How to prepare data and outputs to reinforcement learning

Question

I\'m trying to apply Reinforcement Learning to my project, I was able to find many articles explaining how to use "x almost ready source" but this is not even close to