贝叶斯分类器 scala

为课本上的经典例题，一时觉得用scala可能会比较方便？于是便这样做了。

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

import scala.util.Random


object bayes {

  def randomString(len: Int): String = {
    val rand = new scala.util.Random(System.nanoTime)
    val sb = new StringBuilder(len)
    val ab = "MF"
    for (i <- 0 until len) {
      sb.append(ab(rand.nextInt(ab.length)))
    }
    sb.toString
  }
  def part_prability(data:RDD[String],cuts:String,cutPositon:Int): Double ={
    val data1=data.filter({
      x=>{val words=x.split(" "); words(cutPositon).equals(cuts)}})
      .map({x=>{val words=x.split(" "); (words(cutPositon),words(4))}})
    val a_all=data1.count()
    val a_buy=data1.filter(x=>x._2.equals("是"))
    val a_buynum=a_buy.count()
    val res=a_buynum.toDouble/a_all.toDouble
    res
  }
  def part_prability2(data:RDD[String],cuts:String,cutPositon:Int): Double ={
//    青 高 否 中 否
    val data1=data.filter({
      x=>{val words=x.split(" "); words(4).equals("是")}})
      .map({x=>{val words=x.split(" "); (words(cutPositon),words(4))}})
    val a_all=data1.count()
    val a_buy=data1.filter(x=>x._1.equals(cuts))
    val a_buynum=a_buy.count()
    val res=a_buynum.toDouble/a_all.toDouble
    res
  }
  def apart(data:RDD[String],cuts:String,cutPositon:Int): Double ={
    //    青 高 否 中 否
    val data1=data
      .map({x=>{val words=x.split(" "); (words(cutPositon),1)}})
    val a_all=data1.count()
    val a_buy=data1.filter(x=>x._1.equals(cuts))
    val a_buynum=a_buy.count()
    val res=a_buynum.toDouble/a_all.toDouble
    res
  }
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf()
      .setAppName("WordCount")
      .setMaster("local")
    val sc = new SparkContext(conf)
    val datat=sc.textFile("C:\\Users\\86158\\Desktop\\作业\\spark\\work3\\bayes.txt")
    //    人群总购买率 （买的概率）
    val purchase:Double=0.6429
    println("plez input 0 or 1: 0 for end and 1 for begin")
    var signal=Console.readInt()
    while (signal==1){
      println("请输入年龄变量：")
      var age=Console.readLine()
      println("请输入收入变量：")
      var salary=Console.readLine()
      println("请输入爱好变量：")
      var hobby=Console.readLine()
      println("请输入信用变量：")
      var credit=Console.readLine()
      var ageing=part_prability2(datat,age,0)
      var salaring=part_prability2(datat,salary,1)
      var hobbing=part_prability2(datat,hobby,2)

      var creding=part_prability2(datat,credit,3)
      var apartment=apart(datat,age,0)*apart(datat,salary,1)*apart(datat,hobby,2)*apart(datat,credit,3)

      var result=(ageing)*salaring*hobbing*creding*purchase/apartment
      println(result)
      if (result>0.5){
        println(">0.5 很可能会购买")
      }
      else{
        println("<0.5 购买的概率不大")
      }
      println("plez input your signal:")
      signal=Console.readInt()
    }

  }

}

结果

ps后续：老师布置了用python / java 来完成这个作业，我将代码分享给了同学，所以全班几乎有一半交的这份代码。。。我真是醉了