Python code for evolving complex networks to maximize their robustness.
evolving_network.py: This script uses the parameters specified in <params file> to evolve a seed network until it is maximally robust, and saves the resulting (robust) network in GraphML format.
$ python evolving_network.py <params file>The parameters file (see params.json.sample) has the following JSON format:
{
"seedNetwork" : {"name" : "barabasi_albert_graph",
"args" : {"n" : 100, "m" : 3}},
"mutantsPerEpoch" : 10,
"rewiringsPerMutant" : 10,
"stagnantEpochs" : 30,
"attackStrategy" : "degree",
"sequentialMode" : false,
"maximizeRobustness" : true,
"outFile" : "network.graphml",
"verbose" : true
}where seedNetwork specifies the seed network to evolve (see details below); mutantsPerEpoch specifies the number of mutants to generate in each epoch; rewiringsPerMutant specifies the number of edge rewirings per mutant; stagnantEpochs specifies how many epochs to wait until termination, during which there is no increase in robustness; attackStrategy specifies the attack strategy (betweenness, closeness, degree, eigenvector, or random) to use in order to assess the robustness of the evolving network; sequentialMode specifies whether the mode of attack is sequential or simultaneous; maximizeRobustness specifies whether to maximize or minimize robustness of the evolving network; outfile specifies the name of the output file which will store the evolved network; and verbose specifies whether or not to output the properties of the network as it evolves.
The seedNetwork parameter includes name, which can be any network construction method from NetworkX (barabasi_albert_graph for example), and args, which are the arguments for that method (n and m for example).
plots.py: This script creates plots (as PDF files) showing the time evolution of various network properties (averaged over all the replicates) generated from the results produced by the evolving_network.py program (if verbose is true). The names of the result files, one per replicate, are fed in via STDIN.
$ python plots.py < FILESIf you have any questions about the software, please email [email protected].