Of Notes
Macroscopic Mean Values
Table of Contents
In Basic Graph Theory every quantity we defined belonged to a single node: the degree
To describe the network as a whole we average these microscopic quantities over all
Mean Degree
We take the graph
We sum the degree of every vertex and divide by the number of vertices:
where
We can evaluate the sum by counting edge endpoints instead. Each of the
Degree Distribution
Instead of summing over vertices we can pick a vertex uniformly at random.
Let
which is a normalized distribution since
The mean degree is then just the expectation over this distribution:
which is the same number as
Mean Clustering
We apply the same recipe to the local clustering coefficient of each vertex:
where
Mean Distance
We average the Graph Distance over all pairs of vertices:
where
We sum over pairs with
This is the quantity which behaves as
For a single vertex we drop the outer sum:
and inverting this mean distance is exactly the closeness measure in Centrality.