source: NonGTP/Boost/boost/graph/bc_clustering.hpp @ 857

Revision 857, 5.9 KB checked in by igarcia, 19 years ago (diff)
RevLine 
[857]1// Copyright 2004 The Trustees of Indiana University.
2
3// Use, modification and distribution is subject to the Boost Software
4// License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at
5// http://www.boost.org/LICENSE_1_0.txt)
6
7//  Authors: Douglas Gregor
8//           Andrew Lumsdaine
9#ifndef BOOST_GRAPH_BETWEENNESS_CENTRALITY_CLUSTERING_HPP
10#define BOOST_GRAPH_BETWEENNESS_CENTRALITY_CLUSTERING_HPP
11
12#include <boost/graph/betweenness_centrality.hpp>
13#include <boost/graph/graph_traits.hpp>
14#include <boost/pending/indirect_cmp.hpp>
15#include <algorithm>
16#include <vector>
17#include <boost/property_map.hpp>
18
19namespace boost {
20
21/** Threshold termination function for the betweenness centrality
22 * clustering algorithm.
23 */
24template<typename T>
25struct bc_clustering_threshold
26{
27  typedef T centrality_type;
28
29  /// Terminate clustering when maximum absolute edge centrality is
30  /// below the given threshold.
31  explicit bc_clustering_threshold(T threshold)
32    : threshold(threshold), dividend(1.0) {}
33 
34  /**
35   * Terminate clustering when the maximum edge centrality is below
36   * the given threshold.
37   *
38   * @param threshold the threshold value
39   *
40   * @param g the graph on which the threshold will be calculated
41   *
42   * @param normalize when true, the threshold is compared against the
43   * normalized edge centrality based on the input graph; otherwise,
44   * the threshold is compared against the absolute edge centrality.
45   */
46  template<typename Graph>
47  bc_clustering_threshold(T threshold, const Graph& g, bool normalize = true)
48    : threshold(threshold), dividend(1.0)
49  {
50    if (normalize) {
51      typename graph_traits<Graph>::vertices_size_type n = num_vertices(g);
52      dividend = T((n - 1) * (n - 2)) / T(2);
53    }
54  }
55
56  /** Returns true when the given maximum edge centrality (potentially
57   * normalized) falls below the threshold.
58   */
59  template<typename Graph, typename Edge>
60  bool operator()(T max_centrality, Edge, const Graph&)
61  {
62    return (max_centrality / dividend) < threshold;
63  }
64
65 protected:
66  T threshold;
67  T dividend;
68};
69
70/** Graph clustering based on edge betweenness centrality.
71 *
72 * This algorithm implements graph clustering based on edge
73 * betweenness centrality. It is an iterative algorithm, where in each
74 * step it compute the edge betweenness centrality (via @ref
75 * brandes_betweenness_centrality) and removes the edge with the
76 * maximum betweenness centrality. The @p done function object
77 * determines when the algorithm terminates (the edge found when the
78 * algorithm terminates will not be removed).
79 *
80 * @param g The graph on which clustering will be performed. The type
81 * of this parameter (@c MutableGraph) must be a model of the
82 * VertexListGraph, IncidenceGraph, EdgeListGraph, and Mutable Graph
83 * concepts.
84 *
85 * @param done The function object that indicates termination of the
86 * algorithm. It must be a ternary function object thats accepts the
87 * maximum centrality, the descriptor of the edge that will be
88 * removed, and the graph @p g.
89 *
90 * @param edge_centrality (UTIL/OUT) The property map that will store
91 * the betweenness centrality for each edge. When the algorithm
92 * terminates, it will contain the edge centralities for the
93 * graph. The type of this property map must model the
94 * ReadWritePropertyMap concept. Defaults to an @c
95 * iterator_property_map whose value type is
96 * @c Done::centrality_type and using @c get(edge_index, g) for the
97 * index map.
98 *
99 * @param vertex_index (IN) The property map that maps vertices to
100 * indices in the range @c [0, num_vertices(g)). This type of this
101 * property map must model the ReadablePropertyMap concept and its
102 * value type must be an integral type. Defaults to
103 * @c get(vertex_index, g).
104 */
105template<typename MutableGraph, typename Done, typename EdgeCentralityMap,
106         typename VertexIndexMap>
107void
108betweenness_centrality_clustering(MutableGraph& g, Done done,
109                                  EdgeCentralityMap edge_centrality,
110                                  VertexIndexMap vertex_index)
111{
112  typedef typename property_traits<EdgeCentralityMap>::value_type
113    centrality_type;
114  typedef typename graph_traits<MutableGraph>::edge_iterator edge_iterator;
115  typedef typename graph_traits<MutableGraph>::edge_descriptor edge_descriptor;
116  typedef typename graph_traits<MutableGraph>::vertices_size_type
117    vertices_size_type;
118
119  if (edges(g).first == edges(g).second) return;
120
121  // Function object that compares the centrality of edges
122  indirect_cmp<EdgeCentralityMap, std::less<centrality_type> >
123    cmp(edge_centrality);
124
125  bool is_done;
126  do {
127    brandes_betweenness_centrality(g,
128                                   edge_centrality_map(edge_centrality)
129                                   .vertex_index_map(vertex_index));
130    edge_descriptor e = *max_element(edges(g).first, edges(g).second, cmp);
131    is_done = done(get(edge_centrality, e), e, g);
132    if (!is_done) remove_edge(e, g);
133  } while (!is_done && edges(g).first != edges(g).second);
134}
135
136/**
137 * \overload
138 */
139template<typename MutableGraph, typename Done, typename EdgeCentralityMap>
140void
141betweenness_centrality_clustering(MutableGraph& g, Done done,
142                                  EdgeCentralityMap edge_centrality)
143{
144  betweenness_centrality_clustering(g, done, edge_centrality,
145                                    get(vertex_index, g));
146}
147
148/**
149 * \overload
150 */
151template<typename MutableGraph, typename Done>
152void
153betweenness_centrality_clustering(MutableGraph& g, Done done)
154{
155  typedef typename Done::centrality_type centrality_type;
156  std::vector<centrality_type> edge_centrality(num_edges(g));
157  betweenness_centrality_clustering(g, done,
158    make_iterator_property_map(edge_centrality.begin(), get(edge_index, g)),
159    get(vertex_index, g));
160}
161
162} // end namespace boost
163
164#endif // BOOST_GRAPH_BETWEENNESS_CENTRALITY_CLUSTERING_HPP
Note: See TracBrowser for help on using the repository browser.