Spark学习笔记-KNN算法实现

KNN算法原理可以参考：数据挖掘笔记-分类-KNN-1   

基于Spark简单实现算法代码如下：

object SparkKNN extends Serializable {

  def main(args: Array[String]) {
    if (args.length != 4) {
    	println("error, please input three path.");
    	println("1 train set path.");
		println("2 test set path.");
		println("3 output path.");
		println("4 k value.");
		System.exit(1)
    }
    
    val sc = new SparkContext("spark://centos.host1:7077", "Spark KNN")
    
    val trainSet = sc.textFile(args(0)).map(line => {
      var datas = line.split(" ")
      (datas(0), datas(1), datas(2))
    })
    var bcTrainSet = sc.broadcast(trainSet.collect())
    var bcK = sc.broadcast(args(3).toInt)
    
    val testSet = sc.textFi