code/stats.py

import networkx as nx
import numpy as np


def print_graph_stats(G):
    n = G.number_of_nodes()
    m = G.number_of_edges()

    print(f"Number of nodes: {n}")
    print(f"Number of edges: {m}")

    if n == 0:
        print("Average degree: 0.0")
        print("Connectivity (weak or strong): False")
        print("Is connected as undirected graph: False")
        print("Number of weakly connected components: 0")
        print("Number of strongly connected components: 0")
        print("Number of isolated nodes: 0")
        return

    # Average degree via m/n
    avg_deg = m / n
    print(f"Average degree: {avg_deg}")

    # Connectivity checks for directed graphs
    is_weak = nx.is_weakly_connected(G)
    is_strong = nx.is_strongly_connected(G)
    print(f"Connectivity (weak or strong): {is_weak or is_strong}")

    # Undirected connectivity can raise on empty graphs; guard with try/except just in case
    try:
        und_connected = nx.is_connected(G.to_undirected(as_view=True))
    except Exception:
        und_connected = False
    print(f"Is connected as undirected graph: {und_connected}")

    print(f"Number of weakly connected components: {nx.number_weakly_connected_components(G)}")
    print(f"Number of strongly connected components: {nx.number_strongly_connected_components(G)}")
    print(f"Number of isolated nodes: {len(list(nx.isolates(G)))}")


def audit_adj(adj: np.ndarray):
    has_self = np.all(np.diag(adj) > 0)  # check self-loops
    sym = np.allclose(adj, adj.T, atol=1e-8)
    deg = adj.sum(1)
    isolated = np.sum(deg==0)
    return {"n":adj.shape[0], "m":int(adj.sum()), "sym":sym, "self_loops":has_self,
            "avg_deg":float(deg.mean()), "isolated_nodes":int(isolated)}