Skip to content

Instantly share code, notes, and snippets.

@autodidacticon
Created September 28, 2023 14:11
Show Gist options
  • Save autodidacticon/82a2ba9907ecddfd287e34c40db04a6b to your computer and use it in GitHub Desktop.
Save autodidacticon/82a2ba9907ecddfd287e34c40db04a6b to your computer and use it in GitHub Desktop.
scala.collection.parallel.mutable.ParMap vs java.util.concurrent.ConcurrentHashMap
scala> :load test.sc
val args: Array[String] = Array()
Loading test.sc...
import java.lang.System.nanoTime
import java.util.concurrent.ConcurrentHashMap
import scala.collection.mutable
import scala.collection.parallel.mutable.{ParArray, ParHashMap}
import scala.jdk.CollectionConverters.ConcurrentMapHasAsScala
val javaMap: scala.collection.concurrent.Map[Long,Long] = Map()
val scalaMap: scala.collection.parallel.mutable.ParHashMap[Long,Long] = ParHashMap()
val testSize: Int = 1200000
val iterations: Int = 100
val javaDurations: scala.collection.mutable.ListBuffer[Long] = ListBuffer()
val scalaDurations: scala.collection.mutable.ListBuffer[Long] = ListBuffer()
val javaAvg: Double = 1.48506236771E7
val javaVariance: Double = 3.1534672699274938E13
val javaStdDev: Double = 5615574.120183522
java.util.concurrent.ConcurrentHashMap: average filter time: 1.48506236771E7
java.util.concurrent.ConcurrentHashMap: min filter time: 7929000
java.util.concurrent.ConcurrentHashMap: max filter time: 161625250
java.util.concurrent.ConcurrentHashMap: filter time variance: 3.1534672699274938E13
java.util.concurrent.ConcurrentHashMap: filter time stddev: 5615574.120183522
val scalaAvg: Double = 1.41805124519E7
val scalaVariance: Double = 2.3432521810268447E14
val scalaStdDev: Double = 1.5307684936092867E7
scala.collection.parallel.mutable.ParHashMap: average filter time: 1.41805124519E7
scala.collection.parallel.mutable.ParHashMap: min filter time: 2373333
scala.collection.parallel.mutable.ParHashMap: max filter time: 417851791
scala.collection.parallel.mutable.ParHashMap: filter time variance: 2.3432521810268447E14
scala.collection.parallel.mutable.ParHashMap: filter time stddev: 1.5307684936092867E7
import java.lang.System.nanoTime
import java.util.concurrent.ConcurrentHashMap
import scala.collection.mutable
import scala.collection.parallel.mutable.{ParArray, ParHashMap}
import scala.jdk.CollectionConverters.ConcurrentMapHasAsScala
val javaMap: collection.concurrent.Map[Long, Long] = new ConcurrentHashMap[Long, Long]().asScala
val scalaMap = new ParHashMap[Long, Long]()
val testSize = 1200000
val iterations = 100
(0 until testSize) foreach(i => {
javaMap.put(i, i)
scalaMap.put(i, i)
})
val javaDurations = mutable.ListBuffer.empty[Long]
val scalaDurations = mutable.ListBuffer.empty[Long]
0 until iterations foreach { _ =>
val innerJavaMap = javaMap.clone()
val innerScalaMap = scalaMap.clone()
0 until iterations foreach { i =>
val s = nanoTime()
innerJavaMap.filterInPlace((k, v) => v > i)
javaDurations.addOne(nanoTime() - s)
}
0 until iterations foreach { i =>
val s = nanoTime()
innerScalaMap
.filter { case (key, value) => value <= i }
.foreach { case (key, value) =>
innerScalaMap.remove(key)
}
scalaDurations.addOne(nanoTime() - s)
}
}
val javaAvg = javaDurations.sum / javaDurations.length.toDouble
val javaVariance = javaDurations.map(d => scala.math.pow(d - javaAvg, 2)).sum / javaDurations.length
val javaStdDev = scala.math.sqrt(javaVariance)
println(s"java.util.concurrent.ConcurrentHashMap: average filter time: $javaAvg")
println(s"java.util.concurrent.ConcurrentHashMap: min filter time: ${javaDurations.min}")
println(s"java.util.concurrent.ConcurrentHashMap: max filter time: ${javaDurations.max}")
println(s"java.util.concurrent.ConcurrentHashMap: filter time variance: $javaVariance")
println(s"java.util.concurrent.ConcurrentHashMap: filter time stddev: $javaStdDev")
val scalaAvg = scalaDurations.sum / scalaDurations.length.toDouble
val scalaVariance = scalaDurations.map(d => scala.math.pow(d - javaAvg, 2)).sum / scalaDurations.length
val scalaStdDev = scala.math.sqrt(scalaVariance)
println(s"scala.collection.parallel.mutable.ParHashMap: average filter time: $scalaAvg")
println(s"scala.collection.parallel.mutable.ParHashMap: min filter time: ${scalaDurations.min}")
println(s"scala.collection.parallel.mutable.ParHashMap: max filter time: ${scalaDurations.max}")
println(s"scala.collection.parallel.mutable.ParHashMap: filter time variance: $scalaVariance")
println(s"scala.collection.parallel.mutable.ParHashMap: filter time stddev: $scalaStdDev")
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment