/
bicgstab.hpp
54 lines (45 loc) · 2.68 KB
/
bicgstab.hpp
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
/* Copyright (C) 2005-2022 Massachusetts Institute of Technology.
*
* 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 2 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#ifndef BICGSTAB_H
#define BICGSTAB_H
#include "meep.hpp"
namespace meep {
/* signature of a user callback function, passed to bicgstabL, that computes y = Ax. data is
an opaque pointer to any user data that you need, passed through from the Adata pointer to
bicgstabL (e.g. this could be a data structure that records information about A, such as
its size n). */
typedef void (*bicgstab_op)(const realnum *x, realnum *y, void *data);
/* solve Ax=b by biCGSTAB(L). L is the dimension of the subspace (typically < 10) — larger L
gives more expensive iterations but faster convergence; n is the dimension of the vectors and
matrices (x and b are length n, A is n by n); x is the initial solution guess on input (often
zero) and the approximate solution on output; b is the right hand side; A is a function that
implements A-times-vector (see above) and Adata is an opaque user data pointer passed through to
A; tol is the stopping tolerance |Ax-b| < tol*|b|; on *iters is the maximum allowed number of
iterations and on output it is the actual number; quiet=1 to suppress status messages during the
solver.
You must also allocate a workspace, usually by calling the routine twice: on the first call, pass
NULL for the work pointer, and the return value nwork is the required size of the workspace. Then
allocate work = new realnum[nwork], and call it again.
For non-NULL nwork, returns 0 on success, 1 if the maximum number of iterations was reached, and
-1 if a breakdown in convergence was detected. */
ptrdiff_t bicgstabL(const int L, const size_t n, realnum *x, bicgstab_op A, void *Adata,
const realnum *b, const double tol,
int *iters, // input *iters = max iters, output = actual iters
realnum *work, // if you pass work=NULL, bicgstab returns nwork
const bool quiet);
} // namespace meep
#endif /* BICGSTAB_H */