public interface FullActionModel
Actionobjects that support returning the full Action transition dynamics. Many planning algorithms, such as Dynamic programming methods, require the full transition dynamics, so if you wish to use such an algorithm and it is possible to fully enumerate the transition dynamics, your
Actionimplementation should implement this interface. The required
getTransitions(burlap.oomdp.core.states.State, burlap.oomdp.singleagent.GroundedAction)method provides the full transition dynamics of an
Action. This method should return a list of all transitions from the input
Statethat have non-zero probability of occurring. These transitions are specified with a
TransitionProbabilityobject that is a pair consisting of the next
Stateand a double specifying the probability of transitioning to that state.
FullActionModel.FullActionModelHelperthat has static helper methods that may be useful for working with an
FullActionModel. Specifically, it includes a method for sampling a state from the transition distribution defined with
getTransitions(burlap.oomdp.core.states.State, GroundedAction)and it includes a method for removing
TransitionProbabilityelements from a list that are assigned zero probability.
|Modifier and Type||Interface and Description|
A class with helper methods for working with actions that implement
java.util.List<TransitionProbability> getTransitions(State s, GroundedAction groundedAction)
TransitionProbabilityobjects. The list is only required to contain transitions with non-zero probability.
s- the state from which the transition probabilities when applying this action will be returned.
GroundedActionspecifying the parameters to use