def test_cartpole():
cartpole = QLearning.CartPole()
policy_filename = "policy/policy_9600.npy"
env = gym.make('CartPole-v0')
env = env.unwrapped
env = Monitor(env, "video", force=True)
episodes = 1000
step = 20000
my_test(policy_filename, env, episodes, step, cartpole.discrete_util)
