ROOT::Math::MultiNumGradFunction Class Reference

Description

MultiNumGradFunction class to wrap a normal function in a gradient function using numerical gradient calculation provided by the class Derivator (based on GSL numerical derivation)

Definition at line 49 of file MultiNumGradFunction.h.

Inheritance diagram for ROOT::Math::MultiNumGradFunction:

Collaboration diagram for ROOT::Math::MultiNumGradFunction:

Public Types
typedef T	BackendType
typedef IBaseFunctionMultiDimTempl< T >	BaseFunc
typedef IGradientMultiDimTempl< T >	BaseGrad

Public Member Functions
	MultiNumGradFunction (const IMultiGenFunction &f)
template<class FuncType >
	MultiNumGradFunction (FuncType f, int n)
	~MultiNumGradFunction ()
IMultiGenFunction *	Clone () const
T	Derivative (const T *x, unsigned int icoord=0) const
virtual void	FdF (const T *x, T &f, T *df) const
virtual void	Gradient (const T *x, T *grad) const
unsigned int	NCalls () const
unsigned int	NDim () const
virtual unsigned int	NDim () const=0
T	operator() (const T *x) const
void	SetOwnership (bool on=true)

Static Public Member Functions
static double	GetDerivPrecision ()
	get precision value used for calculating the derivative step-size More...
static void	SetDerivPrecision (double eps)
	precision value used for calculating the derivative step-size h = eps * |x|. The default is 0.001, give a smaller in case function chanes rapidly More...

Private Member Functions
double	DoDerivative (const double *x, unsigned int icoord) const
double	DoEval (const double *x) const

Private Attributes
unsigned int	fDim
const IMultiGenFunction *	fFunc
unsigned int	fNCalls
bool	fOwner

Static Private Attributes
static double	fgEps

Member Typedef Documentation

BackendType

template<class T >

typedef T ROOT::Math::IBaseFunctionMultiDimTempl< T >::BackendType

inherited

Definition at line 66 of file IFunction.h.

BaseFunc

template<class T >

typedef IBaseFunctionMultiDimTempl<T> ROOT::Math::IGradientFunctionMultiDimTempl< T >::BaseFunc

inherited

Definition at line 330 of file IFunction.h.

BaseGrad

template<class T >

typedef IGradientMultiDimTempl<T> ROOT::Math::IGradientFunctionMultiDimTempl< T >::BaseGrad

inherited

Definition at line 331 of file IFunction.h.

Constructor & Destructor Documentation

MultiNumGradFunction() [1/2]

ROOT::Math::MultiNumGradFunction::MultiNumGradFunction ( const IMultiGenFunction &  f )

inline

Constructor from a IMultiGenFunction interface

Definition at line 57 of file MultiNumGradFunction.h.

                                                      :
      fFunc(&f),
      fDim(f.NDim() ),
      fNCalls(0),
      fOwner(false)
   {}

Referenced by Clone().

MultiNumGradFunction() [2/2]

template<class FuncType >

ROOT::Math::MultiNumGradFunction::MultiNumGradFunction ( FuncType  f, int  n  )

inline

Constructor from a generic function (pointer or reference) and number of dimension implementiong operator () (double * x)

Definition at line 70 of file MultiNumGradFunction.h.

                                            :
      fDim( n ),
      fNCalls(0),
      fOwner(true)
   {
      // create a wrapped function
      fFunc = new ROOT::Math::WrappedMultiFunction<FuncType> (f, n);
   }

References fFunc.

~MultiNumGradFunction()

ROOT::Math::MultiNumGradFunction::~MultiNumGradFunction ( )

inline

Destructor (no operations)

Definition at line 82 of file MultiNumGradFunction.h.

                             {
      if (fOwner) delete fFunc;
   }

References fFunc, and fOwner.

Member Function Documentation

Clone()

IMultiGenFunction* ROOT::Math::MultiNumGradFunction::Clone ( ) const

inlinevirtual

Clone a function. Each derived class must implement their version of the Clone method

Implements ROOT::Math::IBaseFunctionMultiDimTempl< T >.

Definition at line 93 of file MultiNumGradFunction.h.

                                     {
      if (!fOwner)
         return new MultiNumGradFunction(*fFunc);
      else {
         // we need to copy the pointer to the wrapped function
         MultiNumGradFunction * f =  new MultiNumGradFunction(*(fFunc->Clone()) );
         f->fOwner = true;
         return f;
      }
   }

References MultiNumGradFunction(), ROOT::Math::IBaseFunctionMultiDimTempl< T >::Clone(), fFunc, and fOwner.

Here is the call graph for this function:

Derivative()

template<class T >

T ROOT::Math::IGradientMultiDimTempl< T >::Derivative ( const T *  x, unsigned int  icoord = 0  ) const

inlineinherited

Return the partial derivative with respect to the passed coordinate

Definition at line 217 of file IFunction.h.

{ return DoDerivative(x, icoord); }

References ROOT::Math::IGradientMultiDimTempl< T >::DoDerivative().

Referenced by ROOT::Math::MinimTransformFunction::DoDerivative(), and ROOT::Math::IGradientFunctionMultiDimTempl< T >::Gradient().

Here is the call graph for this function:

DoDerivative()

double ROOT::Math::MultiNumGradFunction::DoDerivative ( const double *  x, unsigned int  icoord  ) const

private

DoEval()

double ROOT::Math::MultiNumGradFunction::DoEval ( const double *  x ) const

inlineprivate

Definition at line 118 of file MultiNumGradFunction.h.

                                         {
      fNCalls++;
      return (*fFunc)(x);
   }

References fFunc, and fNCalls.

FdF()

template<class T >

virtual void ROOT::Math::IGradientFunctionMultiDimTempl< T >::FdF ( const T *  x, T &  f, T *  df  ) const

inlinevirtualinherited

Optimized method to evaluate at the same time the function value and derivative at a point x. Often both value and derivatives are needed and it is often more efficient to compute them at the same time. Derived class should implement this method if performances play an important role and if it is faster to evaluate value and derivative at the same time

Implements ROOT::Math::IGradientMultiDimTempl< T >.

Definition at line 357 of file IFunction.h.

         {
            f = BaseFunc::operator()(x);
            Gradient(x, df);
         }

References ROOT::Math::IGradientFunctionMultiDimTempl< T >::Gradient(), and ROOT::Math::IBaseFunctionMultiDimTempl< T >::operator()().

Here is the call graph for this function:

GetDerivPrecision()

static double ROOT::Math::MultiNumGradFunction::GetDerivPrecision ( )

static

get precision value used for calculating the derivative step-size

Gradient()

template<class T >

virtual void ROOT::Math::IGradientFunctionMultiDimTempl< T >::Gradient ( const T *  x, T *  grad  ) const

inlinevirtualinherited

Evaluate all the vector of function derivatives (gradient) at a point x. Derived classes must re-implement it if more efficient than evaluting one at a time

Implements ROOT::Math::IGradientMultiDimTempl< T >.

Definition at line 342 of file IFunction.h.

         {
            unsigned int ndim = NDim();
            for (unsigned int icoord  = 0; icoord < ndim; ++icoord)
               grad[icoord] = BaseGrad::Derivative(x, icoord);
         }

References ROOT::Math::IGradientMultiDimTempl< T >::Derivative(), and ROOT::Math::IGradientFunctionMultiDimTempl< T >::NDim().

Referenced by ROOT::Math::IGradientFunctionMultiDimTempl< T >::FdF().

Here is the call graph for this function:

NCalls()

unsigned int ROOT::Math::MultiNumGradFunction::NCalls ( ) const

inline

Definition at line 91 of file MultiNumGradFunction.h.

{ return fNCalls; }

References fNCalls.

NDim() [1/2]

unsigned int ROOT::Math::MultiNumGradFunction::NDim ( ) const

inlinevirtual

Retrieve the dimension of the function

Implements ROOT::Math::IBaseFunctionMultiDimTempl< T >.

Definition at line 89 of file MultiNumGradFunction.h.

{ return fDim; }

References fDim.

NDim() [2/2]

template<class T >

virtual unsigned int ROOT::Math::IBaseFunctionMultiDimTempl< T >::NDim

inherited

Retrieve the dimension of the function

Referenced by ROOT::Math::IGradientFunctionMultiDimTempl< T >::Gradient(), ROOT::Math::MinimTransformFunction::NTot(), and ROOT::Math::GSLMultiMinimizer::Set().

operator()()

template<class T >

T ROOT::Math::IBaseFunctionMultiDimTempl< T >::operator() ( const T *  x ) const

inlineinherited

Evaluate the function at a point x[]. Use the pure virtual private method DoEval which must be implemented by the sub-classes

Definition at line 92 of file IFunction.h.

         {
            return DoEval(x);
         }

References ROOT::Math::IBaseFunctionMultiDimTempl< T >::DoEval().

Referenced by ROOT::Math::IGradientFunctionMultiDimTempl< T >::FdF().

Here is the call graph for this function:

SetDerivPrecision()

static void ROOT::Math::MultiNumGradFunction::SetDerivPrecision ( double  eps )

static

precision value used for calculating the derivative step-size h = eps * |x|. The default is 0.001, give a smaller in case function chanes rapidly

SetOwnership()

void ROOT::Math::MultiNumGradFunction::SetOwnership ( bool  on = true )

inline

Definition at line 105 of file MultiNumGradFunction.h.

{ fOwner = on;  }

References fOwner.

Member Data Documentation

fDim

unsigned int ROOT::Math::MultiNumGradFunction::fDim

private

Definition at line 128 of file MultiNumGradFunction.h.

Referenced by NDim().

fFunc

const IMultiGenFunction* ROOT::Math::MultiNumGradFunction::fFunc

private

Definition at line 127 of file MultiNumGradFunction.h.

Referenced by MultiNumGradFunction(), ~MultiNumGradFunction(), Clone(), and DoEval().

fgEps

double ROOT::Math::MultiNumGradFunction::fgEps

staticprivate

Definition at line 132 of file MultiNumGradFunction.h.

fNCalls

unsigned int ROOT::Math::MultiNumGradFunction::fNCalls

mutableprivate

Definition at line 129 of file MultiNumGradFunction.h.

Referenced by DoEval(), and NCalls().

fOwner

bool ROOT::Math::MultiNumGradFunction::fOwner

private

Definition at line 130 of file MultiNumGradFunction.h.

Referenced by ~MultiNumGradFunction(), Clone(), and SetOwnership().

---

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

• /home/build/builds/o5h8MZZm/0/mlz/bornagain/Fit/3rdparty/RootMinimizers/mathcore/Math/MultiNumGradFunction.h