Rodolphe - 2 years ago 182
Python Question

Issue with modularity and clustering in iGraph [Python]

I have a graph g where vertices have an attribute 'group'. I want to measure the modularity of the graph according to the group attribute. Here's how I proceeced:

``````cl = Clustering(g.vs()['group'])
modul = g2.modularity(cl,weights=None)
``````

The cluster object cl looks like this:

``````Clustering with 134 elements and 26 clusters
[ 0] 52, 53, 54, 55, 56
[ 1] 44, 45, 46, 47
[ 2] 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105
[ 3] 73, 74, 75, 76, 77, 78
[ 4] 36, 37, 38, 39
[ 5] 3
[ 6] 85, 86, 87, 88, 89, 90, 91, 92
[ 7] 11, 48, 49, 50, 51
[ 8] 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119,
120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133
[ 9] 12, 57, 58, 59, 60, 61
[10] 62, 63, 64, 65, 66
[11] 4
[12] 10, 15, 16
[13] 79, 80, 81, 82, 83, 84
[14] 0, 13, 14, 21, 22, 23, 24, 32, 33, 34, 35
[15] 28, 29, 30, 31
[16] 67, 68, 69, 70, 71, 72
[17] 1
[18] 7
[19] 19, 20
[20] 9
[21] 8
[22] 2, 40, 41, 42, 43
[23] 5
[24] 6, 17, 18
[25] 25, 26, 27
``````

Why does iGraph raise this error:

``````ValueError: iterable must yield integers
``````

Am I missing something?

The `modularity` function requires an iterable that yields integers as the first argument, not a `Clustering` object. Incidentally, `g.vs["group"]` gives you exactly that, so there is no need to create a `Clustering`:
``````g.modularity(g.vs["group"])
If you really want a `Clustering` object, you need to construct a `VertexClustering` and then use its `modularity` property (which will in turn call the `modularity` method of the graph):
``````cl = VertexClustering(g, g.vs["group"])