ScaleWeaver: A High-Fidelity Artificial Network of Networks (NoN) Generator

Setup and Installation:

Prerequisites

Ensure your system meets the following requirements:

1. Python >=2.8, 3.2 or above
2. NetworkX library (version >= 1.10)
3. NumPy library (version >= 1.5)

Verification:

To verify your installation, run Python and execute the following commands:

import sys, numpy, networkx
print(sys.version)
print(numpy.__version__)
print(networkx.__version__)

Optional Dependency:

For visualization purposes, it is recommended to install Graphviz.

Citations: [1] https://realpython.com/generative-adversarial-networks/ [2] https://www.v7labs.com/blog/neural-network-architectures-guide [3] https://neptune.ai/blog/6-gan-architectures [4] https://www.acer.europa.eu/electricity/connection-codes/requirements-for-generators [5] https://www.entsoe.eu/network_codes/rfg/

Usage Example:

python ScaleWeaver.py -f data-samples/Nest8W.edges -t edgelist -p "{'node_growth_rate':[0.005, 0.001], 'edge_edit_rate':[0.05, 0.04, 0.03], 'node_edit_rate':[0.07, 0.06, 0.05]}" -o output/test.dot

1. Loads data
2. Increases nodes: 0.9% (level 1), 0.1% (level 0)
3. Edits: 3% edges, 5% nodes (level 2); 4% edges, 6% nodes (level 1); 5% edges, 7% nodes (level 0)
4. Outputs to test.dot

Direct Python call:

import ScaleWeaver
new_G = ScaleWeaver.generate_graph(G, params)

Key Parameters:

• node_growth_rate, edge_growth_rate: Values in (-1,infinity)
• ScaleWeaver: Alternative generators (e.g., AltAlgo.expected_degree_replicate)
• accept_chance_edges: Controls long-distance edge insertion (0.0-1.0)
• deferential_detachment_factor: Modifies edge deletion probability (default: 1)
• locality_bias_correction: Adjusts triangle generation (-1 to 1)
• new_edge_horizon: Depth of scanning for new edges (positive integer)
• preserve_clustering_on_deletion: Enhances clustering fidelity (boolean)

Advanced Options:

• MainAlgo:MainAlgo.ScaleWeaver_snapshots: Sequential editing with snapshots
• MainAlgo:MainAlgo.ScaleWeaver_iterated_cycles: Alternating editing cycles
• MainAlgo:MainAlgo.ScaleWeaver_on_subgraphs: Apply to each component separately
• post_processor: Custom function for final processing

Troubleshooting:

Accelerating the computation

• remove the computation of the metrics with "-M False" argument (all metrics); or in UtilAlloc.py, make some complex metrics optional.
• convert the original graph into a simple format like edgelist
• reduce the new_edge_horizon (reduces fidelity)
• set deferential_detachment_factor close to 0 (reduces fidelity)
• give the authors funding to develop a version in the C language

Clustering is diminished in the replicas

• make sure that the pattern is consistent - there would be and need to be some variance
• reduce the edge editing rate at the finest level: it would improve preservation of all fine-level properties of the graph
• increase locality_bias_correction
• make fine_clustering True
• make preserve_clustering_on_deletion True

Path lengths and related global metrics are diminished in the replicas

• make accept_chance_edges closer to 0. As a compensation to maintain fidelity, increase new_edge_horizon to 10, 20 or larger.

Support for weighted edges

• currently only used in coarsening
• I am planning to include it in future releases

Support for node and edge attributes

• use the parameters 'maintain_node_attributes':True, 'maintain_edge_attributes':True

Error writing DOT file or pygraphviz error:

• pygraphviz is not currently functional in the Windows platform
• specify an alternative output file such as "-o my_output.elist"
• or, try install/ing pydot package

Conclusion:

Thank you for your interest in ScaleWeaver. I hope this tool proves valuable in your network analysis and generation tasks.

Key Features:

• High-fidelity replication of complex network structures
• Multiscale approach for preserving properties across different granularities
• Customizable parameters for fine-tuning network generation
• Support for various input and output formats

Future Development:

I am committed to continually improving ScaleWeaver. Planned enhancements include:

1. Full support for weighted edges
2. Improved performance optimizations
3. Extended documentation and use case examples
4. Integration with additional network analysis tools

Community Contributions:

I welcome contributions from the community. If you encounter issues, have feature requests, or wish to contribute code, please visit our GitHub repository.

Final Notes:

ScaleWeaver is provided as Open-Source, I strive for accuracy and reliability.

I appreciate your feedback and hope ScaleWeaver enhances your network analysis capabilities. Enjoy exploring the intricate world of complex networks with ScaleWeaver!