mc::optim::Optimizer Class Reference

Abstract base class for all optimization algorithms. More...

#include <optimizer.hpp>

Inheritance diagram for mc::optim::Optimizer:

Collaboration diagram for mc::optim::Optimizer:

Public Types
using	StepCallback = std::function< void(const Solution &current_best, size_t iteration)>
	Callback invoked after each step/generation.

Public Member Functions
virtual	~Optimizer ()=default
virtual void	setObjectiveFunction (ObjectiveFunction func)=0
	Set the function to optimize (the "black box").
virtual void	setBounds (const Coordinates &lower_bounds, const Coordinates &upper_bounds)=0
	Define the search space boundaries (hyper-rectangle).
virtual void	setMode (OptimizationMode mode)=0
	Set the optimization goal.
virtual void	setCallback (StepCallback cb)=0
virtual Solution	optimize ()=0
	Run the optimization loop until the stopping criterion is met.
virtual void	step ()=0
	Perform a single iteration/generation of the algorithm.
virtual Solution	getBestSolution () const =0
	Get the best solution found so far.

Detailed Description

Abstract base class for all optimization algorithms.

Enforces a common interface for GA, PSO, etc.

Definition at line 17 of file optimizer.hpp.

Member Typedef Documentation

StepCallback

using mc::optim::Optimizer::StepCallback = std::function<void(const Solution& current_best, size_t iteration)>

Callback invoked after each step/generation.

Parameters

current_best	The best solution so far.
iteration	Zero-based iteration/generation index.

Definition at line 25 of file optimizer.hpp.

Constructor & Destructor Documentation

~Optimizer()

virtual mc::optim::Optimizer::~Optimizer ( )

virtualdefault

Member Function Documentation

getBestSolution()

virtual Solution mc::optim::Optimizer::getBestSolution ( ) const

pure virtual

Get the best solution found so far.

Implemented in mc::optim::GA, and mc::optim::PSO.

optimize()

virtual Solution mc::optim::Optimizer::optimize ( )

pure virtual

Run the optimization loop until the stopping criterion is met.

Returns

The best solution found.

Implemented in mc::optim::GA, and mc::optim::PSO.

setBounds()

virtual void mc::optim::Optimizer::setBounds ( const Coordinates &  lower_bounds, const Coordinates &  upper_bounds  )

pure virtual

Define the search space boundaries (hyper-rectangle).

Parameters

lower_bounds	Minimum coordinate per dimension.
upper_bounds	Maximum coordinate per dimension.

Implemented in mc::optim::GA, and mc::optim::PSO.

setCallback()

virtual void mc::optim::Optimizer::setCallback ( StepCallback  cb )

pure virtual

Implemented in mc::optim::GA, and mc::optim::PSO.

setMode()

virtual void mc::optim::Optimizer::setMode ( OptimizationMode  mode )

pure virtual

Set the optimization goal.

Implemented in mc::optim::GA, and mc::optim::PSO.

setObjectiveFunction()

virtual void mc::optim::Optimizer::setObjectiveFunction ( ObjectiveFunction  func )

pure virtual

Set the function to optimize (the "black box").

Parameters

func	Objective mapping coordinates to scalar value.

Implemented in mc::optim::GA, and mc::optim::PSO.

step()

virtual void mc::optim::Optimizer::step ( )

pure virtual

Perform a single iteration/generation of the algorithm.

Useful for debugging, plotting, or GUI integration.

Implemented in mc::optim::GA, and mc::optim::PSO.

---

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

• src/montecarlo/optimizers/optimizer.hpp