Agent

Module: opytimizer.core.agent

An Agent class for all optimization techniques.

Constructor

Agent(n_variables: int, n_dimensions: int, n_objectives: int, lower_bound: List[Union[int, float]], upper_bound: List[Union[int, float]], mapping: Optional[List[str]] = None, env: opytimizer.core.environment.Environment = None) -> None

Parameters

Parameter	Type	Default	Description
n_variables	int		—
n_dimensions	int		—
n_objectives	int		—
lower_bound	List[Union[int, float]]		—
upper_bound	List[Union[int, float]]		—
mapping	Optional[List[str]]	None	—
env	opytimizer.core.environment.Environment	None	—

Methods

clip_by_bound

clip_by_bound(self) -> None

Clips the agent's decision variables to the bounds limits.

dominates

dominates(self, other: 'Agent') -> bool

Checks if this agent dominates another agent.

Returns: (bool): Whether this agent dominates the other.

Parameters

Parameter	Type	Default	Description
other	Agent		Another agent to be compared.

fill_with_binary

fill_with_binary(self) -> None

Fills the agent's decision variables with a binary distribution.

fill_with_static

fill_with_static(self, values: numpy.ndarray) -> None

Fills the agent's decision variables with static values. Note that this method ignore the agent's bounds, so use it carefully.

Parameters

Parameter	Type	Default	Description
values	numpy.ndarray		Values to be filled (expected shape (n_variables,)).

fill_with_uniform

fill_with_uniform(self) -> None

Fills the agent's decision variables with a uniform distribution based on bounds limits.