I want to generate a random graph to compare it to a real bipartite directed graph. I need a benchmark to compare the real graph against, so I thought of using the Erdős–Rényi G(n, p) model. My idea is to use to generate the random graph the same number (n) of nodes for each type of the real graph. But what probability (p) should I apply to the random graph? Can I derive p from my real graph with a formula?I want to generate a random graph to compare it