def main():
env = Blackjack()
policy = init_policy(env)
v = init_state_map(env)
visits_map = init_state_map(env)
for _ in xrange(20000):
episode = generate_episode(env, policy)
on_policy_state_evaluation(episode, v, visits_map)
env.visualize_state_value(v)
