Main Page   Class Hierarchy   Compound List   File List   Compound Members   File Members  

ITS_BASE Class Reference

Base class for implementing a Taylor series method for computing tight enclosures on the solution. More...

#include <itsbase.h>

Inheritance diagram for ITS_BASE:

TIGHT_ENCL NAME ITS_DIRECT ITS_QR List of all members.

Public Methods

virtual void CompEncl ( INTERVAL_VECTOR & Ytight, INTERVAL_VECTOR & LocErr, const INTERVAL & h, int order, PtrODENumeric ODE, PtrODESolver Solver )
 Computes a tight enclosure.

virtual void Init ( const PtrODENumeric ODE )
 Initializes the method.

 ~ITS_BASE ()
 Destructor.


Protected Methods

 ITS_BASE ( int order, const string & name )
 Constructor. More...

virtual void Prepare ( INTERVAL_VECTOR & LocErr, const INTERVAL & h, int order, PtrODENumeric ODE, PtrODESolver Solver )
 Prepares to compute a tight enclosure. More...

virtual void Compute ( INTERVAL_VECTOR & Ytight, INTERVAL_VECTOR & LocErr, const INTERVAL & h, int order, PtrODENumeric ODE, PtrODESolver Solver ) = 0
 Computes a tight enclosure. More...

void ResizeSpace ( int n )
 Allocates space. More...


Protected Attributes

PtrParall PrevSol
 Pointer to a parallelepiped. More...

INTERVAL_VECTOR YpointEncl
 Enclosure of the point solution.

INTERVAL_MATRIX Sj
 Solution to the variational equation.


Detailed Description

Base class for implementing a Taylor series method for computing tight enclosures on the solution.

This class encapsulates the common part of the direct and QR-factorization methods.

Author(s):
Ned Nedialkov
Date:
19 July 2001


Constructor & Destructor Documentation

ITS_BASE::~ITS_BASE ( ) [inline]
 

Destructor.

ITS_BASE::ITS_BASE ( int order,
const string & name ) [inline, protected]
 

Constructor.

Parameters:
order   maximum order allowed.
name   name of the method

Precondition:
order>0 && order <= TAYLOR_CTRL::MaxOrder && !name.empty()


Member Function Documentation

void ITS_BASE::CompEncl ( INTERVAL_VECTOR & Ytight,
INTERVAL_VECTOR & LocErr,
const INTERVAL & h,
int order,
PtrODENumeric ODE,
PtrODESolver Solver ) [inline, virtual]
 

Computes a tight enclosure.

Reimplemented from TIGHT_ENCL.

void ITS_BASE::Init ( const PtrODENumeric ODE ) [virtual]
 

Initializes the method.

Reimplemented from TIGHT_ENCL.

Reimplemented in ITS_QR.

void ITS_BASE::Prepare ( INTERVAL_VECTOR & LocErr,
const INTERVAL & h,
int order,
PtrODENumeric ODE,
PtrODESolver Solver ) [protected, virtual]
 

Prepares to compute a tight enclosure.

This function encapsulates the common part of the direct and QR-factorization methods.

Parameters:
LocErr   (output) enclosure of the local error
h   enclosure of the stepsize
order   order of the method
ODE   pointer to a problem
Solver   pointer to a solver

Precondition:
NonZero(h) && order>0 && order<=GetMaxOrder() && NotNull(ODE) && NotNull(Solver) && ODE->GetSize()==Dimension(LocErr)

void ITS_BASE::Compute ( INTERVAL_VECTOR & Ytight,
INTERVAL_VECTOR & LocErr,
const INTERVAL & h,
int order,
PtrODENumeric ODE,
PtrODESolver Solver ) [protected, pure virtual]
 

Computes a tight enclosure.

This function is called after Prepare().

Parameters:
Ytight   (output) enclosure of the solution
LocErr   (output) enclosure of the local error
h   enclosure of the stepsize
order   order of the enclosure method
ODE   pointer to a problem
Solver   pointer to a solver

Precondition:
NonZero(h) && order>0 && order<=GetMaxOrder() && NotNull(ODE) && NotNull(Solver) && ODE->GetSize()==Dimension(LocErr) && ODE->GetSize==Dimension(Ytight);

Reimplemented in ITS_DIRECT, and ITS_QR.

void ITS_BASE::ResizeSpace ( int n ) [protected]
 

Allocates space.

This function allocates the necessary matrices and vectors.

Parameters:
n   size of the problem
Precondition:
n>0


Member Data Documentation

PtrParall ITS_BASE::PrevSol [protected]
 

Pointer to a parallelepiped.

It is used to store the solution in the form of a parallelepiped.

INTERVAL_VECTOR ITS_BASE::YpointEncl [protected]
 

Enclosure of the point solution.

INTERVAL_MATRIX ITS_BASE::Sj [protected]
 

Solution to the variational equation.


The documentation for this class was generated from the following file:
Generated at Sun Oct 14 12:45:41 2001 for VNODE by doxygen1.2.0 written by Dimitri van Heesch, © 1997-2000