Qrack
9.13
General classical-emulating-quantum development framework
|
Numeric limits for half-precision floats. More...
#include <half.hpp>
Static Public Member Functions | |
static HALF_CONSTEXPR half_float::half | min () HALF_NOTHROW |
Smallest positive normal value. More... | |
static HALF_CONSTEXPR half_float::half | lowest () HALF_NOTHROW |
Smallest finite value. More... | |
static HALF_CONSTEXPR half_float::half | max () HALF_NOTHROW |
Largest finite value. More... | |
static HALF_CONSTEXPR half_float::half | epsilon () HALF_NOTHROW |
Difference between 1 and next representable value. More... | |
static HALF_CONSTEXPR half_float::half | round_error () HALF_NOTHROW |
Maximum rounding error in ULP (units in the last place). More... | |
static HALF_CONSTEXPR half_float::half | infinity () HALF_NOTHROW |
Positive infinity. More... | |
static HALF_CONSTEXPR half_float::half | quiet_NaN () HALF_NOTHROW |
Quiet NaN. More... | |
static HALF_CONSTEXPR half_float::half | signaling_NaN () HALF_NOTHROW |
Signaling NaN. More... | |
static HALF_CONSTEXPR half_float::half | denorm_min () HALF_NOTHROW |
Smallest positive subnormal value. More... | |
Static Public Attributes | |
static HALF_CONSTEXPR_CONST bool | is_specialized = true |
Is template specialization. More... | |
static HALF_CONSTEXPR_CONST bool | is_signed = true |
Supports signed values. More... | |
static HALF_CONSTEXPR_CONST bool | is_integer = false |
Is not an integer type. More... | |
static HALF_CONSTEXPR_CONST bool | is_exact = false |
Is not exact. More... | |
static HALF_CONSTEXPR_CONST bool | is_modulo = false |
Doesn't provide modulo arithmetic. More... | |
static HALF_CONSTEXPR_CONST bool | is_bounded = true |
Has a finite set of values. More... | |
static HALF_CONSTEXPR_CONST bool | is_iec559 = true |
IEEE conformant. More... | |
static HALF_CONSTEXPR_CONST bool | has_infinity = true |
Supports infinity. More... | |
static HALF_CONSTEXPR_CONST bool | has_quiet_NaN = true |
Supports quiet NaNs. More... | |
static HALF_CONSTEXPR_CONST bool | has_signaling_NaN = true |
Supports signaling NaNs. More... | |
static HALF_CONSTEXPR_CONST float_denorm_style | has_denorm = denorm_present |
Supports subnormal values. More... | |
static HALF_CONSTEXPR_CONST bool | has_denorm_loss = false |
Supports no denormalization detection. More... | |
static HALF_CONSTEXPR_CONST bool | traps = false |
Traps only if HALF_ERRHANDLING_THROW_... is acitvated. More... | |
static HALF_CONSTEXPR_CONST bool | tinyness_before = false |
Does not support no pre-rounding underflow detection. More... | |
static HALF_CONSTEXPR_CONST float_round_style | round_style = half_float::half::round_style |
Rounding mode. More... | |
static HALF_CONSTEXPR_CONST int | digits = 11 |
Significant digits. More... | |
static HALF_CONSTEXPR_CONST int | digits10 = 3 |
Significant decimal digits. More... | |
static HALF_CONSTEXPR_CONST int | max_digits10 = 5 |
Required decimal digits to represent all possible values. More... | |
static HALF_CONSTEXPR_CONST int | radix = 2 |
Number base. More... | |
static HALF_CONSTEXPR_CONST int | min_exponent = -13 |
One more than smallest exponent. More... | |
static HALF_CONSTEXPR_CONST int | min_exponent10 = -4 |
Smallest normalized representable power of 10. More... | |
static HALF_CONSTEXPR_CONST int | max_exponent = 16 |
One more than largest exponent. More... | |
static HALF_CONSTEXPR_CONST int | max_exponent10 = 4 |
Largest finitely representable power of 10. More... | |
Numeric limits for half-precision floats.
See also: Documentation for std::numeric_limits
|
inlinestatic |
Smallest positive subnormal value.
|
inlinestatic |
Difference between 1 and next representable value.
|
inlinestatic |
Positive infinity.
|
inlinestatic |
Smallest finite value.
|
inlinestatic |
Largest finite value.
|
inlinestatic |
Smallest positive normal value.
|
inlinestatic |
Quiet NaN.
|
inlinestatic |
Maximum rounding error in ULP (units in the last place).
|
inlinestatic |
Signaling NaN.
|
static |
Significant digits.
|
static |
Significant decimal digits.
|
static |
Supports subnormal values.
|
static |
Supports no denormalization detection.
|
static |
Supports infinity.
|
static |
Supports quiet NaNs.
|
static |
Supports signaling NaNs.
|
static |
Has a finite set of values.
|
static |
Is not exact.
|
static |
IEEE conformant.
|
static |
Is not an integer type.
|
static |
Doesn't provide modulo arithmetic.
|
static |
Supports signed values.
|
static |
Is template specialization.
|
static |
Required decimal digits to represent all possible values.
|
static |
One more than largest exponent.
|
static |
Largest finitely representable power of 10.
|
static |
One more than smallest exponent.
|
static |
Smallest normalized representable power of 10.
|
static |
Number base.
|
static |
Rounding mode.
|
static |
Does not support no pre-rounding underflow detection.
|
static |
Traps only if HALF_ERRHANDLING_THROW_... is acitvated.