Skip to content

Graph wide algorithms

The first category of algorithms we can run are graph wide, returning one value for the whole graph. There are many useful metrics that fall into this category, of which we run three in the code below:

  • Graph Density - which represents the ratio between the edges present in a graph and the maximum number of edges that the graph could contain.
  • Clustering coefficient - which is a measure of the degree to which nodes in a graph tend to cluster together e.g. how many of your friends are also friends.
  • Reciprocity - which is a measure of the likelihood of nodes in a directed network to be mutually connected e.g. if you follow someone on twitter, whats the change of them following you back.

As you can see below, to run an algorithm you simply need to import the algorithm package, choose an algorithm to run and hand it your graph.

Graph

from raphtory import algorithms as rp

density = rp.directed_graph_density(lotr_graph)
clustering_coefficient = rp.global_clustering_coefficient(lotr_graph)
reciprocity = rp.global_reciprocity(lotr_graph)

print(f"The graph's density is {density}")
print(f"The graph's clustering coefficient is {clustering_coefficient}")
print(f"The graph's reciprocity is {reciprocity}")

Output

The graph's density is 0.03654467687010765
The graph's clustering coefficient is 0.4163023913602468
The graph's reciprocity is 0.19115549215406563