Figure 2.

Plots (a) and (b) show root mean squared errors of the proposed model against power-law attachment spherical surface geometry (including the hyperbolic model), respectively. (c) Effect sizes of degree distributions between model and network (log-normal versus power-law attachment). Dotted lines show the line of parity while ICON is dataset from the Index of Complex Networks and NR is the dataset from the network repository. Plots (d–f) show the surface model parameter plotted against the depth model parameter for the proposed theory, power-law attachment theory and spherical surface geometry theory, respectively. Spearman’s correlation coefficients, $r_s$, and their p-values between the parameters indicate how correlated the model parameters are across the 110 networks.