Local Clustering Coefficient

Supported Graph Characteristics

Unweighted edges

Undirected edges

Homogeneous vertex types

The local clustering coefficient of a vertex in a graph quantifies how close its neighbors are to being a complete graph. In a complete graph, every two distinct vertices are connected.

This algorithm computes the local clustering coefficient of every vertex in a graph.

It is obtained by dividing the number of edges between a vertex’s neighbors by the number of edges that could possibly exist.

For more information, see Clustering Coefficient.


The algorithm does not report the local clustering coefficient for vertices that have only one neighbor.

Local Clustering Coefficient only works with directed graphs if reverse edges are included.


tg_lcc(STRING v_type, STRING e_type,INT top_k=100,
  BOOL print_results = True, STRING result_attribute = "",
  STRING file_path = "", BOOL display_edges = FALSE)

Time complexity

\$O(V)\$, where \$V\$ is the number of vertices in the graph.


Parameter Description Data type

STRING v_type

Vertex type to calculate local clustering coefficient for

(empty string)

STRING e_type

Edge type to traverse. Only vertices that are connected by edges of this type are treated as connected in this algorithm.

(empty string)

INT top_k

Number of the highest local clustering coefficients to report.


BOOL print_results

If true, output JSON to standard output.


STRING result_attribute

If provided, the local clustering coefficient of a vertex will be saved to this attribute.

(empty string)

STRING file_path

If provided, write output to this file in CSV.

(empty string)

BOOL display_edges

If true, the algorithm will also return edges, which help produce better visualized results in GraphStudio.


Run commands

Schema-Free Query

RUN QUERY tg_lcc (<parameters>)

Packaged Template Query

CALL GDBMS_ALGO.community.lcc (<parameters>)


Using the social graph below as an example, Jenny has three neighbors - Tom, Dan and Amily, but only Tom and Dan are connected. Between the three neighbors, the maximum number of edges that could exist is 3. Therefore, the local clustering coefficient for Jenny is 1/3.

On the other hand, Tom has two neighbors that are connected by one edge. Since the maximum number of edges that could exist between two vertices is 1, Tom has a local clustering coefficient of 1/1 = 1.

image (78)

By running the algorithm in GSQL, we can confirm that Tom has the highest local clustering coefficient with a score of 1 and Jenny has a score of 1/3.

RUN QUERY tg_lcc("person", "friendship", _, _, _, _, _)

  "error": false,
  "message": "",
  "version": {
    "schema": 0,
    "edition": "enterprise",
    "api": "v2"
  "results": [{"top_scores": [
      "score": 1,
      "Vertex_ID": "Tom"
      "score": 0.33333,
      "Vertex_ID": "Jenny"
      "score": 0.16667,
      "Vertex_ID": "Dan"
      "score": 0,
      "Vertex_ID": "Nancy"