Spark中组件Mllib的学习17之colStats:以列为基础计算统计量的基本数据

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
colStats:以列为基础计算统计量的基本数据

2.代码：

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

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

/**
  * Created by xubo on 2016/5/23.
  */
object StatisticsColStatsLearning {
  def main(args: Array[String]) {
    val conf = new SparkConf().setMaster("local[4]").setAppName(this.getClass().getSimpleName().filter(!_.equals('$')))
    val sc = new SparkContext(conf)
    //    val rdd = sc.textFile("file/data/mllib/input/basic/MatrixRow.txt") //读取文件
    val rdd = sc.textFile("file/data/mllib/input/basic/stats.txt") //读取文件
      .map(_.split(' ') //按“ ”分割
      .map(_.toDouble)) //转成Double类型
      .map(line => Vectors.dense(line))
    val summary = Statistics.colStats(rdd) //获取Statistics实例

    //    rdd.foreach(each => print(each + " "))
    rdd.foreach(println)
    println("rdd.count:" + rdd.count())
    println()
    println(summary)
    println(summary.max) //最大
    println(summary.min) //最小
    println("count" + summary.count) //个数
    println(summary.numNonzeros) //非零
    println("variance:" + summary.variance) //方差
    println(summary.mean) //计算均值
    println(summary.variance) //计算标准差
    println(summary.normL1) //计算曼哈段距离:相加
    println(summary.normL2) //计算欧几里得距离：平方根


    //    /行向量
    println("\n row Vector:")
    val vec = Vectors.dense(1, 2, 3, 4, 5)
    println(vec)
    println(vec.size)
    println(vec.numActives)
    //    println(vec.variance)//不存在

    sc.stop
  }
}

3.结果：

[1.0]
[2.0]
[3.0]
[4.0]
[5.0]
rdd.count:5

org.apache.spark.mllib.stat.MultivariateOnlineSummarizer@7f9de19a
[5.0]
[1.0]
count5
[5.0]
variance:[2.5]
[3.0]
[2.5]
[15.0]
[7.416198487095663]

 row Vector:
[1.0,2.0,3.0,4.0,5.0]
5
5

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​​