bitrl & cuberl Documentation
Simulation engine for reinforcement learning agents
Loading...
Searching...
No Matches
cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator > Class Template Reference

#include <first_visit_mc.h>

Public Types

typedef EnvType env_type
 The environment type.
 
typedef TrajectoryGenerator trajectory_generator_type
 
typedef DecayLRSchedule decay_lr_schedule_type
 
typedef DiscountGenerator discount_generator_type
 
typedef env_type::time_step_type time_step_type
 The time step type used by the environment.
 

Public Member Functions

 FirstVisitMCSolver (FirstVisitMCSolverConfig solver_config, TrajectoryGenerator &trajectory_gen, DecayLRSchedule &decay_lr_schedule, discount_generator_type &discount_generator)
 Constructor.
 
void actions_before_training_begins (env_type &env)
 actions_before_training_begins. Execute any actions the algorithm needs before starting the iterations
 
void actions_after_training_ends (env_type &)
 actions_after_training_ends. Actions to execute after the training iterations have finisehd
 
void actions_before_episode_begins (env_type &, uint_t)
 actions_before_training_episode
 
void actions_after_episode_ends (env_type &, uint_t, const EpisodeInfo &)
 actions_after_training_episode
 
EpisodeInfo on_training_episode (env_type &env, uint_t episode_idx)
 on_episode Do one on_episode of the algorithm
 
void save (const std::string &filename) const
 save the results
 

Detailed Description

template<typename EnvType, typename TrajectoryGenerator, typename DecayLRSchedule, typename DiscountGenerator>
class cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >
Todo:
write docs

Member Typedef Documentation

◆ decay_lr_schedule_type

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
typedef DecayLRSchedule cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::decay_lr_schedule_type

◆ discount_generator_type

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
typedef DiscountGenerator cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::discount_generator_type

◆ env_type

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
typedef EnvType cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::env_type

The environment type.

◆ time_step_type

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
typedef env_type::time_step_type cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::time_step_type

The time step type used by the environment.

◆ trajectory_generator_type

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
typedef TrajectoryGenerator cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::trajectory_generator_type

Constructor & Destructor Documentation

◆ FirstVisitMCSolver()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::FirstVisitMCSolver ( FirstVisitMCSolverConfig  solver_config,
TrajectoryGenerator &  trajectory_gen,
DecayLRSchedule &  decay_lr_schedule,
discount_generator_type discount_generator 
)

Constructor.

Member Function Documentation

◆ actions_after_episode_ends()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::actions_after_episode_ends ( env_type ,
uint_t  ,
const EpisodeInfo  
)
inline

actions_after_training_episode

◆ actions_after_training_ends()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::actions_after_training_ends ( env_type )
inline

actions_after_training_ends. Actions to execute after the training iterations have finisehd

◆ actions_before_episode_begins()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::actions_before_episode_begins ( env_type ,
uint_t   
)
inline

actions_before_training_episode

◆ actions_before_training_begins()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::actions_before_training_begins ( env_type env)

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

◆ on_training_episode()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
EpisodeInfo cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::on_training_episode ( env_type env,
uint_t  episode_idx 
)

on_episode Do one on_episode of the algorithm

◆ save()

template<typename EnvType , typename TrajectoryGenerator , typename DecayLRSchedule , typename DiscountGenerator >
void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::save ( const std::string &  filename) const

save the results


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