Evolving Networks with memory
In in this page, we describe the software we developed in order to simulate networks with reinforcing walkers, based on the article available at this page. We can actually simulate much more than what we have on the paper. In particular, we can simulate several type of dynamics, destroy links at random, and change various parameters. We can plot plenty of stuff, and also as a function of some variables. We can have directed and undirected graphs, colored (weighted) and have some degree of decision on the starting graph. It took us a while to make it work like this.. but we did it!
If you wanted to receive a copy of the Matlab code package, contact me at "fcaravelli [at] ucl . ac . uk".
This software was developed over more than one year. Thanks to it, we are able to simulate evolving and growing networks, subject to reinforcement walkers (currents) and decay.
We can create videos like this:
and analyze the behavior of the model according to several parameters, and also thermalization times, clustering coefficients and other parameters of the graph.
We have applied for a Marie Curie fellowship, and in the case in which this is approved, this software will be severely improved in order to simulate memristive networks, measure information transfer, synchronization and self-organization.
If you plan to use this software, download it here:
http://figshare.com/articles/Evolving_Networks_with_Memory/1486365
And please cite:
F. Caravelli et al 2015 EPL 109 28006 doi:10.1209/0295-5075/109/28006
Caravelli, Francesco; Hamma, Alioscia; Massimiliano Di Ventra (2015): Evolving Networks with Memory. figshare.
http://dx.doi.org/10.6084/m9.figshare.1486365