Spark RDD算子：RDD分区，HashPartitioner、RangePartitioner、自定义分区

HashPartitioner

设定分区的数量。
Scala版

import org.apache.spark.rdd.RDD
import org.apache.spark.{HashPartitioner, SparkConf, SparkContext}

object HashPartitionScala {
  def main(args: Array[String]): Unit = {
    val conf=new SparkConf().setMaster("local[2]").setAppName("HashPartScala")
    val sc = new SparkContext(conf)
    //构建一个查看分区的方法
    def mapPartIndex(i:Int,it:Iterator[(Int,Int)]):Iterator[(Int,(Int,Int))]={
      var list = List[(Int,(Int,Int))]()
      while (it.hasNext) {
        val next = it.next()
        list=list.::(i,next)
      }
      list.iterator
    }
   def parRdd(rdd:RDD[(Int,Int)])={
     val value = rdd.mapPartitionsWithIndex(mapPartIndex)
     value.foreach(println)
   }
    val rdd = sc.parallelize(List((1,1),(1,2),(2,3),(2,4),(3,1)))
    //分区数设为3
    val partition = rdd.partitionBy(new HashPartitioner(3))
    parRdd(rdd)
  }
}

Java版

package Action;

import org.apache.spark.HashPartitioner;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.api.java.function.Function2;
import org.apache.spark.api.java.function.VoidFunction;
import scala.Tuple2;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.Iterator;
import java.util.List;

public class HashPartition {
    //定义一个打印分区的方法
    private static void printPartRdd(JavaPairRDD<Integer, Integer> pairRDD) {
        JavaRDD<Tuple2<Integer, Tuple2<Integer, Integer>>> tuple2JavaRDD = pairRDD.mapPartitionsWithIndex(new Function2<Integer, Iterator<Tuple2<Integer, Integer>>, Iterator<Tuple2<Integer, Tuple2<Integer, Integer>>>>() {
            @Override
            public Iterator<Tuple2<Integer, Tuple2<Integer, Integer>>> call(Integer i, Iterator<Tuple2<Integer, Integer>> tuple2) throws Exception {
                List<Tuple2<Integer, Tuple2<Integer, Integer>>> list = new ArrayList<>();
                while (tuple2.hasNext()) {
                    Tuple2<Integer, Integer> next = tuple2.next();
                    list.add(new Tuple2<Integer, Tuple2<Integer, Integer>>(i, next));
                }
                return list.iterator();
            }
        }, false);
        tuple2JavaRDD.foreach(new VoidFunction<Tuple2<Integer, Tuple2<Integer, Integer>>>() {
            @Override
            public void call(Tuple2<Integer, Tuple2<Integer, Integer>> itTp2) throws Exception {
                System.out.println(itTp2);
            }
        });
    }
    public static void main(String[] args) {
        SparkConf conf = new SparkConf().setMaster("local[2]").setAppName("HashPartition");
        JavaSparkContext sc = new JavaSparkContext(conf);
        JavaRDD<Tuple2<Integer, Integer>> rdd = sc.parallelize(Arrays.asList(
                new Tuple2<Integer, Integer>(1, 2),
                new Tuple2<Integer, Integer>(1, 3),
                new Tuple2<Integer, Integer>(2, 3),
                new Tuple2<Integer, Integer>(2, 4),
                new Tuple2<Integer, Integer>(3, 1),
                new Tuple2<Integer, Integer>(4, 2)
        ));
        JavaPairRDD<Integer, Integer> rdd2 = JavaPairRDD.fromJavaRDD(rdd);
        JavaPairRDD<Integer, Integer> HashPart = rdd2.partitionBy(new HashPartitioner(3));
        printPartRdd(HashPart);
    }
}

RangePartitioner

范围分区。将范围内的键分配给相应的分区。
scala版

val conf=new SparkConf().setMaster("local[2]").setAppName("HashPartScala")
    val sc = new SparkContext(conf)

    def mapPartIndex(i:Int,it:Iterator[(Int,Int)]):Iterator[(Int,(Int,Int))]={
      var list = List[(Int,(Int,Int))]()
      while (it.hasNext) {
        val next = it.next()
        list=list.::(i,next)
      }
      list.iterator
    }
    def parRdd(rdd:RDD[(Int,Int)])={
      val value = rdd.mapPartitionsWithIndex(mapPartIndex)
      value.foreach(println)
    }
    val rdd = sc.parallelize(List((1,1),(1,2),(4,3),(2,4),(3,1)))
    val partition = rdd.partitionBy(new RangePartitioner(6,rdd))
    parRdd(partition)
  }

自定义分区

实现自定义分区需要继承Partitioner类，并实现相应的方法。
例如：
Scala端
注意包不要导错，否则无法完成重写方法。

import org.apache.spark.rdd.RDD
import org.apache.spark.{Partitioner, SparkConf, SparkContext}

object UDP {
  //自定义规则，继承Partitioner
  class DefinedPartition(n:Int) extends Partitioner {
    override def numPartitions: Int = n

    override def getPartition(key: Any): Int = {
      if (key.toString.toInt<=2) {
        0
      }else if (key.toString.toInt>2&&key.toString.toInt<4) {
        1
      }else{
        2
      }
    }
  }

  def main(args: Array[String]): Unit = {
    val conf=new SparkConf().setMaster("local[2]").setAppName("UDPScala")
    val sc = new SparkContext(conf)

    def mapPartIndex(i:Int,it:Iterator[(Int,Int)]):Iterator[(Int,(Int,Int))]={
      var list = List[(Int,(Int,Int))]()
      while (it.hasNext) {
        val next = it.next()
        list=list.::(i,next)
      }
      list.iterator
    }
    def parRdd(rdd:RDD[(Int,Int)])={
      val value = rdd.mapPartitionsWithIndex(mapPartIndex)
      value.foreach(println)
    }
    val rdd = sc.parallelize(List((1,1),(1,2),(4,3),(2,4),(3,1)))
    val result = rdd.partitionBy(new DefinedPartition(3))
    parRdd(result)
  }

}

java端

import org.apache.spark.Partitioner;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;
import scala.Tuple2;
import java.util.Arrays;
public class UDPJava extends Partitioner {
    int i;

    public UDPJava(int i) {
        this.i = i;
    }

    public UDPJava() {
    }

    @Override
    public int numPartitions() {
        return i;
    }

    @Override
    public int getPartition(Object key) {
        int keyCode = Integer.parseInt(key.toString());
        if (keyCode <= 2) {
            return 0;
        } else if (keyCode < 4) {
            return 1;
        } else {
            return 2;
        }
    }
    public static void main(String[] args) {
        SparkConf conf = new SparkConf().setMaster("local[2]").setAppName("HashPartition");
        JavaSparkContext sc = new JavaSparkContext(conf);
        JavaRDD<Tuple2<Integer, Integer>> rdd = sc.parallelize(Arrays.asList(
                new Tuple2<Integer, Integer>(1, 2),
                new Tuple2<Integer, Integer>(1, 3),
                new Tuple2<Integer, Integer>(2, 3),
                new Tuple2<Integer, Integer>(2, 4),
                new Tuple2<Integer, Integer>(3, 1),
                new Tuple2<Integer, Integer>(4, 2)
        ));
        JavaPairRDD<Integer, Integer> rdd2 = JavaPairRDD.fromJavaRDD(rdd);
        JavaPairRDD<Integer, Integer> DefinePart = rdd2.partitionBy(new UDPJava(3));
        //继承前面写的打印分区的方法
        HashPartition print = new HashPartition();
        print.printPartRdd(DefinePart);

    }
}