1 /* This file is part of libDAI - http://www.libdai.org/
2 *
4 * 2, or (at your option) any later version. libDAI is distributed without any
5 * warranty. See the file COPYING for more details.
6 *
7 * Copyright (C) 2006-2009 Joris Mooij [joris dot mooij at libdai dot org]
9 */
12 /// \file
13 /// \brief Defines class JTree, which implements the junction tree algorithm
16 #ifndef __defined_libdai_jtree_h
17 #define __defined_libdai_jtree_h
20 #include <vector>
21 #include <string>
22 #include <dai/daialg.h>
23 #include <dai/varset.h>
24 #include <dai/regiongraph.h>
25 #include <dai/factorgraph.h>
26 #include <dai/clustergraph.h>
27 #include <dai/weightedgraph.h>
28 #include <dai/enum.h>
29 #include <dai/properties.h>
32 namespace dai {
35 /// Exact inference algorithm using junction tree
36 /** The junction tree algorithm uses message passing on a junction tree to calculate
37 * exact marginal probability distributions ("beliefs") for specified cliques
38 * (outer regions) and separators (intersections of pairs of cliques).
39 *
40 * There are two variants, the sum-product algorithm (corresponding to
41 * finite temperature) and the max-product algorithm (corresponding to
42 * zero temperature).
43 */
44 class JTree : public DAIAlgRG {
45 private:
46 /// Stores the messages
47 std::vector<std::vector<Factor> > _mes;
49 /// Stores the logarithm of the partition sum
50 Real _logZ;
52 public:
53 /// The junction tree (stored as a rooted tree)
54 RootedTree RTree;
56 /// Outer region beliefs
57 std::vector<Factor> Qa;
59 /// Inner region beliefs
60 std::vector<Factor> Qb;
62 /// Parameters of this inference algorithm
63 struct Properties {
64 /// Enumeration of possible JTree updates
65 DAI_ENUM(UpdateType,HUGIN,SHSH);
67 /// Enumeration of inference variants
68 DAI_ENUM(InfType,SUMPROD,MAXPROD);
70 /// Verbosity
71 size_t verbose;
73 /// Type of updates: HUGIN or Shafer-Shenoy
76 /// Type of inference: sum-product or max-product
77 InfType inference;
78 } props;
80 /// Name of this inference algorithm
81 static const char *Name;
83 public:
84 /// \name Constructors/destructors
85 //@{
86 /// Default constructor
87 JTree() : DAIAlgRG(), _mes(), _logZ(), RTree(), Qa(), Qb(), props() {}
89 /// Construct from FactorGraph \a fg and PropertySet \a opts
90 /** \param fg factor graph (which has to be connected);
91 * \param opts parameters;
92 * \param automatic if \c true, construct the junction tree automatically, using the MinFill heuristic.
93 */
94 JTree( const FactorGraph &fg, const PropertySet &opts, bool automatic=true );
95 //@}
98 /// \name General InfAlg interface
99 //@{
100 virtual JTree* clone() const { return new JTree(*this); }
101 virtual std::string identify() const;
102 virtual Factor belief( const Var &n ) const;
103 virtual Factor belief( const VarSet &ns ) const;
104 virtual std::vector<Factor> beliefs() const;
105 virtual Real logZ() const;
106 virtual void init() {}
107 virtual void init( const VarSet &/*ns*/ ) {}
108 virtual Real run();
109 virtual Real maxDiff() const { return 0.0; }
110 virtual size_t Iterations() const { return 1UL; }
111 virtual void setProperties( const PropertySet &opts );
112 virtual PropertySet getProperties() const;
113 virtual std::string printProperties() const;
114 //@}
117 /// \name Additional interface specific for JTree
118 //@{
119 /// Constructs a junction tree based on the cliques \a cl (corresponding to some elimination sequence).
120 /** First, constructs a weighted graph, where the nodes are the elements of \a cl, and
121 * each edge is weighted with the cardinality of the intersection of the state spaces of the nodes.
122 * Then, a maximal spanning tree for this weighted graph is calculated.
123 * Finally, a corresponding region graph is built:
124 * - the outer regions correspond with the cliques and have counting number 1;
125 * - the inner regions correspond with the seperators, i.e., the intersections of two
126 * cliques that are neighbors in the spanning tree, and have counting number -1;
127 * - inner and outer regions are connected by an edge if the inner region is a
128 * seperator for the outer region.
129 */
130 void GenerateJT( const std::vector<VarSet> &cl );
132 /// Returns constant reference to the message from outer region \a alpha to its \a _beta 'th neighboring inner region
133 const Factor & message( size_t alpha, size_t _beta ) const { return _mes[alpha][_beta]; }
134 /// Returns reference to the message from outer region \a alpha to its \a _beta 'th neighboring inner region
135 Factor & message( size_t alpha, size_t _beta ) { return _mes[alpha][_beta]; }
137 /// Runs junction tree algorithm using HUGIN updates
138 /** \note The initial messages may be arbitrary.
139 */
140 void runHUGIN();
142 /// Runs junction tree algorithm using Shafer-Shenoy updates
143 /** \note The initial messages may be arbitrary.
144 */
145 void runShaferShenoy();
147 /// Finds an efficient subtree for calculating the marginal of the variables in \a vs
148 /** First, the current junction tree is reordered such that it gets as root the clique
149 * that has maximal state space overlap with \a vs. Then, the minimal subtree
150 * (starting from the root) is identified that contains all the variables in \a vs
151 * and also the outer region with index \a PreviousRoot (if specified). Finally,
152 * the current junction tree is reordered such that this minimal subtree comes
153 * before the other edges, and the size of the minimal subtree is returned.
154 */
155 size_t findEfficientTree( const VarSet& vs, RootedTree &Tree, size_t PreviousRoot=(size_t)-1 ) const;
157 /// Calculates the marginal of a set of variables (using cutset conditioning, if necessary)
158 /** \pre assumes that run() has been called already
159 */
160 Factor calcMarginal( const VarSet& vs );
162 /// Calculates the joint state of all variables that has maximum probability
163 /** \pre Assumes that run() has been called and that \a props.inference == \c MAXPROD
164 */
165 std::vector<std::size_t> findMaximum() const;
166 //@}
167 };
170 /// Calculates upper bound to the treewidth of a FactorGraph, using the MinFill heuristic
171 /** \relates JTree
172 * \return a pair (number of variables in largest clique, number of states in largest clique)
173 */
174 std::pair<size_t,size_t> boundTreewidth( const FactorGraph & fg );
177 /// Calculates upper bound to the treewidth of a FactorGraph, using the MinFill heuristic
178 /** \deprecated Renamed into boundTreewidth()
179 */
180 std::pair<size_t,size_t> treewidth( const FactorGraph & fg );
183 } // end of namespace dai
186 #endif