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
|
function camd_demo
%CAMD_DEMO a demo of camd, using the can_24 matrix
%
% A demo of CAMD for MATLAB.
%
% Example:
% camd_demo
%
% See also: camd, camd_make
% CAMD, Copyright (c) 2007-2022, Timothy A. Davis, Yanqing Chen, Patrick R.
% Amestoy, and Iain S. Duff. All Rights Reserved.
% SPDX-License-Identifier: BSD-3-clause
% This orders the same matrix as the ANSI C demo, camd_demo.c. It includes an
% additional analysis of the matrix via MATLAB's symbfact routine.
% First, print the help information for CAMD
help camd
% Get the Harwell/Boeing can_24 matrix.
load can_24
A = spconvert (can_24) ;
n = size (A,1) ;
rand ('state', 0) ;
C = irand (6, n) ;
clf
hold off
subplot (2,2,1) ;
spy (A)
title ('HB/can24 matrix') ;
% print the details during CAMD ordering and SYMBFACT
% spparms ('spumoni', 1) ;
% order the matrix. Note that the Info argument is optional.
fprintf ('\nIf the next step fails, then you have\n') ;
fprintf ('not yet compiled the CAMD mexFunction.\n') ;
[p, Info] = camd (A) ; %#ok
% order again, but this time print some statistics
[p, camd_Info] = camd (A, [10 1 1], C) ;
fprintf ('Permutation vector:\n') ;
fprintf (' %2d', p) ;
fprintf ('\n\n') ;
fprintf ('Corresponding constraint sets:\n') ;
if (any (sort (C (p)) ~= C (p)))
error ('Error!') ;
end
for j = 1:n
fprintf (' %2d', C (p (j))) ;
end
fprintf ('\n\n\n') ;
subplot (2,2,2) ;
spy (A (p,p)) ;
title ('Permuted matrix') ;
% The camd_demo.c program stops here.
fprintf ('Analyze A(p,p) with MATLAB symbfact routine:\n') ;
[cn, height, parent, post, R] = symbfact (A(p,p)) ;
subplot (2,2,3) ;
spy (R') ;
title ('Cholesky factor L') ;
subplot (2,2,4) ;
treeplot (parent) ;
title ('etree') ;
% results from symbfact
lnz = sum (cn) ; % number of nonzeros in L, incl. diagonal
cn = cn - 1 ; % get the count of off-diagonal entries
fl = n + sum (cn.^2 + 2*cn) ; % flop count for chol (A (p,p)
fprintf ('number of nonzeros in L (including diagonal): %d\n', lnz) ;
fprintf ('floating point operation count for chol (A (p,p)): %d\n', fl) ;
% approximations from camd:
lnz2 = n + camd_Info (10) ;
fl2 = n + camd_Info (11) + 2 * camd_Info (12) ;
fprintf ('\nResults from CAMD''s approximate analysis:\n') ;
fprintf ('number of nonzeros in L (including diagonal): %d\n', lnz2) ;
fprintf ('floating point operation count for chol (A (p,p)): %d\n\n', fl2) ;
fprintf ('\nNote that the ordering quality is not as good as p=amd(A).\n') ;
fprintf ('This is only because the ordering constraints, C, have been\n') ;
fprintf ('randomly selected.\n') ;
if (lnz2 ~= lnz | fl ~= fl2) %#ok
fprintf ('Note that the nonzero and flop counts from CAMD are slight\n') ;
fprintf ('upper bounds. This is due to the approximate minimum degree\n');
fprintf ('method used, in conjunction with "mass elimination".\n') ;
fprintf ('See the discussion about mass elimination in camd.h and\n') ;
fprintf ('camd_2.c for more details.\n') ;
end
% turn off diagnostic output in MATLAB's sparse matrix routines
% spparms ('spumoni', 0) ;
%-------------------------------------------------------------------------------
function i = irand (n,s)
% irand: return a random vector of size s, with values between 1 and n
if (nargin == 1)
s = 1 ;
end
i = min (n, 1 + floor (rand (1,s) * n)) ;
|
