1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
|
/* deriv/deriv.c
*
* Copyright (C) 2004, 2007 Brian Gough
*
* This program 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.
*
* This program 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.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*/
#include <config.h>
#include <stdlib.h>
#include <gsl/gsl_math.h>
#include <gsl/gsl_errno.h>
#include <gsl/gsl_deriv.h>
static void
central_deriv (const gsl_function * f, double x, double h,
double *result, double *abserr_round, double *abserr_trunc)
{
/* Compute the derivative using the 5-point rule (x-h, x-h/2, x,
x+h/2, x+h). Note that the central point is not used.
Compute the error using the difference between the 5-point and
the 3-point rule (x-h,x,x+h). Again the central point is not
used. */
double fm1 = GSL_FN_EVAL (f, x - h);
double fp1 = GSL_FN_EVAL (f, x + h);
double fmh = GSL_FN_EVAL (f, x - h / 2);
double fph = GSL_FN_EVAL (f, x + h / 2);
double r3 = 0.5 * (fp1 - fm1);
double r5 = (4.0 / 3.0) * (fph - fmh) - (1.0 / 3.0) * r3;
double e3 = (fabs (fp1) + fabs (fm1)) * GSL_DBL_EPSILON;
double e5 = 2.0 * (fabs (fph) + fabs (fmh)) * GSL_DBL_EPSILON + e3;
/* The next term is due to finite precision in x+h = O (eps * x) */
double dy = GSL_MAX (fabs (r3 / h), fabs (r5 / h)) *(fabs (x) / h) * GSL_DBL_EPSILON;
/* The truncation error in the r5 approximation itself is O(h^4).
However, for safety, we estimate the error from r5-r3, which is
O(h^2). By scaling h we will minimise this estimated error, not
the actual truncation error in r5. */
*result = r5 / h;
*abserr_trunc = fabs ((r5 - r3) / h); /* Estimated truncation error O(h^2) */
*abserr_round = fabs (e5 / h) + dy; /* Rounding error (cancellations) */
}
int
gsl_deriv_central (const gsl_function * f, double x, double h,
double *result, double *abserr)
{
double r_0, round, trunc, error;
central_deriv (f, x, h, &r_0, &round, &trunc);
error = round + trunc;
if (round < trunc && (round > 0 && trunc > 0))
{
double r_opt, round_opt, trunc_opt, error_opt;
/* Compute an optimised stepsize to minimize the total error,
using the scaling of the truncation error (O(h^2)) and
rounding error (O(1/h)). */
double h_opt = h * pow (round / (2.0 * trunc), 1.0 / 3.0);
central_deriv (f, x, h_opt, &r_opt, &round_opt, &trunc_opt);
error_opt = round_opt + trunc_opt;
/* Check that the new error is smaller, and that the new derivative
is consistent with the error bounds of the original estimate. */
if (error_opt < error && fabs (r_opt - r_0) < 4.0 * error)
{
r_0 = r_opt;
error = error_opt;
}
}
*result = r_0;
*abserr = error;
return GSL_SUCCESS;
}
static void
forward_deriv (const gsl_function * f, double x, double h,
double *result, double *abserr_round, double *abserr_trunc)
{
/* Compute the derivative using the 4-point rule (x+h/4, x+h/2,
x+3h/4, x+h).
Compute the error using the difference between the 4-point and
the 2-point rule (x+h/2,x+h). */
double f1 = GSL_FN_EVAL (f, x + h / 4.0);
double f2 = GSL_FN_EVAL (f, x + h / 2.0);
double f3 = GSL_FN_EVAL (f, x + (3.0 / 4.0) * h);
double f4 = GSL_FN_EVAL (f, x + h);
double r2 = 2.0*(f4 - f2);
double r4 = (22.0 / 3.0) * (f4 - f3) - (62.0 / 3.0) * (f3 - f2) +
(52.0 / 3.0) * (f2 - f1);
/* Estimate the rounding error for r4 */
double e4 = 2 * 20.67 * (fabs (f4) + fabs (f3) + fabs (f2) + fabs (f1)) * GSL_DBL_EPSILON;
/* The next term is due to finite precision in x+h = O (eps * x) */
double dy = GSL_MAX (fabs (r2 / h), fabs (r4 / h)) * fabs (x / h) * GSL_DBL_EPSILON;
/* The truncation error in the r4 approximation itself is O(h^3).
However, for safety, we estimate the error from r4-r2, which is
O(h). By scaling h we will minimise this estimated error, not
the actual truncation error in r4. */
*result = r4 / h;
*abserr_trunc = fabs ((r4 - r2) / h); /* Estimated truncation error O(h) */
*abserr_round = fabs (e4 / h) + dy;
}
int
gsl_deriv_forward (const gsl_function * f, double x, double h,
double *result, double *abserr)
{
double r_0, round, trunc, error;
forward_deriv (f, x, h, &r_0, &round, &trunc);
error = round + trunc;
if (round < trunc && (round > 0 && trunc > 0))
{
double r_opt, round_opt, trunc_opt, error_opt;
/* Compute an optimised stepsize to minimize the total error,
using the scaling of the estimated truncation error (O(h)) and
rounding error (O(1/h)). */
double h_opt = h * pow (round / (trunc), 1.0 / 2.0);
forward_deriv (f, x, h_opt, &r_opt, &round_opt, &trunc_opt);
error_opt = round_opt + trunc_opt;
/* Check that the new error is smaller, and that the new derivative
is consistent with the error bounds of the original estimate. */
if (error_opt < error && fabs (r_opt - r_0) < 4.0 * error)
{
r_0 = r_opt;
error = error_opt;
}
}
*result = r_0;
*abserr = error;
return GSL_SUCCESS;
}
int
gsl_deriv_backward (const gsl_function * f, double x, double h,
double *result, double *abserr)
{
return gsl_deriv_forward (f, x, -h, result, abserr);
}
|
