clustering-coefficient 0.1.1

Computes the clustering coefficient of nodes as defined by Watts & Strogatz (in their 1998 paper).

Clustering Coefficient

This script allows to compute Watts & Strogatz's clustering coefficient of nodes in a graph $G = (V, E)$. It is defined as the edge density of the graph induced from neighbors of a node, relatively to a clique of comparable size. More precisely, given a node $u \in V$, denoting its set of neighbors as $N_G(u)$, the clustering coefficient of $u$ is equal to:

$C_G(u) = \frac{|E(N_G(u))|}{d(d-1)/2}$

where $d = |N_G(u)|$ is the degree of node $u$ (its number $|N_G(u)|$ of neighbors).

Installing and using the plugin

The library relies on tulip-python, a python binding of the C++ Graph Visualization framework Tulip. Tulip also comes as a GUI.

Several libraries need to be installed prior to using the plugin, that can for instance be installed running poetry install --no-root. The specific dependencies are listed as part of the pyproject.toml file. A simple test script can optionally be run.

The plugin itself is typically used as:

# assuming a graph as already been defined
params = tlp.getDefaultPluginParameters('Clustering Coefficient', graph)
clustering = graph.getDoubleProperty('clustering coeff')
params['result'] = clustering
graph.applyDoubleAlgorithm('Broker score', clustering, params)

Alternatively, the plugin may be used within the Tulip GUI after the script has been loaded and ran.