1 | /*! |
---|
2 | \file |
---|
3 | \brief Mergers for combination of pdfs |
---|
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 MERGER_H |
---|
14 | #define MERGER_H |
---|
15 | |
---|
16 | |
---|
17 | #include "../estim/mixtures.h" |
---|
18 | #include "discrete.h" |
---|
19 | |
---|
20 | namespace bdm { |
---|
21 | using std::string; |
---|
22 | |
---|
23 | //!Merging methods |
---|
24 | enum MERGER_METHOD {ARITHMETIC = 1, GEOMETRIC = 2, LOGNORMAL = 3}; |
---|
25 | |
---|
26 | /*! |
---|
27 | @brief Base class for general combination of pdfs on discrete support |
---|
28 | |
---|
29 | Mixtures of Gaussian densities are used internally. Switching to other densities should be trivial. |
---|
30 | |
---|
31 | The merged pdfs are expected to be of the form: |
---|
32 | \f[ f(x_i|y_i), i=1..n \f] |
---|
33 | where the resulting merger is a density on \f$ \cup [x_i,y_i] \f$ . |
---|
34 | Note that all variables will be joined. |
---|
35 | |
---|
36 | As a result of this feature, each source must be extended to common support |
---|
37 | \f[ f(z_i|y_i,x_i) f(x_i|y_i) f(y_i) i=1..n \f] |
---|
38 | where \f$ z_i \f$ accumulate variables that were not in the original source. |
---|
39 | These extensions are calculated on-the-fly. |
---|
40 | |
---|
41 | However, these operations can not be performed in general. Hence, this class merges only sources on common support, \f$ y_i={}, z_i={}, \forall i \f$. |
---|
42 | For merging of more general cases, use offsprings merger_mix and merger_grid. |
---|
43 | */ |
---|
44 | |
---|
45 | class merger_base : public epdf { |
---|
46 | protected: |
---|
47 | //! Elements of composition |
---|
48 | Array<shared_ptr<pdf> > pdfs; |
---|
49 | |
---|
50 | //! Data link for each pdf in pdfs |
---|
51 | Array<datalink_m2e*> dls; |
---|
52 | |
---|
53 | //! Array of rvs that are not modelled by pdfs at all, \f$ z_i \f$ |
---|
54 | Array<RV> rvzs; |
---|
55 | |
---|
56 | //! Data Links for extension \f$ f(z_i|x_i,y_i) \f$ |
---|
57 | Array<datalink_m2e*> zdls; |
---|
58 | |
---|
59 | //! number of support points |
---|
60 | int Npoints; |
---|
61 | |
---|
62 | //! number of sources |
---|
63 | int Nsources; |
---|
64 | |
---|
65 | //! switch of the methoh used for merging |
---|
66 | MERGER_METHOD METHOD; |
---|
67 | //! Default for METHOD |
---|
68 | static const MERGER_METHOD DFLT_METHOD; |
---|
69 | |
---|
70 | //!Prior on the log-normal merging model |
---|
71 | double beta; |
---|
72 | //! default for beta |
---|
73 | static const double DFLT_beta; |
---|
74 | |
---|
75 | //! Projection to empirical density (could also be piece-wise linear) |
---|
76 | eEmp eSmp; |
---|
77 | |
---|
78 | //! debug or not debug |
---|
79 | bool DBG; |
---|
80 | |
---|
81 | //! debugging file |
---|
82 | it_file* dbg_file; |
---|
83 | public: |
---|
84 | //! \name Constructors |
---|
85 | //! @{ |
---|
86 | |
---|
87 | //! Default constructor |
---|
88 | merger_base () : Npoints ( 0 ), Nsources ( 0 ), DBG ( false ), dbg_file ( 0 ) { |
---|
89 | } |
---|
90 | |
---|
91 | //!Constructor from sources |
---|
92 | merger_base ( const Array<shared_ptr<pdf> > &S ); |
---|
93 | |
---|
94 | //! Function setting the main internal structures |
---|
95 | void set_sources ( const Array<shared_ptr<pdf> > &Sources ); |
---|
96 | |
---|
97 | //! Set support points from rectangular grid |
---|
98 | void set_support ( rectangular_support &Sup ); |
---|
99 | |
---|
100 | //! Set support points from dicrete grid |
---|
101 | void set_support ( discrete_support &Sup ) { |
---|
102 | Npoints = Sup.points(); |
---|
103 | eSmp.set_parameters ( Sup._Spoints() ); |
---|
104 | eSmp.validate(); |
---|
105 | } |
---|
106 | //! set debug file |
---|
107 | void set_debug_file ( const string fname ) { |
---|
108 | if ( DBG ) delete dbg_file; |
---|
109 | dbg_file = new it_file ( fname ); |
---|
110 | if ( dbg_file ) DBG = true; |
---|
111 | } |
---|
112 | //! Set internal parameters used in approximation |
---|
113 | void set_method ( MERGER_METHOD MTH = DFLT_METHOD, double beta0 = DFLT_beta ) { |
---|
114 | METHOD = MTH; |
---|
115 | beta = beta0; |
---|
116 | } |
---|
117 | //! Set support points from a pdf by drawing N samples |
---|
118 | void set_support ( const epdf &overall, int N ) { |
---|
119 | eSmp.set_statistics ( overall, N ); |
---|
120 | Npoints = N; |
---|
121 | } |
---|
122 | |
---|
123 | //! Destructor |
---|
124 | virtual ~merger_base() { |
---|
125 | for ( int i = 0; i < Nsources; i++ ) { |
---|
126 | delete dls ( i ); |
---|
127 | delete zdls ( i ); |
---|
128 | } |
---|
129 | if ( DBG ) delete dbg_file; |
---|
130 | }; |
---|
131 | //!@} |
---|
132 | |
---|
133 | //! \name Mathematical operations |
---|
134 | //!@{ |
---|
135 | |
---|
136 | //!Merge given sources in given points |
---|
137 | virtual void merge (); |
---|
138 | |
---|
139 | //! Merge log-likelihood values in points using method specified by parameter METHOD |
---|
140 | vec merge_points ( mat &lW ); |
---|
141 | |
---|
142 | |
---|
143 | //! sample from merged density |
---|
144 | //! weight w is a |
---|
145 | vec mean() const; |
---|
146 | |
---|
147 | mat covariance() const; |
---|
148 | |
---|
149 | vec variance() const; |
---|
150 | |
---|
151 | //!@} |
---|
152 | |
---|
153 | //! \name Access to attributes |
---|
154 | //! @{ |
---|
155 | |
---|
156 | //! Access function |
---|
157 | eEmp& _Smp() { |
---|
158 | return eSmp; |
---|
159 | } |
---|
160 | |
---|
161 | //! load from setting |
---|
162 | void from_setting ( const Setting& set ) { |
---|
163 | // get support |
---|
164 | // find which method to use |
---|
165 | string meth_str; |
---|
166 | UI::get<string> ( meth_str, set, "method", UI::compulsory ); |
---|
167 | if ( !strcmp ( meth_str.c_str(), "arithmetic" ) ) |
---|
168 | set_method ( ARITHMETIC ); |
---|
169 | else { |
---|
170 | if ( !strcmp ( meth_str.c_str(), "geometric" ) ) |
---|
171 | set_method ( GEOMETRIC ); |
---|
172 | else if ( !strcmp ( meth_str.c_str(), "lognormal" ) ) { |
---|
173 | set_method ( LOGNORMAL ); |
---|
174 | set.lookupValue ( "beta", beta ); |
---|
175 | } |
---|
176 | } |
---|
177 | string dbg_file; |
---|
178 | if ( UI::get ( dbg_file, set, "dbg_file" ) ) |
---|
179 | set_debug_file ( dbg_file ); |
---|
180 | //validate() - not used |
---|
181 | } |
---|
182 | |
---|
183 | void validate() { |
---|
184 | bdm_assert ( eSmp._w().length() > 0, "Empty support, use set_support()." ); |
---|
185 | bdm_assert ( dim == eSmp._samples() ( 0 ).length(), "Support points and rv are not compatible!" ); |
---|
186 | bdm_assert ( isnamed(), "mergers must be named" ); |
---|
187 | } |
---|
188 | //!@} |
---|
189 | }; |
---|
190 | UIREGISTER ( merger_base ); |
---|
191 | SHAREDPTR ( merger_base ); |
---|
192 | |
---|
193 | //! Merger using importance sampling with mixture proposal density |
---|
194 | class merger_mix : public merger_base { |
---|
195 | protected: |
---|
196 | //!Internal mixture of EF models |
---|
197 | MixEF Mix; |
---|
198 | //!Number of components in a mixture |
---|
199 | int Ncoms; |
---|
200 | //! coefficient of resampling [0,1] |
---|
201 | double effss_coef; |
---|
202 | //! stop after niter iterations |
---|
203 | int stop_niter; |
---|
204 | |
---|
205 | //! default value for Ncoms |
---|
206 | static const int DFLT_Ncoms; |
---|
207 | //! default value for efss_coef; |
---|
208 | static const double DFLT_effss_coef; |
---|
209 | |
---|
210 | public: |
---|
211 | //!\name Constructors |
---|
212 | //!@{ |
---|
213 | merger_mix () : Ncoms ( 0 ), effss_coef ( 0 ), stop_niter ( 0 ) { } |
---|
214 | |
---|
215 | merger_mix ( const Array<shared_ptr<pdf> > &S ) : |
---|
216 | Ncoms ( 0 ), effss_coef ( 0 ), stop_niter ( 0 ) { |
---|
217 | set_sources ( S ); |
---|
218 | } |
---|
219 | |
---|
220 | //! Set sources and prepare all internal structures |
---|
221 | void set_sources ( const Array<shared_ptr<pdf> > &S ) { |
---|
222 | merger_base::set_sources ( S ); |
---|
223 | //Nsources = S.length(); |
---|
224 | } |
---|
225 | |
---|
226 | //! Set internal parameters used in approximation |
---|
227 | void set_parameters ( int Ncoms0 = DFLT_Ncoms, double effss_coef0 = DFLT_effss_coef ) { |
---|
228 | Ncoms = Ncoms0; |
---|
229 | effss_coef = effss_coef0; |
---|
230 | } |
---|
231 | //!@} |
---|
232 | |
---|
233 | //! \name Mathematical operations |
---|
234 | //!@{ |
---|
235 | |
---|
236 | //!Merge values using mixture approximation |
---|
237 | void merge (); |
---|
238 | |
---|
239 | //! sample from the approximating mixture |
---|
240 | vec sample () const { |
---|
241 | return Mix.posterior().sample(); |
---|
242 | } |
---|
243 | //! loglikelihood computed on mixture models |
---|
244 | double evallog ( const vec &yt ) const { |
---|
245 | vec dtf = ones ( yt.length() + 1 ); |
---|
246 | dtf.set_subvector ( 0, yt ); |
---|
247 | return Mix.logpred ( dtf ); |
---|
248 | } |
---|
249 | //!@} |
---|
250 | |
---|
251 | //!\name Access functions |
---|
252 | //!@{ |
---|
253 | //! Access function |
---|
254 | MixEF& _Mix() { |
---|
255 | return Mix; |
---|
256 | } |
---|
257 | //! Access function |
---|
258 | emix* proposal() { |
---|
259 | emix* tmp = Mix.epredictor(); |
---|
260 | tmp->set_rv ( rv ); |
---|
261 | return tmp; |
---|
262 | } |
---|
263 | //! from_settings |
---|
264 | void from_setting ( const Setting& set ) { |
---|
265 | merger_base::from_setting ( set ); |
---|
266 | Ncoms=DFLT_Ncoms; |
---|
267 | set.lookupValue ( "ncoms", Ncoms ); |
---|
268 | effss_coef=DFLT_effss_coef; |
---|
269 | set.lookupValue ( "effss_coef", effss_coef ); |
---|
270 | stop_niter=10; |
---|
271 | set.lookupValue ( "stop_niter", stop_niter ); |
---|
272 | } |
---|
273 | |
---|
274 | //! @} |
---|
275 | |
---|
276 | }; |
---|
277 | UIREGISTER ( merger_mix ); |
---|
278 | SHAREDPTR ( merger_mix ); |
---|
279 | |
---|
280 | } |
---|
281 | |
---|
282 | #endif // MER_H |
---|