|
ccgsl 2.7.2
C++wrappersforGnuScientificLibrary
|
This is an abstract base class. More...
#include <multilarge_nlinear_function_fdf.hpp>

Public Member Functions | |
| virtual int | f (gsl::vector const &x, gsl::vector &f)=0 |
| The function. More... | |
| virtual int | df (CBLAS_TRANSPOSE_t TransJ, gsl::vector const &x, gsl::vector const &u, gsl::vector &v, gsl::matrix &JTJ)=0 |
| The derivatives (as Jacobian matrix). More... | |
| virtual size_t | xSize () const =0 |
| This function should return the number of elements of x in f(). More... | |
| virtual size_t | fSize () const =0 |
| This function should return the number of elements of f in f(). More... | |
This is an abstract base class.
Its main purpose is to define the concept of a class whose objects can be used to construct a function object. Any class that implements concept_fdf can be used in the constructor of function. This version allows gsl::vector instead of gsl_vector* with a small overhead.
Definition at line 65 of file multilarge_nlinear_function_fdf.hpp.
|
pure virtual |
The derivatives (as Jacobian matrix).
This can be undefined in which case finite difference approximations are used.
| TransJ | Set to blas::Trans to transpose J |
| x | The argument vector |
| u | An input vector |
| v | An output vector |
| JTJ | (Jacobian) matrix of size \(n\times n\) where \(n\) is the size of x; this holds the return values: may be null. |
|
pure virtual |
The function.
| x | The argument (vector) |
| f | Vector of same size as x to hold the return values. |
|
pure virtual |
This function should return the number of elements of f in f().
|
pure virtual |
This function should return the number of elements of x in f().