapTracked
	HeapTracked ()
	Default constructor.
	HeapTracked (const HeapTracked &)
	Copy constructor.
HeapTracked &	operator= (const HeapTracked &)
	Assignment operator.
virtual	~HeapTracked ()
	Destructor.
bool	__is_on_heap () const
	Return true if this instance is found in heap addresses.

Static Public Member Functions
static double	get_optimal_spacing (unsigned bits)
	Table 3 - Optimum Input Threshold Spacing for a Uniform Digitizer.
Static Public Member Functions inherited from Reference::Able
static size_t	get_instance_count ()
	Returns the current number instances in existence.
Static Public Member Functions inherited from Reference::HeapTracked
static void *	operator new (size_t size, void *ptr=0)
	Records the addresses of new instances allocated on the heap.
static void	operator delete (void location, void ptr)
	Placement delete overload to match the placement new overload.
static void	operator delete (void *location)
	Regular delete overload.
static size_t	get_heap_queue_size ()
	Return the number of heap addresses awaiting processing.

Protected Member Functions
void	set_alpha (double alpha, double Phi)
	Set the inverse width of the gaussian, sort of.
Protected Member Functions inherited from Reference::Able
Handle *	__reference (bool active) const
	Add the address of a reference to this object.
void	__dereference (bool auto_delete=true) const
	Remove the address of a reference to this object.

Protected Attributes
double	threshold
	The sampling threshold.
Reference::To< const Probability >	measured
	The interface to a measured probability distribution of Phi.
double	lo
double	hi
double	A
double	mean_Phi
double	var_Phi

Detailed Description

Statistics of two-bit quantization and artifact correction parameters.

Given either the fraction of digitized samples in the low voltage state, $\Phi$ , or the instantaneous power normalized by the mean power, $\sigma_n=\sigma/\langle\sigma\rangle$ , this class computes

the output levels that minimize signal distortion (dynamic level setting), and
the fractional quantization noise (scattered power correction).

Parameters

Phi fraction of low voltage state samples, or

sigma_n normalized power

Return values

lo the optimal low voltage output level

hi the optimal hi voltage output level

A the fractional quantization noise

Furthermore, given the sampling threshold, , this class computes the mean, variance, and probability distribution of $\Phi$ .

Member Function Documentation

◆ get_optimal_spacing()

double JenetAnderson98::get_optimal_spacing ( unsigned bits )

static

Table 3 - Optimum Input Threshold Spacing for a Uniform Digitizer.

Beyond 8 bits per sample, optimal spacing is extrapolated using an exponential.

Referenced by JenetAnderson98().

◆ get_prob_Phi()

void JenetAnderson98::get_prob_Phi	(	unsigned	L,
		std::vector< float > &	prob_Phi )

Get the theoretical probability distribution of Phi, JA98 Eq.A6.

Parameters

L	is the number of points used to measure Phi

Return values

prob_Phi[i] will contain the probability of Phi = i/L

References get_mean_Phi().

◆ invert_A14()

double JenetAnderson98::invert_A14 ( double sigma_hat )

Given the digitized power, return <Phi>

Uses the van Wijngaarden-Dekker-Brent method to invert Eq.A14 of JA98

References A14().

◆ invert_A4()

double JenetAnderson98::invert_A4 ( double sigma_hat )

Given the digitized power, return Phi.

Uses the van Wijngaarden-Dekker-Brent method to invert Eq.A4 of JA98

References A4().

◆ set_alpha()

void JenetAnderson98::set_alpha	(	double	alpha,
		double	Phi )

protected

Set the inverse width of the gaussian, sort of.

$\alpha= t/(\sqrt{2}\sigma)$ is the term in brackets on the right-hand side of Eq.45 of JA98

Referenced by set_Phi(), and set_sigma_n().

◆ set_mean_Phi()

void JenetAnderson98::set_mean_Phi ( double Phi )

Set the expectation value of Phi.

Given the mean fraction of samples that lay within the thresholds, -t and +t, this method computes the threshold, where threshold = t/sigma.

References threshold.

Referenced by A14().

◆ set_Phi()

void JenetAnderson98::set_Phi ( double Phi )

Set the fraction of samples in the low voltage state.

This method inverts Eq.45 of JA98.

Phi is the left-hand side of Eq.44, the fraction of samples between x2 and x4; alpha is the inverse error function of Phi.

References set_alpha().

Referenced by A4(), and JenetAnderson98::EquationA5::set_nsamp().

◆ set_sigma_n()

void JenetAnderson98::set_sigma_n ( double sigma_n )

Set the normalized, undigitized power.

$\sigma_n = \sigma/\langle\sigma\rangle$ and, as defined in Eq.38, $t = \langle\sigma\rangle s$ , where s is the spacing or optimal threshold.

References set_alpha(), and threshold.

◆ set_threshold()

void JenetAnderson98::set_threshold ( double t = get_optimal_spacing(2) )

Set the sampling threshold.

Given the sampling threshold, t, this method computes the mean fraction of samples that lay within the thresholds, -t and +t, where t = threshold * sigma. This is equivalent to:

Eq.45 (with x4 = -x2 = t [Eq.38]);
Eq.A2 (with xh = -xl = t); or
the expectation value of the binomial distribution, Eq.A6 (cf. http://mathworld.wolfram.com/BinomialDistribution.html, Eqn. 12)

The variance of Phi is the variance of the binomial distribution, Eq.A6. (cf. http://mathworld.wolfram.com/BinomialDistribution.html, Eqn. 17)

References threshold.

Referenced by JenetAnderson98().

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

JenetAnderson98.h
JenetAnderson98.C

Classes

Public Member Functions

Static Public Member Functions

Protected Member Functions

Protected Attributes