power_signal_graph: Simulate a scale-free signal graph structure from an input...

In robertness/signalgraph: Signal Processing with Graphs

Description

Takes an input graph structure, fits a power law model, and simulates a new graph based on the fit. This function uses the Barabasi-Albert model to generate graphs implemented in lucy::power_law_sim, however an alteration is made. The edges of the input graph are reversed before estimating the power law fit, and the simulated graph edges are reversed again before it is returned from the function. This way, the graph is characterized not by preferential attachment of incoming edges, but by preferential attachment of outgoing edges. Further, vertices with 0 degrees are avoided by having high degree vertices 'donate' edges to degreeless vertices.

Usage

power_signal_graph(g, p, start_graph = NULL)

Arguments

g	an igraph object.
p	number of desired vertices in the output graph.
Starter	graph as input for preferencial attachment. Should have reversed edges. See start.graph argument in ba.game/barabasi.game/sample_pa functions in igraph.

Value

A new graph simulated from a power law based on g.

robertness/signalgraph documentation built on May 27, 2019, 10:33 a.m.