Skip to main content
Fig. 6 | Applied Network Science

Fig. 6

From: The block-constrained configuration model

Fig. 6

USairports graph visualisation. USairports graph visualisation. The graph is plotted by means of the force-atlas2 layout with weighted edges, and the size of the vertices reflects their out-degrees. Only the largest connected component of the graph is shown. The visualisations clearly show the block structure that characterises this graph. The vertices in the four visualisations are coloured according to the labels detected applying four community detection algorithms, as described in Table 5. The visualisations are ordered from left to right according to the AIC of the BCCM fitted to observed graph according to the corresponding block structure. From left to right, we see the colours corresponding to the labels obtained from louvain, fast_greedy, infomap and label_propagation detection algorithms respectively. We highlight the fact that the ranking according to AIC corresponds approximately to the ability of the algorithms to detect the separation between high-degree (and low-degree) vertices within the largest cluster, at the top of the visualisations. The reason for this is that within the largest cluster there are clear deviations from what the configuration model predicts, i.e., high-degree vertices tend to connect to each other, and the best BCCMs captures more of these deviations

Back to article page