Spark學習筆記：joinVertices & outJoinVertices

joinVertices
若graph頂點的Id在點集pointTest中，則將graph頂點的屬性使替換爲點集pointTest的點屬性；
若graph頂點Id不在點集pointTest中，則不對graph頂點屬性做任何修改。

1.創建graph

scala> val graph = GraphLoader.edgeListFile(sc, "file:///home/hadoop/1.txt")
scala> graph.triplets.collect.foreach(println(_))

---

2.自定義pointTest點集

scala> val pointTest: RDD[(VertexId,Int)] = sc.parallelize(Array((1L,10),(2L,10),(3L,10)))

---

3.將graph頂點屬性置爲零；joinVertices：

scala> val outputGraph = graph.mapVertices((id,attr) => 0).joinVertices(pointTest)((_,_,optDeg) => optDeg)
scala> outputGraph.triplets.collect.foreach(println(_))

---

4.outJoinVertices

scala> val outputGraph = graph.mapVertices((id,attr) => 0).outerJoinVertices(pointTest)((_,_,optDeg) => optDeg.getOrElse(0))

—————————–

將頂點屬性改爲各點出度：

scala> val outputGraph = graph.mapVertices((id,attr) => 0).outerJoinVertices(graph.outDegrees)((_,_,optDeg) => optDeg.getOrElse(0))