for (istate, s) in enumerate(iterator(statespace))
to iterate through the states when solving, and then state_index() when the policy is queried, the policy will only work if state_index() and the order of states in the iterator happen to match.
This is not required by POMDPs.jl and so should not be relied upon.
since we use
for (istate, s) in enumerate(iterator(statespace))to iterate through the states when solving, and then
state_index()when the policy is queried, the policy will only work ifstate_index()and the order of states in the iterator happen to match.This is not required by POMDPs.jl and so should not be relied upon.