Re: [igraph] Question on MST computation

[Tamas Nepusz]

Dear Rossano,

> I get strange results from the computation of the MST on a weighted  
> undirected graph. Does the graph have to be simple? The graph I need  
> to extract the MST from contains multiple edges.
Unfortunately it looks like the current implementation assumes that  
the graph is simple:

>> g = Graph([(0,1), (0,1), (1,2)])
>> g.es["weight"] = [1,2,3]
>> g2 = g.spanning_tree()
>> g2.get_edgelist()
[(0, 1), (1, 2)]
>> g2.es["weight"]
[2, 3]

Thanks for reporting that, it will be corrected in igraph 0.5.1. Note  
that if you have more edges between vertices v1 and v2, you can safely  
remove all but one (the one possessing the smallest weight) and re-run  
the calculation on the simplified graph.

--
Tamas