State of the Q Network

[abhincho]

Hi,
I have a doubt. In your code the state/observation is a n length vector ( n is the number of nodes in the graph) initialized to zero.
e.g. for a 5 node graph [ 0 0 0 0 0]
And when an action (node) is taken in the MVC environment we get a new state. In your code, the new state is give as this self.observation[node]=1. So if the action (node) 2 is selected than the vector becomes
[0 1 0 0 0]
Why is just the action(node) selected, set to 1 and not all the vertices that is reachable from that action(node) not set to 1. i.e. shouldn't the new state be all the states that have been visited after selecting the action.

Is this correct or am I missing something?