from rl_coach.exploration_policies .additive_noise import AdditiveNoiseParameters

from rl_coach.agents.clipped_ppo_agent import ClippedPPOAgentParameters
from rl_coach.architectures.tensorflow_components.architecture import Dense
from rl_coach.base_parameters import VisualizationParameters, PresetValidationParameters
from rl_coach.environments.environment import MaxDumpMethod, SelectedPhaseOnlyDumpMethod, SingleLevelSelection
from rl_coach.environments.gym_environment import Mujoco, mujoco_v2, MujocoInputFilter
from rl_coach.graph_managers.basic_rl_graph_manager import BasicRLGraphManager
from rl_coach.graph_managers.graph_manager import ScheduleParameters
from rl_coach.schedules import LinearSchedule

from rl_coach.core_types import TrainingSteps, EnvironmentEpisodes, EnvironmentSteps, RunPhase
from rl_coach.filters.observation.observation_normalization_filter import ObservationNormalizationFilter

####################
# Graph Scheduling #
####################

schedule_params = ScheduleParameters()
schedule_params.improve_steps = TrainingSteps(10000000)
schedule_params.steps_between_evaluation_periods = EnvironmentSteps(2048)
schedule_params.evaluation_steps = EnvironmentEpisodes(5)
schedule_params.heatup_steps = EnvironmentSteps(0)

#########
# Agent #
#########
agent_params = ClippedPPOAgentParameters()


agent_params.network_wrappers['main'].learning_rate = 0.0003
agent_params.network_wrappers['main'].input_embedders_parameters['observation'].activation_function = 'tanh'
agent_params.network_wrappers['main'].input_embedders_parameters['observation'].scheme = [Dense([64])]
agent_params.network_wrappers['main'].middleware_parameters.scheme = [Dense([64])]
agent_params.network_wrappers['main'].middleware_parameters.activation_function = 'tanh'
agent_params.network_wrappers['main'].batch_size = 64
agent_params.network_wrappers['main'].optimizer_epsilon = 1e-5
agent_params.network_wrappers['main'].adam_optimizer_beta2 = 0.999

agent_params.algorithm.clip_likelihood_ratio_using_epsilon = 0.2
agent_params.algorithm.clipping_decay_schedule = LinearSchedule(1.0, 0, 1000000)
agent_params.algorithm.beta_entropy = 0
agent_params.algorithm.gae_lambda = 0.95
agent_params.algorithm.discount = 0.99
agent_params.algorithm.optimization_epochs = 10
agent_params.algorithm.estimate_state_value_using_gae = True

agent_params.input_filter = MujocoInputFilter()
agent_params.exploration = AdditiveNoiseParameters()
agent_params.pre_network_filter = MujocoInputFilter()
agent_params.pre_network_filter.add_observation_filter('observation', 'normalize_observation',
                                                       ObservationNormalizationFilter(name='normalize_observation'))

###############
# Environment #
###############
env_params = Mujoco()
env_params.level = SingleLevelSelection(mujoco_v2)

vis_params = VisualizationParameters()
vis_params.video_dump_methods = [SelectedPhaseOnlyDumpMethod(RunPhase.TEST), MaxDumpMethod()]
vis_params.dump_mp4 = False

########
# Test #
########
preset_validation_params = PresetValidationParameters()
preset_validation_params.test = True
preset_validation_params.min_reward_threshold = 400
preset_validation_params.max_episodes_to_achieve_reward = 1000
preset_validation_params.reward_test_level = 'inverted_pendulum'
preset_validation_params.trace_test_levels = ['inverted_pendulum', 'hopper']

graph_manager = BasicRLGraphManager(agent_params=agent_params, env_params=env_params,
                                    schedule_params=schedule_params, vis_params=vis_params,
                                    preset_validation_params=preset_validation_params)