1 /* Copyright (C) 2006-2008 Joris Mooij [joris dot mooij at tuebingen dot mpg dot de]
2 Radboud University Nijmegen, The Netherlands /
3 Max Planck Institute for Biological Cybernetics, Germany
5 This file is part of libDAI.
7 libDAI is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2 of the License, or
10 (at your option) any later version.
12 libDAI is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with libDAI; if not, write to the Free Software
19 Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
23 #ifndef __defined_libdai_hak_h
24 #define __defined_libdai_hak_h
28 #include <dai/daialg.h>
29 #include <dai/regiongraph.h>
31 #include <dai/properties.h>
37 /// HAK provides an implementation of the single and double-loop algorithms by Heskes, Albers and Kappen
38 class HAK
: public DAIAlgRG
{
40 std::vector
<Factor
> _Qa
;
41 std::vector
<Factor
> _Qb
;
42 std::vector
<std::vector
<Factor
> > _muab
;
43 std::vector
<std::vector
<Factor
> > _muba
;
44 /// Maximum difference encountered so far
46 /// Number of iterations needed
55 DAI_ENUM(ClustersType
,MIN
,DELTA
,LOOP
)
56 ClustersType clusters
;
60 /// Name of this inference method
61 static const char *Name
;
64 /// Default constructor
65 HAK() : DAIAlgRG(), _Qa(), _Qb(), _muab(), _muba(), _maxdiff(0.0), _iters(0U), props() {}
67 /// Construct from FactorGraph fg and PropertySet opts
68 HAK( const FactorGraph
&fg
, const PropertySet
&opts
);
70 /// Construct from RegionGraph rg and PropertySet opts
71 HAK( const RegionGraph
&rg
, const PropertySet
&opts
);
74 HAK( const HAK
&x
) : DAIAlgRG(x
), _Qa(x
._Qa
), _Qb(x
._Qb
), _muab(x
._muab
), _muba(x
._muba
), _maxdiff(x
._maxdiff
), _iters(x
._iters
), props(x
.props
) {}
76 /// Clone *this (virtual copy constructor)
77 virtual HAK
* clone() const { return new HAK(*this); }
79 /// Create (virtual default constructor)
80 virtual HAK
* create() const { return new HAK(); }
82 /// Assignment operator
83 HAK
& operator=( const HAK
&x
) {
85 DAIAlgRG::operator=( x
);
90 _maxdiff
= x
._maxdiff
;
97 /// Identifies itself for logging purposes
98 virtual std::string
identify() const;
100 /// Get single node belief
101 virtual Factor
belief( const Var
&n
) const;
103 /// Get general belief
104 virtual Factor
belief( const VarSet
&ns
) const;
107 virtual std::vector
<Factor
> beliefs() const;
109 /// Get log partition sum
110 virtual Real
logZ() const;
112 /// Clear messages and beliefs
115 /// Clear messages and beliefs corresponding to the nodes in ns
116 virtual void init( const VarSet
&ns
);
118 /// The actual approximate inference algorithm
119 virtual double run();
121 /// Return maximum difference between single node beliefs in the last pass
122 virtual double maxDiff() const { return _maxdiff
; }
124 /// Return number of passes over the factorgraph
125 virtual size_t Iterations() const { return _iters
; }
128 Factor
& muab( size_t alpha
, size_t _beta
) { return _muab
[alpha
][_beta
]; }
129 Factor
& muba( size_t alpha
, size_t _beta
) { return _muba
[alpha
][_beta
]; }
130 const Factor
& Qa( size_t alpha
) const { return _Qa
[alpha
]; };
131 const Factor
& Qb( size_t beta
) const { return _Qb
[beta
]; };
134 double doDoubleLoop();
136 void setProperties( const PropertySet
&opts
);
137 PropertySet
getProperties() const;
138 std::string
printProperties() const;
141 void constructMessages();
142 void findLoopClusters( const FactorGraph
&fg
, std::set
<VarSet
> &allcl
, VarSet newcl
, const Var
& root
, size_t length
, VarSet vars
);
146 } // end of namespace dai