Explosive and implosive root concepts: An analysis of music moods rooted by two influential rap artists
Fig 5
Effects of node removal for analyzing contagion-influence associations.
We measure three indicators to observe the effects of node removals from the influence networks. (1) largest path length (panels (A), (D), and (G)); (2) average path length (panels (B), (E), and (H)); and (3) the number of communities detected by the Louvain method (panels (C), (F), and (I)). The numbers of artists who worked with RD’s and NWA’s collaborators are 146 and 162, respectively. Thus, we randomly remove 146 and 162 nodes 10,000 times and report the averages. In the second and third rows, we present the cumulative distributions of results from the randomly removed samples and add the vertical red lines to show the corresponding data for RD and NWA.