scanTrace

def scanTrace() {
 
	import collection.mutable.{ HashMap, MultiMap, Set }
 
	val source = scala.io.Source.fromFile("google-cluster-data-1.csv")
	val map = new HashMap[String, Set[String]] with MultiMap[String, String]
 
	for (i <- source.getLines()) {
	   val fields = i.split(" ")	
	   val parent = fields(1)
	   val child = fields(2)
 
	   map.addBinding(parent, child)
	}
	source.close()

	// mapping from number_of_tasks --> how_many_jobs_have_those_number_of_tasks
	val countingMap = new HashMap[Int,Int]

	for (v <- map.values) {
	   val size = v.size
	   val currSize = countingMap.getOrElse(size, 0)
	   countingMap(size) = currSize + 1
	}

	for (i <- countingMap.keySet.toList.sortWith(_ > _)) 
		println(countingMap.get(i).get + " job has " + i + " tasks")
 
	val jobs = map.keys.size
	val singletonJobs = map.values.foldLeft(0)( (a,b) => if (b.size == 1) a + 1 else a )
	println("jobs:" + jobs)
	print("jobs with 1 task only: " + singletonJobs)
        printf(" (%.2f%s", ((singletonJobs.toDouble/jobs.toDouble) * 100), "%)\n")
	println("largest job:" + map.values.maxBy[Int]( x => x.size ).size + " tasks ")	
}