mlpy.mdp.stateaction.StateActionInfo

class mlpy.mdp.stateaction.StateActionInfo[source]

Bases: object

The models interface.

Contains all relevant information predicted by a model for a given state-action pair. This includes the (predicted) reward and transition probabilities to possible next states.

Attributes

transition_proba (ProbabilityDistribution) The transition probability distribution.
reward_func (RewardFunction) The reward function.
visits (int) The number of times the state-action pair has been visited.
known (bool) Flag indicating whether a reward value is known or not.