748086b7b2
From-SVN: r145841
185 lines
4.9 KiB
C
185 lines
4.9 KiB
C
/* Implementation of the ERFC_SCALED intrinsic, to be included by erfc_scaled.c
|
|
Copyright (c) 2008 Free Software Foundation, Inc.
|
|
|
|
This file is part of the GNU Fortran runtime library (libgfortran).
|
|
|
|
Libgfortran is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU General Public
|
|
License as published by the Free Software Foundation; either
|
|
version 3 of the License, or (at your option) any later version.
|
|
|
|
Libgfortran is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR a PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
Under Section 7 of GPL version 3, you are granted additional
|
|
permissions described in the GCC Runtime Library Exception, version
|
|
3.1, as published by the Free Software Foundation.
|
|
|
|
You should have received a copy of the GNU General Public License and
|
|
a copy of the GCC Runtime Library Exception along with this program;
|
|
see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
|
|
<http://www.gnu.org/licenses/>. */
|
|
|
|
/* This implementation of ERFC_SCALED is based on the netlib algorithm
|
|
available at http://www.netlib.org/specfun/erf */
|
|
|
|
#define TYPE KIND_SUFFIX(GFC_REAL_,KIND)
|
|
#define CONCAT(x,y) x ## y
|
|
#define KIND_SUFFIX(x,y) CONCAT(x,y)
|
|
|
|
#if (KIND == 4)
|
|
|
|
# define EXP(x) expf(x)
|
|
# define TRUNC(x) truncf(x)
|
|
|
|
#elif (KIND == 8)
|
|
|
|
# define EXP(x) exp(x)
|
|
# define TRUNC(x) trunc(x)
|
|
|
|
#else
|
|
|
|
# ifdef HAVE_EXPL
|
|
# define EXP(x) expl(x)
|
|
# endif
|
|
# ifdef HAVE_TRUNCL
|
|
# define TRUNC(x) truncl(x)
|
|
# endif
|
|
|
|
#endif
|
|
|
|
#if defined(EXP) && defined(TRUNC)
|
|
|
|
extern TYPE KIND_SUFFIX(erfc_scaled_r,KIND) (TYPE);
|
|
export_proto(KIND_SUFFIX(erfc_scaled_r,KIND));
|
|
|
|
TYPE
|
|
KIND_SUFFIX(erfc_scaled_r,KIND) (TYPE x)
|
|
{
|
|
/* The main computation evaluates near-minimax approximations
|
|
from "Rational Chebyshev approximations for the error function"
|
|
by W. J. Cody, Math. Comp., 1969, PP. 631-638. This
|
|
transportable program uses rational functions that theoretically
|
|
approximate erf(x) and erfc(x) to at least 18 significant
|
|
decimal digits. The accuracy achieved depends on the arithmetic
|
|
system, the compiler, the intrinsic functions, and proper
|
|
selection of the machine-dependent constants. */
|
|
|
|
int i;
|
|
TYPE del, res, xden, xnum, y, ysq;
|
|
|
|
#if (KIND == 4)
|
|
static TYPE xneg = -9.382, xsmall = 5.96e-8,
|
|
xbig = 9.194, xhuge = 2.90e+3, xmax = 4.79e+37;
|
|
#else
|
|
static TYPE xneg = -26.628, xsmall = 1.11e-16,
|
|
xbig = 26.543, xhuge = 6.71e+7, xmax = 2.53e+307;
|
|
#endif
|
|
|
|
#define SQRPI ((TYPE) 0.56418958354775628695L)
|
|
#define THRESH ((TYPE) 0.46875L)
|
|
|
|
static TYPE a[5] = { 3.16112374387056560l, 113.864154151050156l,
|
|
377.485237685302021l, 3209.37758913846947l, 0.185777706184603153l };
|
|
|
|
static TYPE b[4] = { 23.6012909523441209l, 244.024637934444173l,
|
|
1282.61652607737228l, 2844.23683343917062l };
|
|
|
|
static TYPE c[9] = { 0.564188496988670089l, 8.88314979438837594l,
|
|
66.1191906371416295l, 298.635138197400131l, 881.952221241769090l,
|
|
1712.04761263407058l, 2051.07837782607147l, 1230.33935479799725l,
|
|
2.15311535474403846e-8l };
|
|
|
|
static TYPE d[8] = { 15.7449261107098347l, 117.693950891312499l,
|
|
537.181101862009858l, 1621.38957456669019l, 3290.79923573345963l,
|
|
4362.61909014324716l, 3439.36767414372164l, 1230.33935480374942l };
|
|
|
|
static TYPE p[6] = { 0.305326634961232344l, 0.360344899949804439l,
|
|
0.125781726111229246l, 0.0160837851487422766l,
|
|
0.000658749161529837803l, 0.0163153871373020978l };
|
|
|
|
static TYPE q[5] = { 2.56852019228982242l, 1.87295284992346047l,
|
|
0.527905102951428412l, 0.0605183413124413191l,
|
|
0.00233520497626869185l };
|
|
|
|
y = (x > 0 ? x : -x);
|
|
if (y <= THRESH)
|
|
{
|
|
ysq = 0;
|
|
if (y > xsmall)
|
|
ysq = y * y;
|
|
xnum = a[4]*ysq;
|
|
xden = ysq;
|
|
for (i = 0; i <= 2; i++)
|
|
{
|
|
xnum = (xnum + a[i]) * ysq;
|
|
xden = (xden + b[i]) * ysq;
|
|
}
|
|
res = x * (xnum + a[3]) / (xden + b[3]);
|
|
res = 1 - res;
|
|
res = EXP(ysq) * res;
|
|
return res;
|
|
}
|
|
else if (y <= 4)
|
|
{
|
|
xnum = c[8]*y;
|
|
xden = y;
|
|
for (i = 0; i <= 6; i++)
|
|
{
|
|
xnum = (xnum + c[i]) * y;
|
|
xden = (xden + d[i]) * y;
|
|
}
|
|
res = (xnum + c[7]) / (xden + d[7]);
|
|
}
|
|
else
|
|
{
|
|
res = 0;
|
|
if (y >= xbig)
|
|
{
|
|
if (y >= xmax)
|
|
goto finish;
|
|
if (y >= xhuge)
|
|
{
|
|
res = SQRPI / y;
|
|
goto finish;
|
|
}
|
|
}
|
|
ysq = ((TYPE) 1) / (y * y);
|
|
xnum = p[5]*ysq;
|
|
xden = ysq;
|
|
for (i = 0; i <= 3; i++)
|
|
{
|
|
xnum = (xnum + p[i]) * ysq;
|
|
xden = (xden + q[i]) * ysq;
|
|
}
|
|
res = ysq *(xnum + p[4]) / (xden + q[4]);
|
|
res = (SQRPI - res) / y;
|
|
}
|
|
|
|
finish:
|
|
if (x < 0)
|
|
{
|
|
if (x < xneg)
|
|
res = __builtin_inf ();
|
|
else
|
|
{
|
|
ysq = TRUNC (x*((TYPE) 16))/((TYPE) 16);
|
|
del = (x-ysq)*(x+ysq);
|
|
y = EXP(ysq*ysq) * EXP(del);
|
|
res = (y+y) - res;
|
|
}
|
|
}
|
|
return res;
|
|
}
|
|
|
|
#endif
|
|
|
|
#undef EXP
|
|
#undef TRUNC
|
|
|
|
#undef CONCAT
|
|
#undef TYPE
|
|
#undef KIND_SUFFIX
|