citylearn.agents.marlisa module

class citylearn.agents.marlisa.MARLISA(*args, regression_buffer_capacity: Optional[int] = None, start_regression_time_step: Optional[int] = None, regression_frequency: Optional[int] = None, information_sharing: Optional[bool] = None, pca_compression: Optional[float] = None, iterations: Optional[int] = None, **kwargs)[source]

Bases: citylearn.agents.sac.SAC

add_to_buffer(observations: List[List[float]], actions: List[List[float]], reward: List[float], next_observations: List[List[float]], done: bool)[source]

Update replay buffer.

Parameters

observations (List[List[float]]) – Previous time step observations.
actions (List[List[float]]) – Previous time step actions.
reward (List[float]) – Current time step reward.
next_observations (List[List[float]]) – Current time step observations.
done (bool) – Indication that episode has ended.

property batch_size: int: Batch size.

property coordination_variables_history: List[float]

get_encoded_regression_targets(index: int, observations: List[float]) → float[source]

get_encoded_regression_variables(index: int, observations: List[float]) → List[float][source]

get_exploration_actions(observations: List[List[float]]) → List[List[float]][source]

Return randomly sampled actions from action_space multiplied by action_scaling_coefficient.

Returns: actions – Action values.
Return type: List[List[float]]

get_exploration_actions_with_information_sharing(observations: List[List[float]]) → Tuple[List[List[float]], List[List[float]]][source]

get_exploration_actions_without_information_sharing(observations: List[List[float]]) → Tuple[List[List[float]], List[List[float]]][source]

get_post_exploration_actions(observations: List[List[float]], deterministic: bool) → List[List[float]][source]: Action sampling using policy, post-exploration time step

get_post_exploration_actions_with_information_sharing(observations: List[List[float]], deterministic: bool) → Tuple[List[List[float]], List[List[float]]][source]

get_post_exploration_actions_without_information_sharing(observations: List[List[float]], deterministic: bool) → Tuple[List[List[float]], List[List[float]]][source]

get_regression_variables(index: int, observations: List[float], actions: List[float]) → List[float][source]

property hidden_dimension: List[float]: Hidden dimension.

property information_sharing: bool

property iterations: int

property pca_compression: float

predict_demand(index: int, observations: List[float], actions: List[float]) → float[source]

property regression_buffer_capacity: int

property regression_frequency: int

reset()[source]

Reset environment to initial state.

Calls reset_time_step.

Notes

Override in subclass for custom implementation when reseting environment.

set_energy_coefficients()[source]

set_networks()[source]

set_pca()[source]

set_regression_encoders() → List[List[citylearn.preprocessing.Encoder]][source]

Get observation value transformers/encoders for use in MARLISA agent internal regression model.

The encoder classes are defined in the preprocessing.py module and include PeriodicNormalization for cyclic observations, OnehotEncoding for categorical obeservations, RemoveFeature for non-applicable observations given available storage systems and devices and Normalize for observations with known minimum and maximum boundaries.

Returns: encoders – Encoder classes for observations ordered with respect to active_observations.
Return type: List[Encoder]

property start_regression_time_step: int