Spark中组件Mllib的学习18之corr:两组数据相关关系计算（Pearson、Spearman）

更多代码请见：​​https://github.com/xubo245/SparkLearning​​​
Spark中组件Mllib的学习之基础概念篇
1解释

（1）皮尔森Pearson

皮尔森相似度的原始计算公式为：

书上也有例子：

（2） 斯皮尔曼等级相关 Spearman

di=xi-yi

注意：这里的Xi、Yi是原始数据的等级，也就是排序序号，不是元素数据值例子：

2.代码：

/**
  * @author xubo
  *         ref:Spark MlLib机器学习实战
  *         javascript:void(0)
  *         more code:https://github.com/xubo245/SparkLearning
  *       
  *
  */
package org.apache.spark.mllib.learning.basic

import org.apache.spark.mllib.stat.Statistics
import org.apache.spark.{SparkConf, SparkContext}

/**
  * Created by xubo on 2016/5/23.
  */
object StatisticsCorrLearning {
  def main(args: Array[String]) {
    val conf = new SparkConf().setMaster("local[4]").setAppName(this.getClass().getSimpleName().filter(!_.equals('$')))
    val sc = new SparkContext(conf)

    val rddX = sc.textFile("file/data/mllib/input/basic/StatisticsCorrLearningx.txt") //读取数据
      .flatMap(_.split(' ') //进行分割
      .map(_.toDouble)) //转化为Double类型
    val rddY = sc.textFile("file/data/mllib/input/basic/StatisticsCorrLearningy.txt") //读取数据
        .flatMap(_.split(' ') //进行分割
        .map(_.toDouble)) //转化为Double类型
    println("rddX:")
    rddX.foreach(each => print(each + " "))
    println("\nrddY:")
    rddY.foreach(each => print(each + " "))
    var correlationPearson: Double = Statistics.corr(rddX, rddY) //计算不同数据之间的相关系数:皮尔逊
    println("\ncorrelationPearson：" + correlationPearson) //打印结果

    var correlationSpearman: Double = Statistics.corr(rddX, rddY, "spearman") //使用斯皮尔曼计算不同数据之间的相关系数
    println("correlationSpearman：" + correlationSpearman) //打印结果

    //不读取文件直接生成
    println("\nSecond:")
    val arr1 = (1 to 10).toArray.map(_.toDouble)
    val arr2 = (1 to 10).toArray.map(_.toDouble)
    val rdd1 = sc.parallelize(arr1)
    val rdd2 = sc.parallelize(arr2)
    println("rdd1:")
    rdd1.foreach(each => print(each + " "))
    println("\nrdd2:")
    rdd2.foreach(each => print(each + " "))
    correlationPearson = Statistics.corr(rdd1, rdd2) //计算不同数据之间的相关系数:皮尔逊
    println("\ncorrelationPearson：" + correlationPearson) //打印结果
    correlationSpearman = Statistics.corr(rdd1, rdd2, "spearman") //使用斯皮尔曼计算不同数据之间的相关系数
    println("correlationSpearman：" + correlationSpearman) //打印结果

    println("\nThird:")
    val arr3 = (1 to 5).toArray.map(_.toDouble)
    val arr4 = (2 to 10 by 2).toArray.map(_.toDouble)
    val rdd3 = sc.parallelize(arr3)
    val rdd4 = sc.parallelize(arr4)
    println("rdd3:")
    rdd3.foreach(each => print(each + " "))
    println("\nrdd4:")
    rdd4.foreach(each => print(each + " "))
    val correlationPearson3 = Statistics.corr(rdd3, rdd4) //计算不同数据之间的相关系数:皮尔逊
    println("\ncorrelationPearson3：" + correlationPearson3) //打印结果

    val correlationSpearman3 = Statistics.corr(rdd3, rdd4, "spearman") //使用斯皮尔曼计算不同数据之间的相关系数
    println("correlationSpearman3：" + correlationSpearman3) //打印结果

    println("\nFourth:")
    val rdd5 = sc.parallelize(Array(5.0, 3.0, 2.5))
    val rdd6 = sc.parallelize(Array(4.0, 3.0, 2.0))
    println("rdd5:")
    rdd5.foreach(each => print(each + " "))
    println("\nrdd6:")
    rdd6.foreach(each => print(each + " "))
    val correlationPearson5 = Statistics.corr(rdd5, rdd6) //计算不同数据之间的相关系数:皮尔逊
    println("\ncorrelationPearson5：" + correlationPearson5) //打印结果
    //与链接和书计算结果一致：javascript:void(0) 书：ref

    val correlationSpearman5 = Statistics.corr(rdd5, rdd6, "spearman") //使用斯皮尔曼计算不同数据之间的相关系数
    println("correlationSpearman5：" + correlationSpearman5) //打印结果

    println("\nFifth:")
    val rdd7 = sc.parallelize(Array(170.0, 150.0, 210.0,180.0,160.0))
    val rdd8 = sc.parallelize(Array(180.0, 165.0, 190.0,168.0,172.0))
    println("rdd:")
    rdd7.foreach(each => print(each + " "))
    println("\nrdd:")
    rdd8.foreach(each => print(each + " "))
    val correlationPearson7 = Statistics.corr(rdd7, rdd8) //计算不同数据之间的相关系数:皮尔逊
    println("\ncorrelationPearson：" + correlationPearson7) //打印结果
    //与链接和书计算结果一致：http://baike.baidu.com/link?url=s9aiihE4mMpF2sZLqR33JKz3FNk8R8IWWh9-ZgNFO4aZB5ez9mnADNQZQSApniWXUJGwhr-Ar9mjWEFVwncQlq书：ref

    val correlationSpearman7 = Statistics.corr(rdd7, rdd8, "spearman") //使用斯皮尔曼计算不同数据之间的相关系数
    println("correlationSpearman：" + correlationSpearman7) //打印结果


    sc.stop
  }
}

文件在github

3.结果：

rddX:
2016-05-23 21:25:10 WARN  :139 - Your hostname, xubo-PC resolves to a loopback/non-reachable address: fe80:0:0:0:482:722f:5976:ce1f%20, but we couldn't find any external IP address!
1.0 6.0 7.0 8.0 9.0 10.0 2.0 3.0 4.0 5.0 
rddY:
6.0 7.0 8.0 9.0 10.0 1.0 2.0 3.0 4.0 5.0 2016-05-23 21:25:12 WARN  BLAS:61 - Failed to load implementation from: com.github.fommil.netlib.NativeSystemBLAS
2016-05-23 21:25:12 WARN  BLAS:61 - Failed to load implementation from: com.github.fommil.netlib.NativeRefBLAS

correlationPearson：1.0
correlationSpearman：1.0000000000000009

Second:
rdd1:
3.0 4.0 5.0 1.0 2.0 8.0 9.0 10.0 6.0 7.0 
rdd2:
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 
correlationPearson：1.0
correlationSpearman：1.0000000000000009

Third:
rdd3:
3.0 4.0 5.0 2.0 1.0 
rdd4:
4.0 2.0 6.0 8.0 10.0 
correlationPearson3：0.9999999999999998
correlationSpearman3：0.9999999999999998

Fourth:
rdd5:
3.0 5.0 2.5 
rdd6:
4.0 3.0 2.0 
correlationPearson5：0.944911182523068
correlationSpearman5：1.0

Fifth:
rdd:
150.0 180.0 160.0 210.0 170.0 
rdd:
165.0 190.0 180.0 168.0 172.0 
correlationPearson：0.8171759569273293
correlationSpearman：0.6999999999999998

参考
【1】​​​http://spark.apache.org/docs/1.5.2/mllib-guide.html​​​
【2】​​​http://spark.apache.org/docs/1.5.2/programming-guide.html​​​
【3】​​​https://github.com/xubo245/SparkLearning​​​
【5】​​​http://baike.baidu.com/link?url=s9aiihE4mMpF2sZLqR33JKz3FNk8R8IWWh9-ZgNFO4aZB5ez9mnADNQZQSApniWXUJGwhr-Ar9mjWEFVwncQlq​​