wait for all threads complete in Java

Question

I use a multithread quicksort Java function from cern.colt.ParallelQuickSort library(http://incanter.org/docs/parallelcolt/api/cern/colt/ParallelQuickSort.html and https://github.com/Danimoth/Parallel-Colt/blob/master/src/cern/colt/ParallelQuickSort.java). I want to test how much time elapse when choosing different number of threads. I use System.nanoTime() to keep track of the run time. However, even when the number of threads I choose and unsorted array is the same for multiple runs, the run time are quite different. I think it is because the quicksort() provided in cern.colt.ParallelQuickSort library has no wait for threads to complete. I would like to know how I can write code to wait for all threads to complete so that I can measure the run time outside the function the library provided? Something like below:?

ParallelQuickSort qs=new ParallelQuickSort();
long startTime = System.nanoTime();
qs.quickSort(unsorted_array, 0, array_size, comp, number_threads);

//java code to wait for all threads to complete

long time_elapse= System.nanoTime() - startTime;

Edit: Below is my code: Originally, my code is to run quicksort with threads number from 1 to 15 for array size 2^10, 2^15, 2^20, 2^25 and 2^28 and for each case I run 30 times. In order to debug, I changed my code to only run array size=2^10 using 1 thread and run it for 10 times.

import cern.colt.ParallelQuickSort;
import cern.colt.function.tint.IntComparator;
import cern.colt.Timer;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileWriter;
import java.io.IOException;
import java.lang.System.*;
import java.util.*;


public class quick_sort {

    static void readData(int dst[], int nitems, int num) throws IOException{ 
        String s="mydata"+Integer.toString(num)+".txt";
        //System.out.println(s);
        Scanner scanner = new Scanner(new File(s));
        int i = 0;
        while(scanner.hasNextInt())
        {
            dst[i++] = scanner.nextInt();
        }

    }



public static void main(String [ ] args) throws IOException
{

    //for(int i=0; i<n;i++) dst[i]=n-i;


    /*
    System.out.println("Unsorted: ");
    for(int i=0; i<n; i++) System.out.print(dst[i]+" ");
    System.out.println(" ");
    */

    IntComparator comp=new IntComparator(){
        public int compare(int a, int b){
            if(a>b) return 1;
            else if(a<b) return -1;
            else return 0;
        }
      };




    int iter=10;
    int thread_num=1;
    //FileWriter fw = new FileWriter("out.txt");

    int num[]={10, 15, 20, 25, 28};
for(int m=0; m<1; m++){         
    for(int k=1; k<=thread_num; k++){
    long estimatedTime=0;
    for(int i=0; i<iter; i++){
        int n=1<<num[m];
        int dst[]=new int[n];
        readData(dst, n, num[m]);

        ParallelQuickSort qs=new ParallelQuickSort();

        long startTime = System.nanoTime();

        qs.quickSort(dst, 0, n, comp, k);

        long temp= System.nanoTime() - startTime;
        estimatedTime+=temp;



        System.out.println("Time="+temp*0.000001);

    }
    System.out.println(num[m]+"Wall Clock Time when thread number="+k+": "+estimatedTime/iter*0.000001);
    //fw.write(num[m]+"Wall Clock Time when thread number="+k+": "+estimatedTime/iter*0.000001+'\n');
}   


    //System.out.println("Sorted: ");
    //for(int i=0; i<n; i++) System.out.print(dst[i]+" ");
    //System.out.println(" ");




}
//fw.close();
System.out.println("Finish!");

}   

}

The result is shown below:

Time=0.755289
Time=0.632124
Time=0.502016
Time=0.502922
Time=0.100524
Time=0.076072
Time=0.073657
Time=0.073355
Time=0.074261
Time=0.076374
10Wall Clock Time when thread number=1: 0.286659
Finish!

Answer 1

The function ParallelQuickSort.quicksort returns only once all threads/sub operations are complete. You do not need to manually wait until all threads are finished.

This can be confirmed by looking into the code (look for other.get()) and that it is the only reasonable behavior.

EDIT: Testing performance can be deceivingly hard, see Java Performance Testing and many other places for details.