org.apache.spark.mllib.tree.GradientBoostedTrees Scala Examples

package se.uu.farmbio.cp.alg

import org.apache.spark.mllib.linalg.Vector
import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.mllib.tree.loss.LogLoss
import org.apache.spark.rdd.RDD

import se.uu.farmbio.cp.UnderlyingAlgorithm

//Define a GBTs UnderlyingAlgorithm
private object GBT {
  def trainingProcedure(
      input: RDD[LabeledPoint], 
      numIterations: Int): (Vector => Double) = {
    //Configuration
    val boostingStrategy = BoostingStrategy.defaultParams("Regression")
    boostingStrategy.numIterations = numIterations
    boostingStrategy.treeStrategy.maxDepth = 5
    boostingStrategy.treeStrategy.categoricalFeaturesInfo = Map[Int, Int]()
    boostingStrategy.loss = LogLoss
    //Training
    val remappedInput = input.map(x => new LabeledPoint((x.label * 2) - 1, x.features))
    val model = new GradientBoostedTrees(boostingStrategy)
      .run(input = remappedInput)
    model.predict
  }
}

class GBT(
  private val input: RDD[LabeledPoint],
  private val numIterations: Int)
  extends UnderlyingAlgorithm(
      GBT.trainingProcedure(input,numIterations)) {
  override def nonConformityMeasure(newSample: LabeledPoint) = {
    val score = predictor(newSample.features)
    if (newSample.label == 1.0) {
      -score
    } else {
      score
    }
  }
} 

package org.sparksamples.gradientboosted

import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.rdd.RDD
import org.sparksamples.Util

import scala.collection.Map
import scala.collection.mutable.ListBuffer


object GradientBoostedTreesUtil {

  def getTrainTestData(): (RDD[LabeledPoint], RDD[LabeledPoint]) = {
    val recordsArray = Util.getRecords()
    val records = recordsArray._1
    val first = records.first()
    val numData = recordsArray._2

    println(numData.toString())
    records.cache()
    print("Mapping of first categorical feature column: " +  get_mapping(records, 2))
    var list = new ListBuffer[Map[String, Long]]()
    for( i <- 2 to 9){
      val m = get_mapping(records, i)
      list += m
    }
    val mappings = list.toList
    var catLen = 0
    mappings.foreach( m => (catLen +=m.size))

    val numLen = records.first().slice(11, 15).size
    val totalLen = catLen + numLen

    val data = {
      records.map(r => LabeledPoint(Util.extractLabel(r), Util.extractFeatures(r, catLen, mappings)))
    }


    val splits = data.randomSplit(Array(0.8, 0.2), seed = 11L)
    val training = splits(0).cache()
    val test = splits(1)
    return (training, test)
  }

  def get_mapping(rdd :RDD[Array[String]], idx: Int) : Map[String, Long] = {
    return rdd.map( fields=> fields(idx)).distinct().zipWithIndex().collectAsMap()
  }



  def evaluate(train: RDD[LabeledPoint],test: RDD[LabeledPoint], iterations:Int, maxDepth:Int,
               maxBins: Int): Double ={

    var boostingStrategy = BoostingStrategy.defaultParams("Regression")
    boostingStrategy.setNumIterations(iterations)
    boostingStrategy.treeStrategy.setMaxDepth(maxDepth)
    boostingStrategy.treeStrategy.setMaxBins(maxBins)

    val model = GradientBoostedTrees.train(train, boostingStrategy)
//
//    @classmethod
//    @since("1.3.0")
//    def trainRegressor(cls, data, categoricalFeaturesInfo,
//                       loss="leastSquaresError", numIterations=100, learningRate=0.1, maxDepth=3,
//                       maxBins=32):

    val true_vs_predicted = test.map(p => (p.label, model.predict(p.features)))
    val rmsle = Math.sqrt(true_vs_predicted.map{ case(t, p) => Util.squaredLogError(t, p)}.mean())
    return rmsle
  }

} 

package org.sparksamples.gradientboosted

import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.rdd.RDD
import org.sparksamples.Util

import scala.collection.Map
import scala.collection.mutable.ListBuffer


object GradientBoostedTreesApp{

  def get_mapping(rdd :RDD[Array[String]], idx: Int) : Map[String, Long] = {
    return rdd.map( fields=> fields(idx)).distinct().zipWithIndex().collectAsMap()
  }

  def main(args: Array[String]) {
    //val conf = new SparkConf().setMaster("local").setAppName("GradientBoostedTreesRegressionApp")
    val sc = Util.sc

    // we take the raw data in CSV format and convert it into a set of records
    // of the form (user, product, price)
    val rawData = sc.textFile("../data/hour_noheader.csv")
    val numData = rawData.count()
    val records = rawData.map(line => line.split(","))
    records.cache()
    var list = new ListBuffer[Map[String, Long]]()
    for( i <- 2 to 9){
      val m = get_mapping(records, i)
      list += m
    }
    val mappings = list.toList
    var catLen = 0
    mappings.foreach( m => (catLen +=m.size))

    val numLen = records.first().slice(11, 15).size
    val totalLen = catLen + numLen

    print("Feature vector length for categorical features:"+ catLen)
    print("Feature vector length for numerical features:" + numLen)
    print("Total feature vector length: " + totalLen)

    val data = {
      records.map(r => LabeledPoint(Util.extractLabel(r), Util.extractFeatures(r, catLen, mappings)))
    }
    val first_point = data.first()
    println("Gradient Boosted Trees Model feature vector:" + first_point.features.toString)
    println("Gradient Boosted Trees Model feature vector length: " + first_point.features.size)


    var boostingStrategy = BoostingStrategy.defaultParams("Regression")
    boostingStrategy.setNumIterations(3)// Note: Use more iterations in practice.
    boostingStrategy.treeStrategy.setMaxDepth(5)


    val model = GradientBoostedTrees.train(data, boostingStrategy)
    val true_vs_predicted = data.map(p => (p.label, model.predict(p.features)))
    val true_vs_predicted_take5 = true_vs_predicted.take(5)
    for(i <- 0 until 5) {
      println("True vs Predicted: " + "i :" + true_vs_predicted_take5(i))
    }
    val save = true
    if(save){
      val true_vs_predicted_csv = data.map(p => p.label + " ,"  + model.predict(p.features))
      val format = new java.text.SimpleDateFormat("dd-MM-yyyy-hh-mm-ss")
      val date = format.format(new java.util.Date())
      true_vs_predicted_csv.saveAsTextFile("./output/gradient_boosted_trees_" + date + ".csv")
    }
    val mse = true_vs_predicted.map{ case(t, p) => Util.squaredError(t, p)}.mean()
    val mae = true_vs_predicted.map{ case(t, p) => Util.absError(t, p)}.mean()
    val rmsle = Math.sqrt(true_vs_predicted.map{ case(t, p) => Util.squaredLogError(t, p)}.mean())

    println("Gradient Boosted Trees - Mean Squared Error: "  + mse)
    println("Gradient Boosted Trees - Mean Absolute Error: " + mae)
    println("Gradient Boosted Trees - Root Mean Squared Log Error:" + rmsle)
  }
} 

package org.apache.spark.examples.mllib

import org.apache.spark.SparkConf
import org.apache.spark.SparkContext
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.{BoostingStrategy, Algo}
import org.apache.spark.mllib.util.MLUtils

    // 加载数据
    val data = MLUtils.loadLibSVMFile(sc, "../data/mllib/rf_libsvm_data.txt")
    // 将数据随机分配为两份,一份用于训练,一份用于测试
    val splits = data.randomSplit(Array(0.7, 0.3))
    //数据分成训练和测试数据集
    val (trainingData, testData) = (splits(0), splits(1))
    //创建一个分类的提升策略并设置迭代次数为3(随机森林也支持回归)
    val boostingStrategy =BoostingStrategy.defaultParams("Classification")
        boostingStrategy.numIterations = 3
    //梯度提升决策树:综合多个决策树,消除噪声,避免过拟合
    val model = GradientBoostedTrees.train(trainingData,boostingStrategy)
    //基于测试实例评估模型并计算测试错误
    val testErr = testData.map { point =>
            //预测
            val prediction = model.predict(point.features)
            if (point.label == prediction) 
                1.0 
            else 0.0}.mean()//平均数
    //检查模型
    println("Test Error = " + testErr)
    println("Learned Random Forest:n" + model.toDebugString)
  }
} 

// scalastyle:off println
package com.intel.hibench.sparkbench.ml

import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.mllib.tree.model.GradientBoostedTreesModel
import org.apache.spark.mllib.util.MLUtils
import org.apache.spark.rdd.RDD
import org.apache.spark.mllib.regression.LabeledPoint

object GradientBoostingTree {
  def main(args: Array[String]): Unit = {
    var inputPath = ""
    var numIterations: Int = 3
    val numClasses: Int = 2
    val maxDepth: Int = 5

    if (args.length == 2) {
      inputPath = args(0)
      numIterations = args(1).toInt
    }

    val conf = new SparkConf()
        .setAppName("GradientBoostingTree")
    val sc = new SparkContext(conf)

    // Load and parse the data file.
    //val data = MLUtils.loadLibSVMFile(sc, "data/mllib/sample_libsvm_data.txt")
    val data: RDD[LabeledPoint] = sc.objectFile(inputPath)

    // Split the data into training and test sets (30% held out for testing)
    val splits = data.randomSplit(Array(0.7, 0.3))
    val (trainingData, testData) = (splits(0), splits(1))

    // Train a GradientBoostedTrees model.
    // The defaultParams for Classification use LogLoss by default.
    val boostingStrategy = BoostingStrategy.defaultParams("Classification")
    boostingStrategy.numIterations = numIterations
    boostingStrategy.treeStrategy.numClasses = numClasses
    boostingStrategy.treeStrategy.maxDepth = maxDepth
    // Empty categoricalFeaturesInfo indicates all features are continuous.
    boostingStrategy.treeStrategy.categoricalFeaturesInfo = Map[Int, Int]()

    val model = GradientBoostedTrees.train(trainingData, boostingStrategy)

    // Evaluate model on test instances and compute test error
    val labelAndPreds = testData.map { point =>
      val prediction = model.predict(point.features)
      (point.label, prediction)
    }
    val testErr = labelAndPreds.filter(r => r._1 != r._2).count.toDouble / testData.count()
    println("Test Error = " + testErr)

    sc.stop()
  }
} 

package com.intel.hibench.sparkbench.ml

import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.mllib.tree.model.GradientBoostedTreesModel
import org.apache.spark.rdd.RDD
import org.apache.spark.mllib.regression.LabeledPoint

import scopt.OptionParser

object GradientBoostedTree {

  case class Params(
    numClasses: Int = 2,
    maxDepth: Int = 30,
    maxBins: Int = 32,
    numIterations: Int = 20,
    learningRate: Double = 0.1,
    dataPath: String = null
  )

  def main(args: Array[String]): Unit = {
    val defaultParams = Params()

    val parser = new OptionParser[Params]("GBT"){
      head("GBT: an example of Gradient Boosted Tree for classification")
      opt[Int]("numClasses")
        .text(s"numClasses, default: ${defaultParams.numClasses}")
        .action((x,c) => c.copy(numClasses = x))
      opt[Int]("maxDepth")
        .text(s"maxDepth, default: ${defaultParams.maxDepth}")
        .action((x,c) => c.copy(maxDepth = x))
      opt[Int]("maxBins")
        .text(s"maxBins, default: ${defaultParams.maxBins}")
        .action((x,c) => c.copy(maxBins = x))
      opt[Int]("numIterations")
        .text(s"numIterations, default: ${defaultParams.numIterations}")
        .action((x,c) => c.copy(numIterations = x))
      opt[Double]("learningRate")
        .text(s"learningRate, default: ${defaultParams.learningRate}")
        .action((x,c) => c.copy(learningRate = x))
      arg[String]("<dataPath>")
        .required()
        .text("data path for Gradient Boosted Tree")
        .action((x,c) => c.copy(dataPath = x))
    }
    parser.parse(args, defaultParams) match {
      case Some(params) => run(params)
      case _ => sys.exit(1)
    }
  }

  def run(params: Params): Unit = {
    val conf = new SparkConf().setAppName(s"Gradient Boosted Tree with $params")
    val sc = new SparkContext(conf)

    val dataPath = params.dataPath
    val numClasses = params.numClasses
    val maxDepth = params.maxDepth
    val maxBins = params.maxBins
    val numIterations = params.numIterations
    val learningRate = params.learningRate

    // Load  data file.
    val data: RDD[LabeledPoint] = sc.objectFile(dataPath)

    // Split the data into training and test sets (30% held out for testing)
    val splits = data.randomSplit(Array(0.7, 0.3))
    val (trainingData, testData) = (splits(0), splits(1))

    // Train a GradientBoostedTrees model.
    val boostingStrategy = BoostingStrategy.defaultParams("Classification")
    boostingStrategy.numIterations = numIterations
    boostingStrategy.learningRate = learningRate
    boostingStrategy.treeStrategy.numClasses = numClasses
    boostingStrategy.treeStrategy.maxDepth = maxDepth
    boostingStrategy.treeStrategy.maxBins = maxBins
    // Empty categoricalFeaturesInfo indicates all features are continuous.
    boostingStrategy.treeStrategy.categoricalFeaturesInfo = Map[Int, Int]()

    val model = GradientBoostedTrees.train(trainingData, boostingStrategy)

    // Evaluate model on test instances and compute test error
    val labelAndPreds = testData.map { point =>
      val prediction = model.predict(point.features)
      (point.label, prediction)
    }
    val testErr = labelAndPreds.filter(r => r._1 != r._2).count.toDouble / testData.count()
    println("Test Error = " + testErr)

    sc.stop()
  }
} 

package spark.ml.cookbook.chapter10

import org.apache.log4j.{Level, Logger}
import org.apache.spark.mllib.evaluation.MulticlassMetrics
import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.model.GradientBoostedTreesModel
import org.apache.spark.rdd.RDD
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.sql.SparkSession


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

    Logger.getLogger("org").setLevel(Level.ERROR)

    val spark = SparkSession
      .builder
      .master("local[*]")
      .appName("MyGradientBoostedTreesClassification")
      .config("spark.sql.warehouse.dir", ".")
      .getOrCreate()

    val rawData = spark.sparkContext.textFile("../data/sparkml2/chapter10/breast-cancer-wisconsin.data")
    val data = rawData.map(_.trim)
      .filter(text => !(text.isEmpty || text.startsWith("#") || text.indexOf("?") > -1))
      .map { line =>
        val values = line.split(',').map(_.toDouble)
        val slicedValues = values.slice(1, values.size)
        val featureVector = Vectors.dense(slicedValues.init)
        val label = values.last / 2 -1
        LabeledPoint(label, featureVector)

      }


    val splits = data.randomSplit(Array(0.7, 0.3))
    val (trainingData, testData) = (splits(0), splits(1))

    println("Training Data count:"+trainingData.count())
    println("Test Data Count:"+testData.count())


    val algo = "Classification"
    val numIterations = 3
    val numClasses = 2
    val maxDepth   = 5
    val maxBins  = 32
    val categoricalFeatureInfo = Map[Int,Int]()


    val boostingStrategy = BoostingStrategy.defaultParams(algo)

    boostingStrategy.setNumIterations(numIterations)
    boostingStrategy.treeStrategy.setNumClasses(numClasses)
    boostingStrategy.treeStrategy.setMaxDepth(maxDepth)
    boostingStrategy.treeStrategy.setMaxBins(maxBins)
    boostingStrategy.treeStrategy.categoricalFeaturesInfo = categoricalFeatureInfo

    evaluate(trainingData, testData, boostingStrategy)
    println("===================")

    spark.stop()
  }

  def evaluate(
                trainingData: RDD[LabeledPoint],
                testData: RDD[LabeledPoint],
                boostingStrategy : BoostingStrategy
                ) :Unit = {



    val model = GradientBoostedTrees.train(trainingData, boostingStrategy)

    val metrics = getMetrics(model, testData)
    println("Confusion Matrix :")
    println(metrics.confusionMatrix)
    println("Model Accuracy: "+metrics.precision)
    println("Model Error: "+ (1-metrics.precision))
//    (0 until boostingStrategy.treeStrategy.getNumClasses()).map(
//      category => (metrics.precision(category), metrics.recall(category))
//    ).foreach(println)
//    println("My Classification GBT model:\n" + model.toDebugString)
  }

  def getMetrics(model: GradientBoostedTreesModel, data: RDD[LabeledPoint]): MulticlassMetrics = {
    val predictionsAndLabels = data.map(example =>
      (model.predict(example.features), example.label)
    )
    new MulticlassMetrics(predictionsAndLabels)
  }
} 

package spark.ml.cookbook.chapter10

import org.apache.log4j.{Level, Logger}
import org.apache.spark.mllib.evaluation.RegressionMetrics
import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.model.GradientBoostedTreesModel
import org.apache.spark.rdd.RDD
import org.apache.spark.mllib.tree.GradientBoostedTrees
import org.apache.spark.mllib.tree.configuration.BoostingStrategy
import org.apache.spark.sql.SparkSession



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


    Logger.getLogger("org").setLevel(Level.ERROR)

    val spark = SparkSession
      .builder
      .master("local[*]")
      .appName("MyGradientBoostedTreesRegression")
      .config("spark.sql.warehouse.dir", ".")
      .getOrCreate()

    val rawData = spark.sparkContext.textFile("../data/sparkml2/chapter10/breast-cancer-wisconsin.data")
    val data = rawData.map(_.trim)
      .filter(text => !(text.isEmpty || text.startsWith("#") || text.indexOf("?") > -1))
      .map { line =>
        val values = line.split(',').map(_.toDouble)
        val slicedValues = values.slice(1, values.size)
        val featureVector = Vectors.dense(slicedValues.init)
        val label = values.last / 2 -1
        LabeledPoint(label, featureVector)

      }

    val splits = data.randomSplit(Array(0.7, 0.3))
    val (trainingData, testData) = (splits(0), splits(1))


    println("Training Data count:"+trainingData.count())
    println("Test Data Count:"+testData.count())

    val algo = "Regression"
    val numIterations = 3
    val maxDepth   = 5
    val maxBins  = 32
    val categoricalFeatureInfo = Map[Int,Int]()


    val boostingStrategy = BoostingStrategy.defaultParams(algo)

    boostingStrategy.setNumIterations(numIterations)

    boostingStrategy.treeStrategy.setMaxDepth(maxDepth)
    boostingStrategy.treeStrategy.setMaxBins(maxBins)

    boostingStrategy.treeStrategy.categoricalFeaturesInfo = categoricalFeatureInfo

    val model = GradientBoostedTrees.train(trainingData, boostingStrategy)


    val metrics = getMetrics(model, testData)

    println("Test Mean Squared Error = " + metrics.meanSquaredError)
    println("My regression GBT model:\n" + model.toDebugString)
    spark.stop()
  }
  def getMetrics(model: GradientBoostedTreesModel, data: RDD[LabeledPoint]): RegressionMetrics = {
    val predictionsAndLabels = data.map(example =>
      (model.predict(example.features), example.label)
    )
    new RegressionMetrics(predictionsAndLabels)
  }
}