1 | #include "particles.h" |
---|
2 | |
---|
3 | namespace bdm { |
---|
4 | |
---|
5 | using std::endl; |
---|
6 | |
---|
7 | void PF::bayes_gensmp(){ |
---|
8 | #pragma parallel for |
---|
9 | for (int i = 0; i < n; i++ ) { |
---|
10 | _samples ( i ) = par->samplecond ( _samples ( i ) ); |
---|
11 | lls(i) = 0; |
---|
12 | } |
---|
13 | } |
---|
14 | |
---|
15 | void PF::bayes_weights(){ |
---|
16 | // |
---|
17 | double mlls=max(lls); |
---|
18 | // compute weights |
---|
19 | for (int i = 0; i < n; i++ ) { |
---|
20 | _w ( i ) *= exp ( lls ( i ) - mlls ); // multiply w by likelihood |
---|
21 | } |
---|
22 | |
---|
23 | //renormalize |
---|
24 | double sw=sum(_w); |
---|
25 | if (!std::isfinite(sw)) { |
---|
26 | for (int i=0;i<n;i++){ |
---|
27 | if (!std::isfinite(_w(i))) {_w(i)=0;} |
---|
28 | } |
---|
29 | sw = sum(_w); |
---|
30 | if (!std::isfinite(sw)) { |
---|
31 | bdm_error("Particle filter is lost; no particle is good enough."); |
---|
32 | } |
---|
33 | } |
---|
34 | _w /= sw; |
---|
35 | } |
---|
36 | |
---|
37 | void PF::bayes ( const vec &dt ) { |
---|
38 | int i; |
---|
39 | // generate samples - time step |
---|
40 | bayes_gensmp(); |
---|
41 | // weight them - data step |
---|
42 | for ( i = 0; i < n; i++ ) { |
---|
43 | lls ( i ) += obs->evallogcond ( dt, _samples ( i ) ); //+= because lls may have something from gensmp! |
---|
44 | } |
---|
45 | |
---|
46 | bayes_weights(); |
---|
47 | |
---|
48 | if (do_resampling()) { |
---|
49 | est.resample ( resmethod); |
---|
50 | } |
---|
51 | |
---|
52 | } |
---|
53 | |
---|
54 | // void PF::set_est ( const epdf &epdf0 ) { |
---|
55 | // int i; |
---|
56 | // |
---|
57 | // for ( i=0;i<n;i++ ) { |
---|
58 | // _samples ( i ) = epdf0.sample(); |
---|
59 | // } |
---|
60 | // } |
---|
61 | |
---|
62 | void MPF::bayes ( const vec &dt ) { |
---|
63 | // follows PF::bayes in most places!!! |
---|
64 | |
---|
65 | int i; |
---|
66 | int n=pf->__w().length(); |
---|
67 | vec &lls = pf->_lls(); |
---|
68 | |
---|
69 | // generate samples - time step |
---|
70 | pf->bayes_gensmp(); |
---|
71 | // weight them - data step |
---|
72 | #pragma parallel for |
---|
73 | for ( i = 0; i < n; i++ ) { |
---|
74 | BMs(i) -> condition(pf->posterior()._sample(i)); |
---|
75 | BMs(i) -> bayes(dt); |
---|
76 | lls ( i ) += BMs(i)->_ll(); |
---|
77 | } |
---|
78 | |
---|
79 | pf->bayes_weights(); |
---|
80 | |
---|
81 | ivec ind; |
---|
82 | if (pf->do_resampling()) { |
---|
83 | pf->resample ( ind); |
---|
84 | |
---|
85 | #pragma omp parallel for |
---|
86 | for ( i = 0; i < n; i++ ) { |
---|
87 | if ( ind ( i ) != i ) {//replace the current Bm by a new one |
---|
88 | delete BMs ( i ); |
---|
89 | BMs ( i ) = BMs ( ind ( i ) )->_copy_(); //copy constructor |
---|
90 | } |
---|
91 | }; |
---|
92 | } |
---|
93 | }; |
---|
94 | |
---|
95 | |
---|
96 | } |
---|
97 | //MPF::MPF:{} |
---|