首先準備如下社交圖形數據：

打開spark-shell；

導入相關包：

import org.apache.spark._
import org.apache.spark.graphx._
import org.apache.spark.rdd.RDD

創建如上graph對象：

// Create an RDD for the vertices
val users: RDD[(VertexId, (String, Boolean))] =sc.parallelize(Array((1L, ("Li Yapeng", true)), (2L, ("Wang Fei", false)), (3L, ("Xie Tingfeng", true)), (4L, ("Zhang Bozhi", false)), (5L, ("Chen Guanxi", true))))
// Create an RDD for edges
val relationships: RDD[Edge[String]] =sc.parallelize(Array(Edge(1L, 2L, "spouse"),    Edge(2L, 3L, "spouse"),Edge(3L, 4L, "spouse"), Edge(4L, 5L, "friend"), Edge(3L, 5L, "friend")))
// Define a default user in case there are relationship with missing user
val defaultUser = ("Who?", false)
// Build the initial Graph
val graph = Graph(users, relationships, defaultUser)

嘗試打印出所有的男藝人：

graph.vertices.filter(_._2._2).collectres8: Array[(org.apache.spark.graphx.VertexId, (String, Boolean))] = Array((1,(Li Yapeng,true)), (3,(Xie Tingfeng,true)), (5,(Chen Guanxi,true)))

按焦點程度逆序打印出藝人的ID：

graph.degrees.sortBy(_._2,false).collectres12: Array[(org.apache.spark.graphx.VertexId, Int)] = Array((3,3), (2,2), (4,2), (5,2), (1,1))

但是沒辦法知道藝人的名字，join一下：

graph.degrees.leftJoin(graph.vertices)((vid, vd1, vd2)=>(vd2.get._1,vd1)).sortBy(_._2._2, false).collectres35: Array[(org.apache.spark.graphx.VertexId, (String, Int))] = Array((3,(Xie Tingfeng,3)), (2,(Wang Fei,2)), (4,(Zhang Bozhi,2)), (5,(Chen Guanxi,2)), (1,(Li Yapeng,1)))

將sortBy放在后面，是因為sortBy會返回MappedRDD，丟失了VertexRDD的一些方法

直接拿vd2.get有點冒險，下面改成安全版本：

graph.degrees.leftJoin(graph.vertices)((vid, vd1, vd2)=>(vd2 match {case Some(vvd2) => (vvd2._1, vd1); case None => ("", vd1)})).sortBy(_._2._2, false).collect

測試一下消息機制，發送配偶信息給每個人：

graph.aggregateMessages({
(ctx:EdgeContext[(String, Boolean),String,String])=>
if(ctx.attr=="spouse"){
ctx.sendToSrc(ctx.dstAttr._1);
ctx.sendToDst(ctx.srcAttr._1)
}
}, ((s1:String,s2:String)=>s1+"|"+s2)).collectres46: Array[(org.apache.spark.graphx.VertexId, String)] = Array((1,Wang Fei), (2,Li Yapeng|Xie Tingfeng), (3,Wang Fei|Zhang Bozhi), (4,Xie Tingfeng))

輸出pagerank值：

import org.apache.spark.graphx.lib
graph.pageRank(0.01).vertices.collect.foreach(println)(1,0.15)
(2,0.27749999999999997)
(3,0.38587499999999997)
(4,0.313996875)
(5,0.58089421875)

數一數每個藝人所處的三角關系：

graph.triangleCount.vertices.collectres48: Array[(org.apache.spark.graphx.VertexId, Int)] = Array((1,0), (2,0), (3,1), (4,1), (5,1))

轉載于:https://www.cnblogs.com/bluejoe/p/5115846.html

總結

以上是生活随笔為你收集整理的学习GraphX的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。