-
Notifications
You must be signed in to change notification settings - Fork 59
/
cyclic_reduction.html
312 lines (274 loc) · 8.51 KB
/
cyclic_reduction.html
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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
<html>
<head>
<title>
CYCLIC_REDUCTION - A Direct Solution Method for Tridiagonal Linear Systems.
</title>
</head>
<body bgcolor="#EEEEEE" link="#CC0000" alink="#FF3300" vlink="#000055">
<h1 align = "center">
CYCLIC_REDUCTION <br> A Direct Solution Method for Tridiagonal Linear Systems.
</h1>
<hr>
<p>
<b>CYCLIC_REDUCTION</b>
is a FORTRAN90 library which
applies the cyclic reduction method to solve a tridiagonal system of
linear equations A*x=b.
</p>
<p>
The matrix is assumed to be diagonally dominant - that is, for every row,
we require that the magnitude of the diagonal entry be at least as great
as the sum of the magnitudes of the two off-diagonal elements. This is
(just barely) true for the "-1, 2, -1" matrix, for instance.
</p>
<p>
Other methods for solving tridiagonal linear systems include:
<ul>
<li>
Gauss elimination with pivoting;
</li>
<li>
the Thomas algorithm, (Gauss elimination without pivoting);
</li>
<li>
the Jacobi, Gauss-Seidel, and SOR iterative methods;
</li>
</ul>
</p>
<p>
Cyclic reduction is a form of Gauss elimination. It proceeds by first
eliminating half of the variables simultaneously, then half of the remainder,
and so on. This amounts to more work, but the work in each elimination
step can be done in parallel. Thus, unlike the Gauss and Thomas algorithms,
cyclic reduction offers a procedure for the direct solution of a tridiagonal
linear system that can take advantage of parallelism.
</p>
<p>
Cyclic reduction can also be adapted to the block tridiagonal linear systems
that arise when Poisson's equation is discretized over a 2D region.
</p>
<h3 align = "center">
Licensing:
</h3>
<p>
The computer code and data files described and made available on this web page
are distributed under
<a href = "../../txt/gnu_lgpl.txt">the GNU LGPL license.</a>
</p>
<h3 align = "center">
Languages:
</h3>
<p>
<b>CYCLIC_REDUCTION</b> is available in
<a href = "../../c_src/cyclic_reduction/cyclic_reduction.html">a C version</a> and
<a href = "../../cpp_src/cyclic_reduction/cyclic_reduction.html">a C++ version</a> and
<a href = "../../f77_src/cyclic_reduction/cyclic_reduction.html">a FORTRAN77 version</a> and
<a href = "../../f_src/cyclic_reduction/cyclic_reduction.html">a FORTRAN90 version</a> and
<a href = "../../m_src/cyclic_reduction/cyclic_reduction.html">a MATLAB version</a>.
</p>
<h3 align = "center">
Related Data and Programs:
</h3>
<p>
<a href = "../../c_src/csparse/csparse.html">
CSPARSE</a>,
a C library which
carries out the direct solution of sparse linear systems.
</p>
<p>
<a href = "../../f_src/dlap/dlap.html">
DLAP</a>,
a FORTRAN90 library which
carries out the iterative solution of sparse linear systems.
</p>
<p>
<a href = "../../f77_src/lapack_examples/lapack_examples.html">
LAPACK_EXAMPLES</a>,
a FORTRAN77 program which
demonstrates the use of the LAPACK linear algebra library.
</p>
<p>
<a href = "../../f_src/linpack/linpack.html">
LINPACK</a>,
a FORTRAN90 library which
factors and solves systems of linear equations in a variety of
formats and arithmetic types.
</p>
<p>
<a href = "../../f_src/linplus/linplus.html">
LINPLUS</a>,
a FORTRAN90 library which
carries out simple manipulations of matrices in a variety of formats.
</p>
<p>
<a href = "../../f_src/mgmres/mgmres.html">
MGMRES</a>,
a FORTRAN90 library which
applies the restarted GMRES algorithm to solve a sparse linear system.
</p>
<p>
<a href = "../../f_src/sparsekit/sparsekit.html">
SPARSEKIT</a>,
a FORTRAN90 library which
carries out operations on sparse matrices, including conversion between various formats.
</p>
<p>
<a href = "../../c_src/super_lu/super_lu.html">
SUPER_LU</a>,
a C library which
implements some very fast direct
solvers for systems of sparse linear equations.
</p>
<p>
<a href = "../../f_src/test_mat/test_mat.html">
TEST_MAT</a>,
a FORTRAN90 library which
defines test matrices, some of
which have known determinants, eigenvalues and eigenvectors,
inverses and so on.
</p>
<h3 align = "center">
Reference:
</h3>
<p>
<ol>
<li>
Gene Golub, Charles VanLoan,<br>
Matrix Computations,<br>
Third Edition,<br>
Johns Hopkins, 1996,<br>
ISBN: 0-8018-4513-X,<br>
LC: QA188.G65.
</li>
<li>
Roger Hockney,<br>
A fast direct solution of Poisson's equation using Fourier Analysis,<br>
Journal of the ACM,<br>
Volume 12, Number 1, pages 95-113, January 1965.
</li>
</ol>
</p>
<h3 align = "center">
Source Code:
</h3>
<p>
<ul>
<li>
<a href = "cyclic_reduction.f90">cyclic_reduction.f90</a>, the source code.
</li>
<li>
<a href = "cyclic_reduction.sh">cyclic_reduction.sh</a>,
commands to compile the source code.
</li>
</ul>
</p>
<h3 align = "center">
Examples and Tests:
</h3>
<p>
<ul>
<li>
<a href = "cyclic_reduction_prb.f90">cyclic_reduction_prb.f90</a>,
a sample calling program.
</li>
<li>
<a href = "cyclic_reduction_prb.sh">cyclic_reduction_prb.sh</a>,
commands to compile and run the sample program.
</li>
<li>
<a href = "cyclic_reduction_prb_output.txt">cyclic_reduction_prb_output.txt</a>,
the output file.
</li>
</ul>
</p>
<h3 align = "center">
List of Routines:
</h3>
<p>
<ul>
<li>
<b>C83_CR_FA</b> decomposes a C83 matrix using cyclic reduction.
</li>
<li>
<b>C83_CR_SL</b> solves a linear system factored by C83_CR_FA.
</li>
<li>
<b>C83_CR_SLS</b> solves several linear systems factored by C83_CR_FA.
</li>
<li>
<b>C83_INDICATOR</b> sets up a C83 indicator matrix.
</li>
<li>
<b>C83_MXV</b> multiplies a C83 matrix times a C8VEC.
</li>
<li>
<b>C83_PRINT</b> prints a C83 matrix.
</li>
<li>
<b>C83_PRINT_SOME</b> prints some of a C83 matrix.
</li>
<li>
<b>C8MAT_PRINT</b> prints a C8MAT.
</li>
<li>
<b>C8MAT_PRINT_SOME</b> prints some of a C8MAT.
</li>
<li>
<b>C8VEC_INDICATOR</b> sets a C8VEC to an "indicator" vector.
</li>
<li>
<b>C8VEC_PRINT</b> prints a C8VEC, with an optional title.
</li>
<li>
<b>C8VEC_PRINT_SOME</b> prints some of a C8VEC.
</li>
<li>
<b>R83_CR_FA</b> decomposes an R83 matrix using cyclic reduction.
</li>
<li>
<b>R83_CR_SL</b> solves a linear systems factored by R83_CR_FA.
</li>
<li>
<b>R83_CR_SLS</b> solves several linear systems factored by R83_CR_FA.
</li>
<li>
<b>R83_PRINT</b> prints an R83 matrix.
</li>
<li>
<b>R83_PRINT_SOME</b> prints some of an R83 matrix.
</li>
<li>
<b>R8MAT_PRINT</b> prints an R8MAT.
</li>
<li>
<b>R8MAT_PRINT_SOME</b> prints some of an R8MAT.
</li>
<li>
<b>R8VEC_INDICATOR</b> sets an R8VEC to the indicator vector.
</li>
<li>
<b>R83_MXV</b> multiplies an R83 matrix times an R8VEC.
</li>
<li>
<b>R8VEC_PRINT</b> prints an R8VEC, with an optional title.
</li>
<li>
<b>R8VEC_PRINT_SOME</b> prints "some" of an R8VEC.
</li>
<li>
<b>TIMESTAMP</b> prints the current YMDHMS date as a time stamp.
</li>
</ul>
</p>
<p>
You can go up one level to <a href = "../f_src.html">
the FORTRAN90 source codes</a>.
</p>
<hr>
<i>
Last revised on 06 May 2010.
</i>
<!-- John Burkardt -->
</body>
<!-- Initial HTML skeleton created by HTMLINDEX. -->
</html>