MUQ  0.4.3
Static Public Member Functions | List of all members
muq::Inference::KalmanSmoother Class Reference

#include <KalmanSmoother.h>

Detailed Description

Implements the Rauch–Tung–Striebel smoother.

Definition at line 14 of file KalmanSmoother.h.

Static Public Member Functions

static std::pair< Eigen::VectorXd, Eigen::MatrixXd > Analyze (std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &currDist_t, std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &nextDist_t, std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &nextDist_n, std::shared_ptr< muq::Modeling::LinearOperator > F)
 
template<typename MatrixType >
static std::pair< Eigen::VectorXd, Eigen::MatrixXd > Analyze (std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &currDist_t, std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &nextDist_t, std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &nextDist_n, MatrixType const &F)
 

Member Function Documentation

◆ Analyze() [1/2]

template<typename MatrixType >
static std::pair<Eigen::VectorXd, Eigen::MatrixXd> muq::Inference::KalmanSmoother::Analyze ( std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  currDist_t,
std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  nextDist_t,
std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  nextDist_n,
MatrixType const &  F 
)
inlinestatic
Template Parameters
MatrixTypeA type that can be converted to a MUQ LinearOperator. Examples include Eigen::MatrixXd and Eigen::SparseMatrix

Definition at line 34 of file KalmanSmoother.h.

References Analyze().

◆ Analyze() [2/2]

std::pair< Eigen::VectorXd, Eigen::MatrixXd > KalmanSmoother::Analyze ( std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  currDist_t,
std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  nextDist_t,
std::pair< Eigen::VectorXd, Eigen::MatrixXd > const &  nextDist_n,
std::shared_ptr< muq::Modeling::LinearOperator >  F 
)
static
Parameters
[in]currDist_tThe distribution at time t after the forward Kalman filtering step (i.e., using all data up to and including time t).
[in]nextDist_tThe distribution at time t+1 from the prediction phase of the Kalman filtering step (i.e., using all data up to time t).
[in]nextDist_nThe distribution at time t+1 after the RTS smoothing step (i.e., using all data).
[in]FThe linear operator acting on the state at time t, to produce the state at time t+1
Returns
A distribution (mean and covariance) at time t after accounting for all data.

Definition at line 9 of file KalmanSmoother.cpp.

Referenced by Analyze(), and muq::Inference::PythonBindings::KalmanWrapper().

◆ ComputeC()

Eigen::MatrixXd KalmanSmoother::ComputeC ( Eigen::MatrixXd const &  currDist_t_cov,
Eigen::MatrixXd const &  nextDist_t_cov,
std::shared_ptr< muq::Modeling::LinearOperator > const &  F 
)
staticprivate

Definition at line 26 of file KalmanSmoother.cpp.

The documentation for this class was generated from the following files: