bdm::diffbifn Class Reference

Class representing a differentiable function of two variables $f(x,u)$

#include <libFN.h>

List of all members.

Public Member Functions

vec eval (const vec &cond)

Evaluates $f(x0,u0)$ (VS: Do we really need common eval? ).

virtual vec eval (const vec &x0, const vec &u0)

Evaluates .

virtual void dfdx_cond (const vec &x0, const vec &u0, mat &A, bool full=true)

Evaluates $A=\frac{d}{dx}f(x,u)|_{x0,u0}$ and writes result into A .

virtual void dfdu_cond (const vec &x0, const vec &u0, mat &A, bool full=true)

Evaluates $A=\frac{d}{du}f(x,u)|_{x0,u0}$ and writes result into A .

diffbifn ()

Default constructor (dimy is not set!).

int _dimx () const

access function

int _dimu () const

access function

virtual void condition (const vec &val)

function substitutes given value into an appropriate position

int dimension () const

access function

Protected Attributes

RV rvx

Indentifier of the first rv.

RV rvu

Indentifier of the second rv.

int dimx

cache for rvx.count()

int dimu

cache for rvu.count()

int dimy

Length of the output vector.

Detailed Description

Class representing a differentiable function of two variables $f(x,u)$

.

Function of two variables.

TODO: 1) Technically, it could have a common parent (e.g. fnc ) with other functions. For now, we keep it as it is. 2) It could be generalized into multivariate form, (which was original meaning of fnc ).

Member Function Documentation

virtual void bdm::diffbifn::dfdu_cond	(	const vec &	x0,
		const vec &	u0,
		mat &	A,
		bool	full = `true`
	)			`[inline, virtual]`

Evaluates $A=\frac{d}{du}f(x,u)|_{x0,u0}$ and writes result into A .

Parameters:

	full	denotes that even unchanged entries are to be rewritten. When, false only the changed elements are computed.
	x0	numeric value of ,
	u0	numeric value of
	A	a place where the result will be stored.

Reimplemented in IMk1, IMpmsm, IMpmsm2o, IMpmsmStat, and bdm::bilinfn.

virtual void bdm::diffbifn::dfdx_cond	(	const vec &	x0,
		const vec &	u0,
		mat &	A,
		bool	full = `true`
	)			`[inline, virtual]`

Evaluates $A=\frac{d}{dx}f(x,u)|_{x0,u0}$ and writes result into A .

Parameters:

	full	denotes that even unchanged entries are to be rewritten. When, false only the changed elements are computed.
	x0	numeric value of ,
	u0	numeric value of
	A	a place where the result will be stored.

Reimplemented in IMk1, OMk1, IMpmsm, IMpmsm2o, IMpmsmStat, OMpmsm, and bdm::bilinfn.

Referenced by bdm::EKF< sq_T >::bayes(), bdm::EKFCh::bayes(), bdm::EKFfull::bayes(), bdm::EKF< sq_T >::set_parameters(), bdm::EKFCh::set_parameters(), and bdm::EKFfull::set_parameters().

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

libFN.h


Public Member Functions
vec	eval (const vec &cond)
	Evaluates (VS: Do we really need common eval? ).
virtual vec	eval (const vec &x0, const vec &u0)
	Evaluates .
virtual void	dfdx_cond (const vec &x0, const vec &u0, mat &A, bool full=true)
	Evaluates $A=\frac{d}{dx}f(x,u)\|_{x0,u0}$ and writes result into `A` .
virtual void	dfdu_cond (const vec &x0, const vec &u0, mat &A, bool full=true)
	Evaluates $A=\frac{d}{du}f(x,u)\|_{x0,u0}$ and writes result into `A` .
	diffbifn ()
	Default constructor (dimy is not set!).
int	_dimx () const
	access function
int	_dimu () const
	access function
virtual void	condition (const vec &val)
	function substitutes given value into an appropriate position
int	dimension () const
	access function
Protected Attributes
RV	rvx
	Indentifier of the first rv.
RV	rvu
	Indentifier of the second rv.
int	dimx
	cache for rvx.count()
int	dimu
	cache for rvu.count()
int	dimy
	Length of the output vector.