0

I'm analyzing a large graph (382 219 nodes, 15 038 083 edges) with 1 721 connected components. One of these components has 374 511 nodes and 15 014 839 edges. Since I don't have the resources (CPU & RAM) to analyze such a large graph I decided to remove the large connected component from the graph.

So far I managed to generate a graph for each connected component and make a list containing them all except the large one.

Now I'd like to generate a graph containing all these small components.

crocefisso
  • 793
  • 2
  • 14
  • 29
  • 1
    I believe you're looking for an algorithm to compute a graph sum? https://mathworld.wolfram.com/GraphSum.html – shynjax287 Sep 04 '20 at 20:01
  • I guess it is. Is it implemented in Networkx or Python? – crocefisso Sep 04 '20 at 20:09
  • 1
    The solution is there : [compose_all](https://networkx.github.io/documentation/stable/reference/algorithms/generated/networkx.algorithms.operators.all.compose_all.html#networkx.algorithms.operators.all.compose_all) – crocefisso Sep 04 '20 at 20:18

0 Answers0