目录
一、解释
1. subgraph:
方法的定义:
def subgraph(
epred: EdgeTriplet[VD, ED] => Boolean = (x => true),
vpred: (VertexId, VD) => Boolean = ((v, d) => true)
): Graph[VD, ED]
返回的对象是一个图,图中包含着的顶点和边分别要满足vpred和epred两个函数。(要注意,顶点和边是完全不同的概念,如果一个边被砍掉了,这个边关联的两个顶点并不会受影响)
要注意,在图里,如果一个顶点没了,其对应的边也就没了,但边没了之后,点不会受影响。
所以,subgraph一般用于:restrict the graph to the vertices and edges of interest或者eliminate broken links.
2.reverse:
return a new graph with all edge directions reversed.
调用方法:
val reverseGraph = graph.reverse
3. groupEdges:
groupEdges用于将多重图中的相同的顶点之间的边做合并(除了属性其实没其他可以合并的)。
def groupEdges(merge: (ED, ED) => ED): Graph[VD, ED]
详解来自:https://www.jianshu.com/p/d9170a0723e4
4.mask:
mask用于创建一个子图,子图中包含在输入的图中也包含的顶点和边。该方法通常跟subgraph方法一起使用,来根据另一个关联的图来过滤当前的图中展示的数据。
二、代码综合实现:
package homeWork
import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.graphx.{Edge, Graph, VertexId}
import org.apache.spark.rdd.RDD
object MapGraphX2 {
def main(args: Array[String]): Unit = {
//设置运行环境
val conf = new SparkConf().setAppName("SimpleGraphX").setMaster("local")
val sc = new SparkContext(conf)
sc.setLogLevel("WARN")
//设置users顶点
val users: RDD[(VertexId, (String, Int))] =
sc.parallelize(Array(
(1L, ("Alice", 28)),
(2L, ("Bob", 27)),
(3L, ("Charlie", 65)),
(4L, ("David", 42)),
(5L, ("Ed", 55)),
(6L, ("Fran", 50))
))
//设置relationships边
val relationships: RDD[Edge[Int]] =
sc.parallelize(Array(
Edge(2L, 4L, 7),
Edge(2L, 4L, 2),
Edge(3L, 2L, 4),
Edge(3L, 6L, 3),
Edge(4L, 1L, 1),
Edge(5L, 2L, 2),
Edge(5L, 3L, 8),
Edge(5L, 6L, 3)
))
// 定义默认的作者,以防与不存在的作者有relationship边
val defaultUser = ("John Doe", 0)
println("(1)通过上面的项点数据和边数据创建图对象")
// Build the initial Graph
val graph: Graph[(String, Int), Int] = Graph(users, relationships,
defaultUser)
graph.edges.collect.foreach(println(_))
graph.vertices.collect.foreach(println(_))
//设置users顶点
val users2: RDD[(VertexId, (String, Int))] =
sc.parallelize(Array(
(1L, ("Alice2", 28)),
(2L, ("Bob2", 27)),
(3L, ("Charlie2", 65)),
(4L, ("David2", 42))
))
//设置relationships边
val relationships2: RDD[Edge[Int]] =
sc.parallelize(Array(
Edge(2L, 3L, 7),
Edge(2L, 4L, 2),
Edge(3L, 2L, 4),
Edge(4L, 1L, 1)
))
// 定义默认的作者,以防与不存在的作者有relationship边
val defaultUser2 = ("Missing", 0)
println("(1.1)通过上面的项点数据和边数据创建图对象")
// Build the initial Graph
val graph2: Graph[(String, Int), Int] = Graph(users2, relationships2,
defaultUser2)
graph2.edges.collect.foreach(println(_))
graph2.vertices.collect.foreach(println(_))
println("(2)对上述的顶点数据和边数据进行修改,再创建一张图2,使得图2中有一些点和边在图1中不存在,然后调用图1的mask方法,传入图2作为参数,观察返回的图3的信息")
println("mask效果:返回公共子图")
val graph3 = graph.mask(graph2)
graph3.edges.collect.foreach(println(_))
graph3.vertices.collect.foreach(println(_))
println("(3)基于上面的数据进行修改练习subgraph 、groupEdges 、reverses方法")
println("-----------subgraph")
var graph4: Graph[(String, Int), Int] = graph.subgraph(epred = (ed) => !
(ed.srcId == 1L || ed.dstId == 2L)
, vpred = (id, attr) => id != 3L)
graph4.edges.collect.foreach(println(_))
graph4.vertices.collect.foreach(println(_))
println("-----------groupEdges")
var graph5: Graph[(String, Int), Int]= graph.groupEdges(merge = (ed1,
ed2) =>
(ed1 + ed2 + 100000000))
graph5.edges.collect.foreach(println(_))
// graph5.vertices.collect.foreach(println(_))
println("----------reverses")
graph2.edges.foreach(println(_))
println("-------")
graph2.reverse.edges.foreach(println(_))
}
}