rint.3m (2010 09)

r
rint(3M) rint(3M)
NAME
rint(), rintf(), rintl(), rintw(), rintq(), nearbyint(), nearbyintf(), nearbyintl(), nearbyintw(), nearbyintq()
- round to nearest integer functions
SYNOPSIS
#include <math.h>
double rint(double x);
double nearbyint(double x);
HP Integrity Server Only
float rintf(float x);
long double rintl(long double x);
extended rintw(extended x);
quad rintq(quad x);
float nearbyintf(float x);
long double nearbyintl(long double x);
extended nearbyintw(extended x);
quad nearbyintq(quad x);
DESCRIPTION
rint() and nearbyint() return the integer (represented as a floating-point number) nearest x
according to the current rounding direction mode. These functions meet the requirement of the IEEE-754
standard for an operation that rounds to integer in floating-point format.
The two functions are identical except that
rint() raises the inexact exception if the result differs in
value from the argument, while nearbyint() does not.
In the default rounding direction (round to nearest),
rint(x) is the integer nearest x with the additional
stipulation that if |
rint(x) − x|=1/2, then rint(x) is even. (The same statement applies to
nearbyint(x).)
If the current rounding direction is toward negative infinity, rint() and nearbyint() return the
same value as floor(). If the current rounding direction is toward positive infinity,
rint() and
nearbyint() return the same value as ceil().
Integrity Server Only
rintf()and nearbyintf() are float versions of rint() and nearbyint() respectively; they
take a float argument and return a float result.
rintl() and nearbyintl() are long double versions of rint() and nearbyint() respec-
tively; they take a long double argument and return a long double result.
rintw() and nearbyintw() are extended versions of rint() and nearbyint() respectively;
they take an extended argument and return an extended result.
rintq() and nearbyintq() are equivalent to rintl() and nearbyintl() respectively on HP-UX
systems.
USAGE
To use these functions, compile either with the default
-Ae option or with the -Aa and
-D_HPUX_SOURCE options. To use (for Integrity servers) ceilw(), nearbyintw(), ceilq(),or
nearbyintq(), compile with the -fpwidetypes option. Make sure your program includes
<math.h>. Link in the math library by specifying -lm on the compiler or linker command line.
For more information, see the HP-UX floating-point guide for HP Integrity servers at the following site:
http://www.hp.com/go/fp.
RETURN VALUE
If x is ±INFINITY or ±zero,
rint() and nearbyint() return x.
If x is NaN, the
rint() and nearbyint() functions return NaN.
HP-UX 11i Version 3: September 2010 − 1 − Hewlett-Packard Company 1