MCTreeSearchBase. More...

#include <mc_tree_search_solver.h>

Inheritance diagram for cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >:

[legend]

Collaboration diagram for cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >:

[legend]

Public Types
typedef EnvTp	env_type
	env_type

typedef NodeTp	node_type
	node_type

typedef env_type::time_step_type	time_step_type
	The time step type.

Public Types inherited from cuberl::rl::algos::RLSolverBase< EnvTp >
typedef EnvTp	env_type

Public Member Functions
	MCTSSolver (MCTreeSearchConfig config)
	MCTreeSearchBase.

virtual void	actions_before_training_begins (env_type &) override final
	actions_before_training_begins. Execute any actions the algorithm needs before starting the iterations

virtual void	actions_after_training_ends (env_type &) override final
	actions_after_training_ends. Actions to execute after the training iterations have finisehd

virtual void	actions_before_episode_begins (env_type &, uint_t)
	actions_before_training_episode

virtual void	actions_after_episode_ends (env_type &, uint_t, const EpisodeInfo &)
	actions_after_training_episode

virtual EpisodeInfo	on_training_episode (env_type &, uint_t) override final
	on_episode Do one on_episode of the algorithm

time_step_type	simulate_node (std::shared_ptr< node_type > node, env_type &env)
	simulate_node

void	expand_node (std::shared_ptr< node_type > node, env_type &env)
	expand_node

void	backprop (std::shared_ptr< node_type > node)
	backprop

uint_t	max_depth_tree () const noexcept
	max_depth_tree

Public Member Functions inherited from cuberl::rl::algos::RLSolverBase< EnvTp >
virtual	~RLSolverBase ()=default
	Destructor.

virtual void	actions_before_episode_begins (env_type &, uint_t)
	actions_before_training_episode

virtual void	actions_after_episode_ends (env_type &, uint_t, const EpisodeInfo &)
	actions_after_training_episode

virtual EpisodeInfo	on_training_episode (env_type &, uint_t)=0
	on_episode Do one on_episode of the algorithm

Protected Attributes
IterationMixin	itr_mix_
	itr_mix_

std::shared_ptr< node_type >	root_
	root_

uint_t	max_depth_tree_
	max_depth_tree_

real_t	temperature_
	temperature_

Additional Inherited Members
Protected Member Functions inherited from cuberl::rl::algos::RLSolverBase< EnvTp >
	RLSolverBase ()=default
	Constructor.

Detailed Description

template<typename EnvTp, typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>
class cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >

MCTreeSearchBase.

Member Typedef Documentation

◆ env_type

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

typedef EnvTp cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::env_type

env_type

◆ node_type

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

typedef NodeTp cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::node_type

node_type

◆ time_step_type

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

typedef env_type::time_step_type cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::time_step_type

The time step type.

Constructor & Destructor Documentation

◆ MCTSSolver()

template<typename EnvTp , typename NodeTp >

cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::MCTSSolver ( MCTreeSearchConfig config )

MCTreeSearchBase.

Parameters

config

Member Function Documentation

◆ actions_after_episode_ends()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

virtual void cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::actions_after_episode_ends	(	env_type &	,
		uint_t	,
		const EpisodeInfo &
	)

inlinevirtual

actions_after_training_episode

◆ actions_after_training_ends()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

virtual void cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::actions_after_training_ends ( env_type & )

inlinefinaloverridevirtual

actions_after_training_ends. Actions to execute after the training iterations have finisehd

Implements cuberl::rl::algos::RLSolverBase< EnvTp >.

◆ actions_before_episode_begins()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

virtual void cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::actions_before_episode_begins	(	env_type &	,
		uint_t
	)

inlinevirtual

actions_before_training_episode

◆ actions_before_training_begins()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

virtual void cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::actions_before_training_begins ( env_type & )

inlinefinaloverridevirtual

actions_before_training_begins. Execute any actions the algorithm needs before starting the iterations

Implements cuberl::rl::algos::RLSolverBase< EnvTp >.

◆ backprop()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

void cubeai::rl::algos::MCTSSolver::backprop ( std::shared_ptr< node_type > node )

backprop

◆ expand_node()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

void cubeai::rl::algos::MCTSSolver::expand_node	(	std::shared_ptr< node_type >	node,
		env_type &	env
	)

expand_node

Parameters

node

◆ max_depth_tree()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

uint_t cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::max_depth_tree ( ) const

inlinenoexcept

max_depth_tree

Returns

◆ on_training_episode()

template<typename EnvTp , typename NodeTp >

EpisodeInfo cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::on_training_episode	(	env_type &	env,
		uint_t
	)

finaloverridevirtual

on_episode Do one on_episode of the algorithm

◆ simulate_node()

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

MCTSSolver< Env, NodeTp >::time_step_type cubeai::rl::algos::MCTSSolver::simulate_node	(	std::shared_ptr< node_type >	node,
		env_type &	env
	)

simulate_node

Parameters

node

Member Data Documentation

◆ itr_mix_

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

IterationMixin cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::itr_mix_

protected

itr_mix_

◆ max_depth_tree_

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

uint_t cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::max_depth_tree_

protected

max_depth_tree_

◆ root_

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

std::shared_ptr<node_type> cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::root_

protected

root_

◆ temperature_

template<typename EnvTp , typename NodeTp = MCTSNodeBase<typename EnvTp::action_type, typename EnvTp::state_type>>

real_t cubeai::rl::algos::mc::MCTSSolver< EnvTp, NodeTp >::temperature_

protected

temperature_

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

libs/cuberl/include/cuberl/rl/algorithms/mc/mc_tree_search_solver.h

Public Types

Public Member Functions

Protected Attributes

Additional Inherited Members

Detailed Description

Member Typedef Documentation

◆ env_type

◆ node_type

◆ time_step_type

Constructor & Destructor Documentation

◆ MCTSSolver()

Member Function Documentation

◆ actions_after_episode_ends()

◆ actions_after_training_ends()

◆ actions_before_episode_begins()

◆ actions_before_training_begins()

◆ backprop()

◆ expand_node()

◆ max_depth_tree()

◆ on_training_episode()

◆ simulate_node()

Member Data Documentation

◆ itr_mix_

◆ max_depth_tree_

◆ root_

◆ temperature_