MLlib数据统计基本概念

备注：kimi.txt中的内容如下：
1
2
3
4
5

一.求数据的均值和标准差
import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.stat.Statistics
import org.apache.spark.{SparkConf, SparkContext}
object testVector { def main(args: Array[String]): Unit = {
val conf = new SparkConf().setMaster("local")
.setAppName("testVector");
val sc = new SparkContext(conf);
var rdd = sc.textFile("kimi.txt")
.map(_.split(' ')
.map(_.toDouble))
.map(line => Vectors.dense(line));
var summary = Statistics.colStats(rdd);
println(summary.mean);//计算均值
println(summary.variance);//计算标准差
}
}
程序结果：[3.0][2.5]
二.距离计算
1.欧几里得距离（normL1）:指在m维空间中两个点之间的真实距离，或者向量的自然长度（即该点到原点的距离）。
2.曼哈段距离（normL2）：两个点在标准坐标系上的绝对轴距总和。

import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.stat.Statistics
import org.apache.spark.{SparkConf, SparkContext}
object testVector {
def main(args: Array[String]): Unit = {
val conf = new SparkConf().setMaster("local")
.setAppName("testVector");
val sc = new SparkContext(conf);
var rdd = sc.textFile("kimi.txt")
.map(_.split(' ')
.map(_.toDouble))
.map(line => Vectors.dense(line));
var summary = Statistics.colStats(rdd);
println(summary.normL1);
println(summary.normL2);
}
}

程序结果：
[15.0]
[7.416198487095663]
三.相关系数
x.txt，y.txt内容：
1 2 3 4 5
2 4 6 8 10

import org.apache.spark.mllib.stat.Statistics
import org.apache.spark.{SparkConf, SparkContext}
object testVector {
def main(args: Array[String]): Unit = {
val conf = new SparkConf().setMaster("local")
.setAppName("testVector");
val sc = new SparkContext(conf);
var rddX = sc.textFile("x.txt")
.flatMap(_.split(' ')
.map(_.toDouble));
var rddY = sc.textFile("y.txt")
.flatMap(_.split(' ')
.map(_.toDouble));
var correlation: Double = Statistics.corr(rddX,rddY);//皮尔逊相关系数 1.0
println(correlation);
val correlation2: Double = Statistics.corr(rddX,rddY,"spearman");//斯皮尔曼相关系数 1.0000000000000009
println(correlation2);
}
}
单个数据集相关系数的计算

import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.stat.Statistics
import org.apache.spark.{SparkConf, SparkContext}
object testVector {
def main(args: Array[String]): Unit = {
val conf = new SparkConf().setMaster("local")
.setAppName("testVector");
val sc = new SparkContext(conf);
var rdd = sc.textFile("x.txt")
.map(_.split(' ')
.map(_.toDouble))
.map(line => Vectors.dense(line))
println(Statistics.corr(rdd,"spearman"));
}
}

1.0                 0.9999999999999998  0.9999999999999998  ... (5 total)
0.9999999999999998  1.0                 0.9999999999999998  ...
0.9999999999999998  0.9999999999999998  1.0                 ...
0.9999999999999998  0.9999999999999998  0.9999999999999998  ...
0.9999999999999998  0.9999999999999998  0.9999999999999998  ...