def test_acrobot():
acrobot = QLearning.Acrobot()
policy_filename = "policy/policy_9900.npy"
env = gym.make('Acrobot-v1')
env = env.unwrapped
#env = Monitor(env, "video", force=True)
episodes = 1000
step = 2000
my_test(policy_filename, env, episodes, step, acrobot.discrete_util)
