Program function: Collect the location of the neighbor pointing to the vertex
/** Locate the region of the neighbor pointed to by each vertex */import Org. apache. spark. sparkcontextimport Org. apache. spark. sparkcontext. _ import Org. apache. spark. sparkconfimport Scala. collection. mutable. mapimport Org. apache. spark. _ import Org. apache. spark. graphx. _ import Org. apache. spark. RDD. rddobject testapp {def main (ARGs: array [String]) {Val conf = new sparkconf (). setappname ("testapp") Val SC = new sparkcontext (CONF) Val graph = graphloader. edgelistfile (SC, "/home/spark/graphx/data/followers.txt") // when an edge is loaded, the vertex Val users = SC appears on the edge. textfile ("/home/spark/graphx/data/users.txt "). map {line => Val fields = line. split (",") (fields (0 ). tolong, (fields (1), fields (2) // parse vertex data: ID (must be converted to long type), name, Region} Val mygraph = graph. apply (users, graph. edges) // re-graph. The vertex data is subject to users Val vertices = mygraph. mapreducetriplets [map [String, int] (// collect the region of the neighbor pointing to each point, triplet => iterator (triplet. srcid, map [String, int] (triplet. dstattr. _ 2-> 1), // map function one-way sends a message to the Source Vertex (A, B) with the directed edge) ==>{ // reduce function collects the message var mymap = map [String, int] () for (K, v) <-a) {If (B. contains (k) {var T = a (k) + B (k) mymap + = (k-> T )} else mymap + = (k-> A (k)} mymap // return the pooled result}) vertices. collect. foreach (A => Print (a + "\ n") // print the location of the collected neighbors }}
Users.txt vertex data: ID, name, Region
1,BarackObama,American2,ladygaga,American3,John,American4,xiaoming,Beijing6,Hanmeimei,Beijing7,Polly,American8,Tom,American
Followers.txt Edge Data: only source and target vertices are separated by spaces in the middle. Redundant columns are useless. For example, 2 1 Other has three columns of data, but graphx only reads the first two columns.
2 14 11 26 37 37 66 73 7
Result:
(4,Map(American -> 1))(6,Map(American -> 2))(2,Map(American -> 1))(1,Map(American -> 1))(3,Map(American -> 1))(7,Map(American -> 1))
Project directory structure:
./test.sbt./src./src/main./src/main/scala./src/main/scala/testApp.scala
Test. SBT content:
name := "test Project"version := "1.0"scalaVersion := "2.10.4"libraryDependencies += "org.apache.spark" %% "spark-core" % "1.0.1"libraryDependencies += "org.apache.spark" %% "spark-graphx" %"1.0.1"resolvers += "Akka Repository" at "http://repo.akka.io/releases/"
Execute the following command line for compilation:
sbt package
Execute the following command line to submit to the cluster (on the premise that the cluster is enabled). Here, my project is created in the spark installation directory under apps/testapp, so the following starts ../../
../../bin/spark-submit --class "testApp" --master local[4] target/scala-2.10/test-project_2.10-1.0.jar