mlpy.mdp.stateaction.Experience

class mlpy.mdp.stateaction.Experience(state, action, next_state, reward=None)

Bases: object

Experience base class.

Representation of an experience occurring from acting in the environment.

Parameters:

state : MDPState

The representation of the current state.

action : MDPAction

The executed action.

next_state : MDPState

The representation of the state following from acting with action in state state.

reward : int or float

The reward awarded by the environment for the state-action pair.

Attributes

state	(MDPState) The experienced state
action	(MDPAction) The experienced action.
next_state	(MDPState) The experienced next state.
reward	(float) The experienced reward.