知识图谱学习笔记－图操作

一、自定义图

import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.graphx.{Edge, VertexId}
import org.apache.spark.rdd.RDD
import org.apache.spark.graphx.Graph

object PropertiesGraph {

  def main(args: Array[String]): Unit = {

    val conf = new SparkConf().setAppName("graph").setMaster("local[*]")
    val sc = new SparkContext(conf)
    sc.setLogLevel("OFF")

    //定义点
    val users:RDD[(VertexId,(String,String))] = sc.parallelize(
      Array(
        (3L,("s1","student")),
        (6L,("s2","teacher")),
        (5L,("s3","pro")),
        (2L,("s4","pro"))
      )
    )

    //定义边
    val relationships:RDD[Edge[String]] = sc.parallelize(
      Array(
        Edge(3L,6L,"cooperate"),
        Edge(5L,6L,"agvisor"),
        Edge(2L,5L,"colleague")
      )
    )

    val defaultUser = ("Jack ma","defaultUser")
    val graph = Graph(users,relationships,defaultUser)


  }


}

二、图基本操作

import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.graphx.{Edge, VertexId}
import org.apache.spark.rdd.RDD
import org.apache.spark.graphx.Graph

object PropertiesGraph {

  def main(args: Array[String]): Unit = {

    val conf = new SparkConf().setAppName("graph").setMaster("local[*]")
    val sc = new SparkContext(conf)
    sc.setLogLevel("OFF")

    //定义点
    val users:RDD[(VertexId,(String,String))] = sc.parallelize(
      Array(
        (3L,("s1","student")),
        (6L,("s2","teacher")),
        (5L,("s3","pro")),
        (2L,("s4","pro"))
      )
    )

    //定义边
    val relationships:RDD[Edge[String]] = sc.parallelize(
      Array(
        Edge(3L,6L,"cooperate"),
        Edge(5L,6L,"advisor"),
        Edge(2L,5L,"colleague")
      )
    )

    val defaultUser = ("Jack ma","defaultUser")
    val graph = Graph(users,relationships,defaultUser)


    /**********************  图操作  ***************************/
    
    //找到途中属性是student的点
    graph.vertices.filter{case (id,(name,work))=>work=="student"}
      .collect.foreach{case (id,(name,work))=>println(s"$name is $work")}

    //找到途中属性为advisor的边
    graph.edges.filter(x => x.attr=="advisor")
      .collect().foreach(x=>println(s"${x.srcId} to ${x.dstId} 属性为 ${x.attr}"))

    //出度和入度操作
    def max(a:(VertexId,Int),b:(VertexId,Int)):(VertexId,Int) = {
      if(a._2 > b._2) a else b
    }

    println("最大的出度" + graph.outDegrees.reduce(max))
    println("最大的入度" + graph.inDegrees.reduce(max))
    println("最大的度数" + graph.degrees.reduce(max))


    // 给每一个点的职业属性上加一个字符串
    graph.mapVertices{case (id,(name,work))=>(id,(name,work+"_spark"))}
      .vertices.collect.foreach(println)

    // triplet:a关系b  a -> b
    // 给图中每个元组的Edge的属性值设置为源的属性值+边的属性值+目标点的属性值
    graph.mapTriplets(x=>x.srcAttr._2 + "+" + x.attr + "+" + x.dstAttr._2)
      .edges.collect().foreach(println)

    // 两者之间的关系
    graph.triplets.map(x=>x.srcAttr._1 + " is the " + x.attr + " of " + x.dstAttr._1).foreach(println)

    // 删除节点，构建子图
    val validGraphx = graph.subgraph(vpred = (id,attr)=> attr._2 != "pro")
    validGraphx.vertices.foreach(println)
    validGraphx.triplets.map(x=>x.srcAttr._1 + " is the " + x.attr + " of " + x.dstAttr._1).foreach(println)

    // 构建职业是professor的子图
    val subGraph = graph.subgraph(vpred = (id, attr)=>attr._2=="pro")
    subGraph.vertices.collect().foreach(x=>println(s"${x._2._1} is ${x._2._2}"))



  }


}

三、Spark加载neo4j

import org.apache.spark.SparkConf
import org.apache.spark.api.java.JavaSparkContext
import org.apache.spark.ml.tree.InternalNode
import org.neo4j.spark.Neo4j

object SparkGraphxConnector {

  def main(args: Array[String]): Unit = {

    val conf = new SparkConf()
      .setAppName("neo4j")
      .setMaster("local[*]")
      .set("spark.neo4j.bolt.url","bolt://ip")
      .set("spark.neo4j.bolt.user","user_name")
      .set("spark.neo4j.bolt.password","password")

    val sc = new JavaSparkContext(conf)
    sc.setLogLevel("OFF")
    val neo4j = Neo4j(sc)
    val rdd = neo4j.cypher("match (n:Person) return n").loadRowRdd

    val personRDD = rdd.map(row=>{
      val map = row.get(0).asInstanceOf[InternalNode]
      new Person(map.get("home").asString(),
        map.get("name").asString(),
        map.get("personID").asString())
    })

    personRDD.foreach(println)


    // 图查询

    val graphQuery = "match (n:Person)-[r]-(a:Person) return id(p) as source,id(a) as target,type(r) as value"

    val graph:Graph[String,String] = neo4j.rels(graphQuery).loadGraph
    graph.edges.foreach(println(_))


  }

  case class Person(
                   val home:String,
                   val name:String,
                   val personID:String
                   )


}