I am having a confusion of these two terms 'observation_space' and 'state', and I do not see a purpose of even having the 'observation_space' in my code in the first place. I have seen other answers, but I dove deeper into the code of RL algorithms like keras-rl DDPGAgent and I do not even see a mere usage of this 'observation_space'.
The project that I am working on employs a double DQN, and it takes in a state and outputs an action to take based on the highest Q-value output from the model. From this, can someone shed some light on the use of the 'observation_space' in this application of a double DQN? I am trying to create a standardised environment with gym.Env inheritance, and this 'space' is annoying me.
If there is a code source out there that even uses this 'observation_space', please do share it too!