Use a fixed step during the descent. More...
#include <cathier/minTools.h>
Public Types | |
| typedef TFunctor::argument_type | input_type |
| typedef value_type_of< input_type >::type | input_prec |
| typedef TFunctor::result_type | output_prec |
Public Member Functions | |
| FixedStep (input_prec stepsize) | |
| output_prec | operator() (input_type &p, const input_type &dir) |
Detailed Description
template<typename TFunctor>
class til::FixedStep< TFunctor >
Use a fixed step during the descent.
This does not check if we are actually optimizing the criterion. Yet this can be useful in some cases, e.g. when doing alternate minimization, for which we know that it's not really worth to super-fine one part of the arguments while the other is fixed. So step should be small. OR, the size of the direction argument has actually already been chosen wisely. In this case a fixed step (say, of 1) could be used.
Definition at line 270 of file minTools.h.
Member Typedef Documentation
◆ input_prec
| typedef value_type_of<input_type>::type til::FixedStep< TFunctor >::input_prec |
Definition at line 275 of file minTools.h.
◆ input_type
| typedef TFunctor::argument_type til::FixedStep< TFunctor >::input_type |
Definition at line 274 of file minTools.h.
◆ output_prec
| typedef TFunctor::result_type til::FixedStep< TFunctor >::output_prec |
Definition at line 276 of file minTools.h.
Constructor & Destructor Documentation
◆ FixedStep()
|
inlineexplicit |
Definition at line 280 of file minTools.h.
Member Function Documentation
◆ operator()()
|
inline |
Definition at line 286 of file minTools.h.
References til::expr::_1, til::expr::_2, and til::detail::loop_xx().
The documentation for this class was generated from the following file:
- cathier/minTools.h
