Java Cookbook

Program: Timing Comparisons

New developers sometimes worry about the overhead of these collections and think they should use arrays instead of data structures. To investigate, I wrote a program that creates and accesses 250,000 objects, once through a Java array and again through an ArrayList. This is a lot more objects than most programs use. First the code for the Array version:

import com.darwinsys.util.MutableInteger;

/** Time a bunch of creates and gets through an Array */
public class Array {
    public static final int MAX = 250000;
    public static void main(String[] args) {
        System.out.println(new Array().run(  ));
    }
    public int run(  ) {
        MutableInteger list[] = new MutableInteger[MAX];
        for (int i=0; i<list.length; i++) {
            list[i] = new MutableInteger(i);
        }
        int sum = 0;
        for (int i=0; i<list.length; i++) {
            sum += list[i].getValue(  );
        }
        return sum;
    }
}

And the ArrayList version:

import java.util.ArrayList;

import com.darwinsys.util.MutableInteger;

/** Time a bunch of creates and gets through an Array */
public class ArrayLst {
    public static final int MAX = 250000;
    public static void main(String[] args) {
        System.out.println(new ArrayLst().run(  ));
    }
    public int run(  ) {
        ArrayList list = new ArrayList(  );
        for (int i=0; i<MAX; i++) {
            list.add(new MutableInteger(i));
        }
        int sum = 0;
        for (int i=0; i<MAX; i++) {
            sum += ((MutableInteger)list.get(i)).getValue(  );
        }
        return sum;
    }
}

The Vector-based version, ArrayVec , is sufficiently similar that I don’t feel the need to kill a tree reprinting ...