SimTK::OptimizerSystem Class Reference

Abstract class which defines an objective/cost function which is optimized by and Optimizer object. More...

#include <Optimizer.h>

List of all members.

Public Member Functions
	OptimizerSystem ()
	OptimizerSystem (int nParameters)
virtual	~OptimizerSystem ()
virtual int	objectiveFunc (const Vector &parameters, bool new_parameters, Real &f) const
	Objective/cost function which is to be optimized; return 0 when successful.
virtual int	gradientFunc (const Vector &parameters, bool new_parameters, Vector &gradient) const
	Computes the gradient of the objective function; return 0 when successful.
virtual int	constraintFunc (const Vector &parameters, bool new_parameters, Vector &constraints) const
	Computes the value of the constraints; return 0 when successful.
virtual int	constraintJacobian (const Vector &parameters, bool new_parameters, Matrix &jac) const
	Computes Jacobian of the constraints; return 0 when successful.
virtual int	hessian (const Vector &parameters, bool new_parameters, Vector &gradient) const
	Computes Hessian of the objective function; return 0 when successful.
void	setNumParameters (const int nParameters)
	Sets the number of parameters in the objective function.
void	setNumEqualityConstraints (const int n)
	Sets the number of equality constraints.
void	setNumInequalityConstraints (const int n)
	Sets the number of inequality constraints.
void	setNumLinearEqualityConstraints (const int n)
	Sets the number of lineaer equality constraints.
void	setNumLinearInequalityConstraints (const int n)
	Sets the number of lineaer inequality constraints.
void	setParameterLimits (const Vector &lower, const Vector &upper)
	Set the upper and lower bounds on the paramters.
int	getNumParameters () const
	Returns the number of parameters, that is, the number of variables that the Optimizer may adjust while searching for a solution.
int	getNumConstraints () const
	Returns the total number of constraints.
int	getNumEqualityConstraints () const
	Returns the number of equality constraints.
int	getNumInequalityConstraints () const
	Returns the number of inequality constraints.
int	getNumLinearEqualityConstraints () const
	Returns the number of linear equality constraints.
int	getNumNonlinearEqualityConstraints () const
	Returns the number of nonlinear equality constraints.
int	getNumLinearInequalityConstraints () const
	Returns the number of linear inequality constraints.
int	getNumNonlinearInequalityConstraints () const
	Returns the number of linear inequality constraints.
bool	getHasLimits () const
	Returns true if there are limits on the parameters.
void	getParameterLimits (Real **lower, Real **upper) const
	Returns the limits on the allowed values of each parameter, as an array of lower bounds and an array of upper bounds, with assumed lengths matching the number of parameters.

---

Detailed Description

Abstract class which defines an objective/cost function which is optimized by and Optimizer object.

The OptimizerSystem also defines any constraints which must be satisfied.

---

Constructor & Destructor Documentation

SimTK::OptimizerSystem::OptimizerSystem

(

)

[inline]

SimTK::OptimizerSystem::OptimizerSystem

(

int

nParameters

)

[inline, explicit]

virtual SimTK::OptimizerSystem::~OptimizerSystem

(

)

[inline, virtual]

---

Member Function Documentation

virtual int SimTK::OptimizerSystem::objectiveFunc	(	const Vector &	parameters,
		bool	new_parameters,
		Real &	f
	)		const [inline, virtual]

Objective/cost function which is to be optimized; return 0 when successful.

This method must be supplied by concrete class.

virtual int SimTK::OptimizerSystem::gradientFunc	(	const Vector &	parameters,
		bool	new_parameters,
		Vector &	gradient
	)		const [inline, virtual]

Computes the gradient of the objective function; return 0 when successful.

This method does not have to be supplied if a numerical gradient is used.

virtual int SimTK::OptimizerSystem::constraintFunc	(	const Vector &	parameters,
		bool	new_parameters,
		Vector &	constraints
	)		const [inline, virtual]

Computes the value of the constraints; return 0 when successful.

This method must be supplied if the objective function has constraints.

virtual int SimTK::OptimizerSystem::constraintJacobian	(	const Vector &	parameters,
		bool	new_parameters,
		Matrix &	jac
	)		const [inline, virtual]

Computes Jacobian of the constraints; return 0 when successful.

This method does not have to be supplied if a numerical jacobian is used.

virtual int SimTK::OptimizerSystem::hessian	(	const Vector &	parameters,
		bool	new_parameters,
		Vector &	gradient
	)		const [inline, virtual]

Computes Hessian of the objective function; return 0 when successful.

This method does not have to be supplied if limited memory is used.

void SimTK::OptimizerSystem::setNumParameters

(

const int

nParameters

)

[inline]

Sets the number of parameters in the objective function.

void SimTK::OptimizerSystem::setNumEqualityConstraints

(

const int

n

)

[inline]

Sets the number of equality constraints.

void SimTK::OptimizerSystem::setNumInequalityConstraints

(

const int

n

)

[inline]

Sets the number of inequality constraints.

void SimTK::OptimizerSystem::setNumLinearEqualityConstraints

(

const int

n

)

[inline]

Sets the number of lineaer equality constraints.

void SimTK::OptimizerSystem::setNumLinearInequalityConstraints

(

const int

n

)

[inline]

Sets the number of lineaer inequality constraints.

void SimTK::OptimizerSystem::setParameterLimits	(	const Vector &	lower,
		const Vector &	upper
	)		[inline]

Set the upper and lower bounds on the paramters.

int SimTK::OptimizerSystem::getNumParameters

(

)

const [inline]

Returns the number of parameters, that is, the number of variables that the Optimizer may adjust while searching for a solution.

int SimTK::OptimizerSystem::getNumConstraints

(

)

const [inline]

Returns the total number of constraints.

int SimTK::OptimizerSystem::getNumEqualityConstraints

(

)

const [inline]

Returns the number of equality constraints.

int SimTK::OptimizerSystem::getNumInequalityConstraints

(

)

const [inline]

Returns the number of inequality constraints.

int SimTK::OptimizerSystem::getNumLinearEqualityConstraints

(

)

const [inline]

Returns the number of linear equality constraints.

int SimTK::OptimizerSystem::getNumNonlinearEqualityConstraints

(

)

const [inline]

Returns the number of nonlinear equality constraints.

int SimTK::OptimizerSystem::getNumLinearInequalityConstraints

(

)

const [inline]

Returns the number of linear inequality constraints.

int SimTK::OptimizerSystem::getNumNonlinearInequalityConstraints

(

)

const [inline]

Returns the number of linear inequality constraints.

bool SimTK::OptimizerSystem::getHasLimits

(

)

const [inline]

Returns true if there are limits on the parameters.

void SimTK::OptimizerSystem::getParameterLimits	(	Real **	lower,
		Real **	upper
	)		const [inline]

Returns the limits on the allowed values of each parameter, as an array of lower bounds and an array of upper bounds, with assumed lengths matching the number of parameters.

---

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

• Optimizer.h