louvain-enhanced 0.2

Community detection package using louvain's algorithm

Louvain-Enhanced

This package has some functions taken from python-louvain package

Louvain-Enhanced is a Python package for community detection in large networks using the Louvain method. This package provides enhanced functionalities and optimizations for efficient and accurate community detection.

Features

• Efficient implementation of the Louvain method for community detection.
• Support for weighted and unweighted graphs.
• Easy integration with NetworkX.
• Randomized node evaluation for different partitions at each call.
• Modular and extensible design.

Installation

You can install Louvain-Enhanced using pip:

pip install louvain-enhanced

Usage

Importing the Package

import networkx as nx

from louvain_enhanced import (
    get_partition_at_level,
    calculate_modularity,
    find_best_partition,
    create_dendrogram,
    create_induced_graph,
    load_binary_graph,
)

Creating a Graph

G = nx.karate_club_graph()

Finding the Best Partition

partition = find_best_partition(G)
print(partition)

Calculating Modularity

modularity = calculate_modularity(partition, G)
print(f"Modularity: {modularity}")

Creating a Dendrogram

dendrogram = create_dendrogram(G)
print(dendrogram)

Getting Partition at a Specific Level

level = 1
partition_at_level = get_partition_at_level(dendrogram, level)
print(partition_at_level)

Creating an Induced Graph

induced_graph = create_induced_graph(partition, G)
print(induced_graph)

Loading a Binary Graph

binary_graph = load_binary_graph("path_to_binary_file")
print(binary_graph)

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Author

Himangshu Singh