Table 5.

A feature vector consists of the parameters degree all (DG_all), average degree (AvgDG_nb), cluster coefficient all (CluC_all), cluster coefficient of 2nd neighbors (CluC₂), variation coefficient of neighbor degree (VC_DG), and locality (Loc; for a definition of parameters, see Echtermeyer et al., 2011).

Component	DGall	AvgDGnb	CluCall	CluC2	VCDG	Loc	Share
1	0.392	0.448	0.457	−0.297	−0.375	0.456	58.893%
2	−0.529	0.266	0.412	0.425	−0.45	−0.312	20.035%
3	0.33	−0.382	−0.107	0.693	−0.324	0.384	13.459%
4	0.164	0.43	0.251	0.483	0.697	0.073	5.918%
5	−0.46	−0.471	0.491	−0.122	0.254	0.496	1.468%
6	0.468	−0.419	0.533	−0.049	0.04	−0.544	0.328%

For each principal component, the contribution of the feature vector is given for each feature, and by its Share (percentage of the variance of a component of the total variance of the 6-dimensional feature space; Echtermeyer et al., 2011).