|
bitrl & cuberl Documentation
Simulation engine for reinforcement learning agents
|
#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 | |
| typedef DecayLRSchedule cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::decay_lr_schedule_type |
| typedef DiscountGenerator cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::discount_generator_type |
| typedef EnvType cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::env_type |
The environment type.
| 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.
| typedef TrajectoryGenerator cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::trajectory_generator_type |
| 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.
|
inline |
actions_after_training_episode
|
inline |
actions_after_training_ends. Actions to execute after the training iterations have finisehd
|
inline |
actions_before_training_episode
| 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
| 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
| void cuberl::rl::algos::mc::FirstVisitMCSolver< EnvType, TrajectoryGenerator, DecayLRSchedule, DiscountGenerator >::save | ( | const std::string & | filename | ) | const |
save the results