fabs

fabs is a mathematical function in the C and C++ standard libraries that returns the absolute value of a floating-point number. The name stands for floating-point absolute value. In C, fabs is declared in math.h with the signature double fabs(double x), and it has float and long double variants: fabsf(float x) and fabsl(long double x). In C++, the function is available in <cmath> and typically accessed as std::fabs with similar overloads for double, float, and long double.

Behavior and semantics: For finite input, fabs returns the nonnegative value with the same magnitude as the input; if x is negative, fabs returns -x, and if x is already nonnegative, fabs returns x. If the input is NaN, the result is NaN. If the input is ±∞, the result is +∞. For -0.0, fabs(-0.0) yields +0.0. The function does not raise domain errors; it simply returns a value according to IEEE 754 rules.

Usage notes: fabs is distinct from the integer absolute value functions such as labs and llabs. In C++, std::abs is commonly overloaded to handle floating-point types as well, so both std::fabs and std::abs can be used for floating arguments, depending on the namespace and headers included.

Implementation considerations: On modern platforms, fabs is typically implemented as a fast operation that clears the sign bit of the floating-point representation.