#include <arx.h>


Public Member Functions | |
| ARX (const RV &rv, const mat &V0, const double &nu0, const double frg0=1.0) | |
| Full constructor. | |
| ARX (const ARX &A0) | |
| Copy constructor. | |
| ARX * | _copy_ (bool changerv=false) |
| Auxiliary function. | |
| void | set_parameters (const ldmat &V0, const double &nu0) |
| Set sufficient statistics. | |
| void | set_statistics (const BMEF *BM0) |
| get statistics from another model | |
| void | get_parameters (mat &V0, double &nu0) |
| Returns sufficient statistics. | |
| void | bayes (const vec &dt, const double w) |
Here . | |
| void | bayes (const vec &dt) |
| Incremental Bayes rule. | |
| const epdf & | _epdf () const |
| Returns a pointer to the epdf representing posterior density on parameters. Use with care! | |
| double | logpred (const vec &dt) const |
| void | flatten (const BMEF *B) |
| Flatten the posterior according to the given BMEF (of the same type!). | |
| enorm< ldmat > * | predictor (const RV &rv) |
| Constructs a predictive density (marginal density on data). | |
| ivec | structure_est (egiw Eg0) |
| Brute force structure estimation. | |
| virtual void | flatten (double nu0) |
| Flatten the posterior as if to keep nu0 data. | |
| virtual void | bayesB (const mat &Dt) |
| Batch Bayes rule (columns of Dt are observations). | |
| vec | logpred_m (const mat &dt) const |
| Matrix version of logpred. | |
| const RV & | _rv () const |
| access function | |
| double | _ll () const |
| access function | |
| void | set_evalll (bool evl0) |
| access function | |
Protected Attributes | |
| egiw | est |
Posterior estimate of in the form of Normal-inverse Wishart density. | |
| ldmat & | V |
| cached value of est.V | |
| double & | nu |
| cached value of est.nu | |
| double | frg |
| forgetting factor | |
| double | last_lognc |
cached value of lognc() in the previous step (used in evaluation of ll ) | |
| RV | rv |
| Random variable of the posterior. | |
| double | ll |
| Logarithm of marginalized data likelihood. | |
| bool | evalll |
If true, the filter will compute likelihood of the data record and store it in ll . Set to false if you want to save computational time. | |
Regression of the following kind:
where unknown parameters rv are
, regression vector
is a known function of past outputs and exogeneous variables
. Distrubances
are supposed to be normally distributed:
Extension for time-variant parameters
may be achived using exponential forgetting (Kulhavy and Zarrop, 1993). In such a case, the forgetting factor frg
should be given in the constructor. Time-invariant parameters are estimated for frg = 1.
| void ARX::bayes | ( | const vec & | dt | ) | [inline, virtual] |
| double ARX::logpred | ( | const vec & | dt | ) | const [virtual] |
Evaluates predictive log-likelihood of the given data record I.e. marginal likelihood of the data with the posterior integrated out.
Reimplemented from BM.
References egiw::_nu(), egiw::_V(), est, BM::evalll, BMEF::frg, BMEF::last_lognc, egiw::lognc(), nu, ldmat::opupdt(), egiw::pow(), and V.
| ivec ARX::structure_est | ( | egiw | Eg0 | ) |
Brute force structure estimation.
References RV::count(), est, egiw::lognc(), and BM::rv.
1.5.6