comparison toolboxes/FullBNT-1.0.7/KPMtools/optimalMatching.m @ 0:e9a9cd732c1e tip

first hg version after svn
author wolffd
date Tue, 10 Feb 2015 15:05:51 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:e9a9cd732c1e
1 % MATCH - Solves the weighted bipartite matching (or assignment)
2 % problem.
3 %
4 % Usage: a = match(C);
5 %
6 % Arguments:
7 % C - an m x n cost matrix; the sets are taken to be
8 % 1:m and 1:n; C(i, j) gives the cost of matching
9 % items i (of the first set) and j (of the second set)
10 %
11 % Returns:
12 %
13 % a - an m x 1 assignment vector, which gives the
14 % minimum cost assignment. a(i) is the index of
15 % the item of 1:n that was matched to item i of
16 % 1:m. If item i (of 1:m) was not matched to any
17 % item of 1:n, then a(i) is zero.
18
19 % Copyright (C) 2002 Mark A. Paskin
20 %
21 % This program is free software; you can redistribute it and/or modify
22 % it under the terms of the GNU General Public License as published by
23 % the Free Software Foundation; either version 2 of the License, or
24 % (at your option) any later version.
25 %
26 % This program is distributed in the hope that it will be useful, but
27 % WITHOUT ANY WARRANTY; without even the implied warranty of
28 % MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
29 % General Public License for more details.
30 %
31 % You should have received a copy of the GNU General Public License
32 % along with this program; if not, write to the Free Software
33 % Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
34 % USA.
35 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36
37 function [a] = optimalMatching(C)
38
39 % Trivial cases:
40 [p, q] = size(C);
41 if (p == 0)
42 a = [];
43 return;
44 elseif (q == 0)
45 a = zeros(p, 1);
46 return;
47 end
48
49
50 if 0
51 % First, reduce the problem by making easy optimal matches. If two
52 % elements agree that they are the best match, then match them up.
53 [x, a] = min(C, [], 2);
54 [y, b] = min(C, [], 1);
55 u = find(1:p ~= b(a(:)));
56 a(u) = 0;
57 v = find(1:q ~= a(b(:))');
58 C = C(u, v);
59 if (isempty(C)) return; end
60 end
61
62 % Get the (new) size of the two sets, u and v.
63 [m, n] = size(C);
64
65 %mx = realmax;
66 mx = 2*max(C(:));
67 mn = -2*min(C(:));
68 % Pad the affinity matrix to be square
69 if (m < n)
70 C = [C; mx * ones(n - m, n)];
71 elseif (n < m)
72 C = [C, mx * ones(m, m - n)];
73 end
74
75 % Run the Hungarian method. First replace infinite values by the
76 % largest (or smallest) finite values.
77 C(find(isinf(C) & (C > 0))) = mx;
78 C(find(isinf(C) & (C < 0))) = mn;
79 %fprintf('running hungarian\n');
80 [b, cost] = hungarian(C');
81
82 % Extract only the real assignments
83 ap = b(1:m)';
84 ap(find(ap > n)) = 0;
85
86 a = ap;
87 %% Incorporate this sub-assignment into the complete assignment
88 % k = find(ap);
89 % a(u(k)) = v(ap(k));
90