Hands-On Reinforcement Learning for Games

Testing Gym

In the next exercise, we will write code to test Gym and an environment called FrozenLake, which also happens to be our test environment for this chapter. Open the Chapter_2_4.py code example and follow the exercise:

1. For reference, the code is shown as follows:

from os import system, nameimport timeimport gymimport numpy as npenv = gym.make('FrozenLake-v0')env.reset()def clear():    if name == 'nt':         _ = system('cls')        else:         _ = system('clear')for _ in range(1000):    clear()    env.render()    time.sleep(.5)    env.step(env.action_space.sample()) # take a random actionenv.close()

2. At the top, we have the imports to load the system modules as well as gym, time, and numpy. numpy is a helper library we use to construct tensors. Tensors are ...