example_networkx_wrapper_kwargs.py

#!/usr/bin/env python3
"""
NetworkX wrapper kwargs demonstration.

Shows how to pass keyword arguments through `monoplex_nx_wrapper` for weighted
centrality computations. Prerequisites: py3plex installed; NetworkX is already
available through py3plex dependencies.
"""

from __future__ import annotations

from py3plex.core import multinet


def build_weighted_network() -> multinet.multi_layer_network:
    """Create a simple weighted multilayer network."""
    network = multinet.multi_layer_network()
    edges = [
        {"source": "A", "target": "B", "source_type": "layer1", "target_type": "layer1", "weight": 2.0},
        {"source": "B", "target": "C", "source_type": "layer1", "target_type": "layer1", "weight": 3.0},
        {"source": "C", "target": "D", "source_type": "layer1", "target_type": "layer1", "weight": 1.0},
        {"source": "A", "target": "D", "source_type": "layer1", "target_type": "layer1", "weight": 1.5},
        {"source": "B", "target": "D", "source_type": "layer1", "target_type": "layer1", "weight": 2.5},
    ]
    network.add_edges(edges)
    return network


def main() -> int:
    """Run keyword-argument demonstrations."""
    network = build_weighted_network()

    print("=" * 70)
    print("Demonstrating monoplex_nx_wrapper with kwargs support")
    print("=" * 70)

    print("\n1. Degree Centrality (unweighted):")
    print("-" * 50)
    degree_cent = network.monoplex_nx_wrapper("degree_centrality")
    for node, centrality in sorted(degree_cent.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n2. Betweenness Centrality (unweighted):")
    print("-" * 50)
    betweenness_unweighted = network.monoplex_nx_wrapper("betweenness_centrality")
    for node, centrality in sorted(betweenness_unweighted.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n3. Betweenness Centrality (weighted):")
    print("-" * 50)
    print("   Using kwargs={'weight': 'weight'} to consider edge weights")
    betweenness_weighted = network.monoplex_nx_wrapper(
        "betweenness_centrality",
        kwargs={"weight": "weight"},
    )
    for node, centrality in sorted(betweenness_weighted.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n4. Betweenness Centrality (weighted + not normalized):")
    print("-" * 50)
    print("   Using kwargs={'weight': 'weight', 'normalized': False}")
    betweenness_custom = network.monoplex_nx_wrapper(
        "betweenness_centrality",
        kwargs={"weight": "weight", "normalized": False},
    )
    for node, centrality in sorted(betweenness_custom.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n5. Closeness Centrality (using weight as distance):")
    print("-" * 50)
    print("   Using kwargs={'distance': 'weight'}")
    closeness_weighted = network.monoplex_nx_wrapper(
        "closeness_centrality",
        kwargs={"distance": "weight"},
    )
    for node, centrality in sorted(closeness_weighted.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n6. PageRank (custom damping factor):")
    print("-" * 50)
    print("   Using kwargs={'alpha': 0.90} instead of default 0.85")
    pagerank_custom = network.monoplex_nx_wrapper(
        "pagerank",
        kwargs={"alpha": 0.90},
    )
    for node, centrality in sorted(pagerank_custom.items(), key=lambda x: x[1], reverse=True)[:3]:
        print(f"   {node}: {centrality:.4f}")

    print("\n" + "=" * 70)
    print("All examples completed successfully!")
    print("=" * 70)
    print("\nKey takeaway:")
    print("  The kwargs parameter now allows you to pass any NetworkX function")
    print("  parameters, enabling weighted centrality calculations and custom")
    print("  configurations for multiplex networks.")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())