GRASS GIS 8 Programmer's Manual  8.5.0dev(2024)-36359e2344
solvers_classic_iter.c
Go to the documentation of this file.
1 /*****************************************************************************
2  *
3  * MODULE: Grass numerical math interface
4  * AUTHOR(S): Soeren Gebbert, Berlin (GER) Dec 2006
5  * soerengebbert <at> googlemail <dot> com
6  *
7  * PURPOSE: linear equation system solvers
8  * part of the gmath library
9  *
10  * COPYRIGHT: (C) 2010 by the GRASS Development Team
11  *
12  * This program is free software under the GNU General Public
13  * License (>=v2). Read the file COPYING that comes with GRASS
14  * for details.
15  *
16  *****************************************************************************/
17 
18 #include <assert.h>
19 #include <math.h>
20 #include <unistd.h>
21 #include <stdio.h>
22 #include <string.h>
23 #include <grass/gis.h>
24 #include <grass/glocale.h>
25 #include <grass/gmath.h>
26 
27 /*!
28  * \brief The iterative jacobi solver for sparse matrices
29  *
30  * The Jacobi solver solves the linear equation system Ax = b
31  * The result is written to the vector x.
32  *
33  * The parameter <i>maxit</i> specifies the maximum number of iterations. If the
34  * maximum is reached, the solver will abort the calculation and writes the
35  * current result into the vector x. The parameter <i>err</i> defines the error
36  * break criteria for the solver.
37  *
38  * \param Asp G_math_spvector ** -- the sparse matrix
39  * \param x double * -- the vector of unknowns
40  * \param b double * -- the right side vector
41  * \param rows int -- number of rows
42  * \param maxit int -- the maximum number of iterations
43  * \param sor double -- defines the successive overrelaxion parameter [0:1]
44  * \param error double -- defines the error break criteria
45  * \return int -- 1=success, -1=could not solve the les
46  *
47  * */
48 int G_math_solver_sparse_jacobi(G_math_spvector **Asp, double *x, double *b,
49  int rows, int maxit, double sor, double error)
50 {
51  unsigned int i, j, center, finished = 0;
52 
53  int k;
54 
55  double *Enew;
56 
57  double E, err = 0;
58 
59  assert(rows >= 0);
60 
61  Enew = G_alloc_vector(rows);
62 
63  for (k = 0; k < maxit; k++) {
64  err = 0;
65  {
66  if (k == 0) {
67  for (j = 0; j < (unsigned int)rows; j++) {
68  Enew[j] = x[j];
69  }
70  }
71  for (i = 0; i < (unsigned int)rows; i++) {
72  E = 0;
73  center = 0;
74  for (j = 0; j < Asp[i]->cols; j++) {
75  E += Asp[i]->values[j] * x[Asp[i]->index[j]];
76  if (Asp[i]->index[j] == i)
77  center = j;
78  }
79  Enew[i] = x[i] - sor * (E - b[i]) / Asp[i]->values[center];
80  }
81  for (j = 0; j < (unsigned int)rows; j++) {
82  err += (x[j] - Enew[j]) * (x[j] - Enew[j]);
83 
84  x[j] = Enew[j];
85  }
86  }
87 
88  G_message(_("sparse Jacobi -- iteration %5i error %g\n"), k, err);
89 
90  if (err < error) {
91  finished = 1;
92  break;
93  }
94  }
95 
96  G_free(Enew);
97 
98  return finished;
99 }
100 
101 /*!
102  * \brief The iterative gauss seidel solver for sparse matrices
103  *
104  * The Jacobi solver solves the linear equation system Ax = b
105  * The result is written to the vector x.
106  *
107  * The parameter <i>maxit</i> specifies the maximum number of iterations. If the
108  * maximum is reached, the solver will abort the calculation and writes the
109  * current result into the vector x. The parameter <i>err</i> defines the error
110  * break criteria for the solver.
111  *
112  * \param Asp G_math_spvector ** -- the sparse matrix
113  * \param x double * -- the vector of unknowns
114  * \param b double * -- the right side vector
115  * \param rows int -- number of rows
116  * \param maxit int -- the maximum number of iterations
117  * \param sor double -- defines the successive overrelaxion parameter [0:2]
118  * \param error double -- defines the error break criteria
119  * \return int -- 1=success, -1=could not solve the les
120  *
121  * */
122 int G_math_solver_sparse_gs(G_math_spvector **Asp, double *x, double *b,
123  int rows, int maxit, double sor, double error)
124 {
125  unsigned int i, j, finished = 0;
126 
127  int k;
128 
129  double *Enew;
130 
131  double E, err = 0;
132 
133  int center;
134 
135  assert(rows >= 0);
136 
137  Enew = G_alloc_vector(rows);
138 
139  for (k = 0; k < maxit; k++) {
140  err = 0;
141  {
142  if (k == 0) {
143  for (j = 0; j < (unsigned int)rows; j++) {
144  Enew[j] = x[j];
145  }
146  }
147  for (i = 0; i < (unsigned int)rows; i++) {
148  E = 0;
149  center = 0;
150  for (j = 0; j < Asp[i]->cols; j++) {
151  E += Asp[i]->values[j] * Enew[Asp[i]->index[j]];
152  if (Asp[i]->index[j] == i)
153  center = j;
154  }
155  Enew[i] = x[i] - sor * (E - b[i]) / Asp[i]->values[center];
156  }
157  for (j = 0; j < (unsigned int)rows; j++) {
158  err += (x[j] - Enew[j]) * (x[j] - Enew[j]);
159 
160  x[j] = Enew[j];
161  }
162  }
163 
164  G_message(_("sparse SOR -- iteration %5i error %g\n"), k, err);
165 
166  if (err < error) {
167  finished = 1;
168  break;
169  }
170  }
171 
172  G_free(Enew);
173 
174  return finished;
175 }
176 
177 /*!
178  * \brief The iterative jacobi solver for quadratic matrices
179  *
180  * The Jacobi solver solves the linear equation system Ax = b
181  * The result is written to the vector x.
182  *
183  * The parameter <i>maxit</i> specifies the maximum number of iterations. If the
184  * maximum is reached, the solver will abort the calculation and writes the
185  * current result into the vector x. The parameter <i>err</i> defines the error
186  * break criteria for the solver.
187  *
188  * \param A double ** -- the dense matrix
189  * \param x double * -- the vector of unknowns
190  * \param b double * -- the right side vector
191  * \param rows int -- number of rows
192  * \param maxit int -- the maximum number of iterations
193  * \param sor double -- defines the successive overrelaxion parameter [0:1]
194  * \param error double -- defines the error break criteria
195  * \return int -- 1=success, -1=could not solve the les
196  *
197  * */
198 int G_math_solver_jacobi(double **A, double *x, double *b, int rows, int maxit,
199  double sor, double error)
200 {
201  int i, j, k;
202 
203  double *Enew;
204 
205  double E, err = 0;
206 
207  Enew = G_alloc_vector(rows);
208 
209  for (j = 0; j < rows; j++) {
210  Enew[j] = x[j];
211  }
212 
213  for (k = 0; k < maxit; k++) {
214  for (i = 0; i < rows; i++) {
215  E = 0;
216  for (j = 0; j < rows; j++) {
217  E += A[i][j] * x[j];
218  }
219  Enew[i] = x[i] - sor * (E - b[i]) / A[i][i];
220  }
221  err = 0;
222  for (j = 0; j < rows; j++) {
223  err += (x[j] - Enew[j]) * (x[j] - Enew[j]);
224  x[j] = Enew[j];
225  }
226  G_message(_("Jacobi -- iteration %5i error %g\n"), k, err);
227  if (err < error)
228  break;
229  }
230 
231  return 1;
232 }
233 
234 /*!
235  * \brief The iterative gauss seidel solver for quadratic matrices
236  *
237  * The Jacobi solver solves the linear equation system Ax = b
238  * The result is written to the vector x.
239  *
240  * The parameter <i>maxit</i> specifies the maximum number of iterations. If the
241  * maximum is reached, the solver will abort the calculation and writes the
242  * current result into the vector x. The parameter <i>err</i> defines the error
243  * break criteria for the solver.
244  *
245  * \param A double ** -- the dense matrix
246  * \param x double * -- the vector of unknowns
247  * \param b double * -- the right side vector
248  * \param rows int -- number of rows
249  * \param maxit int -- the maximum number of iterations
250  * \param sor double -- defines the successive overrelaxion parameter [0:2]
251  * \param error double -- defines the error break criteria
252  * \return int -- 1=success, -1=could not solve the les
253  *
254  * */
255 int G_math_solver_gs(double **A, double *x, double *b, int rows, int maxit,
256  double sor, double error)
257 {
258  int i, j, k;
259 
260  double *Enew;
261 
262  double E, err = 0;
263 
264  Enew = G_alloc_vector(rows);
265 
266  for (j = 0; j < rows; j++) {
267  Enew[j] = x[j];
268  }
269 
270  for (k = 0; k < maxit; k++) {
271  for (i = 0; i < rows; i++) {
272  E = 0;
273  for (j = 0; j < rows; j++) {
274  E += A[i][j] * Enew[j];
275  }
276  Enew[i] = x[i] - sor * (E - b[i]) / A[i][i];
277  }
278  err = 0;
279  for (j = 0; j < rows; j++) {
280  err += (x[j] - Enew[j]) * (x[j] - Enew[j]);
281  x[j] = Enew[j];
282  }
283  G_message(_("SOR -- iteration %5i error %g\n"), k, err);
284  if (err < error)
285  break;
286  }
287 
288  return 1;
289 }
void G_free(void *)
Free allocated memory.
Definition: gis/alloc.c:150
void G_message(const char *,...) __attribute__((format(printf
double * G_alloc_vector(size_t)
Vector matrix memory allocation.
Definition: dalloc.c:39
#define _(str)
Definition: glocale.h:10
#define assert(condition)
Definition: lz4.c:291
double b
Definition: r_raster.c:39
int G_math_solver_sparse_jacobi(G_math_spvector **Asp, double *x, double *b, int rows, int maxit, double sor, double error)
The iterative jacobi solver for sparse matrices.
int G_math_solver_sparse_gs(G_math_spvector **Asp, double *x, double *b, int rows, int maxit, double sor, double error)
The iterative gauss seidel solver for sparse matrices.
int G_math_solver_jacobi(double **A, double *x, double *b, int rows, int maxit, double sor, double error)
The iterative jacobi solver for quadratic matrices.
int G_math_solver_gs(double **A, double *x, double *b, int rows, int maxit, double sor, double error)
The iterative gauss seidel solver for quadratic matrices.
The row vector of the sparse matrix.
Definition: gmath.h:54
double * values
Definition: gmath.h:55
unsigned int cols
Definition: gmath.h:56
unsigned int * index
Definition: gmath.h:57
SYMBOL * err(FILE *fp, SYMBOL *s, char *msg)
Definition: symbol/read.c:216
#define x