PIO-208: Update examples

examples/scala-parallel-classification/README.md

@@ -17,4 +17,4 @@ limitations under the License.
 
 This is based on Classification Engine Template v0.14.0.
 
-Please refer to http://predictionio.apache.org/templates/classification/how-to/
+Please refer to https://predictionio.apache.org/templates/classification/how-to/

examples/scala-parallel-classification/add-algorithm/src/main/scala/NaiveBayesAlgorithm.scala

@@ -37,6 +37,7 @@ class NaiveBayesAlgorithm(val ap: AlgorithmParams)
 
   @transient lazy val logger = Logger[this.type]
 
+  override
   def train(sc: SparkContext, data: PreparedData): NaiveBayesModel = {
     // MLLib NaiveBayes cannot handle empty training data.
     require(data.labeledPoints.take(1).nonEmpty,
@@ -47,6 +48,7 @@ class NaiveBayesAlgorithm(val ap: AlgorithmParams)
     NaiveBayes.train(data.labeledPoints, ap.lambda)
   }
 
+  override
   def predict(model: NaiveBayesModel, query: Query): PredictedResult = {
     val label = model.predict(Vectors.dense(
       Array(query.attr0, query.attr1, query.attr2)

examples/scala-parallel-classification/add-algorithm/src/main/scala/PrecisionEvaluation.scala

@@ -25,6 +25,7 @@ case class Precision(label: Double)
   extends OptionAverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header: String = s"Precision(label = $label)"
 
+  override
   def calculate(query: Query, predicted: PredictedResult, actual: ActualResult)
   : Option[Double] = {
     if (predicted.label == label) {

examples/scala-parallel-classification/add-algorithm/src/main/scala/Preparator.scala

@@ -29,6 +29,7 @@ class PreparedData(
 
 class Preparator extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(trainingData.labeledPoints)
   }

examples/scala-parallel-classification/add-algorithm/src/main/scala/RandomForestAlgorithm.scala

@@ -42,6 +42,7 @@ class RandomForestAlgorithm(val ap: RandomForestAlgorithmParams) // CHANGED
   Query, PredictedResult] {
 
   // CHANGED
+  override
   def train(sc: SparkContext, data: PreparedData): RandomForestModel = {
     // CHANGED
     // Empty categoricalFeaturesInfo indicates all features are continuous.
@@ -57,6 +58,7 @@ class RandomForestAlgorithm(val ap: RandomForestAlgorithmParams) // CHANGED
       ap.maxBins)
   }
 
+  override
   def predict(
     model: RandomForestModel, // CHANGED
     query: Query): PredictedResult = {

examples/scala-parallel-classification/reading-custom-properties/src/main/scala/NaiveBayesAlgorithm.scala

@@ -37,6 +37,7 @@ class NaiveBayesAlgorithm(val ap: AlgorithmParams)
 
   @transient lazy val logger = Logger[this.type]
 
+  override
   def train(sc: SparkContext, data: PreparedData): NaiveBayesModel = {
     // MLLib NaiveBayes cannot handle empty training data.
     require(data.labeledPoints.take(1).nonEmpty,
@@ -47,6 +48,7 @@ class NaiveBayesAlgorithm(val ap: AlgorithmParams)
     NaiveBayes.train(data.labeledPoints, ap.lambda)
   }
 
+  override
   def predict(model: NaiveBayesModel, query: Query): PredictedResult = {
     val label = model.predict(Vectors.dense(
       // MODIFIED

examples/scala-parallel-classification/reading-custom-properties/src/main/scala/PrecisionEvaluation.scala

@@ -25,6 +25,7 @@ case class Precision(label: Double)
   extends OptionAverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header: String = s"Precision(label = $label)"
 
+  override
   def calculate(query: Query, predicted: PredictedResult, actual: ActualResult)
   : Option[Double] = {
     if (predicted.label == label) {

examples/scala-parallel-classification/reading-custom-properties/src/main/scala/Preparator.scala

@@ -29,6 +29,7 @@ class PreparedData(
 
 class Preparator extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(trainingData.labeledPoints)
   }

examples/scala-parallel-ecommercerecommendation/README.md

@@ -17,4 +17,4 @@ limitations under the License.
 
 This is based on E-Commerce Recommendation Template v0.14.0.
 
-Please refer to http://predictionio.apache.org/templates/ecommercerecommendation/how-to/
+Please refer to https://predictionio.apache.org/templates/ecommercerecommendation/how-to/

examples/scala-parallel-ecommercerecommendation/adjust-score/src/main/scala/ECommAlgorithm.scala

@@ -87,6 +87,7 @@ class ECommAlgorithm(val ap: ECommAlgorithmParams)
 
   @transient lazy val logger = Logger[this.type]
 
+  override
   def train(sc: SparkContext, data: PreparedData): ECommModel = {
     require(!data.viewEvents.take(1).isEmpty,
       s"viewEvents in PreparedData cannot be empty." +
@@ -239,6 +240,7 @@ class ECommAlgorithm(val ap: ECommAlgorithmParams)
     buyCountsRDD.collectAsMap.toMap
   }
 
+  override
   def predict(model: ECommModel, query: Query): PredictedResult = {
 
     val userFeatures = model.userFeatures

examples/scala-parallel-ecommercerecommendation/adjust-score/src/main/scala/Preparator.scala

@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
 class Preparator
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(
       users = trainingData.users,

examples/scala-parallel-recommendation/README.md

@@ -17,4 +17,4 @@ limitations under the License.
 
 This is based on Recommendation Template v0.14.0.
 
-Please refer to http://predictionio.apache.org/templates/recommendation/how-to/
+Please refer to https://predictionio.apache.org/templates/recommendation/how-to/

examples/scala-parallel-recommendation/blacklist-items/src/main/scala/ALSAlgorithm.scala

@@ -48,6 +48,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       s"To remedy it, set lower numIterations or checkpoint parameters.")
   }
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     // MLLib ALS cannot handle empty training data.
     require(!data.ratings.take(1).isEmpty,
@@ -92,6 +93,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       itemStringIntMap = itemStringIntMap)
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
     // Convert String ID to Int index for Mllib
     model.userStringIntMap.get(query.user).map { userInt =>

examples/scala-parallel-recommendation/blacklist-items/src/main/scala/ALSModel.scala

@@ -59,6 +59,7 @@ class ALSModel(
     scored.top(num)(Ordering.by(_._2))
   }
 
+  override
   def save(id: String, params: ALSAlgorithmParams,
     sc: SparkContext): Boolean = {
 

examples/scala-parallel-recommendation/blacklist-items/src/main/scala/Evaluation.scala

@@ -35,6 +35,7 @@ case class PrecisionAtK(k: Int, ratingThreshold: Double = 2.0)
 
   override def header = s"Precision@K (k=$k, threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Option[Double] = {
     val positives: Set[String] = a.ratings.filter(_.rating >= ratingThreshold).map(_.item).toSet
 
@@ -53,6 +54,7 @@ case class PositiveCount(ratingThreshold: Double = 2.0)
     extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header = s"PositiveCount (threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Double = {
     a.ratings.filter(_.rating >= ratingThreshold).size
   }

examples/scala-parallel-recommendation/blacklist-items/src/main/scala/Preparator.scala

@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
 class Preparator
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(ratings = trainingData.ratings)
   }

examples/scala-parallel-recommendation/customize-data-prep/src/main/scala/ALSAlgorithm.scala

@@ -48,6 +48,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       s"To remedy it, set lower numIterations or checkpoint parameters.")
   }
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     // MLLib ALS cannot handle empty training data.
     require(!data.ratings.take(1).isEmpty,
@@ -92,6 +93,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       itemStringIntMap = itemStringIntMap)
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
     // Convert String ID to Int index for Mllib
     model.userStringIntMap.get(query.user).map { userInt =>

examples/scala-parallel-recommendation/customize-data-prep/src/main/scala/ALSModel.scala

@@ -39,6 +39,7 @@ class ALSModel(
   extends MatrixFactorizationModel(rank, userFeatures, productFeatures)
   with PersistentModel[ALSAlgorithmParams] {
 
+  override
   def save(id: String, params: ALSAlgorithmParams,
     sc: SparkContext): Boolean = {
 

examples/scala-parallel-recommendation/customize-data-prep/src/main/scala/Evaluation.scala

@@ -35,6 +35,7 @@ case class PrecisionAtK(k: Int, ratingThreshold: Double = 2.0)
 
   override def header = s"Precision@K (k=$k, threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Option[Double] = {
     val positives: Set[String] = a.ratings.filter(_.rating >= ratingThreshold).map(_.item).toSet
 
@@ -53,6 +54,7 @@ case class PositiveCount(ratingThreshold: Double = 2.0)
     extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header = s"PositiveCount (threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Double = {
     a.ratings.filter(_.rating >= ratingThreshold).size
   }

examples/scala-parallel-recommendation/customize-data-prep/src/main/scala/Preparator.scala

@@ -34,6 +34,7 @@ case class CustomPreparatorParams(
 class Preparator(pp: CustomPreparatorParams) // ADDED CustomPreparatorParams
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     val noTrainItems = Source.fromFile(pp.filepath).getLines.toSet // CHANGED
     val ratings = trainingData.ratings.filter( r =>

examples/scala-parallel-recommendation/customize-serving/src/main/scala/ALSAlgorithm.scala

@@ -48,6 +48,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       s"To remedy it, set lower numIterations or checkpoint parameters.")
   }
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     // MLLib ALS cannot handle empty training data.
     require(!data.ratings.take(1).isEmpty,
@@ -92,6 +93,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       itemStringIntMap = itemStringIntMap)
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
     // Convert String ID to Int index for Mllib
     model.userStringIntMap.get(query.user).map { userInt =>

examples/scala-parallel-recommendation/customize-serving/src/main/scala/ALSModel.scala

@@ -39,6 +39,7 @@ class ALSModel(
   extends MatrixFactorizationModel(rank, userFeatures, productFeatures)
   with PersistentModel[ALSAlgorithmParams] {
 
+  override
   def save(id: String, params: ALSAlgorithmParams,
     sc: SparkContext): Boolean = {
 

examples/scala-parallel-recommendation/customize-serving/src/main/scala/Evaluation.scala

@@ -35,6 +35,7 @@ case class PrecisionAtK(k: Int, ratingThreshold: Double = 2.0)
 
   override def header = s"Precision@K (k=$k, threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Option[Double] = {
     val positives: Set[String] = a.ratings.filter(_.rating >= ratingThreshold).map(_.item).toSet
 
@@ -53,6 +54,7 @@ case class PositiveCount(ratingThreshold: Double = 2.0)
     extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header = s"PositiveCount (threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Double = {
     a.ratings.filter(_.rating >= ratingThreshold).size
   }

examples/scala-parallel-recommendation/customize-serving/src/main/scala/Preparator.scala

@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
 class Preparator
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(ratings = trainingData.ratings)
   }

examples/scala-parallel-recommendation/reading-custom-events/src/main/scala/ALSAlgorithm.scala

@@ -48,6 +48,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       s"To remedy it, set lower numIterations or checkpoint parameters.")
   }
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     // MLLib ALS cannot handle empty training data.
     require(!data.ratings.take(1).isEmpty,
@@ -92,6 +93,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       itemStringIntMap = itemStringIntMap)
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
     // Convert String ID to Int index for Mllib
     model.userStringIntMap.get(query.user).map { userInt =>

examples/scala-parallel-recommendation/reading-custom-events/src/main/scala/ALSModel.scala

@@ -39,6 +39,7 @@ class ALSModel(
   extends MatrixFactorizationModel(rank, userFeatures, productFeatures)
   with PersistentModel[ALSAlgorithmParams] {
 
+  override
   def save(id: String, params: ALSAlgorithmParams,
     sc: SparkContext): Boolean = {
 

examples/scala-parallel-recommendation/reading-custom-events/src/main/scala/Evaluation.scala

@@ -35,6 +35,7 @@ case class PrecisionAtK(k: Int, ratingThreshold: Double = 2.0)
 
   override def header = s"Precision@K (k=$k, threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Option[Double] = {
     val positives: Set[String] = a.ratings.filter(_.rating >= ratingThreshold).map(_.item).toSet
 
@@ -53,6 +54,7 @@ case class PositiveCount(ratingThreshold: Double = 2.0)
     extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header = s"PositiveCount (threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Double = {
     a.ratings.filter(_.rating >= ratingThreshold).size
   }

examples/scala-parallel-recommendation/reading-custom-events/src/main/scala/Preparator.scala

@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
 class Preparator
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(ratings = trainingData.ratings)
   }

examples/scala-parallel-recommendation/train-with-view-event/src/main/scala/ALSAlgorithm.scala

@@ -48,6 +48,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       s"To remedy it, set lower numIterations or checkpoint parameters.")
   }
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     // MLLib ALS cannot handle empty training data.
     require(!data.ratings.take(1).isEmpty,
@@ -93,6 +94,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
       itemStringIntMap = itemStringIntMap)
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
     // Convert String ID to Int index for Mllib
     model.userStringIntMap.get(query.user).map { userInt =>

examples/scala-parallel-recommendation/train-with-view-event/src/main/scala/ALSModel.scala

@@ -39,6 +39,7 @@ class ALSModel(
   extends MatrixFactorizationModel(rank, userFeatures, productFeatures)
   with PersistentModel[ALSAlgorithmParams] {
 
+  override
   def save(id: String, params: ALSAlgorithmParams,
     sc: SparkContext): Boolean = {
 

examples/scala-parallel-recommendation/train-with-view-event/src/main/scala/Evaluation.scala

@@ -35,6 +35,7 @@ case class PrecisionAtK(k: Int, ratingThreshold: Double = 2.0)
 
   override def header = s"Precision@K (k=$k, threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Option[Double] = {
     val positives: Set[String] = a.ratings.filter(_.rating >= ratingThreshold).map(_.item).toSet
 
@@ -53,6 +54,7 @@ case class PositiveCount(ratingThreshold: Double = 2.0)
     extends AverageMetric[EmptyEvaluationInfo, Query, PredictedResult, ActualResult] {
   override def header = s"PositiveCount (threshold=$ratingThreshold)"
 
+  override
   def calculate(q: Query, p: PredictedResult, a: ActualResult): Double = {
     a.ratings.filter(_.rating >= ratingThreshold).size
   }

examples/scala-parallel-recommendation/train-with-view-event/src/main/scala/Preparator.scala

@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
 class Preparator
   extends PPreparator[TrainingData, PreparedData] {
 
+  override
   def prepare(sc: SparkContext, trainingData: TrainingData): PreparedData = {
     new PreparedData(ratings = trainingData.ratings)
   }

examples/scala-parallel-similarproduct/README.md

@@ -17,4 +17,4 @@ limitations under the License.
 
 This is based on Similar Product Template v0.14.0.
 
-Please refer to http://predictionio.apache.org/templates/similarproduct/how-to/
+Please refer to https://predictionio.apache.org/templates/similarproduct/how-to/

examples/scala-parallel-similarproduct/multi-events-multi-algos/src/main/scala/ALSAlgorithm.scala

@@ -62,6 +62,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
 
   @transient lazy val logger = Logger[this.type]
 
+  override
   def train(sc: SparkContext, data: PreparedData): ALSModel = {
     require(!data.viewEvents.take(1).isEmpty,
       s"viewEvents in PreparedData cannot be empty." +
@@ -133,6 +134,7 @@ class ALSAlgorithm(val ap: ALSAlgorithmParams)
     )
   }
 
+  override
   def predict(model: ALSModel, query: Query): PredictedResult = {
 
     val productFeatures = model.productFeatures

examples/scala-parallel-similarproduct/multi-events-multi-algos/src/main/scala/CooccurrenceAlgorithm.scala

@@ -44,6 +44,7 @@ class CooccurrenceModel(
 class CooccurrenceAlgorithm(val ap: CooccurrenceAlgorithmParams)
   extends P2LAlgorithm[PreparedData, CooccurrenceModel, Query, PredictedResult] {
 
+  override
   def train(sc: SparkContext, data: PreparedData): CooccurrenceModel = {
 
     val itemStringIntMap = BiMap.stringInt(data.items.keys)
@@ -103,6 +104,7 @@ class CooccurrenceAlgorithm(val ap: CooccurrenceAlgorithmParams)
     topCooccurrences
   }
 
+  override
   def predict(model: CooccurrenceModel, query: Query): PredictedResult = {
 
     // convert items to Int index