[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
## Re: [igraph] label propagation algorithm

**From**: |
Tamas Nepusz |

**Subject**: |
Re: [igraph] label propagation algorithm |

**Date**: |
Fri, 8 May 2015 14:59:04 +0200 |

**User-agent**: |
Mutt/1.5.23 (2014-03-12) |

>* what dose the input graph , shoud be undirected to make sense mean , in*
>* label propagation algorithm?*
In my opinion, yes, it should be undirected because communities are not
well-defined for directed graphs anyway. (None of the classical, informal
definitions of a community take edge directions into account as far as I know).
>* so what is the output when you run on a directed graph?*
Labels will be propagated forward along the edges, but the convergence of the
algorithm is not guaranteed; for instance, nodes without inbound edges are free
to choose their labels in an arbitrary manner because no other nodes influence
them.
T.