.\" Copyright (c) 1991 The Regents of the University of California. .\" All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 3. All advertising materials mentioning features or use of this software .\" must display the following acknowledgement: .\" This product includes software developed by the University of .\" California, Berkeley and its contributors. .\" 4. Neither the name of the University nor the names of its contributors .\" may be used to endorse or promote products derived from this software .\" without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .\" from: @(#)acos.3 5.1 (Berkeley) 5/2/91 .\" $Id: math.3,v 1.2 2003/08/17 20:36:47 scp Exp $ .\" .Dd March 20, 2007 .Dt MATH 3 .Os .Sh NAME .Nm math .Nd mathematical library functions .Sh SYNOPSIS .Fd #include <math.h> .Sh DESCRIPTION The header file math.h provides function prototypes and macros for working with C99 floating point values. .Pp Each math.h function is provided in three variants: single, double and extended precision. The single and double precision variants operate on IEEE-754 single and double precision values, which correspond to the C types .Ft float and .Ft double , respectively. .Pp On Intel Macs, the C type .Ft long double corresponds to 80-bit IEEE-754 double extended precision. On PowerPC Macs, the C type .Ft long double corresponds by default to ordinary double precision, but a 128-bit non-IEEE-754 long double type is also available via the compiler flag .Ft -mlong-double-128 and linker flag .Ft -lmx . .Pp Details of the floating point formats can be found via "man float". .Pp Users who need to repeatedly perform the same calculation on a large set of data will probably find that the vector math library (composed of vMathLib and vForce) yields better performance for their needs than sequential calls to the libm. .Pp Users who need to perform mathematical operations on complex floating-point numbers should consult the man pages for the complex portion of the math library, via "man complex". .Sh LIST OF FUNCTIONS Each of the functions that use floating-point values are provided in single, double, and extended precision; the double precision prototypes are listed here. The man pages for the individual functions provide more details on their use, special cases, and prototypes for their single and extended precision versions. .Pp .Ft int .Fn fpclassify "double" .br .Ft int .Fn isfinite "double" .br .Ft int .Fn isinf "double" .br .Ft int .Fn isnan "double" .br .Ft int .Fn isnormal "double" .br .Ft int .Fn signbit "double" .Pp These function-like macros are used to classify a single floating-point argument. .Pp .Ft double .Fn copysign "double, double" .br .Ft double .Fn nextafter "double, double" .Pp .Fn copysign "x, y" returns the value equal in magnitude to .Fa x with the sign of .Fa y . .Fn nextafter "x, y" returns the next floating-point number after .Fa x in the direction of .Fa y . Both are correctly-rounded. .Pp .Ft double .Fn nan "const char *tag" .Pp The .Fn nan function returns a quiet NaN, without raising the invalid flag. .Pp .Ft double .Fn ceil "double" .br .Ft double .Fn floor "double" .br .Ft double .Fn nearbyint "double" .br .Ft double .Fn rint "double" .br .Ft double .Fn round "double" .br .Ft long int .Fn lrint "double" .br .Ft long int .Fn lround "double" .br .Ft long long int .Fn llrint "double" .br .Ft long long int .Fn llround "double" .br .Ft double .Fn trunc "double" .Pp These functions provide various means to round floating-point values to integral values. They are correctly rounded. .Pp .Ft double .Fn fmod "double, double" .br .Ft double .Fn remainder "double, double" .br .Ft double .Fn remquo "double x, double y, int *" .Pp These return a remainder of the division of x by y with an integral quotient. .Fn remquo additionally provides access to a few lower bits of the quotient. They are correctly rounded. .Pp .Ft double .Fn fdim "double, double" .br .Ft double .Fn fmax "double, double" .br .Ft double .Fn fmin "double, double" .Pp .Fn fmax "x, y" and .Fn fmin "x, y" return the maximum and minimum of .Fa x and .Fa y , respectively. .Fn fdim "x, y" returns the positive difference of .Fa x and .Fa y . All are correctly rounded. .Pp .Ft double .Fn fma "double x, double y, double z" .Pp .Fn fma "x, y, z" computes the value (x*y) + z as though without intermediate rounding. It is correctly rounded. .Pp .Ft double .Fn fabs "double" .br .Ft double .Fn sqrt "double" .br .Ft double .Fn cbrt "double" .br .Ft double .Fn hypot "double, double" .Pp .Fn fabs "x", .Fn sqrt "x", and .Fn cbrt "x" return the absolute value, square root, and cube root of .Fa x , respectively. .Fn hypot "x, y" returns sqrt(x*x + y*y). .Fn fabs and .Fn sqrt are correctly rounded. .Pp .Ft double .Fn exp "double" .br .Ft double .Fn exp2 "double" .br .Ft double .Fn expm1 "double" .Pp .Fn exp "x" , .Fn exp2 "x" , and .Fn expm1 "x" return e**x, 2**x, and e**x - 1, respectively. .Pp .Ft double .Fn log "double" .br .Ft double .Fn log2 "double" .br .Ft double .Fn log10 "double" .br .Ft double .Fn log1p "double" .Pp .Fn log "x" , .Fn log2 "x" , and .Fn log10 "x" return the natural, base-2, and base-10 logarithms of .Fa x , respectively. .Fn log1p "x" returns the natural log of 1+x. .Pp .Ft double .Fn logb "double" .br .Ft int .Fn ilogb "double" .Pp .Fn logb "x" and .Fn ilogb "x" return the exponent of .Fa x . .Pp .Ft double .Fn modf "double, double *" .br .Ft double .Fn frexp "double, int *" .Pp .Fn modf "x, &y" returns the fractional part of .Fa x and stores the integral part in .Fa y . .Fn frexp "x, &n" returns the mantissa of .Fa x and stores the exponent in .Fa n . They are correctly rounded. .Pp .Ft double .Fn ldexp "double, int" .br .Ft double .Fn scalbn "double, int" .br .Ft double .Fn scalbln "double, long int" .Pp .Fn ldexp "x, n" , .Fn scalbn "x, n" , and .Fn scalbln "x, n" return x*2**n. They are correctly rounded. .Pp .Ft double .Fn pow "double, double" .Pp .Fn pow "x,y" returns x raised to the power y. .Pp .Ft double .Fn cos "double" .br .Ft double .Fn sin "double" .br .Ft double .Fn tan "double" .Pp .Fn cos "x" , .Fn sin "x" , and .Fn tan "x" return the cosine, sine and tangent of .Fa x , respectively. .Pp .Ft double .Fn cosh "double" .br .Ft double .Fn sinh "double" .br .Ft double .Fn tanh "double" .Pp .Fn cosh "x" , .Fn sinh "x" , and .Fn tanh "x" return the hyperbolic cosine, hyperbolic sine and hyperbolic tangent of .Fa x , respectively. .Pp .Ft double .Fn acos "double" .br .Ft double .Fn asin "double" .br .Ft double .Fn atan "double" .br .Ft double .Fn atan2 "double, double" .Pp .Fn acos "x" , .Fn asin "x" , and .Fn atan "x" return the inverse cosine, inverse sine and inverse tangent of .Fa x , respectively. .Fn atan2 "y, x" returns the inverse tangent of y/x, with sign chosen according to the quadrant of (x,y). .Pp .Ft double .Fn acosh "double" .br .Ft double .Fn asinh "double" .br .Ft double .Fn atanh "double" .Pp .Fn acosh "x" , .Fn asinh "x" , and .Fn atanh "x" return the inverse hyperbolic cosine, inverse hyperbolic sine and inverse hyperbolic tangent of .Fa x , respectively. .Pp .Ft double .Fn tgamma "double" .br .Ft double .Fn lgamma "double" .Pp .Fn tgamma "x" and .Fn lgamma "x" return the values of the gamma function and its logarithm evalutated at .Fa x , respectively. .Pp .Ft double .Fn j0 "double" .br .Ft double .Fn j1 "double" .br .Ft double .Fn jn "double" .br .Ft double .Fn y0 "double" .br .Ft double .Fn y1 "double" .br .Ft double .Fn yn "double" .Pp .Fn j0 "x" , .Fn j1 "x" , and .Fn jn "x" return the values of the zeroth, first, and nth Bessel function of the first kind evaluated at .Fa x , respectively. .Fn y0 "x" , .Fn y1 "x" , and .Fn yn "x" return the values of the zeroth, first, and nth Bessel function of the second kind evaluated at .Fa x , respectively. .Pp .Ft double .Fn erf "double" .br .Ft double .Fn erfc "double" .Pp .Fn erf "x" and .Fn erfc "x" return the values of the error function and the complementary error function evaluated at .Fa x , respectively. .Sh MATHEMATICAL CONSTANTS In addition to the functions listed above, math.h defines a number of useful constants, listed below. All are defined as C99 floating-point constants. .nf .ta \w'M_2_SQRTPI'u+6 .sp 1 CONSTANT VALUE M_E base of natural logarithm, e M_LOG2E log2(e) M_LOG10E log10(e) M_LN2 ln(2) M_LN10 ln(10) M_PI pi M_PI_2 pi / 2 M_PI_4 pi / 4 M_1_PI 1 / pi M_2_PI 2 / pi M_2_SQRTPI 2 / sqrt(pi) M_SQRT2 sqrt(2) M_SQRT1_2 sqrt(1/2) .ta .fi .Sh IEEE STANDARD 754 FLOATING\-POINT ARITHMETIC The libm functions declared in math.h provide mathematical library functions in single-, double-, and extended-precision IEEE-754 floating-point formats on Intel macs, and in single- and double-precision IEEE-754 floating-point formats on PowerPC macs. .Sh SEE ALSO .Xr float 3 , .Xr complex 3 .Sh STANDARDS The <math.h> functions conform to the ISO/IEC 9899:1999(E) standard.