On 20.12.2017 19:14, Rafael Pereira wrote:
I am not a physicists nor a mathematician, so I am really just a layman when it comes to the mathematical calculations of network science. Sorry if my question doesn't make sense. However, I've read in a few papers* suggesting that the calculation of centrality measures for multilayer networks are a bit different than for single layer networks. I would like to know if graph-tool incorporates this advances in the literature, does it?
If you look closely, you will notice that in most cases the definition of most centrality measures for multilayer networks is identical to the classical "monolayer" case, with the only difference that we need to differentiate between copies of the same node in different layers, and the connections between layers. So all you need to do is use the exact same algorithms for monolayer networks, but represent the multilayer graph as an edge-annotated monolayer network, where each node appears multiple times for each layer it belongs, and their copies are connected by intra-layer edges. -- Tiago de Paula Peixoto <tiago@skewed.de>