ianoc

import com.twitter.algebird.Semigroup


case class ClockCache[K, V](cacheSize: Int, maxHits: Int = 10 )
              (implicit sg: Semigroup[V]) {
  private class CacheElement {
    var hits: Int = 0
    var cacheHit: Int = 0
    var key: Any = null

ianoc

/*
GUI settings used:
{
  "server": "127.0.0.1:5556",
  "server_type": "ws",
  "workspaces": [
    {
      "name": "Riemann",
      "view": {
        "type": "Balloon",

ianoc

[info] Formatting 1 Scala source {file:/Users/ioconnell/workspace/github/algebird/}algebird-caliper(test) ...
[info] Formatting 1 Scala source {file:/Users/ioconnell/workspace/github/algebird/}algebird-core(compile) ...
[info] Compiling 1 Scala source to /Users/ioconnell/workspace/github/algebird/algebird-core/target/scala-2.10/classes...
[info] Compiling 6 Scala sources to /Users/ioconnell/workspace/github/algebird/algebird-caliper/target/scala-2.10/test-classes...
[info] Running com.google.caliper.Runner com.twitter.algebird.caliper.HLLPresentBenchmark
[info]  0% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=5, max=10, numHLL=10} 1217.34 ns; σ=60.94 ns @ 10 trials
[info]  2% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=10, max=10, numHLL=10} 1065.77 ns; σ=21.95 ns @ 10 trials
[info]  3% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=17, max=10, numHLL=10} 1127.64 ns; σ=7.16 ns @ 3 trials
[info]  5% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=20, max=10, numHLL=10} 1054.

ianoc

[info]  0% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=5, max=10, numHLL=10} 1549.60 ns; σ=14.49 ns @ 3 trials
[info]  2% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=10, max=10, numHLL=10} 1747.76 ns; σ=59.56 ns @ 10 trials
[info]  3% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=17, max=10, numHLL=10} 1706.26 ns; σ=1.00 ns @ 3 trials
[info]  5% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=20, max=10, numHLL=10} 1672.16 ns; σ=33.41 ns @ 10 trials
[info]  7% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=5, max=100, numHLL=10} 1289.51 ns; σ=12.45 ns @ 3 trials
[info]  8% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=10, max=100, numHLL=10} 1687.68 ns; σ=35.01 ns @ 10 trials
[info] 10% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=17, max=100, numHLL=10} 1744.20 ns; σ=30.19 ns @ 10 trials
[info] 12% Scenario{vm=java, trial=0, benchmark=BatchCreate, bits=20, max=100, numHLL=10} 1675.37 ns; σ=30.35 ns @ 10 trials
[info] 13% Scenario{vm=java, trial

ianoc

val mappedName = "com.blah.X.y.MyType"
    val splitName = mappedName.split('.')
    val exceptLast = splitName.dropRight(1)

    val terms = exceptLast.map(newTermName(_))

    val (first, second) = (terms(0), terms(1))

    val accessorSelect = terms.drop(2).foldLeft(Select(Ident(first), second)) {
      case (existing, next) =>

ianoc

object exp {

  sealed trait ExprTpe[T]
sealed trait ExprPrimitiveTpe[T] extends ExprTpe[T]
sealed trait ExprContainerTpe[T] extends ExprTpe[T] {
  def mapping: List[(String, ExprTpe[_])]
}

case class ContainerTpe2[A, B](_1: ExprTpe[A], nme1: String,  _2: ExprTpe[B], nme2: String) extends ExprContainerTpe[(A, B)] {
  override def toString = s"""Type container of ${_1} and ${_2}"""

ianoc

import com.twitter.algebird.{Aggregator, Semigroup}
import com.twitter.scalding._

import scala.util.Random

/**
 * This job is a tutorial of sorts for scalding's Execution[T] abstraction.
 * It is a simple implementation of Lloyd's algorithm for k-means on 2D data.
 *
 * http://en.wikipedia.org/wiki/K-means_clustering

ianoc

import qualified Turtle as T
import qualified System.IO as IO
import Control.DeepSeq

mergeUntilTwoBlankLines :: (() -> IO (Maybe String)) -> IO [String]
mergeUntilTwoBlankLines fn = do
    rd <- fn ()
    processFunc [] rd
  where processFunc ("" : existing) (Just "") = return existing
        processFunc existing (Just nxt) = do

ianoc

import qualified Turtle as T

main :: IO ()
main = do
  ln <- T.readline
  putStrLn $ show ln

ianoc

def getTiming(ebScaldVariant:String, idx: Int, runName: String, tp: TypedPipe[_]) = {
    val ms = System.currentTimeMillis
    val cntrs = execute(tp.filter(_ => false).toIterableExecution.getAndResetCounters.map(_._2))
    val after = System.currentTimeMillis
    val cpuMS = cntrs.get(StatKey("CPU_MILLISECONDS", "org.apache.hadoop.mapreduce.TaskCounter")).get
    val gcMS = cntrs.get(StatKey("GC_TIME_MILLIS", "org.apache.hadoop.mapreduce.TaskCounter")).get
    val millisMap = cntrs.get(StatKey("MILLIS_MAPS", "org.apache.hadoop.mapreduce.JobCounter")).get
    (ebScaldVariant, idx, runName, cpuMS, gcMS, millisMap, after - ms)
}