BALL::RombergIntegrator< Function, DataType > Class Template Reference
[Function classes]

#include <BALL/MATHS/rombergIntegrator.h>

Inheritance diagram for BALL::RombergIntegrator< Function, DataType >:

List of all members.

Public Member Functions
Constructors and destructor.
	RombergIntegrator (float epsilon=1E-5, Size nsteps=1000)
	Default constructor.
	RombergIntegrator (const RombergIntegrator &romint)
	Copy constructor.
	~RombergIntegrator ()
	Destructor.
Assignment
const RombergIntegrator &	operator= (const RombergIntegrator &romint)
	Assignment operator.
virtual void	clear ()
	Clear method.
void	setEpsilon (float eps)
	Set the upper bound for the error we want to allow.
void	setMaxNumSteps (Size mns)
	Set the maximum number of steps we want to use in computation.
Predicates
bool	operator== (const RombergIntegrator &romint) const
	Equality operator.
Accessors
DataType	integrate (DataType from, DataType to)
DataType	trapezoid (DataType h, DataType from, DataType to)
Protected Attributes
float	epsilon_
Size	maxNumSteps_
vector< DataType >	result_

---

Detailed Description

template<typename Function, typename DataType>
class BALL::RombergIntegrator< Function, DataType >

A numerical integration of a BALL-Function using a Romberg integration scheme.

---

Constructor & Destructor Documentation

template<typename Function , typename DataType >

BALL_INLINE BALL::RombergIntegrator< Function, DataType >::RombergIntegrator	(	float	epsilon = 1E-5,
		Size	nsteps = 1000
	)			[inline]

Default constructor.

References BALL::RombergIntegrator< Function, DataType >::maxNumSteps_, and BALL::RombergIntegrator< Function, DataType >::result_.

template<typename Function , typename DataType >

BALL_INLINE BALL::RombergIntegrator< Function, DataType >::RombergIntegrator

(

const RombergIntegrator< Function, DataType > &

romint

)

[inline]

Copy constructor.

template<typename Function , typename DataType >

BALL_INLINE BALL::RombergIntegrator< Function, DataType >::~RombergIntegrator

(

)

[inline]

Destructor.

---

Member Function Documentation

template<typename Function , typename DataType >

BALL_INLINE void BALL::RombergIntegrator< Function, DataType >::clear

(

)

[inline, virtual]

Clear method.

template<typename Function , typename DataType >

BALL_INLINE DataType BALL::RombergIntegrator< Function, DataType >::integrate	(	DataType	from,
		DataType	to
	)			[inline]

Integrate the function numerically.

Parameters:

	from	lower limit of the integration
	to	upper limit of the integration

Returns:

the value of the integral

References BALL::Constants::h.

template<typename Function , typename DataType >

BALL_INLINE const RombergIntegrator< Function, DataType > & BALL::RombergIntegrator< Function, DataType >::operator=

(

const RombergIntegrator< Function, DataType > &

romint

)

[inline]

Assignment operator.

template<typename Function , typename DataType >

BALL_INLINE bool BALL::RombergIntegrator< Function, DataType >::operator==

(

const RombergIntegrator< Function, DataType > &

romint

)

const [inline]

Equality operator.

template<typename Function , typename DataType >

BALL_INLINE void BALL::RombergIntegrator< Function, DataType >::setEpsilon

(

float

eps

)

[inline]

Set the upper bound for the error we want to allow.

References BALL::RombergIntegrator< Function, DataType >::epsilon_.

template<typename Function , typename DataType >

BALL_INLINE void BALL::RombergIntegrator< Function, DataType >::setMaxNumSteps

(

Size

mns

)

[inline]

Set the maximum number of steps we want to use in computation.

References BALL::RombergIntegrator< Function, DataType >::maxNumSteps_, and BALL::RombergIntegrator< Function, DataType >::result_.

template<typename Function , typename DataType >

BALL_INLINE DataType BALL::RombergIntegrator< Function, DataType >::trapezoid	(	DataType	h,
		DataType	from,
		DataType	to
	)			[inline]

Integrate the function numerically using a simple trapezoid integration. This function is intended as a helper function for the computation of the romberg integration, but it can be used as a regular integrator as well, if speed is more important than reliability.

Parameters:

	h	gives the width of each step
	from	lower limit of the integration
	to	upper limit of the integration

Returns:

the value of the integral

References BALL::NumericalIntegrator< Function, DataType >::function_.

---

Member Data Documentation

template<typename Function, typename DataType>

float BALL::RombergIntegrator< Function, DataType >::epsilon_ [protected]

Referenced by BALL::RombergIntegrator< Function, DataType >::setEpsilon().

template<typename Function, typename DataType>

Size BALL::RombergIntegrator< Function, DataType >::maxNumSteps_ [protected]

Referenced by BALL::RombergIntegrator< Function, DataType >::RombergIntegrator(), and BALL::RombergIntegrator< Function, DataType >::setMaxNumSteps().

template<typename Function, typename DataType>

vector<DataType> BALL::RombergIntegrator< Function, DataType >::result_ [protected]

Referenced by BALL::RombergIntegrator< Function, DataType >::RombergIntegrator(), and BALL::RombergIntegrator< Function, DataType >::setMaxNumSteps().