numeric.float32 ≡
inline bool isnan(float32 x) §source
Check x is not a number.
inline bool isinf(float32 x) §source
Check x is positive or negative infinity.
inline bool isfinite(float32 x) §source
inline bool isnormal(float32 x) §source
inline bool signbit(float32 x) §source
Return sign of x.
template <Denorm denorm_mode> inline float32 add(float32 in1, float32 in2) §source
Addition.
template <Denorm denorm_mode> inline float32 mul(float32 in1, float32 in2) §source
Multiplication.
template <Denorm denorm_mode> inline bool eq(float32 in1, float32 in2) §source
Compare two float32 values for equality.
template <Denorm denorm_mode> inline bool lt(float32 in1, float32 in2) §source
Check in1 less than in2.
template <Denorm denorm_mode> inline bool gt(float32 in1, float32 in2) §source
Check in1 greater than in2.
template <Denorm denorm_mode> inline float32 sub(float32 in1, float32 in2) §source
Subtract in2 from in1.
inline float32 rcp(float32 x) §source
Reciprocal 1 / x.
inline float32 exp2(float32 x) §source
2x
inline float32 exp(float32 x) §source
ex
inline float32 expm1(float32 x) §source
Calculate ex − 1. Denorm
inputs and outputs are supported. For smaller values of x,
expm1(x) may be more accurate than
exp(x)-1.
inline float32 log2(float32 x) §source
log2(x)
inline float32 log(float32 x) §source
ln(x)
inline float32 log10(float32 x) §source
log10(x)
inline float32 log1p(float32 value) §source
Calculate ln(x + 1). Denorm
inputs and outputs are supported. For small values of x,
log1p may be more accurate than log(1+x).
inline float32 logb(float32 x) §source
Calculate the floating-point value of the unbiased exponent part of
log(x). If x is a subnormal value, the result
of log2(mantissa part of x)-126 is returned.
inline int32 ilogb(float32 x) §source
Return the signed integral value of the unbiased exponent part of
log(x). If x is a subnormal value, the result
of log2(mantissa part of x)-126 is returned.
inline float32 atan(float32 x) §source
Return the principal values of arctangent. Denorm inputs and outputs are supported.
inline float32 pow(float32 base, float32 expo) §source
baseexpo
inline float32 sin(float32 x) §source
sin(x)
with x in radians.
inline float32 cos(float32 x) §source
cos(x)
with x in radians.
template <Denorm denorm_mode> inline float32 min(float32 x, float32 y) §source
Return the smaller of x and y. A NaN is
treated as missing data, and the other value will be returned.
template <Denorm denorm_mode> inline float32 max(float32 x, float32 y) §source
Return the larger of x and y. A NaN is
treated as missing data, and the other value will be returned.
template <Denorm denorm_mode> inline float32 ceil(float32 x) §source
Return the smallest integer not less than x: ⌈x⌉
template <Denorm denorm_mode> inline float32 floor(float32 x) §source
Return the largest integer not greater than x: ⌊x⌋
template <Denorm denorm_mode> inline float32 round(float32 x) §source
Return the nearest integer to x, rounding halfway away
from zero.
template <Denorm denorm_mode> inline float32 trunc(float32 x) §source
Return the nearest integer not greater in magnitude than
x.
template <Denorm denorm_mode> inline float32 dim(float32 x, float32 y) §source
The positive difference between x and y:
max(x - y, 0). Return NaN if x or
y is NaN.
template <Denorm denorm_mode> inline float32 abs(float32 x) §source
Return the absolute value of x: |x|
template <auto N> inline float32 from_int(int<N> value) §source
Convert a signed integer to a float32.
Parameters
-
auto NWidth of
value.
template <auto N> inline float32 from_uint(uint<N> value) §source
Convert an unsigned integer to a float32.
Parameters
-
auto NWidth of
value.
template <auto N> inline optional<int<N>> to_int(float32 value) §source
Convert a float32 to a signed integer.
Parameters
-
auto NWidth of the signed integer.
template <auto N> inline optional<uint<N>> to_uint(float32 value) §source
Convert a float32 to an unsigned integer
Parameters
-
auto NWidth of the unsigned integer.