Hands-On Q-Learning with Python

Choosing an action manually

We will run these following two lines in a notebook:

env.step(1)env.render()

The following is the result for the preceding code:

env.step(1) returns the following four variables:

• observation: This refers to the new state that we are in (that is, state 382).
• reward: This refers to the reward that we have received (-1).
• done: This tells us whether we have successfully dropped off the passenger at the correct location (False).
• info: This provides any additional information that we may need for debugging.

Note that we chose an argument of 1 for env.step() specifically, because we wanted the agent to move north. Choosing ...