@c ----------------------------------------------------------------------
@node cos, math
@findex cos
@subheading Syntax

@example
#include <math.h>

double cos(double x);
@end example

@subheading Description

This function computes the cosine of @var{x} (which should be given in
radians).

@subheading Return Value

The cosine of @var{x}.  If the absolute value of @var{x} is finite but
greater than or equal to @math{2^63}, the value is 1 (since for
arguments that large each bit of the mantissa is more than @code{Pi}).
If the value of @var{x} is infinite or @code{NaN}, the return value is
@code{NaN} and @code{errno} is set to @code{EDOM}.

@subheading Portability

@portability ansi, posix

@subheading Accuracy

In general, this function's relative accuracy is about
@math{1.7*10^(-16)}, which is close to the machine precision for a
@code{double}.  However, for arguments very close to @code{Pi/2} and its
odd multiples, the relative accuracy can be many times worse, due to
loss of precision in the internal FPU computations.  Since
@math{cos(Pi/2)} is zero, the absolute accuracy is still very good; but
if your program needs to preserve high @emph{relative} accuracy for such
arguments, link with @samp{-lm} and use the version of @code{cos} from
@file{libm.a} which does elaborate argument reduction, but is about
three times slower.