Created
December 16, 2016 20:45
-
-
Save ldacosta/b475008959a1009a0ede64469528ce45 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| package com.mediative.mpn.brain.datascience.platform | |
| package nationals.recommendations.strategy | |
| import com.cra.figaro.library.atomic.continuous.{ AtomicNormal, Normal } | |
| import com.holdenkarau.spark.testing.{ DatasetGenerator, DatasetSuiteBase } | |
| import com.mediative.mpn.brain.datascience.platform.nationals.generators | |
| import org.apache.spark.sql.Dataset | |
| import org.scalacheck.Gen | |
| import org.scalatest.FreeSpec | |
| import org.scalatest.prop.GeneratorDrivenPropertyChecks | |
| case class MyCC(mean: Double, standardDev: Double) { | |
| val probabilityD: AtomicNormal = Normal(mean, math.pow(standardDev, 2)) | |
| // 95% interval: | |
| val interval95: (Double, Double) = | |
| (probabilityD.mean - 2 * probabilityD.standardDeviation, probabilityD.mean + 2 * probabilityD.standardDeviation) | |
| } | |
| object AtomicNormalTest { | |
| val MyCCGen: Gen[MyCC] = for { | |
| m <- Gen.choose(1.0, 10.0) | |
| s <- Gen.choose(1.0, 10.0) | |
| } yield MyCC(mean = m, standardDev = s) | |
| } | |
| import com.mediative.mpn.brain.datascience.platform.nationals.recommendations.strategy.AtomicNormalTest._ | |
| class AtomicNormalTest extends FreeSpec with DatasetSuiteBase with GeneratorDrivenPropertyChecks { | |
| implicit override val generatorDrivenConfig = PropertyCheckConfig(minSuccessful = 5) | |
| "test " - { | |
| "direct sampling" in { | |
| forAll(MyCCGen) { cc => | |
| val (iMin, iMax) = cc.interval95 | |
| assert(iMin + iMax >= iMin) // whatever condition. I just need no-one to be lazy. | |
| } | |
| } | |
| "dataset sampling" in { | |
| val myCCDatasetGen: Gen[Dataset[MyCC]] = { | |
| import spark.implicits._ | |
| DatasetGenerator | |
| .genDataset[MyCC](sqlContext)(MyCCGen) | |
| .suchThat(_.count > 0) | |
| } | |
| forAll(myCCDatasetGen) { aSet => | |
| // aSet.show() | |
| assert(aSet.count() >= 0) | |
| } | |
| } | |
| } | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The test called "direct sampling" never fails. The test called "dataset sampling" fails occasionally (ie, not consistently) with a stack trace that looks like the following:
org.scalatest.exceptions.GeneratorDrivenPropertyCheckFailedException: SparkException was thrown during property evaluation. Message: Job aborted due to stage failure: Task 15 in stage 6.0 failed 1 times, most recent failure: Lost task 15.0 in stage 6.0 (TID 84, localhost): java.lang.ArrayIndexOutOfBoundsException: 259 at com.cra.figaro.util.HashSelectableSet.contains(SelectableSet.scala:142) at scala.collection.mutable.SetLike$class.add(SetLike.scala:81) at com.cra.figaro.util.HashSelectableSet.add(SelectableSet.scala:37) at com.cra.figaro.language.Universe.activate(Universe.scala:175) at com.cra.figaro.language.Element.<init>(Element.scala:480) at com.cra.figaro.library.atomic.continuous.AtomicNormal.<init>(Normal.scala:24) at com.cra.figaro.library.atomic.continuous.Normal$.apply(Normal.scala:140) at com.mediative.mpn.brain.datascience.platform.nationals.recommendations.strategy.MyCC.<init>(AtomicNormalTest.scala:13) at com.mediative.mpn.brain.datascience.platform.nationals.recommendations.strategy.AtomicNormalTest$$anonfun$2$$anonfun$apply$1.apply(AtomicNormalTest.scala:24) at com.mediative.mpn.brain.datascience.platform.nationals.recommendations.strategy.AtomicNormalTest$$anonfun$2$$anonfun$apply$1.apply(AtomicNormalTest.scala:23) at scala.Option.map(Option.scala:146)