fpclassify.3

.TH FPCLASSIFY 3  "December 18, 2009"
.UC 4
.SH NAME
fpclassify, isfinite, isinf, isnan, isnormal, signbit \- classify floating point numbers
.SH SYNOPSIS
.nf
.ft B
#include <math.h>

int fpclassify(double \fIx\fP):
int isfinite(double \fIx\fP);
int isinf(double \fIx\fP);
int isnan(double \fIx\fP);
int isnormal(double \fIx\fP);
int signbit(double \fIx\fP);
.fi
.SH DESCRIPTION
These functions provide information about the specified floating point number 
\fIx\fP. fpclassify returns one of the values FP_INFINITE, FP_NAN, FP_NORMAL,
FP_SUBNORMAL and FP_ZERO depending on the type of number provided. The isinf,
isinf, isnan and isnormal test for specific number classes, returning a 
non-zero value is and only if the specified number belongs to the class 
specified by the function name. The signbit function returns a non-zero value
if and only if the sign bit is set, which for non-NaN values (including zero)
means that the number is negative.