Specialization: fxd::fixed is similar to floating-point. More...

#include <limits.hpp>

Public Types
using	float_type = fxd::detail::select_float_t< digits >
	Extension: lossless conversion to a floating-point type.

Static Public Member Functions
static constexpr fxd::fixed< Int, Frac, Raw >	min () noexcept
	Smallest positive value; same semantics as floating-point.

static constexpr fxd::fixed< Int, Frac, Raw >	lowest () noexcept
	Closest value to -∞.

static constexpr fxd::fixed< Int, Frac, Raw >	max () noexcept
	Maximum value.

static constexpr fxd::fixed< Int, Frac, Raw >	epsilon () noexcept
	Smallest increment.

static constexpr fxd::fixed< Int, Frac, Raw >	round_error () noexcept
	Maximum possible error.

static constexpr fxd::fixed< Int, Frac, Raw >	infinity () noexcept
	Not meaningful, `has_infinity` is false.

static constexpr fxd::fixed< Int, Frac, Raw >	quiet_NaN () noexcept
	Not meaningful, `has_quiet_NaN` is false.

static constexpr fxd::fixed< Int, Frac, Raw >	signaling_NaN () noexcept
	Not meaningful, `has_signaling_NaN` is false.

static constexpr fxd::fixed< Int, Frac, Raw >	denorm_min () noexcept
	Same as `min()` since there are no denormalized values.

static constexpr fxd::fixed< Int, Frac, Raw >	max_pow2 () noexcept
	Extension: maximum power of two.

Static Public Attributes
static constexpr bool	is_specialized = true
	This is a valid specialization.

static constexpr bool	is_signed = numeric_limits<Raw>::is_signed
	Sign depends on the underlying integer type.

static constexpr bool	is_integer = Frac <= 0
	If this represents integers only.

static constexpr bool	is_exact = true
	All arithmetic is done with integers, so `fxd::fixed` is exact.

static constexpr bool	has_infinity = false
	No representation for ∞ .

static constexpr bool	has_quiet_NaN = false
	No NaN.

static constexpr bool	has_signaling_NaN = false
	No NaN.

static constexpr float_denorm_style	has_denorm = denorm_absent
	No denormals.

static constexpr bool	has_denorm_loss = false
	No denormals.

static constexpr float_round_style	round_style = round_toward_zero
	Default rounding is towards zero, just like with integers.

static constexpr bool	is_iec559 = false
	Not a standard floating point type.

static constexpr bool	is_bounded = numeric_limits<Raw>::is_bounded
	It's bounded by the underlying integer type.

static constexpr bool	is_modulo = numeric_limits<Raw>::is_modulo
	It's modulo if the underlying integer is modulo.

static constexpr int	radix = numeric_limits<Raw>::radix
	Same radix as the underlying integer.

static constexpr int	digits
	`fxd::fixed::bits` - `is_signed`.

static constexpr int	digits10
	Same estimation used for floating-point.

static constexpr int	max_digits10
	Same estimation used for floating-point.

static constexpr int	min_exponent = 1 - Frac
	Minimum power-of-two that can be stored.

static constexpr int	min_exponent10
	Same estimation used for floating-point.

static constexpr int	max_exponent = Int - is_signed
	Maximum power-of-two that can be stored.

static constexpr int	max_exponent10
	Same estimation used for floating-point.

static constexpr bool	traps = numeric_limits<Raw>::traps
	Depends on the underlying integer.

static constexpr bool	tinyness_before = false
	Not meaningful.

static constexpr int	max_bit = Int - is_signed
	Extension: maximum logical bit.

static constexpr int	min_bit = - Frac
	Extension: minimum logical bit.

Detailed Description

template<int Int, int Frac, typename Raw>
requires (numeric_limits<Raw>::is_specialized)
struct std::numeric_limits< fxd::fixed< Int, Frac, Raw > >

Specialization: fxd::fixed is similar to floating-point.

Definition at line 30 of file limits.hpp.

Member Typedef Documentation

◆ float_type

using float_type = fxd::detail::select_float_t<digits>

Extension: lossless conversion to a floating-point type.

Definition at line 111 of file limits.hpp.

Member Function Documentation

◆ denorm_min()

static constexpr fxd::fixed< Int, Frac, Raw > denorm_min ( )

inlinestaticconstexprnoexcept

Same as min() since there are no denormalized values.

Definition at line 206 of file limits.hpp.

◆ epsilon()

static constexpr fxd::fixed< Int, Frac, Raw > epsilon ( )

inlinestaticconstexprnoexcept

Smallest increment.

Definition at line 156 of file limits.hpp.

◆ infinity()

static constexpr fxd::fixed< Int, Frac, Raw > infinity ( )

inlinestaticconstexprnoexcept

Not meaningful, has_infinity is false.

Definition at line 176 of file limits.hpp.

◆ lowest()

static constexpr fxd::fixed< Int, Frac, Raw > lowest ( )

inlinestaticconstexprnoexcept

Closest value to -∞.

Definition at line 127 of file limits.hpp.

◆ max()

static constexpr fxd::fixed< Int, Frac, Raw > max ( )

inlinestaticconstexprnoexcept

Maximum value.

Definition at line 141 of file limits.hpp.

◆ max_pow2()

static constexpr fxd::fixed< Int, Frac, Raw > max_pow2 ( )

inlinestaticconstexprnoexcept

Extension: maximum power of two.

Definition at line 216 of file limits.hpp.

◆ min()

static constexpr fxd::fixed< Int, Frac, Raw > min ( )

inlinestaticconstexprnoexcept

Smallest positive value; same semantics as floating-point.

Definition at line 117 of file limits.hpp.

◆ quiet_NaN()

static constexpr fxd::fixed< Int, Frac, Raw > quiet_NaN ( )

inlinestaticconstexprnoexcept

Not meaningful, has_quiet_NaN is false.

Definition at line 186 of file limits.hpp.

◆ round_error()

static constexpr fxd::fixed< Int, Frac, Raw > round_error ( )

inlinestaticconstexprnoexcept

Maximum possible error.

Definition at line 166 of file limits.hpp.

◆ signaling_NaN()

static constexpr fxd::fixed< Int, Frac, Raw > signaling_NaN ( )

inlinestaticconstexprnoexcept

Not meaningful, has_signaling_NaN is false.

Definition at line 196 of file limits.hpp.

Member Data Documentation

◆ digits

constexpr int digits

staticconstexpr

fxd::fixed::bits - is_signed.

Definition at line 78 of file limits.hpp.

◆ digits10

constexpr int digits10

staticconstexpr

Same estimation used for floating-point.

Definition at line 81 of file limits.hpp.

◆ has_denorm

constexpr float_denorm_style has_denorm = denorm_absent

staticconstexpr

No denormals.

Definition at line 57 of file limits.hpp.

◆ has_denorm_loss

constexpr bool has_denorm_loss = false

staticconstexpr

No denormals.

Definition at line 60 of file limits.hpp.

◆ has_infinity

constexpr bool has_infinity = false

staticconstexpr

No representation for ∞ .

Definition at line 48 of file limits.hpp.

◆ has_quiet_NaN

constexpr bool has_quiet_NaN = false

staticconstexpr

No NaN.

Definition at line 51 of file limits.hpp.

◆ has_signaling_NaN

constexpr bool has_signaling_NaN = false

staticconstexpr

No NaN.

Definition at line 54 of file limits.hpp.

◆ is_bounded

constexpr bool is_bounded = numeric_limits<Raw>::is_bounded

staticconstexpr

It's bounded by the underlying integer type.

Definition at line 69 of file limits.hpp.

◆ is_exact

constexpr bool is_exact = true

staticconstexpr

All arithmetic is done with integers, so fxd::fixed is exact.

Definition at line 45 of file limits.hpp.

◆ is_iec559

constexpr bool is_iec559 = false

staticconstexpr

Not a standard floating point type.

Definition at line 66 of file limits.hpp.

◆ is_integer

constexpr bool is_integer = Frac <= 0

staticconstexpr

If this represents integers only.

Definition at line 42 of file limits.hpp.

◆ is_modulo

constexpr bool is_modulo = numeric_limits<Raw>::is_modulo

staticconstexpr

It's modulo if the underlying integer is modulo.

Definition at line 72 of file limits.hpp.

◆ is_signed

constexpr bool is_signed = numeric_limits<Raw>::is_signed

staticconstexpr

Sign depends on the underlying integer type.

Definition at line 39 of file limits.hpp.

◆ is_specialized

constexpr bool is_specialized = true

staticconstexpr

This is a valid specialization.

Definition at line 36 of file limits.hpp.

◆ max_bit

constexpr int max_bit = Int - is_signed

staticconstexpr

Extension: maximum logical bit.

Definition at line 105 of file limits.hpp.

◆ max_digits10

constexpr int max_digits10

staticconstexpr

Same estimation used for floating-point.

Definition at line 84 of file limits.hpp.

◆ max_exponent

constexpr int max_exponent = Int - is_signed

staticconstexpr

Maximum power-of-two that can be stored.

Definition at line 93 of file limits.hpp.

◆ max_exponent10

constexpr int max_exponent10

staticconstexpr

Same estimation used for floating-point.

Definition at line 96 of file limits.hpp.

◆ min_bit

constexpr int min_bit = - Frac

staticconstexpr

Extension: minimum logical bit.

Definition at line 108 of file limits.hpp.

◆ min_exponent

constexpr int min_exponent = 1 - Frac

staticconstexpr

Minimum power-of-two that can be stored.

Definition at line 87 of file limits.hpp.

◆ min_exponent10

constexpr int min_exponent10

staticconstexpr

Same estimation used for floating-point.

Definition at line 90 of file limits.hpp.

◆ radix

constexpr int radix = numeric_limits<Raw>::radix

staticconstexpr

Same radix as the underlying integer.

Definition at line 75 of file limits.hpp.

◆ round_style

constexpr float_round_style round_style = round_toward_zero

staticconstexpr

Default rounding is towards zero, just like with integers.

Definition at line 63 of file limits.hpp.

◆ tinyness_before

constexpr bool tinyness_before = false

staticconstexpr

Not meaningful.

Definition at line 102 of file limits.hpp.

◆ traps

constexpr bool traps = numeric_limits<Raw>::traps

staticconstexpr

Depends on the underlying integer.

Definition at line 99 of file limits.hpp.

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

limits.hpp

Public Types

Static Public Member Functions

Static Public Attributes

Detailed Description

Member Typedef Documentation

◆ float_type

Member Function Documentation

◆ denorm_min()

◆ epsilon()

◆ infinity()

◆ lowest()

◆ max()

◆ max_pow2()

◆ min()

◆ quiet_NaN()

◆ round_error()

◆ signaling_NaN()

Member Data Documentation

◆ digits

◆ digits10

◆ has_denorm

◆ has_denorm_loss

◆ has_infinity

◆ has_quiet_NaN

◆ has_signaling_NaN

◆ is_bounded

◆ is_exact

◆ is_iec559

◆ is_integer

◆ is_modulo

◆ is_signed

◆ is_specialized

◆ max_bit

◆ max_digits10

◆ max_exponent

◆ max_exponent10

◆ min_bit

◆ min_exponent

◆ min_exponent10

◆ radix

◆ round_style

◆ tinyness_before

◆ traps