Spark ML（6）：PCA

一、环境配置

1.spark2.1.0-cdh5.7.0（自编译）

2.cdh5.7.0

3.scala2.11.8

4.centos6.4

二、环境准备

参考https://blog.csdn.net/u010886217/article/details/90312617

三、实现代码

1.测试集iris.data描述

5.1,3.5,1.4,0.2,Iris-setosa
4.9,3.0,1.4,0.2,Iris-setosa
4.7,3.2,1.3,0.2,Iris-setosa
4.6,3.1,1.5,0.2,Iris-setosa
5.0,3.6,1.4,0.2,Iris-setosa
5.4,3.9,1.7,0.4,Iris-setosa
4.6,3.4,1.4,0.3,Iris-setosa
5.0,3.4,1.5,0.2,Iris-setosa
4.4,2.9,1.4,0.2,Iris-setosa
4.9,3.1,1.5,0.1,Iris-setosa
5.4,3.7,1.5,0.2,Iris-setosa
4.8,3.4,1.6,0.2,Iris-setosa
4.8,3.0,1.4,0.1,Iris-setosa
4.3,3.0,1.1,0.1,Iris-setosa
5.8,4.0,1.2,0.2,Iris-setosa
...

2.PCA代码

package sparktest

import org.apache.spark.SparkConf
import org.apache.spark.ml.classification.{DecisionTreeClassifier, NaiveBayes}
import org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator
import org.apache.spark.ml.feature.{PCA, VectorAssembler}
import org.apache.spark.sql.SparkSession

import scala.util.Random

object pca {
  def main(args: Array[String]): Unit = {

    val conf = new SparkConf().setMaster("local").setAppName("iris")
    val spark = SparkSession.builder().config(conf).getOrCreate()
    spark.sparkContext.setLogLevel("WARN") ///日志级别

    val file = spark.read.format("csv").load("iris.data")
    //file.show()

    import spark.implicits._
    val random = new Random()
    val data = file.map(row =>{
      val label =  row.getString(4) match {
        case "Iris-setosa" => 0
        case "Iris-versicolor" => 1
        case "Iris-virginica" => 2
      }

      (row.getString(0).toDouble,
      row.getString(1).toDouble,
      row.getString(2).toDouble,
      row.getString(3).toDouble,
      label,
      random.nextDouble())
    }).toDF("_c0","_c1","_c2","_c3","label","rand").sort("rand")//.where("label = 1 or label = 0")

    val assembler = new VectorAssembler().setInputCols(Array("_c0","_c1","_c2","_c3")).setOutputCol("features")

    val pca = new PCA()
      .setInputCol("features")
      .setOutputCol("features2")
      .setK(3)
    val dataset = assembler.transform(data)
    val pcaModel = pca.fit(dataset)
    val dataset2 = pcaModel.transform(dataset)
    val Array(train,test) = dataset2.randomSplit(Array(0.8,0.2))

    val dt = new DecisionTreeClassifier().setFeaturesCol("features2").setLabelCol("label")
    val model = dt.fit(train)
    val result = model.transform(test)
    result.show(false)
    val evaluator = new MulticlassClassificationEvaluator()
      .setLabelCol("label")
      .setPredictionCol("prediction")
      .setMetricName("accuracy")
    val accuracy = evaluator.evaluate(result)
    println(s"""accuracy is $accuracy""")
  }
}