This repository provides the codes and data used for the partition-based spread identification performed for the paper "An efficient Partition-Based Approach to Identify and Scatter Multiple Relevant Spreaders in Complex Networks".
We have uploaded the paper to XX and has been published. It is available at https://. The Authors are Jedidiah Yanez-Sierra, Arturo Diaz-Perez and Victor Sosa-Sosa.
The repository is organized as follows:
- Algorithms: source code containing the implementations of the PBSI method, as well as the reference methods.
- graphs: the folder containing the input graphs used in each experiment (due to size limits, only small graphs are included).
- Plots: output folder.
- SprModel: source code containing the implementations of the SIR propagation process.
The root folder contains the following scripts correlated with each experiment performed in the paper:
- Exp1.py Script containing the execution pipeline to compute the Final Spreading Scope achieved for a set of spreaders identified by the selected methods for a range of spreading probabilities. It receives as input the path of a complex network, the name of the set of methods, and the number of spreaders.
- Exp2.py Implementation to compute the Final Spreading Scope achieved by a range of spreaders identified by the selected methods for three distinct spreading probabilities. It takes as input a complex network, the range of spreaders, and the name of the set of methods.
- Exp3.py Script used to compute the Average Shortest Path Length among the source spreaders selected by the methods selected. It takes as input a complex network, the range of spreaders, and the name of the set of methods.
- Exp4.py Performs a comparison of the Final Spreading Scope achieved for each method using the distribution strategy used by PBSI. It takes as input a complex network, the spreading probability, and the name of the set of methods.
- Input graphs must be in GraphML format.
- Graphs must be connected with only one component. When more than one component is present, only the giant must be kept.
- Additional graphs used in the paper can be downloaded from:
For each script, you need to specify all the above-described parameters, optional parameters to modify the number of Monte Carlo simulations, and SIR recovering probability are also available. The results are stored in the Plot directory.
Python and libraries versions currently used:
| Library | Version |
|---|---|
| python | 2.7.16 |
| python-igraph | 0.8.3 |
| numpy | 1.16.5 |
| matplotlib | 2.2.3 |