OPTPP::NLP0 Class Reference

Base Class for NonLinear Programming Problem For NLP0 the only assumption on the objective function is that it be continuous. More...

Inheritance diagram for OPTPP::NLP0:

[legend]

Collaboration diagram for OPTPP::NLP0:

[legend]

List of all members.

Public Member Functions
NEWMAT::ColumnVector	BDGrad (const NEWMAT::ColumnVector &, const NEWMAT::ColumnVector &, double &, NEWMAT::ColumnVector &)
NEWMAT::ColumnVector	CDGrad (const NEWMAT::ColumnVector &, const NEWMAT::ColumnVector &, double &, NEWMAT::ColumnVector &)
NEWMAT::Matrix	CONBDGrad (const NEWMAT::ColumnVector &)
NEWMAT::Matrix	CONCDGrad (const NEWMAT::ColumnVector &)
NEWMAT::Matrix	CONFDGrad (const NEWMAT::ColumnVector &)
	Finite-difference gradient and Hessian of nonlinear constraints.
virtual void	eval ()=0
virtual void	evalC (const NEWMAT::ColumnVector &x)=0
virtual NEWMAT::ColumnVector	evalCF (const NEWMAT::ColumnVector &x)=0
	Evaluate the constraint at x.
virtual NEWMAT::Matrix	evalCG (const NEWMAT::ColumnVector &x)=0
	Evaluate the constraint gradient at x.
virtual OptppArray < NEWMAT::SymmetricMatrix >	evalCH (NEWMAT::ColumnVector &x, int darg)=0
	Evaluate the constraint Hessian at x.
virtual NEWMAT::SymmetricMatrix	evalCH (NEWMAT::ColumnVector &x)=0
virtual real	evalF (const NEWMAT::ColumnVector &x)=0
virtual real	evalF ()=0
virtual NEWMAT::ColumnVector	evalG (const NEWMAT::ColumnVector &x)=0
virtual NEWMAT::ColumnVector	evalG ()=0
	Evaluate a finite-difference gradient and Hessian.
virtual NEWMAT::SymmetricMatrix	evalH (NEWMAT::ColumnVector &x)=0
virtual NEWMAT::SymmetricMatrix	evalH ()=0
virtual real	evalLagrangian (const NEWMAT::ColumnVector &x, NEWMAT::ColumnVector &mult, const NEWMAT::ColumnVector &type)=0
	Evaluate the Lagrangian.
virtual NEWMAT::ColumnVector	evalLagrangianGradient (const NEWMAT::ColumnVector &x, const NEWMAT::ColumnVector &mult, const NEWMAT::ColumnVector &type)=0
	Evaluate the Lagrangian gradient.
NEWMAT::SymmetricMatrix	FD2Hessian (NEWMAT::ColumnVector &)
NEWMAT::ColumnVector	FDGrad (const NEWMAT::ColumnVector &, const NEWMAT::ColumnVector &, double &, NEWMAT::ColumnVector &)
	Finite-difference gradient and Hessian.
virtual void	fPrintState (ostream , char )
	Print state of the function to a file.
CompoundConstraint *	getConstraints ()
NEWMAT::ColumnVector	getConstraintValue () const
virtual bool	getDebug () const
DerivOption	getDerivOption () const
virtual int	getDim () const
virtual real	getF () const
virtual NEWMAT::ColumnVector	getFcnAccrcy () const
virtual real	getFcnTime () const
virtual int	getFevals () const
virtual int	getIsExpensive () const
virtual bool	getModeOverride () const
int	getNumOfCons () const
int	getNumOfNLCons () const
SpecOption	getSpecOption () const
virtual NEWMAT::ColumnVector	getXc () const
bool	hasConstraints ()
virtual void	initFcn ()=0
	NLP0 (int ndim, CompoundConstraint *constraint)
	NLP0 (int ndim, int nlncons)
	NLP0 (int ndim)
	NLP0 ()
	Default Constructor.
void	printConstraints ()
	Print constraints to standard output.
virtual void	printState (char *)
	Print the function.
virtual void	reset ()=0
void	saveState ()
	Save current state of the function.
void	setConstraints (CompoundConstraint *constraintSet)
	Set a pointer to a CompoundConstraint object.
void	setConstraintValue (const NEWMAT::ColumnVector &cfx)
	Set the constraint value.
virtual void	setDebug ()
void	setDerivOption (DerivOption d)
virtual void	setF (const real &fx)
virtual void	setFcnAccrcy (const NEWMAT::ColumnVector &accrcy)
virtual void	setFcnAccrcy (const int i, const real &accrcy)
virtual void	setIsExpensive (const int e)
virtual void	setModeOverride (bool override_mode)
void	setSpecOption (SpecOption SpecEval)
virtual void	setX (const NEWMAT::ColumnVector &x)
virtual void	setX (const int i, const real &x)
	Functions for setting various properties of this NLP problem.
virtual	~NLP0 ()
	Destructor.
Protected Attributes
Appl_Data	application
CompoundConstraint *	constraint_
	Pointer to constraints.
NEWMAT::ColumnVector	constraint_value
	Constraint residual.
bool	debug_
	Print debug statements.
int	dim
	Dimension of the problem.
DerivOption	finitediff
	User-specified derivative option.
double	function_time
	Function compute time.
real	fvalue
	Objective function value at mem_xc.
int	is_expensive
	Is this an expensive function?
NEWMAT::ColumnVector	mem_fcn_accrcy
	Accuracy available from function.
NEWMAT::ColumnVector	mem_xc
	Current point.
bool	modeOverride
int	ncnln
	Number of nonlinear constraints.
int	nfevals
	Number of function evaluations.
char	nlp_name [80]
	Optional identifying name.
NEWMAT::ColumnVector	partial_grad
double	specF
SpecOption	SpecFlag
	Speculative gradient information.

Detailed Description

Base Class for NonLinear Programming Problem For NLP0 the only assumption on the objective function is that it be continuous.

No derivative information is available. Note that NLP0-2 are abstract data types

Author:: J.C. Meza, Sandia National Laboratories, meza@ca.sandia.gov

Definition at line 28 of file NLP0.h.

Constructor & Destructor Documentation

OPTPP::NLP0::NLP0 ( ) [inline]

Default Constructor.

See also:: NLP0(int dim); NLP0(int dim, int nlncons); NLP0(int dim, CompoundConstraint* constraint)

Definition at line 115 of file NLP0.h.

References finitediff, mem_fcn_accrcy, mem_xc, and SpecFlag.

OPTPP::NLP0::NLP0 ( int ndim ) [inline]

Parameters:

ndim

an int

See also:: NLP0(); NLP0(int dim, int nlncons); NLP0(int dim, CompoundConstraint* constraint)

Definition at line 127 of file NLP0.h.

References finitediff, mem_fcn_accrcy, mem_xc, and SpecFlag.

OPTPP::NLP0::NLP0	(	int	ndim,
		int	nlncons
	)			`[inline]`

Parameters:

	ndim	an int
	nlncons	an int

See also:: NLP0(); NLP0(int dim); NLP0(int dim, CompoundConstraint* constraint)

Definition at line 140 of file NLP0.h.

References constraint_value, finitediff, mem_fcn_accrcy, mem_xc, and SpecFlag.

OPTPP::NLP0::NLP0	(	int	ndim,
		CompoundConstraint *	constraint
	)			`[inline]`

Parameters:

	ndim	an int
	constraint	pointer to a CompoundConstraint

See also:: NLP0(); NLP0(int dim); NLP0(int dim, int nlncons)

Definition at line 154 of file NLP0.h.

References finitediff, mem_fcn_accrcy, mem_xc, and SpecFlag.

virtual OPTPP::NLP0::~NLP0 ( ) [inline, virtual]

Destructor.

Definition at line 263 of file NLP0.h.

Member Function Documentation

NEWMAT::Matrix OPTPP::NLP0::CONFDGrad ( const NEWMAT::ColumnVector & )

Finite-difference gradient and Hessian of nonlinear constraints.

virtual NEWMAT::ColumnVector OPTPP::NLP0::evalCF ( const NEWMAT::ColumnVector & x ) [pure virtual]

Evaluate the constraint at x.

Implements OPTPP::NLPBase.

Implemented in OPTPP::LSQNLF, OPTPP::NLF0, OPTPP::NLF1, OPTPP::FDNLF1, OPTPP::NLF0APP, and OPTPP::FDNLF1APP.

virtual NEWMAT::Matrix OPTPP::NLP0::evalCG ( const NEWMAT::ColumnVector & x ) [pure virtual]

Evaluate the constraint gradient at x.

Implements OPTPP::NLPBase.

Implemented in OPTPP::LSQNLF, OPTPP::NLF1, OPTPP::FDNLF1, and OPTPP::FDNLF1APP.

virtual OptppArray<NEWMAT::SymmetricMatrix> OPTPP::NLP0::evalCH	(	NEWMAT::ColumnVector &	x,
		int	darg
	)			`[pure virtual]`

Evaluate the constraint Hessian at x.

Implements OPTPP::NLPBase.

Implemented in OPTPP::NLF1.

virtual NEWMAT::ColumnVector OPTPP::NLP0::evalG ( ) [pure virtual]

Evaluate a finite-difference gradient and Hessian.

Implements OPTPP::NLPBase.

Implemented in OPTPP::LSQNLF, OPTPP::NLF0, OPTPP::NLF1, OPTPP::NLF2, OPTPP::FDNLF1, OPTPP::NLF0APP, and OPTPP::FDNLF1APP.

virtual real OPTPP::NLP0::evalLagrangian	(	const NEWMAT::ColumnVector &	x,
		NEWMAT::ColumnVector &	mult,
		const NEWMAT::ColumnVector &	type
	)			`[pure virtual]`

Evaluate the Lagrangian.

Implemented in OPTPP::LSQNLF, OPTPP::NLF0, OPTPP::NLF1, OPTPP::FDNLF1, OPTPP::NLF0APP, and OPTPP::FDNLF1APP.

virtual NEWMAT::ColumnVector OPTPP::NLP0::evalLagrangianGradient	(	const NEWMAT::ColumnVector &	x,
		const NEWMAT::ColumnVector &	mult,
		const NEWMAT::ColumnVector &	type
	)			`[pure virtual]`