0 # initial state
[2] # terminal state
(0,0,'!B',ConstantRewardFunction(1))
(0,1,'B',ConstantRewardFunction(1))
(1,1,'!R',ConstantRewardFunction(1))
(1,2,'R',ConstantRewardFunction(1))
(2,2,'!Y',ConstantRewardFunction(1))
(2,0,'Y',ConstantRewardFunction(1)) 