Reward function Question: Implement the reward function described in the setup. Specifically, given a - dimensional vector 'avg', return - dimensional vector 'rew' such that []=[]+[] where [](0,) where is the identity matrix of size