Reciprocal distance sum RDS(v) = Σ 1/d(v,u) for all u≠v. Vertex-level Harary: H(G) = ½ Σ RDS(v). High RDS = central, accessible vertex. Low RDS = peripheral. Close vertex analogue of closeness centrality.
How to Use
Select graph:
- RDS(v): Per vertex
- Max: Most central
- H = ½Σ: Harary
Centrality Measure
RDS(v) is a closeness centrality variant. Classical closeness: n/Σd(v,u). RDS: Σ1/d(v,u). RDS is more robust: doesn't blow up with one distant vertex. Harmonic centrality.
Node Identification
Highest RDS: graph center (most accessible). Lowest RDS: periphery. RDS ranking identifies structurally important vertices for drug targets, network hubs, central atoms.
Step-by-Step Instructions
- 1Select graph.
- 2For each vertex v: sum 1/d(v,u).
- 3Rank vertices by RDS.
- 4Identify central nodes.
- 5Sum for Harary.