mlpy.mdp.stateaction.StateData

class mlpy.mdp.stateaction.StateData(state_id, actions)[source]

Bases: object

State information interface.

Information about the state can be accessed here.

Parameters:

state_id : int

The unique id of the state

actions : list[Action]

List of actions that can be taken in this state.

Attributes

id (int) The unique id of the state.
models (dict) The reward and transition models for each action.
q (dict) The q-table, containing a q-value for each action.
steps_away (int) The number of steps the state is away from its closest neighbor.