1: Schematic illustration of renormalization (Eq. 1.34) to eliminate a single node n from a Markov chain parameterized by the transition probability matrix T. The transition probabilities T of the renormalized Markov chain account for transitions that occur indirectly, via the "censored" state (here, the n-th node). Thus, the reduced model features a γ ← β transition that is not present in the original network, which corresponds to the family of transitions γ ← n ← . . . ← n ← β, where an arbitrary number of n ← n transitions occur. Similarly, the reduced Markov chain contains a β ← β transition, and the probabilities of the β ← δ and γ ← δ transitions have increased (indicated by +) to account for paths that proceed via the eliminated node n.