Compute automatically a gradient with finite differences. More...
#include <roboptim/core/finite-difference-gradient.hh>
Public Member Functions | |
FiniteDifferenceGradient (const Function &f, value_type e=finiteDifferenceEpsilon) throw () | |
Instantiate a finite differences gradient. | |
~FiniteDifferenceGradient () throw () | |
![]() | |
size_type | gradientSize () const throw () |
Return the gradient size. | |
jacobianSize_t | jacobianSize () const throw () |
Return the jacobian size as a pair. | |
bool | isValidGradient (const gradient_t &gradient) const throw () |
Check if the gradient is valid (check size). | |
bool | isValidJacobian (const jacobian_t &jacobian) const throw () |
Check if the jacobian is valid (check sizes). | |
jacobian_t | jacobian (const argument_t &argument) const throw () |
Computes the jacobian. | |
void | jacobian (jacobian_t &jacobian, const argument_t &argument) const throw () |
Computes the jacobian. | |
gradient_t | gradient (const argument_t &argument, size_type functionId=0) const throw () |
Computes the gradient. | |
void | gradient (gradient_t &gradient, const argument_t &argument, size_type functionId=0) const throw () |
Computes the gradient. | |
virtual std::ostream & | print (std::ostream &o) const throw () |
Display the function on the specified output stream. | |
![]() | |
bool | isValidResult (const result_t &result) const throw () |
Check the given result size is valid. | |
size_type | inputSize () const throw () |
Return the input size (i.e. argument's vector size). | |
size_type | outputSize () const throw () |
Return the output size (i.e. result's vector size). | |
virtual | ~Function () throw () |
Trivial destructor. | |
result_t | operator() (const argument_t &argument) const throw () |
Evaluate the function at a specified point. | |
void | operator() (result_t &result, const argument_t &argument) const throw () |
Evaluate the function at a specified point. | |
const std::string & | getName () const throw () |
Get function name. |
Protected Member Functions | |
void | impl_compute (result_t &, const argument_t &) const throw () |
Function evaluation. | |
void | impl_gradient (gradient_t &, const argument_t &argument, size_type=0) const throw () |
Gradient evaluation. | |
![]() | |
DerivableFunction (size_type inputSize, size_type outputSize=1, std::string name=std::string()) throw () | |
Concrete class constructor should call this constructor. | |
virtual void | impl_jacobian (jacobian_t &jacobian, const argument_t &arg) const throw () |
Jacobian evaluation. | |
![]() | |
Function (size_type inputSize, size_type outputSize=1, std::string name=std::string()) throw () | |
Concrete class constructor should call this constructor. |
Protected Attributes | |
const Function & | adaptee_ |
Reference to the wrapped function. | |
const value_type | epsilon_ |
Additional Inherited Members | |
![]() | |
typedef vector_t | gradient_t |
Gradient type. | |
typedef matrix_t | jacobian_t |
Jacobian type. | |
typedef std::pair< value_type, value_type > | jacobianSize_t |
Jacobian size type (pair of values). | |
![]() | |
typedef double | value_type |
Values type. | |
typedef std::size_t | size_type |
Size type. | |
typedef ublas::vector< value_type > | vector_t |
Basic vector type. | |
typedef ublas::matrix< value_type > | matrix_t |
Basic matrix type. | |
typedef vector_t | result_t |
Type of a function evaluation result. | |
typedef vector_t | argument_t |
Type of a function evaluation argument. | |
typedef std::pair< value_type, value_type > | interval_t |
Interval type (lower, upper). Use negative or positive infinity to respectively disable the lower or upper bound. | |
typedef std::vector< interval_t > | intervals_t |
Vector of intervals. | |
typedef boost::tuple < value_type, value_type, value_type > | discreteInterval_t |
Types representing a discrete interval. A discrete interval is a triplet of values: |
Compute automatically a gradient with finite differences.
Finite difference gradient is a method to approximate a function's gradient. It is particularly useful in RobOptim to avoid the need to compute the analytical gradient manually. This class takes a Function as its input and wraps it into a derivable function. The one dimensional formula is:
where is a constant given when calling the class constructor.
roboptim::FiniteDifferenceGradient< FdgPolicy >::FiniteDifferenceGradient | ( | const Function & | f, |
value_type | e = finiteDifferenceEpsilon |
||
) | throw () |
Instantiate a finite differences gradient.
Instantiate a derivable function that will wraps a non derivable function and compute automatically its gradient using finite differences.
f | function that will e wrapped |
e | epsilon used in finite difference computation |
roboptim::FiniteDifferenceGradient< FdgPolicy >::~FiniteDifferenceGradient | ( | ) | throw () |
|
protectedvirtual |
Function evaluation.
Evaluate the function, has to be implemented in concrete classes.
result | result will be stored in this vector |
argument | point at which the function will be evaluated |
Implements roboptim::Function.
|
protectedvirtual |
Gradient evaluation.
Compute the gradient, has to be implemented in concrete classes. The gradient is computed for a specific sub-function which id is passed through the functionId argument.
gradient | gradient will be store in this argument |
argument | point where the gradient will be computed |
functionId | evaluated function id in the split representation |
Implements roboptim::DerivableFunction.
|
protected |
Reference to the wrapped function.
|
protected |