1 | /*! |
---|
2 | \file |
---|
3 | \brief Probability distributions for Exponential Family models. |
---|
4 | \author Vaclav Smidl. |
---|
5 | |
---|
6 | ----------------------------------- |
---|
7 | BDM++ - C++ library for Bayesian Decision Making under Uncertainty |
---|
8 | |
---|
9 | Using IT++ for numerical operations |
---|
10 | ----------------------------------- |
---|
11 | */ |
---|
12 | |
---|
13 | #ifndef EF_H |
---|
14 | #define EF_H |
---|
15 | |
---|
16 | #include <itpp/itbase.h> |
---|
17 | #include "libDC.h" |
---|
18 | #include "libBM.h" |
---|
19 | //#include <std> |
---|
20 | |
---|
21 | using namespace itpp; |
---|
22 | |
---|
23 | /*! |
---|
24 | * \brief General conjugate exponential family posterior density. |
---|
25 | |
---|
26 | * More?... |
---|
27 | */ |
---|
28 | class eEF : public epdf { |
---|
29 | |
---|
30 | public: |
---|
31 | virtual void tupdate( double phi, mat &vbar, double nubar ) {}; |
---|
32 | virtual void dupdate( mat &v,double nu=1.0 ) {}; |
---|
33 | }; |
---|
34 | |
---|
35 | class mEF : public mpdf { |
---|
36 | |
---|
37 | public: |
---|
38 | |
---|
39 | }; |
---|
40 | |
---|
41 | /*! |
---|
42 | * \brief General exponential family density |
---|
43 | |
---|
44 | * More?... |
---|
45 | */ |
---|
46 | template<class sq_T> |
---|
47 | class enorm : public eEF { |
---|
48 | int dim; |
---|
49 | vec mu; |
---|
50 | sq_T R; |
---|
51 | public: |
---|
52 | enorm( RV &rv, vec &mu, sq_T &R ); |
---|
53 | enorm(); |
---|
54 | void tupdate( double phi, mat &vbar, double nubar ); |
---|
55 | void dupdate( mat &v,double nu=1.0 ); |
---|
56 | vec sample(); |
---|
57 | mat sample(int N); |
---|
58 | double eval( const vec &val ); |
---|
59 | Normal_RNG RNG; |
---|
60 | }; |
---|
61 | |
---|
62 | /*! |
---|
63 | \brief |
---|
64 | */ |
---|
65 | template<class sq_T> |
---|
66 | class mlnorm : public mEF { |
---|
67 | enorm<sq_T> epdf; |
---|
68 | mat A; |
---|
69 | public: |
---|
70 | //! Constructor |
---|
71 | mlnorm( RV &rv,RV &rvc, mat &A, sq_T &R ); |
---|
72 | //!Generate one sample of the posterior |
---|
73 | vec samplecond( vec &cond, double &lik ); |
---|
74 | mat samplecond( vec &cond, vec &lik, int n ); |
---|
75 | void condition( vec &cond ); |
---|
76 | }; |
---|
77 | |
---|
78 | //////////////////////// |
---|
79 | |
---|
80 | template<class sq_T> |
---|
81 | enorm<sq_T>::enorm( RV &rv, vec &mu0, sq_T &R0 ) { |
---|
82 | dim = rv.count(); |
---|
83 | mu = mu0; |
---|
84 | R = R0; |
---|
85 | }; |
---|
86 | |
---|
87 | template<class sq_T> |
---|
88 | void enorm<sq_T>::dupdate( mat &v, double nu ) { |
---|
89 | // |
---|
90 | }; |
---|
91 | |
---|
92 | template<class sq_T> |
---|
93 | void enorm<sq_T>::tupdate( double phi, mat &vbar, double nubar ) { |
---|
94 | // |
---|
95 | }; |
---|
96 | |
---|
97 | template<class sq_T> |
---|
98 | vec enorm<sq_T>::sample() { |
---|
99 | vec x( dim ); |
---|
100 | RNG.sample_vector( dim,x ); |
---|
101 | vec smp = R.sqrt_mult( x ); |
---|
102 | |
---|
103 | smp += mu; |
---|
104 | return smp; |
---|
105 | }; |
---|
106 | |
---|
107 | template<class sq_T> |
---|
108 | mat enorm<sq_T>::sample( int N ) { |
---|
109 | mat X( dim,N ); |
---|
110 | vec x( dim ); |
---|
111 | vec pom; |
---|
112 | int i; |
---|
113 | for ( i=0;i<N;i++ ) { |
---|
114 | RNG.sample_vector( dim,x ); |
---|
115 | pom = R.sqrt_mult( x ); |
---|
116 | pom +=mu; |
---|
117 | X.set_col( i, pom); |
---|
118 | } |
---|
119 | return X; |
---|
120 | }; |
---|
121 | |
---|
122 | template<class sq_T> |
---|
123 | double enorm<sq_T>::eval( const vec &val ) { |
---|
124 | // |
---|
125 | }; |
---|
126 | |
---|
127 | |
---|
128 | template<class sq_T> |
---|
129 | enorm<sq_T>::enorm() {}; |
---|
130 | |
---|
131 | template<class sq_T> |
---|
132 | mlnorm<sq_T>::mlnorm( RV &rv,RV &rvc, mat &A, sq_T &R ) { |
---|
133 | int dim = rv.count(); |
---|
134 | vec mu( dim ); |
---|
135 | |
---|
136 | epdf = enorm<sq_T>( rv,mu,R ); |
---|
137 | } |
---|
138 | |
---|
139 | template<class sq_T> |
---|
140 | vec mlnorm<sq_T>::samplecond( vec &cond, double &lik ) { |
---|
141 | this->condition( cond ); |
---|
142 | vec smp = epdf.sample(); |
---|
143 | lik = epdf.eval( smp ); |
---|
144 | return smp; |
---|
145 | } |
---|
146 | |
---|
147 | template<class sq_T> |
---|
148 | mat mlnorm<sq_T>::samplecond( vec &cond, vec &lik, int n ) { |
---|
149 | int i; |
---|
150 | int dim = rv.count(); |
---|
151 | mat Smp( dim,n ); |
---|
152 | vec smp( dim ); |
---|
153 | this->condition( cond ); |
---|
154 | for ( i=0; i<dim; i++ ) { |
---|
155 | smp = epdf.sample(); |
---|
156 | lik( i ) = epdf.eval( smp ); |
---|
157 | Smp.set_col( i ,smp ); |
---|
158 | } |
---|
159 | return Smp; |
---|
160 | } |
---|
161 | |
---|
162 | template<class sq_T> |
---|
163 | void mlnorm<sq_T>::condition( vec &cond ) { |
---|
164 | epdf.mu = A*cond; |
---|
165 | //R is already assigned; |
---|
166 | } |
---|
167 | |
---|
168 | #endif //EF_H |
---|