* Copyright (c) 1985 Regents of the University of California.
* Redistribution and use in source and binary forms are permitted
* provided that the above copyright notice and this paragraph are
* duplicated in all such forms and that any documentation,
* advertising materials, and other materials related to such
* distribution and use acknowledge that the software was developed
* by the University of California, Berkeley. The name of the
* University may not be used to endorse or promote products derived
* from this software without specific prior written permission.
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
* WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.
* All recipients should regard themselves as participants in an ongoing
* research project and hence should feel obligated to report their
* experiences (good or bad) with these elementary function codes, using
* the sendbug(8) program, to the authors.
static char sccsid
[] = "@(#)sinh.c 5.4 (Berkeley) %G%";
* RETURN THE HYPERBOLIC SINE OF X
* DOUBLE PRECISION (VAX D format 56 bits, IEEE DOUBLE 53 BITS)
* CODED IN C BY K.C. NG, 1/8/85;
* REVISED BY K.C. NG on 2/8/85, 3/7/85, 3/24/85, 4/16/85.
* Required system supported functions :
* Required kernel functions:
* expm1(x) ...return exp(x)-1
* 1. reduce x to non-negative by sinh(-x) = - sinh(x).
* expm1(x) + expm1(x)/(expm1(x)+1)
* 0 <= x <= lnovfl : sinh(x) := --------------------------------
* lnovfl <= x <= lnovfl+ln2 : sinh(x) := expm1(x)/2 (avoid overflow)
* lnovfl+ln2 < x < INF : overflow to INF
* sinh(x) is x if x is +INF, -INF, or NaN.
* only sinh(0)=0 is exact for finite argument.
* sinh(x) returns the exact hyperbolic sine of x nearly rounded. In
* a test run with 1,024,000 random arguments on a VAX, the maximum
* observed error was 1.93 ulps (units in the last place).
* The hexadecimal values are the intended ones for the following constants.
* The decimal values may be used, provided that the compiler will convert
* from decimal to binary accurately enough to produce the hexadecimal values
vc(mln2hi
, 8.8029691931113054792E1
,0f33
,43b0
,2bdb
,c7e2
, 7, .B00F33C7E22BDB
)
vc(mln2lo
,-4.9650192275318476525E-16 ,1b60
,a70f
,582a
,279e
, -50,-.8F1B60279E582A
)
vc(lnovfl
, 8.8029691931113053016E1
,0f33
,43b0
,2bda
,c7e2
, 7, .B00F33C7E22BDA
)
ic(mln2hi
, 7.0978271289338397310E2
, 10, 1.62E42FEFA39EF
)
ic(mln2lo
, 2.3747039373786107478E-14, -45, 1.ABC9E3B39803F
)
ic(lnovfl
, 7.0978271289338397310E2
, 9, 1.62E42FEFA39EF
)
#define mln2hi vccast(mln2hi)
#define mln2lo vccast(mln2lo)
#define lnovfl vccast(lnovfl)
#if defined(vax)||defined(tahoe)
#else /* defined(vax)||defined(tahoe) */
#endif /* defined(vax)||defined(tahoe) */
static const double one
=1.0, half
=1.0/2.0 ;
#if !defined(vax)&&!defined(tahoe)
if(x
!=x
) return(x
); /* x is NaN */
#endif /* !defined(vax)&&!defined(tahoe) */
{t
=expm1(x
); return(copysign((t
+t
/(one
+t
))*half
,sign
));}
/* subtract x by ln(2^(max+1)) and return 2^max*exp(x)
to avoid unnecessary overflow */
return(copysign(scalb(one
+expm1((x
-mln2hi
)-mln2lo
),max
),sign
));
else /* sinh(+-INF) = +-INF, sinh(+-big no.) overflow to +-INF */