Changeset 1077 for library/bdm/estim/arx.h
- Timestamp:
- 06/10/10 21:40:09 (14 years ago)
- Files:
-
- 1 modified
Legend:
- Unmodified
- Added
- Removed
-
library/bdm/estim/arx.h
r1064 r1077 27 27 Regression of the following kind: 28 28 \f[ 29 y_t = \theta_1 \psi_1 + \theta_2 + \psi_2 +\ldots + \theta_n \psi_n + r e_t29 y_t = heta_1 \psi_1 + heta_2 + \psi_2 +\ldots + heta_n \psi_n + r e_t 30 30 \f] 31 where unknown parameters \c rv are \f$[ \theta r]\f$, regression vector \f$\psi=\psi(y_{1:t},u_{1:t})\f$ is a known function of past outputs and exogeneous variables \f$u_t\f$. Distrubances \f$e_t\f$ are supposed to be normally distributed:31 where unknown parameters \c rv are \f$[ heta r]\f$, regression vector \f$\psi=\psi(y_{1:t},u_{1:t})\f$ is a known function of past outputs and exogeneous variables \f$u_t\f$. Distrubances \f$e_t\f$ are supposed to be normally distributed: 32 32 \f[ 33 33 e_t \sim \mathcal{N}(0,1). … … 39 39 \include arx_simple.cpp 40 40 41 \todo sort out constant terms - bayes should accept vec without additional 1s41 odo sort out constant terms - bayes should accept vec without additional 1s 42 42 */ 43 43 class ARX: public BMEF { … … 132 132 //!@} 133 133 134 /*! UI for ARX estimator134 /*! Create object from the following structure 135 135 136 136 \code 137 137 class = 'ARX'; 138 yrv = RV({names_of_dt} ) // description of output variables 139 rgr = RV({names_of_regressors}, [-1,-2]} // description of regressor variables 140 constant = 1; // 0/1 switch if the constant term is modelled or not 141 142 --- optional --- 143 prior = {class='egiw',...}; // Prior density, when given default is used instead 144 alternative = {class='egiw',...}; // Alternative density in stabilized estimation, when not given prior is used 145 146 frg = 1.0; // forgetting, default frg=1.0 147 148 rv = RV({names_of_parameters}} // description of parametetr names 149 // default: [""] 138 rgr = RV({'names',...},[sizes,...],[times,...]); % description of regressor variables 139 --- optional fields --- 140 prior = configuration of bdm::egiw; % any offspring of eqiw for prior density, bdm::egiw::from_setting 141 alternative = configuration of bdm::egiw; % any offspring of eqiw for alternative density in stabilized estimation of prior density 142 constant = []; % 0/1 switch if the constant term is modelled or not 143 --- inherited fields --- 144 bdm::BMEF::from_setting 145 \endcode 146 If the optional fields are not given, they will be filled as follows: 147 \code 148 prior = posterior; % when prior is not given the posterior is used (TODO it is unclear) 149 alternative = prior; % when alternative is not given the prior is used 150 constant = 1; % constant term is modelled on default 150 151 \endcode 151 152 */ … … 208 209 SHAREDPTR ( ARX ); 209 210 210 /*! ARX model conditined by knowledge of the forgetting factor211 \f[ f( \theta| d_1 \ldots d_t , \phi_t) \f]211 /*! \brief ARX model conditined by knowledge of the forgetting factor 212 \f[ f( heta| d_1 \ldots d_t , \phi_t) \f] 212 213 213 214 The symbol \f$ \phi \f$ is assumed to be the last of the conditioning variables.