filters::statespace::NoiseCovMat< stateDim, size, paramSize > Class Template Reference

Represents a square noise covariance matrix. More...

#include <StateSpaceUtil.h>

Public Types
using	state_t = Eigen::Matrix<double, stateDim, 1>
using	param_t = Eigen::Matrix<double, paramSize, 1>

Public Member Functions
	NoiseCovMat (const Eigen::Matrix< double, size, 1 > &stdDevs)
	Create a time-invariant noise covariance matrix modelling independent noise with the given standard deviations.
	NoiseCovMat (const Eigen::Matrix< double, size, size > &mat)
	Create a time-invariant noise covariance matrix equal to the given matrix.
	NoiseCovMat (const std::function< Eigen::Matrix< double, size, size >(const state_t &, const param_t &)> &func)
	Create a time-varying noise covariance matrix.
Eigen::Matrix< double, size, size >	get (const state_t &x, const param_t &param)
	Gets the noise covariance matrix, given the current state and additonal parameter.

Detailed Description

template<int stateDim, int size, int paramSize>
class filters::statespace::NoiseCovMat< stateDim, size, paramSize >

Represents a square noise covariance matrix.

Returning the zero matrix can sometimes be dangerous depending on your model, as it may cause numerical instability or incorrect computations.

Template Parameters

stateDim	The dimension of the state space for this system. This is the number of elements in the state vector.
size	The size of this matrix.
paramSize	The dimension of the vector accepted by get() in addition to a state vector.

Constructor & Destructor Documentation

NoiseCovMat() [1/3]

template<int stateDim, int size, int paramSize>

filters::statespace::NoiseCovMat< stateDim, size, paramSize >::NoiseCovMat ( const Eigen::Matrix< double, size, 1 > & stdDevs )

inlineexplicit

Create a time-invariant noise covariance matrix modelling independent noise with the given standard deviations.

Parameters

stdDevs

The standard deviations of each element.

NoiseCovMat() [2/3]

template<int stateDim, int size, int paramSize>

filters::statespace::NoiseCovMat< stateDim, size, paramSize >::NoiseCovMat ( const Eigen::Matrix< double, size, size > & mat )

inline

Create a time-invariant noise covariance matrix equal to the given matrix.

Parameters

mat	The noise covariance matrix.

NoiseCovMat() [3/3]

template<int stateDim, int size, int paramSize>

filters::statespace::NoiseCovMat< stateDim, size, paramSize >::NoiseCovMat ( const std::function< Eigen::Matrix< double, size, size >(const state_t &, const param_t &)> & func )

inline

Create a time-varying noise covariance matrix.

At runtime, the matrix will be calculated when needed using the supplied function.

Parameters

func	The function that supplies the noise covariance matrix, given the state vector and one additional vector. For process noise, this is usually the input vector. For output noise, this is usually the output vector.

Member Function Documentation

get()

template<int stateDim, int size, int paramSize>

Eigen::MatrixXd filters::statespace::NoiseCovMat< stateDim, size, paramSize >::get ( const state_t & x, const param_t & param )

inline

Gets the noise covariance matrix, given the current state and additonal parameter.

The matrix may be time-invariant, which case the values of x and param do not matter.

Parameters

x	The current state vector.
param	The parameter vector, as defined by the use of this matrix. For process noise, this is usually the input vector. For output noise this is usually the output vector.

Returns

The noise covariance matrix.

---

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

• src/filters/StateSpaceUtil.h
• src/filters/StateSpaceUtil.cpp