| package org.template.classification |
| |
| import io.prediction.controller.AverageMetric |
| import io.prediction.controller.EmptyEvaluationInfo |
| import io.prediction.controller.EngineParams |
| import io.prediction.controller.EngineParamsGenerator |
| import io.prediction.controller.Evaluation |
| import io.prediction.controller.Workflow |
| import org.apache.spark.SparkContext |
| import org.apache.spark.SparkContext._ |
| import org.apache.spark.rdd.RDD |
| |
| case class Precision |
| 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) |
| } |
| |
| object PrecisionEvaluation extends Evaluation { |
| // Define Engine and Metric used in Evaluation |
| engineMetric = (ClassificationEngine(), new Precision()) |
| } |
| |
| object EngineParamsList extends EngineParamsGenerator { |
| // Define list of EngineParams used in Evaluation |
| |
| // First, we define the base engine params. It specifies the appId from which |
| // the data is read, and a evalK parameter is used to define the |
| // cross-validation. |
| private[this] val baseEP = EngineParams( |
| dataSourceParams = DataSourceParams(appId = 18, evalK = Some(5))) |
| |
| // Second, we specify the engine params list by explicitly listing all |
| // algorithm parameters. In this case, we evaluate 3 engine params, each with |
| // a different algorithm params value. |
| engineParamsList = Seq( |
| baseEP.copy(algorithmParamsList = Seq(("naive", AlgorithmParams(10.0)))), |
| baseEP.copy(algorithmParamsList = Seq(("naive", AlgorithmParams(100.0)))), |
| baseEP.copy(algorithmParamsList = Seq(("naive", AlgorithmParams(1000.0))))) |
| } |