Re: [graph-tool] shortest_distance anomalies?