Computing Adamic-Adar index for a specific node in a graph

link_prediction.py

import networkx as nx
import pandas as pd


def main():
    print("Creating graph...")
    G = nx.dense_gnm_random_graph(4000, 20000)
    print("Created graph: ", G)

    print("nodes", list(G))
    nodes = list(G)

    u = 0
    nodes.remove(u)
    ebunch = [(u, x) for x in nodes]

    print("Ebunch", ebunch)

    print("Making link predictions...")
    preds = nx.adamic_adar_index(G, ebunch=ebunch)
    print("Made link predictions.")

    preds_df = pd.DataFrame(preds, columns=["node_1", "node_2", "adamic_adar_index"])

    print("20 Best link predictions for node 0: ")
    sorted_preds_df = preds_df.sort_values(by="adamic_adar_index", ascending=False)
    print(sorted_preds_df[:20])


if __name__ == "__main__":
    main()