ipsdk::math::NonLinearSolver Class Reference

Class implementing a non linear solver based on Levenberg Marquardt algorithm. More...

#include <NonLinearSolver.h>

Inheritance diagram for ipsdk::math::NonLinearSolver:

Public Member Functions
eOptStatus	getStatus () const
	check whether solve has been done
OptResult	solve (const CostFunctionConstPtr &pCostFunction, const Vector &initParams)
	solve of optimization process More...
ipUInt32	getNbIteration () const
	retrieve number of iterations of optimization process More...
const Vector &	getEstimParams () const
	retrieve estimated parameters More...
void	clearResults ()
	method allowing to clear previous optimization process results
Constructors and destructor
	NonLinearSolver ()
	~NonLinearSolver ()
const ipUInt32	getNbMaxIteration () const
	access to maximum number of iteration for optimization process More...
void	setNbMaxIteration (const ipUInt32 nbMaxIteration)
	access to maximum number of iteration for optimization process More...

Static Protected Member Functions
static ipInt32	minPackProxy (void *pThis, ipInt32 m, ipInt32 n, ipReal64 *x, ipReal64 *fvec, ipReal64 *fjac, ipInt32 ldfjac, ipInt32 iflag)
	function called by minpack to ensure function computation

Protected Attributes
ipUInt32	_nbMaxIteration
	maximum number of iteration for optimization process More...
eOptStatus	_optStatus
	status associated to object
ipUInt32	_nbIteration
	number of iterations associated to optimization process
Vector	_estimParams
	estimated parameters
CostFunctionConstPtr	_pCostFunction
	cost function used during optimization process
std::string	_errorMsg
	internal function error reason
Vector	_funValue
Vector	_wa2
Vector	_wa4
Matrix	_jacValue

Detailed Description

Class implementing a non linear solver based on Levenberg Marquardt algorithm.

Author

E. Noirfalise

Date

2016/02/18

Copyright

2016 Reactiv'IP, all right reserved.

Member Function Documentation

getNbMaxIteration()

const ipUInt32 ipsdk::math::NonLinearSolver::getNbMaxIteration ( ) const

inline

access to maximum number of iteration for optimization process

Note

default : 100

setNbMaxIteration()

void ipsdk::math::NonLinearSolver::setNbMaxIteration ( const ipUInt32  nbMaxIteration )

inline

access to maximum number of iteration for optimization process

Note

default : 100

solve()

OptResult ipsdk::math::NonLinearSolver::solve	(	const CostFunctionConstPtr &	pCostFunction,
		const Vector &	initParams
	)

solve of optimization process

Exceptions

ipsdk::math::IPSDKMathException	if pCostFunction.get() == 0
ipsdk::math::IPSDKMathException	if pCostFunction->getNbParameters() != initParameters.size()

Returns

optimization result

getNbIteration()

ipUInt32 ipsdk::math::NonLinearSolver::getNbIteration

(

)

const

retrieve number of iterations of optimization process

Exceptions

ipsdk::math::IPSDKMathException

if getStatus() == eOptStatus::eOS_Failed

getEstimParams()

const Vector& ipsdk::math::NonLinearSolver::getEstimParams

(

)

const

retrieve estimated parameters

Exceptions

ipsdk::math::IPSDKMathException

if getStatus() == eOptStatus::eOS_Failed

Member Data Documentation

_nbMaxIteration

ipUInt32 ipsdk::math::NonLinearSolver::_nbMaxIteration

protected

maximum number of iteration for optimization process

Note

default : 100

_funValue

Vector ipsdk::math::NonLinearSolver::_funValue

protected

internal working arrays

_wa2

Vector ipsdk::math::NonLinearSolver::_wa2

protected

internal working arrays

_wa4

Vector ipsdk::math::NonLinearSolver::_wa4

protected

internal working arrays

_jacValue

Matrix ipsdk::math::NonLinearSolver::_jacValue

protected

internal working arrays

---

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

• /media/reactivipdev/Nightly/IPSDK_Patch/IPSDK_4_1/IPSDK_Nightly/include/public/IPSDKMath/Optimization/NonLinearSolver.h