include/dai/mr.h

   1 /*  This file is part of libDAI - http://www.libdai.org/
   2  *
   3  *  libDAI is licensed under the terms of the GNU General Public License version
   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) 2007       Bastian Wemmenhove
   8  *  Copyright (C) 2007-2009  Joris Mooij         [joris dot mooij at libdai dot org]
   9  *  Copyright (C) 2007       Radboud University Nijmegen, The Netherlands
  10  */
  11
  12
  13 /// \file
  14 /// \brief Defines class MR, which implements loop corrections as proposed by Montanari and Rizzo
  15
  16
  17 #ifndef __defined_libdai_mr_h
  18 #define __defined_libdai_mr_h
  19
  20
  21 #include <vector>
  22 #include <string>
  23 #include <dai/factorgraph.h>
  24 #include <dai/daialg.h>
  25 #include <dai/enum.h>
  26 #include <dai/properties.h>
  27 #include <dai/exceptions.h>
  28 #include <boost/dynamic_bitset.hpp>
  29
  30
  31 namespace dai {
  32
  33
  34 /// Approximate inference algorithm by Montanari and Rizzo [\ref MoR05]
  35 /** \author Bastian Wemmenhove wrote the original implementation before it was merged into libDAI
  36  */
  37 class MR : public DAIAlgFG {
  38     private:
  39         /// Is the underlying factor graph supported?
  40         bool supported;
  41         /// con[i] = connectivity of spin \a i
  42         std::vector<size_t>                             con;
  43         /// nb[i] are the neighbours of spin \a i
  44         std::vector<std::vector<size_t> >               nb;
  45         /// tJ[i][_j] is the hyperbolic tangent of the interaction between spin \a i and its neighbour nb[i][_j]
  46         std::vector<std::vector<Real> >                 tJ;
  47         /// theta[i] is the local field on spin \a i
  48         std::vector<Real>                               theta;
  49         /// M[i][_j] is \f$ M^{(i)}_j \f$
  50         std::vector<std::vector<Real> >                 M;
  51         /// The \a _j 'th neighbour of spin \a i has spin \a i as its kindex[i][_j]'th neighbour
  52         std::vector<std::vector<size_t> >               kindex;
  53         /// Cavity correlations
  54         std::vector<std::vector<std::vector<Real> > >   cors;
  55         /// Maximum connectivity
  56         static const size_t kmax = 31;
  57         /// Type used for managing a subset of neighbors
  58         typedef boost::dynamic_bitset<> sub_nb;
  59         /// Number of variables (spins)
  60         size_t N;
  61         /// Magnetizations
  62         std::vector<Real> Mag;
  63         /// Maximum difference encountered so far
  64         Real _maxdiff;
  65         /// Number of iterations needed
  66         size_t _iters;
  67
  68     public:
  69         /// Parameters of this inference algorithm
  70         struct Properties {
  71             /// Enumeration of different types of update equations
  72             /** The possible update equations are:
  73              *  - FULL full updates, slow but accurate
  74              *  - LINEAR linearized updates, faster but less accurate
  75              */
  76             DAI_ENUM(UpdateType,FULL,LINEAR);
  77
  78             /// Enumeration of different ways of initializing the cavity correlations
  79             /** The possible cavity initializations are:
  80              *  - RESPPROP using response propagation ("linear response")
  81              *  - CLAMPING using clamping and BP
  82              *  - EXACT using JunctionTree
  83              */
  84             DAI_ENUM(InitType,RESPPROP,CLAMPING,EXACT);
  85
  86             /// Verbosity
  87             size_t verbose;
  88
  89             /// Tolerance
  90             Real tol;
  91
  92             /// Update equations
  93             UpdateType updates;
  94
  95             /// How to initialize the cavity correlations
  96             InitType inits;
  97         } props;
  98
  99         /// Name of this inference method
 100         static const char *Name;
 101
 102     public:
 103         /// Default constructor
 104         MR() : DAIAlgFG(), supported(), con(), nb(), tJ(), theta(), M(), kindex(), cors(), N(), Mag(), _maxdiff(), _iters(), props() {}
 105
 106         /// Construct from FactorGraph \a fg and PropertySet \a opts
 107         /** \note This implementation only deals with binary variables and pairwise interactions.
 108          *  \throw NOT_IMPLEMENTED if \a fg has factors depending on three or more variables or has variables with more than two possible states.
 109          */
 110         MR( const FactorGraph &fg, const PropertySet &opts );
 111
 112
 113     /// \name General InfAlg interface
 114     //@{
 115         virtual MR* clone() const { return new MR(*this); }
 116         virtual std::string identify() const;
 117         virtual Factor belief( const Var &n ) const;
 118         virtual Factor belief( const VarSet &/*ns*/ ) const { DAI_THROW(NOT_IMPLEMENTED); return Factor(); }
 119         virtual std::vector<Factor> beliefs() const;
 120         virtual Real logZ() const { DAI_THROW(NOT_IMPLEMENTED); return 0.0; }
 121         virtual void init() {}
 122         virtual void init( const VarSet &/*ns*/ ) { DAI_THROW(NOT_IMPLEMENTED); }
 123         virtual Real run();
 124         virtual Real maxDiff() const { return _maxdiff; }
 125         virtual size_t Iterations() const { return _iters; }
 126         virtual void setProperties( const PropertySet &opts );
 127         virtual PropertySet getProperties() const;
 128         virtual std::string printProperties() const;
 129     //@}
 130
 131     private:
 132         /// Returns the signum of \a a
 133         Real sign(Real a) { return (a >= 0) ? 1.0 : -1.0; }
 134
 135         /// Initialize N, con, nb, tJ, theta
 136         void init(size_t Nin, Real *_w, Real *_th);
 137
 138         /// Initialize kindex
 139         void makekindex();
 140
 141         /// Initialize cors
 142         void init_cor();
 143
 144         /// Calculate cors using response propagation
 145         Real init_cor_resp();
 146
 147         /// Iterate update equations for cavity fields
 148         void solvemcav();
 149
 150         /// Calculate magnetizations
 151         void solveM();
 152
 153         /// Calculate the product of all tJ[i][_j] for _j in A
 154         /** \param i variable index
 155          *  \param A subset of neighbors of variable \a i
 156          */
 157         Real _tJ(size_t i, sub_nb A);
 158
 159         /// Calculate \f$ \Omega^{(i)}_{j,l} \f$ as defined in [\ref MoR05] eqn. (2.15)
 160         Real Omega(size_t i, size_t _j, size_t _l);
 161
 162         /// Calculate \f$ T^{(i)}_A \f$ as defined in [\ref MoR05] eqn. (2.17) with \f$ A = \{l_1,l_2,\dots\} \f$
 163         /** \param i variable index
 164          *  \param A subset of neighbors of variable \a i
 165          */
 166         Real T(size_t i, sub_nb A);
 167
 168         /// Calculates \f$ T^{(i)}_j \f$ where \a j is the \a _j 'th neighbor of \a i
 169         Real T(size_t i, size_t _j);
 170
 171         /// Calculates \f$ \Gamma^{(i)}_{j,l_1l_2} \f$ as defined in [\ref MoR05] eqn. (2.16)
 172         Real Gamma(size_t i, size_t _j, size_t _l1, size_t _l2);
 173
 174         /// Calculates \f$ \Gamma^{(i)}_{l_1l_2} \f$ as defined in [\ref MoK07] on page 1141
 175         Real Gamma(size_t i, size_t _l1, size_t _l2);
 176
 177         /// Approximates moments of variables in \a A
 178         /** Calculate the moment of variables in \a A from M and cors, neglecting higher order cumulants,
 179          *  defined as the sum over all partitions of A into subsets of cardinality two at most of the
 180          *  product of the cumulants (either first order, i.e. M, or second order, i.e. cors) of the
 181          *  entries of the partitions.
 182          *
 183          *  \param i variable index
 184          *  \param A subset of neighbors of variable \a i
 185          */
 186         Real appM(size_t i, sub_nb A);
 187
 188         /// Calculate sum over all even/odd subsets B of \a A of _tJ(j,B) appM(j,B)
 189         /** \param j variable index
 190          *  \param A subset of neighbors of variable \a j
 191          *  \param sum_even on return, will contain the sum over all even subsets
 192          *  \param sum_odd on return, will contain the sum over all odd subsets
 193          */
 194         void sum_subs(size_t j, sub_nb A, Real *sum_even, Real *sum_odd);
 195 };
 196
 197
 198 } // end of namespace dai
 199
 200
 201 #endif