Fortran Library Reference

`libm` Math Functions

The following functions and subroutines are part of the math library libm. Some routines are intrinsics and return the same data type (single precision, double precision, or quad precision) as their argument. The rest are non-intrinsics that take a specific data type as an argument and return the same. These non-intrinsics do have to be declared in the routine referencing them.

`libm` Intrinsic Functions

Here is a list of the intrinsic functions in libm. You need not put them in a type statement. These functions take single, double, or quad precision data as arguments and return the same.

sqrt(x)	asin(x)	cosd(x)
log(x)	acos(x)	asind(x)
log10(x)	atan(x)	acosd(x)
exp(x)	atan2(x,y)	atand(x)
x**y	sinh(x)	atan2d(x,y)
sin(x)	cosh(x)	aint(x)
cos(x)	tanh(x)	anint(x)
tan(x)	sind(x)	nint(x)

The functionssind(x), cosd(x), asind(x), acosd(x), atand(x), atan2d(x,y) are not considered intrinsics by the FORTRAN 77 standard.

`libm_double`: Double-Precision Functions

The following subprograms are double-precision libm functions and subroutines.

In general, these functions do not correspond to standard FORTRAN generic intrinsic functions--data types are determined by the usual data typing rules.

Example: Subroutine and non-Intrinsic double-precision functions:

    DOUBLE PRECISION c, d_acosh, d_hypot, d_infinity, s, x, y, z 
    ... 
    z = d_acosh( x ) 
    i = id_finite( x ) 
    z = d_hypot( x, y ) 
    z = d_infinity() 
    CALL d_sincos( x, s, c )

These DOUBLE PRECISION functions need to appear in a DOUBLE PRECISION statement.

Refer to the C library man pages for details: the man page for d_acos(x) is acos(3M)

Table 1-5 Double Precision libm Functions


`d_acos( x )` `d_acosd( x )` `d_acosh( x )` `d_acosp(` `x` `)` `d_acospi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`arc cosine` -- `arc cosh` -- --
`d_atan( x )` `d_atand( x )` `d_atanh( x )` `d_atanp( x )` `d_atanpi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`arc` `tangent` -- `arc` `tanh` -- --
`d_asin( x )` `d_asind( x )` `d_asinh( x )` `d_asinp( x )` `d_asinpi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`arc` `sine` -- `arc` `sinh` -- --
`d_atan2(( y, x )` `d_atan2d( y, x )` `d_atan2pi( y, x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function`	`arc` `tangent` -- --
`d_cbrt(` `x` `)` `d_ceil(` `x` `)` `d_copysign(` `x,` `x` `)`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function`	`cube` `root` `ceiling` --
`d_cos( x )` `d_cosd( x )` `d_cosh( x )` `d_cosp( x )` `d_cospi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`cosine` -- `hyperb` `cos` -- --
`d_erf( x )` `d_erfc( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function`	`error` `func` --
`d_expm1( x )` `d_floor( x )` `d_hypot( x, y )` `d_infinity( )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function`	`(e**x)-1` `floor` `hypotenuse` --
`d_j0( x )` `d_j1( x )` `d_jn( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function`	`Bessel` -- --
`id_finite( x )` `id_fp_class( x )` `id_ilogb( x )` `id_irint( x )` `id_isinf( x )` `id_isnan( x )` `id_isnormal( x )` `id_issubnormal( x )` `id_iszero( x )` `id_signbit( x )`	`INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`
`d_addran()` `d_addrans(x, p, l, u)` `d_lcran()` `d_lcrans(x, p, l, u )` `d_shufrans(x, p, l,u)`	`DOUBLE PRECISION` n/a `DOUBLE PRECISION` n/a n/a	`Function` `Subroutine` `Function` `Subroutine` `Subroutine`	`random` `number` `generators`
`d_lgamma( x )` `d_logb( x )` `d_log1p( x )` `d_log2( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function`	`log` `gamma` -- -- --
`d_max_normal()` `d_max_subnormal()` `d_min_normal()` `d_min_subnormal()` `d_nextafter( x, y )` `d_quiet_nan( n )` `d_remainder( x, y )` `d_rint( x )` `d_scalb( x, y )` `d_scalbn( x, n )` `d_signaling_nan( n )` `d_significand( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`
`d_sin( x )` `d_sind( x )` `d_sinh( x )` `d_sinp( x )` `d_sinpi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`sine` -- `hyperb` `sine` -- --
`d_sincos( x, s, c )` `d_sincosd( x, s, c )` `d_sincosp( x, s, c )` `d_sincospi( x, s, c )`	n/a n/a n/a n/a	`Subroutine` `Subroutine` `Subroutine` `Subroutine`	`sine` `and` `cosine` -- --
`d_tan( x )` `d_tand( x )` `d_tanh( x )` `d_tanp( x )` `d_tanpi( x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function` `Function` `Function`	`tangent` -- `hyperb` `tan` -- --
`d_y0( x )` `d_y1( x )` `d_yn( n, x )`	`DOUBLE PRECISION` `DOUBLE PRECISION` `DOUBLE PRECISION`	`Function` `Function` `Function`	`bessel` -- --

Variables c, l, p, s, u, x, and y are of type DOUBLE PRECISION.

Explicitly type these functions on a DOUBLE PRECISION statement or with an appropriate IMPLICIT statement).

sind(x), asind(x), ... take degrees rather than radians.

See also: intro(3M) and the Numerical Computation Guide.

`libm_quadruple`: Quad-Precision Functions

These subprograms are quadruple-precision (REAL*16) libm functions and subroutines (SPARC only).

In general, these do not correspond to standard generic intrinsic functions; data types are determined by the usual data typing rules.

Samples: Quadruple precision functions:

    REAL*16 c, q_acosh, q_hypot, q_infinity, s, x, y, z 
    ... 
    z = q_acosh( x ) 
    i = iq_finite( x ) 
    z = q_hypot( x, y ) 
    z = q_infinity() 
    CALL q_sincos( x, s, c )

The quadruple precision functions must appear in a REAL*16 statement

Table 1-6 Quadruple-Precision libm Functions


`q_copysign( x, y )` `q_fabs( x )` `q_fmod( x )` `q_infinity( )`	`REAL16` `REAL16` `REAL16` `REAL16`	`Function` `Function` `Function` `Function`
`iq_finite( x )` `iq_fp_class( x )` `iq_ilogb( x )` `iq_isinf( x )` `iq_isnan( x )` `iq_isnormal( x )` `iq_issubnormal( x )` `iq_iszero( x )` `iq_signbit( x )`	`INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`
`q_max_normal()` `q_max_subnormal()` `q_min_normal()` `q_min_subnormal()` `q_nextafter( x, y )` `q_quiet_nan( n )` `q_remainder( x, y )` `q_scalbn( x, n )` `q_signaling_nan( n )`	`REAL16` `REAL16` `REAL16` `REAL16` `REAL16` `REAL16` `REAL16` `REAL16` `REAL*16`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`

The variables c, l, p, s, u, x, and y are of type quadruple precision.

Explicitly type these functions with a REAL*16 statement or with an appropriate IMPLICIT statement.

sind(x), asind(x), ... take degrees rather than radians.

If you need to use any other quadruple-precision libm function, you can call it using $PRAGMA C(fcn) before the call. For details, see the chapter on the C-FORTRAN interface in the Fortran Programming Guide.

`libm_single`: Single-Precision Functions

These subprograms are single-precision libm functions and subroutines.

In general, the functions below provide access to single-precision libm functions that do not correspond to standard FORTRAN generic intrinsic functions--data types are determined by the usual data typing rules.

Samples: Single-precision libm functions:

    REAL c, s, x, y, z 
    ..
    z = r_acosh( x ) 
    i = ir_finite( x ) 
    z = r_hypot( x, y ) 
    z = r_infinity() 
    CALL r_sincos( x, s, c )

These functions need not be explicitly typed with a REAL statement as long as default typing holds. (Variables beginning with "r" are REAL, with "i" are INTEGER.)

For details on these routines, see the C math library man pages (3M). For example, for r_acos(x) see the acos(3M) man page.

Table 1-7 Single-Precision libm functions


`r_acos( x )` `r_acosd( x )` `r_acosh( x )` `r_acosp( x )` `r_acospi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`arc` `cosine` -- `arc` `cosh` -- --
`r_atan( x )` `r_atand( x )` `r_atanh( x )` `r_atanp( x )` `r_atanpi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`arc` `tangent` -- `arc` `tanh` -- --
`r_asin( x )` `r_asind( x )` `r_asinh( x )` `r_asinp( x )` `r_asinpi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`arc` `sine` -- `arc` `sinh` -- --
`r_atan2(( y, x )` `r_atan2d( y, x )` `r_atan2pi( y, x )`	`REAL` `REAL` `REAL`	`Function` `Function` `Function`	`arc` `tangent` -- --
`r_cbrt( x )` `r_ceil( x )` `r_copysign( x, y )`	`REAL` `REAL` `REAL`	`Function` `Function` `Function`	`cube` `root` `ceiling` --
`r_cos( x )` `r_cosd( x )` `r_cosh( x )` `r_cosp( x )` `r_cospi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`cosine` -- `hyperb` `cos` -- --
`r_erf( x )` `r_erfc( x )`	`REAL` `REAL`	`Function` `Function`	`err` `function` --
`r_expm1( x )` `r_floor( x )` `r_hypot( x, y )` `r_infinity( )` `r_j0( x )` `r_j1( x )` `r_jn( x )`	`REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function` `Function` `Function`	`(e**x)-1` `floor` `hypotenuse` `bessel` -- -- --
`ir_finite( x )` `ir_fp_class( x )` `ir_ilogb( x )` `ir_irint( x )` `ir_isinf( x )` `ir_isnan( x )` `ir_isnormal( x )` `ir_issubnormal( x )` `ir_iszero( x )` `ir_signbit( x )`	`INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER` `INTEGER`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`	-- -- -- -- -- -- -- -- -- --
`r_addran()` `r_addrans( x, p, l, u )` `r_lcran()` `r_lcrans( x, p, l, u )` `r_shufrans(x, p, l, u)`	`REAl` n/a `REAL` n/a n/a	`Function` `Subroutine` `Function` `Subroutine` `Subroutine`	`random` `number` -- -- -- --
`r_lgamma( x )` `r_logb( x )` `r_log1p( x )` `r_log2( x )`	`REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function`	`log` `gamma` -- -- --
`r_max_normal()` `r_max_subnormal()` `r_min_normal()` `r_min_subnormal()` `r_nextafter( x, y )` `r_quiet_nan( n )` `r_remainder( x, y )` `r_rint( x )` `r_scalb( x, y )` `r_scalbn( x, n )` `r_signaling_nan( n )` `r_significand( x )`	`REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function` `Function`
`r_sin( x )` `r_sind( x )` `r_sinh( x )` `r_sinp( x )` `r_sinpi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`sine` -- `hyperb sin` -- --
`r_sincos( x, s, c )` `r_sincosd( x, s, c )` `r_sincosp( x, s, c )` `r_sincospi( x, s, c )`	n/a n/a n/a n/a	`Subroutine` `Subroutine` `Subroutine` `Subroutine`	`sine` `&` `cosine` -- -- --
`r_tan( x )` `r_tand( x )` `r_tanh( x )` `r_tanp( x )` `r_tanpi( x )`	`REAL` `REAL` `REAL` `REAL` `REAL`	`Function` `Function` `Function` `Function` `Function`	`tangent` -- `hyperb` `tan` -- --
`r_y0( x )` `r_y1( x )` `r_yn( n, x )`	`REAL` `REAL` `REAL`	`Function` `Function` `Function`	`bessel` -- --

Variables c, l, p, s, u, x, and y are of type REAL.

Type these functions as explicitly REAL if an IMPLICIT statement is in effect that types names starting with "r" to some other date type.

sind(x), asind(x), ... take degrees rather than radians.

See also: intro(3M) and the Numerical Computation Guide.

libm Math Functions

libm Intrinsic Functions

libm_double: Double-Precision Functions

libm_quadruple: Quad-Precision Functions

libm_single: Single-Precision Functions

`libm` Math Functions

`libm` Intrinsic Functions

`libm_double`: Double-Precision Functions

`libm_quadruple`: Quad-Precision Functions

`libm_single`: Single-Precision Functions