1: \begin{abstract}

2: In this paper, we describe the general framework to describe the diffusion operators associated to a positive matrix.

3: We define the equations associated to diffusion operators and present some general properties of their state vectors.

4: We show how this can be applied to prove and improve the convergence of a fixed point problem associated to the matrix iteration scheme.

5: The approach can be understood as a decomposition of the matrix-vector product operation in elementary operation

6: at the vector entry level.

7: \end{abstract}