6. Monte Carlo Methods

Activity 6.01: Exploring the Frozen Lake Problem – the Reward Function

1. Import the necessary libraries:

   import gym
   import numpy as np
   from collections import defaultdict

2. Select the environment as FrozenLake. is_slippery is set to False. The environment is reset with the line env.reset() and rendered with the line env.render():

   env = gym.make("FrozenLake-v0", is_slippery=False)
   env.reset()
   env.render()

   You will get the following output:

   Figure 6.15: Frozen Lake state rendered

   This is a text grid with the letters S, F, G, and H used to represent the current environment of FrozenLake. The highlighted cell S is the current state of the agent.

3. Print the possible values in the observation space and the number of action values using the print(env.observation_space) and print(env.action_space) functions respectively:

   print(env.observation_space)
   print(env.action_space)
   name_action = {0:'Left',1:'Down',2:'Right'...