Wrong distances in valued/weighted social networks

[Dimitris Kalamaras]

Due to a bug in our dijkstra implementation, there are cases where the distances in valued/weighted graphs are not all correct.
For instance,

1. Import the following undirected, valued network of 4 actors via a csv file

0 1 0 0
1 0 10 0
0 10 0 1
0 0 1 0

2. Compute Distance Matrix.

The result will be
      v 1 2 3 4

      1 0 1 11 12
      2 1 0 10 11
      3 11 10 0 1
      4 inf 11 1 0

where DM(4,1)=inf is not correct, since 4 and 1 are reachable and their distance is actually the same as DM(1,4) = 12.