SPARK转换算子简单的实现示例（map,filter,flatmap,groupByKey）

package day05
import org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}

class MyTransformation_scala {

  //使用map对集合中的每一个元素乘 2
  def myMap(sc: SparkContext): Unit ={
    val array: Array[Int] = Array(1,2,3,4,5,6,7,8,9)
    val pRdd: RDD[Int] = sc.parallelize(array)
    val mapRDD: RDD[Int] = pRdd.map(_ * 2)
    mapRDD.collect()
    println(mapRDD.collect().toList.toString)
  }

  //过滤整型集合中出现的奇数 保留偶数
  def myFilter(sc: SparkContext): Unit ={
    val array: Array[Int] = Array(1,2,3,4,5,6,7,8,9)
    val pRdd: RDD[Int] = sc.parallelize(array)
    val filterRDD: RDD[Int] = pRdd.filter(_ % 2 == 0)
    println(filterRDD.collect().toList.toString())
  }
  //将多维的集合压扁成一维的
  def myFlatMap(sc : SparkContext){
    val array: Array[String] =
      Array("Hello World" , "Hello scala" , "Hello Java")
    val parallelizeRDD: RDD[String] = sc.parallelize(array)
    val flatMapRDD: RDD[String] = parallelizeRDD.flatMap(_.split(" "))
    println(flatMapRDD.collect().toList.toString())
  }

  //按照key分组
  def myGroupByKey(sc : SparkContext): Unit ={
    val array: Array[(String, Int)] = Array(Tuple2("class1",80),
      Tuple2("class2",60),Tuple2("class1",75),Tuple2("class2",92))
    val parallelizeRDD: RDD[(String, Int)] = sc.parallelize(array)
    val groupByKeyRDD: RDD[(String, Iterable[Int])] =
      parallelizeRDD.groupByKey()
    println(groupByKeyRDD.collect().toList.toString())
  }
}


object MyTransformation_scala{
  val tfs: MyTransformation_scala = new MyTransformation_scala
  def main(args: Array[String]) {
    val conf =  new SparkConf().
      setMaster("local").setAppName("MyTransformation_scala")
    val sc: SparkContext = new SparkContext(conf)
    //    tfs.myMap(sc)
    //    tfs.myFilter(sc)
    //    tfs.myFlatMap(sc)
    tfs.myGroupByKey(sc)
    sc.stop()
  }
}