FP_NORMAL, FP_SUBNORMAL, FP_ZERO, FP_INFINITE, FP_NAN (original) (raw)
| Defined in header | | | | ----------------------------------------------------------------------- | | ------------- | | #define FP_NORMAL /* implementation defined */ | | (since C++11) | | #define FP_SUBNORMAL /* implementation defined */ | | (since C++11) | | #define FP_ZERO /* implementation defined */ | | (since C++11) | | #define FP_INFINITE /* implementation defined */ | | (since C++11) | | #define FP_NAN /* implementation defined */ | | (since C++11) |
The FP_NORMAL, FP_SUBNORMAL, FP_ZERO, FP_INFINITE, FP_NAN macros each represent a distinct category of floating-point numbers. They all expand to an integer constant expression.
| Constant | Explanation |
|---|---|
| FP_NORMAL | indicates that the value is normal, i.e. not an infinity, subnormal, not-a-number or zero |
| FP_SUBNORMAL | indicates that the value is subnormal |
| FP_ZERO | indicates that the value is positive or negative zero |
| FP_INFINITE | indicates that the value is not representable by the underlying type (positive or negative infinity) |
| FP_NAN | indicates that the value is not-a-number (NaN) |
[edit] Example
#include #include #include auto show_classification(double x) { switch (std::fpclassify(x)) { case FP_INFINITE: return "Inf"; case FP_NAN: return "NaN"; case FP_NORMAL: return "normal"; case FP_SUBNORMAL: return "subnormal"; case FP_ZERO: return "zero"; default: return "unknown"; } } int main() { std::cout << "1.0/0.0 is " << show_classification(1 / 0.0) << '\n' << "0.0/0.0 is " << show_classification(0.0 / 0.0) << '\n' << "DBL_MIN/2 is " << show_classification(DBL_MIN / 2) << '\n' << "-0.0 is " << show_classification(-0.0) << '\n' << "1.0 is " << show_classification(1.0) << '\n'; }
Output:
1.0/0.0 is Inf 0.0/0.0 is NaN DBL_MIN/2 is subnormal -0.0 is zero 1.0 is normal