Spark Transformations之mapPartitions

mapPartitions(func) Similar to map, but runs separately on each partition (block) of the RDD, so func must be of type Iterator<T> => Iterator<U> when running on an RDD of type T.

mapPartitions

mapPartitions是map的一个变种。map的输入函数是应用于RDD中每个元素，而mapPartitions的输入函数是应用于每个分区，也就是把每个分区中的内容作为整体来处理的。 
它的函数定义为：

def mapPartitions[U: ClassTag](f: Iterator[T] => Iterator[U], preservesPartitioning: Boolean = false): RDD[U]

f即为输入函数，它处理每个分区里面的内容。每个分区中的内容将以Iterator[T]传递给输入函数f，f的输出结果是Iterator[U]。最终的RDD由所有分区经过输入函数处理后的结果合并起来的。

举例：

scala> val a = sc.parallelize(1 to 9, 3)
scala> def myfunc[T](iter: Iterator[T]) : Iterator[(T, T)] = {
    var res = List[(T, T)]() 
    var pre = iter.next while (iter.hasNext) {
        val cur = iter.next; 
        res .::= (pre, cur) pre = cur;
    } 
    res.iterator
}
scala> a.mapPartitions(myfunc).collect
res0: Array[(Int, Int)] = Array((2,3), (1,2), (5,6), (4,5), (8,9), (7,8))

上述例子中的函数myfunc是把分区中一个元素和它的下一个元素组成一个Tuple。因为分区中最后一个元素没有下一个元素了，所以(3,4)和(6,7)不在结果中。 
mapPartitions还有些变种，比如mapPartitionsWithContext，它能把处理过程中的一些状态信息传递给用户指定的输入函数。还有mapPartitionsWithIndex，它能把分区的index传递给用户指定的输入函数。

在scala eclipse中运行

package yanan.spark.core.transformations.example

import org.apache.spark.SparkConf
import org.apache.spark.SparkContext

object TransformationsTest {
  def mapPartitionsFunc[T](iter: Iterator[T]): Iterator[(T, T)] = {
    var res = List[(T, T)]()
    var pre = iter.next
    while (iter.hasNext) {
      val cur = iter.next
      res.::=(pre, cur)
      pre = cur;
    }
    res.iterator
  }
  def mapPartitionsTest(sc: SparkContext) = {
    val a = sc.parallelize(1 to 9, 3)
    a.mapPartitions(mapPartitionsFunc).collect.foreach(println)
  }

  def main(args: Array[String]) {
    val conf = new SparkConf().setAppName(s"Book example: Scala").setMaster("local[2]")
    val sc = new SparkContext(conf)
    mapPartitionsTest(sc)
    sc.stop()
  }
}

pom.xml文件：

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
	xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
	<modelVersion>4.0.0</modelVersion>

	<groupId>com.yanan.spark_maven</groupId>
	<artifactId>spark1.3.1</artifactId>
	<version>0.0.1-SNAPSHOT</version>
	<packaging>jar</packaging>

	<name>spark_maven</name>
	<url>http://maven.apache.org</url>

	<properties>
		<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
		<jackson.version>1.9.13</jackson.version>
		<!-- <spark.version>1.3.1</spark.version> -->
		<spark.version>1.4.0</spark.version>
	</properties>

	<dependencies>
		<dependency>
			<groupId>junit</groupId>
			<artifactId>junit</artifactId>
			<version>3.8.1</version>
			<scope>test</scope>
		</dependency>
		<dependency>
			<groupId>org.scala-lang</groupId>
			<artifactId>scala-library</artifactId>
			<version>2.10.4</version>
		</dependency>
		<dependency>
			<groupId>org.apache.spark</groupId>
			<artifactId>spark-core_2.10</artifactId>
			<version>${spark.version}</version>
		</dependency>
		<dependency>
			<groupId>org.apache.spark</groupId>
			<artifactId>spark-mllib_2.10</artifactId>
			<version>${spark.version}</version>
		</dependency>
		<!--<dependency> <groupId>org.apache.spark</groupId> <artifactId>spark-sql_2.10</artifactId> 
			<version>1.3.1</version> </dependency> <dependency> <groupId>org.apache.spark</groupId> 
			<artifactId>spark-hive_2.10</artifactId> <version>1.3.1</version> </dependency> 
			<dependency> <groupId>org.apache.spark</groupId> <artifactId>spark-bagel_2.10</artifactId> 
			<version>1.3.1</version> </dependency>
		 <dependency>
			<groupId>org.apache.spark</groupId>
			<artifactId>spark-graphx_2.10</artifactId>
			<version>1.3.1</version>
		</dependency> -->
		
		<!-- specify the version for json_truple <dependency> <groupId>org.codehaus.jackson</groupId> 
			<artifactId>jackson-core-asl</artifactId> <version>${jackson.version}</version> 
			</dependency> <dependency> <groupId>org.codehaus.jackson</groupId> <artifactId>jackson-mapper-asl</artifactId> 
			<version>${jackson.version}</version> </dependency> -->

	</dependencies>


	<build>
		<plugins>
			<plugin>
				<groupId>org.scala-tools</groupId>
				<artifactId>maven-scala-plugin</artifactId>
				<executions>
					<execution>
						<goals>
							<goal>compile</goal>
							<goal>testCompile</goal>
						</goals>
					</execution>
				</executions>
			</plugin>
		</plugins>
	</build>
	<pluginRepositories>
		<pluginRepository>
			<id>scala-tools.org</id>
			<name>Scala-tools Maven2 Repository</name>
			<url>http://scala-tools.org/repo-releases</url>
		</pluginRepository>
	</pluginRepositories>

	<repositories>
		<repository>
			<id>cloudera-repo-releases</id>
			<url>https://repository.cloudera.com/artifactory/repo/</url>
		</repository>
	</repositories>
</project>