0 # initial state
[2] # terminal state
(0,0,'B|I',ConstantRewardFunction(1))
(0,1,'A',ConstantRewardFunction(1))
(0,2,'D',ConstantRewardFunction(1))
(1,1,'A|I',ConstantRewardFunction(1))
(1,2,'B',ConstantRewardFunction(1))
(1,2,'D',ConstantRewardFunction(1))
(2,2,'A|B|D|I',ConstantRewardFunction(1))
