Merge pull request apache#3 from detrevid/develop

Cleanup

src/main/scala/DataSource.scala

@@ -2,13 +2,10 @@ package org.template.classification
 
 import io.prediction.controller.PDataSource
 import io.prediction.controller.EmptyEvaluationInfo
-import io.prediction.controller.EmptyActualResult
 import io.prediction.controller.Params
-import io.prediction.data.storage.Event
 import io.prediction.data.store.PEventStore
 
 import org.apache.spark.SparkContext
-import org.apache.spark.SparkContext._
 import org.apache.spark.rdd.RDD
 import org.apache.spark.mllib.regression.LabeledPoint
 import org.apache.spark.mllib.linalg.Vectors
@@ -60,7 +57,7 @@ class DataSource(val dsp: DataSourceParams)
  override
  def readEval(sc: SparkContext)
  : Seq[(TrainingData, EmptyEvaluationInfo, RDD[(Query, ActualResult)])] = {
- require(!dsp.evalK.isEmpty, "DataSourceParams.evalK must not be None")
+ require(dsp.evalK.nonEmpty, "DataSourceParams.evalK must not be None")
 
  // The following code reads the data from data store. It is equivalent to
  // the readTraining method. We copy-and-paste the exact code here for
@@ -95,7 +92,7 @@ class DataSource(val dsp: DataSourceParams)
 
  // K-fold splitting
  val evalK = dsp.evalK.get
- val indexedPoints: RDD[(LabeledPoint, Long)] = labeledPoints.zipWithIndex
+ val indexedPoints: RDD[(LabeledPoint, Long)] = labeledPoints.zipWithIndex()
 
  (0 until evalK).map { idx =>
  val trainingPoints = indexedPoints.filter(_._2 % evalK != idx).map(_._1)

src/main/scala/Engine.scala

@@ -1,6 +1,6 @@
 package org.template.classification
 
-import io.prediction.controller.IEngineFactory
+import io.prediction.controller.EngineFactory
 import io.prediction.controller.Engine
 
 class Query(
@@ -15,7 +15,7 @@ class ActualResult(
  val label: Double
 ) extends Serializable
 
-object ClassificationEngine extends IEngineFactory {
+object ClassificationEngine extends EngineFactory {
  def apply() = {
  new Engine(
  classOf[DataSource],

src/main/scala/Evaluation.scala

@@ -6,7 +6,7 @@ import io.prediction.controller.EngineParams
 import io.prediction.controller.EngineParamsGenerator
 import io.prediction.controller.Evaluation
 
-case class Accuracy
+case class Accuracy()
  extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
  def calculate(query: Query, predicted: PredictedResult, actual: ActualResult)
  : Double = (if (predicted.label == actual.label) 1.0 else 0.0)

src/main/scala/NaiveBayesAlgorithm.scala

@@ -22,10 +22,10 @@ class NaiveBayesAlgorithm(val ap: AlgorithmParams)
 
  def train(sc: SparkContext, data: PreparedData): NaiveBayesModel = {
  // MLLib NaiveBayes cannot handle empty training data.
- require(!data.labeledPoints.take(1).isEmpty,
- s"RDD[labeldPoints] in PreparedData cannot be empty." +
+ require(data.labeledPoints.take(1).nonEmpty,
+ s"RDD[labeledPoints] in PreparedData cannot be empty." +
  " Please check if DataSource generates TrainingData" +
- " and Preprator generates PreparedData correctly.")
+ " and Preparator generates PreparedData correctly.")
 
  NaiveBayes.train(data.labeledPoints, ap.lambda)
  }

src/main/scala/PrecisionEvaluation.scala

@@ -17,7 +17,7 @@ case class Precision(label: Double)
  Some(0.0) // False positive
  }
  } else {
- None // Unrelated case for calcuating precision
+ None // Unrelated case for calculating precision
  }
  }
 }

src/main/scala/Preparator.scala

@@ -3,7 +3,6 @@ package org.template.classification
 import io.prediction.controller.PPreparator
 
 import org.apache.spark.SparkContext
-import org.apache.spark.SparkContext._
 import org.apache.spark.rdd.RDD
 import org.apache.spark.mllib.regression.LabeledPoint