1.code inside "A formal introduction to Deep Reinforcement Learning"

main.py

import gym
env = gym.make('CartPole-v1')
# initialize metrics
total_reward = 0
steps = 0
current_state = env.reset() # obtain first state
done = False
while not done: # when done is True the episode ends
    action = env.action_space.sample() # get a random action from A of the environment
    next_state, reward, done, _ = env.step(action) # perform the action
    print(f'Transition from state {current_state} to state {next_state}, '
    + f'I earned reward: {reward} and now the episode is done is {done}')
    # update metrics
    total_reward += reward
    steps += 1
    current_state = next_state # update current_state
print(f'Episode done in {steps} steps, total reward {total_reward}')