0 # initial state
(0,0,'!b',ConstantRewardFunction(0))
(0,1,'b',ConstantRewardFunction(0))
(1,1,'!d',ConstantRewardFunction(0))
(1,2,'d',ConstantRewardFunction(0))
(2,2,'!b',ConstantRewardFunction(0))
(2,3,'b',ConstantRewardFunction(0))
(3,3,'!g',ConstantRewardFunction(0))
(3,4,'g',ConstantRewardFunction(1))
(4,4,'True',ConstantRewardFunction(0))
