RDD编程实践2-编写独立应用程序实现数据去重，实现求平均值问题

（1）编写独立应用程序实现数据去重

package my.scala
import org.apache.spark.{SparkConf, SparkContext}
object case2 {
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf().setMaster("local").setAppName("reduce")
    val sc = new SparkContext(conf)
    sc.setLogLevel("ERROR")
    //获取数据
    val two = sc.textFile("file:///usr/local/spark/text_4/sec")
    two.filter(_.trim().length>0) //trim()函数返回空格个数
        .map(line=>(line.trim,""))//全部值当key，(key value,"")
          .groupByKey()//groupByKey,过滤重复的key value ，发送到总机器上汇总
              .sortByKey() //按key value的自然顺序排序
                  .keys.collect().foreach(println)//collect是将结果转换为数组的形式
  }
}

（2）编写独立应用程序实现求平均值问题

package my.scala
import org.apache.spark.{SparkConf, SparkContext}
object pingjunzhi {
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf().setMaster("local").setAppName("reduce")
    val sc = new SparkContext(conf)
    sc.setLogLevel("ERROR")
   
val fourth = sc.textFile("file:///usr/local/spark/text_4/thi")
  
val res = fourth.filter(_.trim().length>0).map(line=>(line.split("\t")(0).trim(),line.split("\t")(1).trim().toInt)).groupByKey().map(x => {
   var num = 0.0
   var sum = 0
   for(i <- x._2){
    sum = sum + i
    num = num +1
   }
   val avg = sum/num
   val format = f"$avg%1.2f".toDouble
   (x._1,format)
 }).collect.foreach(x => println(x._1+"\t"+x._2))
  }
}