Spark GraphX之全局聚类系数、局部聚类系数、网络平均聚类系数

修行修心

已于 2022-11-07 11:31:15 修改

阅读量9.2k

点赞数

分类专栏： Spark 图文章标签： Spark 图论

于 2018-07-18 20:57:31 首次发布

本文链接：https://blog.csdn.net/yitengtongweishi/article/details/75039682

版权

本文探讨了Spark GraphX中的全局聚类系数、局部聚类系数和网络平均聚类系数，用于衡量图的连通性。通过实例展示了如何计算这些系数，并解释了GraphX在处理重复顶点和边时的处理方式。

摘要由CSDN通过智能技术生成

Global clustering coefficient（全局聚类系数）

Another way to measure connectedness, the global clustering coefficient, is better in that it always returns a number between 0 and 1, making it possible to compare the connectedness of different sized graphs.

全局聚类系数通常定义如下：closed triplets / total triplets (open or closed)

A triplet in this case is a set of three vertices that have two or three edges among them. If there are three edges, then it’s a triangle, and this is called a closed triplet. If there are only two edges, then it’s called an open triplet. Triplets are counted for each vertex and then added all together; this means that a triangle will count as three closed triplets, because each of the three vertices will h