RandomizedStrategy< T > Class Template Reference

Expansion strategy that shuffles the label order randomly each cycle. More...

#include <RandomizedStrategy.hpp>

Inheritance diagram for RandomizedStrategy< T >:

Collaboration diagram for RandomizedStrategy< T >:

Public Member Functions
	RandomizedStrategy (int max_cycles=100, unsigned int seed=42)
	Constructs the strategy.
int	execute (AlphaExpansion< T > &optimizer, EnergyModel< T > &model) const override
	Runs randomized alpha-expansion until convergence or max_cycles.
Public Member Functions inherited from ExpansionStrategy< T >
virtual	~ExpansionStrategy ()=default

Detailed Description

template<typename T>
class RandomizedStrategy< T >

Expansion strategy that shuffles the label order randomly each cycle.

Works the same as SequentialStrategy, but the label order is randomly shuffled at the start of each cycle using a Mersenne Twister RNG. Good for studying how the label order affects convergence speed.

Template Parameters

T	Numeric cost type.

Constructor & Destructor Documentation

RandomizedStrategy()

template<typename T >

RandomizedStrategy< T >::RandomizedStrategy ( int  max_cycles = 100, unsigned int  seed = 42  )

inline

Constructs the strategy.

Parameters

max_cycles	Maximum number of cycles before stopping (default: 100).
seed	Seed for the Mersenne Twister RNG (default: 42).

Member Function Documentation

execute()

template<typename T >

int RandomizedStrategy< T >::execute ( AlphaExpansion< T > &  optimizer, EnergyModel< T > &  model  ) const

inlineoverridevirtual

Runs randomized alpha-expansion until convergence or max_cycles.

Returns

Number of full cycles completed.

Implements ExpansionStrategy< T >.

---

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

• src/strategies/RandomizedStrategy.hpp