amlgym.modeling package

Submodules

amlgym.modeling.UPEnv module

class amlgym.modeling.UPEnv.UPEnv(domain_path, problem_path)

Bases: Env

A Gymnasium environment which simulates a PDDL problem through the unified-planning sequential simulator.

Set environment state and seed through reset() for reproducibility

__init__(domain_path, problem_path)

Set environment state and seed through reset() for reproducibility

applicable_actions(state)

Returns the set of actions applicable in the current state s :type state: :param state: the current state s :rtype: Dict[str, Set[str]] :return: the list of actions applicable in the current state s

apply(state, action)

Return the state \(s'\) reached after executing action \(a\) in state \(s\). :type state: :param state: current state \(s\) :type action: :param action: action \(a\) to be executed :return: future state \(s'\)

property ground_actions: Dict[str, Any]

Return a list of all ground actions for the current environment. :return: ground actions list

problem: Problem

amlgym.modeling.env module

class amlgym.modeling.env.Env

Bases: ABC

An AMLGym environment interface that is required for measuring predictive power

metrics. An example of AMLGym environment is a unified-planning sequential simulator that implements the apply() and applicable_actions() methods.

abstract applicable_actions(state)

Returns the set of actions applicable in the current state s :type state: TypeVar(StateType) :param state: the current state s :rtype: List[TypeVar(ActionType)] :return: the list of actions applicable in the current state s

abstract apply(s, a)

Return the state obtained after exeucuting action a in the current state s :type s: TypeVar(StateType) :param s: the current state s :type a: TypeVar(ActionType) :param a: the action a to execute :rtype: TypeVar(StateType) :return: the state s’ reached after executing action a in state s

amlgym.modeling.trajectory module

class amlgym.modeling.trajectory.Trajectory(states, actions)

Bases: object

actions: List[ActionInstance]

states: List[UPState]

write(file_path)

Module contents