igraph plot for multiple layers partition does not return expected output

[vtraag]

@balandongiv

However, plotting these partition does not result the expected output

Could you indicate what output you would expect exactly?

@balandongiv

G_layers, G_interslice, G = la.time_slices_to_layers ( [G_1, G_2, G_3], interslice_weight=1,
                                                       slice_attr='slice', vertex_id_attr='id',
                                                       edge_type_attr='type', weight_attr='weight' )

partition = la.find_partition(G, la.CPMVertexPartition,resolution_parameter = 0.05);

Please note that this may be incorrect. You try to detect clusters in the overall graph containing both all layers and the interslice coupling links. You typically want to use a resolution parameter of 0 for the interslice coupling links however. Please see the documentation and the included example for more details. Note that there’s also the simpler find_partition_temporal, which does most of this stuff automatically correctly.

@balandongiv

I suspect this is due to wrongly assign the vertex_label

ig.plot(partition,vertex_label = ['A','B','C','D','E','F','G'])

I’m not sure exactly what the expected output would be, but in principle this might not give correct results. The partition is defined on a graph with 3 * 7 = 21 nodes, and the vertex_label that you provide just has 7 entries. For some reason, the entries seem to be recycled across all nodes, but I’m not 100% sure whether the labels refer to the “correct” label always. Instead, you can use

ig.plot(partition, vertex_label = G.vs['id'])

which is guaranteed to have the correct labels.