1. I am trying to compare two social network friend of friend graphs taken at 2 different locations with different # of vertices and edges. I want to see overall similarity between the structure of the graphs. I was thinking of using "graph-matching" to do this. Is there another option?2.
2. For comparing graphs of different orders, there is a mention giving weight -1 for complementer, and then doing match_vertices. However, the current match vertices algorithm you have uses only adjacency matrices, and when one has negative weights, it can't handle it. Explanation and any code will be appreciated.
3. When one adds graph to its complementer, it doubles the size of the adjacency matrix by adding new vertices with complementary edges. For large networks, this is a hurdle. Anything else I could do?
Thanks.