plr

Data Structures
class	RobProb::plr
	Represents the normal-domain probabilities of the two outcomes of a single binary event as a ratio. More...
Functions
	RobProb::plr::plr (void)
	Default constructor.
	RobProb::plr::plr (double x)
	Constructor allowing implicit conversion from the double data type.
bool	RobProb::plr::plus_infinity (void) const
	Determines if the normal-domain probability ratio is equal to plus infinity.
void	RobProb::plr::set_to_plus_infinity (void)
	Sets the normal-domain probability ratio to plus infinity.
plr	RobProb::plr::operator+ (const plr &rhs) const
	Addition operator.
plr &	RobProb::plr::operator+= (const plr &rhs)
	Addition assignment operator.
plr	RobProb::plr::operator- (const plr &rhs) const
	Subtraction operator.
plr &	RobProb::plr::operator-= (const plr &rhs)
	Subtraction assignment operator.
plr	RobProb::plr::operator * (const plr &rhs) const
	Multiplication operator.
plr &	RobProb::plr::operator *= (const plr &rhs)
	Multiplication assignment operator.
plr	RobProb::plr::operator/ (const plr &rhs) const
	Division operator.
plr &	RobProb::plr::operator/= (const plr &rhs)
	Division assignment operator.
bool	RobProb::plr::operator== (const plr &rhs) const
	Test for equality operator.
bool	RobProb::plr::operator!= (const plr &rhs) const
	Test for non-equality operator.
bool	RobProb::plr::operator> (const plr &rhs) const
	Test for greater than operator.
bool	RobProb::plr::operator< (const plr &rhs) const
	Test for less than operator.
bool	RobProb::plr::operator>= (const plr &rhs) const
	Test for greater than or equal to operator.
bool	RobProb::plr::operator<= (const plr &rhs) const
	Test for less than or equal to operator.
plr	RobProb::to_plr (double x, double temp_infinity=infinity)
	Conversion from the double data type.
double	RobProb::to_double (const plr &x, double temp_infinity=infinity)
	Conversion to the double data type.
pvalues	RobProb::to_pvalues (const plr &x)
	Conversion to the pvalues data type.
bin	RobProb::hard (const plr &x)
	Hard decision.
llr	RobProb::log (const plr &x)
	Logarithm, allowing conversion to the llr data type.
plr	RobProb::add (const plr &lhs, const plr &rhs, JacobianType temp_jacobian_type=jacobian_type)
	Adds two normal-domain probability ratios, allowing the specification of how to correct the Jacobian approximation.
istream &	RobProb::operator>> (istream &in, plr &rhs)
	Input operator.
ostream &	RobProb::operator<< (ostream &out, const plr &rhs)
	Output operator.
double	RobProb::entropy (const plr &x)
	Calculates the entropy associated with a normal-domain probability ratio.
double	RobProb::mutual_information (const plr &x, const plr &source_plr)
	Calculates the mutual information associated with a normal-domain probability ratio for a non-equiprobable source.
double	RobProb::mutual_information (const plr &x)
	Calculates the mutual information associated with a normal-domain probability ratio for an equiprobable source.
plr	RobProb::generate_gaussian_plr (bin bit, double mutual_information)
	Generates a Gaussian-distributed normal-domain probability ratio with a given mutual information for a bit from an equiprobable source.
plr	RobProb::generate_bec_plr (bin bit, double mutual_information)
	Generates a BEC-distributed normal-domain probability ratio with a given mutual information for a bit from an equiprobable source.

---

Function Documentation

RobProb::plr::plr

(

void

)

[inherited]

Default constructor.

The plr variable is set to a normal-domain probability ratio of unity.

RobProb::plr::plr

(

double

x

)

[inherited]

Constructor allowing implicit conversion from the double data type.

Parameters:

x

The normal-domain probability ratio that the plr variable is set to. This must be greater than or equal to zero.

bool RobProb::plr::plus_infinity

(

void

)

const [inherited]

Determines if the normal-domain probability ratio is equal to plus infinity.

Returns:

The result of the test.

void RobProb::plr::set_to_plus_infinity

(

void

)

[inherited]

Sets the normal-domain probability ratio to plus infinity.

plr RobProb::plr::operator+

(

const plr &

rhs

)

const [inherited]

Addition operator.

Note that behind the scenes, addition in the normal-domain is achieved by applying the Jacobian operator in the logarithmic-domain. For this reason, the value of the global variable jacobian_type is used to specify how to correct the Jacobian approximation. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the addition.

plr& RobProb::plr::operator+=

(

const plr &

rhs

)

[inherited]

Addition assignment operator.

Note that behind the scenes, addition in the normal-domain is achieved by applying the Jacobian operator in the logarithmic-domain. For this reason, the value of the global variable jacobian_type is used to specify how to correct the Jacobian approximation. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The plr variable itself, which now stores the result of the addition.

plr RobProb::plr::operator-

(

const plr &

rhs

)

const [inherited]

Subtraction operator.

Note that behind the scenes, subtraction in the normal-domain is achieved by applying the Jacobian operator in the logarithmic-domain. In this case, the Jacobian approximation is always corrected. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio. This must have a value which is less than or equal to the left-hand-side operand.

Returns:

The result of the subtraction.

plr& RobProb::plr::operator-=

(

const plr &

rhs

)

[inherited]

Subtraction assignment operator.

Note that behind the scenes, subtraction in the normal-domain is achieved by applying the Jacobian operator in the logarithmic-domain. In this case, the Jacobian approximation is always corrected. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio. This must have a value which is less than or equal to the left-hand-side operand.

Returns:

The plr variable itself, which now stores the result of the subtraction.

plr RobProb::plr::operator *

(

const plr &

rhs

)

const [inherited]

Multiplication operator.

Note that behind the scenes, multiplication in the normal-domain is achieved by addition in the logarithmic-domain. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the multiplication.

plr& RobProb::plr::operator *=

(

const plr &

rhs

)

[inherited]

Multiplication assignment operator.

Note that behind the scenes, multiplication in the normal-domain is achieved by addition in the logarithmic-domain. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The plr variable itself, which now stores the result of the mutliplication.

plr RobProb::plr::operator/

(

const plr &

rhs

)

const [inherited]

Division operator.

Note that behind the scenes, division in the normal-domain is achieved by subtraction in the logarithmic-domain. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the division.

plr& RobProb::plr::operator/=

(

const plr &

rhs

)

[inherited]

Division assignment operator.

Note that behind the scenes, division in the normal-domain is achieved by subtraction in the logarithmic-domain. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The pvalue variable itself, which now stores the result of the division.

bool RobProb::plr::operator==

(

const plr &

rhs

)

const [inherited]

Test for equality operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

bool RobProb::plr::operator!=

(

const plr &

rhs

)

const [inherited]

Test for non-equality operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

bool RobProb::plr::operator>

(

const plr &

rhs

)

const [inherited]

Test for greater than operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

bool RobProb::plr::operator<

(

const plr &

rhs

)

const [inherited]

Test for less than operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

bool RobProb::plr::operator>=

(

const plr &

rhs

)

const [inherited]

Test for greater than or equal to operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

bool RobProb::plr::operator<=

(

const plr &

rhs

)

const [inherited]

Test for less than or equal to operator.

Parameters:

rhs

The right-hand-side normal-domain probability ratio.

Returns:

The result of the test.

plr RobProb::to_plr	(	double	x,
		double	temp_infinity = infinity
	)

Conversion from the double data type.

Parameters:

	x	The normal-domain probability ratio that the new plr variable is set to. This must be greater than or equal to zero.
	temp_infinity	Specifies the absolute value used to represent infinity when converting from the double type. If a value of 0.0 is specified, no clipping is employed. If not provided, the value of the global variable infinity is used by default.

Returns:

The new plr variable.

double RobProb::to_double	(	const plr &	x,
		double	temp_infinity = infinity
	)

Conversion to the double data type.

Parameters:

	x	The plr variable.
	temp_infinity	Specifies the absolute value used to represent infinity when converting to the double type. If a value of 0.0 is specified, no clipping is employed. In this case, an error is generated in the event of converting a positive or negative infinite value to the double data type. If not provided, the value of the global variable infinity is used by default.

Returns:

The normal-domain probability ratio of the plr variable.

pvalues RobProb::to_pvalues

(

const plr &

x

)

Conversion to the pvalues data type.

Parameters:

x

The normal-domain probability ratio.

Returns:

A two-entry normal-domain probabilities array, providing the unnormalized probabilties of the zero- and unity-valued event outcomes.

bin RobProb::hard

(

const plr &

x

)

Hard decision.

Parameters:

x

The normal-domain probability ratio.

Returns:

The index of the most likely outcome of the single binary event.

llr RobProb::log

(

const plr &

x

)

Logarithm, allowing conversion to the llr data type.

Note that behind the scenes, this conversion is achieved by a change of interface rather than a calculation.

Parameters:

x

The normal-domain probability ratio.

Returns:

, where is the normal-domain probability ratio.

plr RobProb::add	(	const plr &	lhs,
		const plr &	rhs,
		JacobianType	temp_jacobian_type = jacobian_type
	)

Adds two normal-domain probability ratios, allowing the specification of how to correct the Jacobian approximation.

This is required because, behind the scenes, addition in the normal-domain is achieved by applying the Jacobian operator in the logarithmic-domain. Also note that behind the scenes, normal-domain probability ratios of plus infinity are handled as a special case.

Parameters:

	lhs	The left-hand-side normal-domain probability ratio.
	rhs	The right-hand-side normal-domain probability ratio.
	temp_jacobian_type	Specifies how to correct the Jacobian approximation. If not provided, the value of the global variable jacobian_type is used by default.

Returns:

The result of the addition.

istream & RobProb::operator>>	(	istream &	in,
		plr &	rhs
	)

Input operator.

Parameters:

	in	The input stream containing the normal-domain probability ratio that the right-hand-side operand is set to. This must be greater than or equal to zero.
	rhs	The right-hand-side normal-domain probability ratio.

Returns:

The input stream itself.

ostream & RobProb::operator<<	(	ostream &	out,
		const plr &	rhs
	)

Output operator.

Parameters:

	out	The output stream to which the normal-domain probability ratio of the right-hand-side operand is written.
	rhs	The right-hand-side normal-domain probability ratio.

Returns:

The output stream itself.

double RobProb::entropy

(

const plr &

x

)

Calculates the entropy associated with a normal-domain probability ratio.

Preforms a conversion to the pvalues data type and calls the associated entropy function.

Parameters:

x

The normal-domain probability ratio.

Returns:

The result of the calculation.

double RobProb::mutual_information	(	const plr &	x,
		const plr &	source_plr
	)

Calculates the mutual information associated with a normal-domain probability ratio for a non-equiprobable source.

Note that truely APP decoders should be used to obtain accurate mutual information measurements.

Parameters:

	x	The normal-domain probability ratio.
	source_plr	The normal-domain source probability ratio.

Returns:

, where is the entropy of the normal-domain source probability ratio and is the entropy of the normal-domain probability ratio.

double RobProb::mutual_information

(

const plr &

x

)

Calculates the mutual information associated with a normal-domain probability ratio for an equiprobable source.

Note that truely APP decoders should be used to obtain accurate mutual information measurements.

Parameters:

x

The normal-domain probability ratio.

Returns:

, where is the entropy of the normal-domain probability ratio.

plr RobProb::generate_gaussian_plr	(	bin	bit,
		double	mutual_information
	)

Generates a Gaussian-distributed normal-domain probability ratio with a given mutual information for a bit from an equiprobable source.

Parameters:

	bit	The bit.
	mutual_information	The mutual information.

Returns:

The generated normal-domain probability ratio.

plr RobProb::generate_bec_plr	(	bin	bit,
		double	mutual_information
	)

Generates a BEC-distributed normal-domain probability ratio with a given mutual information for a bit from an equiprobable source.

Parameters:

	bit	The bit.
	mutual_information	The mutual information.

Returns:

The generated normal-domain probability ratio.

---