Java Array, Finding Duplicates

Question

I have an array, and am looking for duplicates.

duplicates=false;
for (j=0;j<zipcodeList.length;j++)
  for (k=0;k<zipcodeList.length;k++)
    if (zipcodeList[k]==zipcodeList[j]){
      duplicates=true;
     }

However, this code doesnt work when there are no duplicates. Whys that?

Answer 1

To check for duplicates you need to compare distinct pairs.

Answer 2

Cause you are comparing the first element of the array against itself so It finds that there are duplicates even where there aren't.

Answer 3

Let's see how your algorithm works:

an array of unique values:

[1, 2, 3]

check 1 == 1. yes, there is duplicate, assigning duplicate to true.
check 1 == 2. no, doing nothing.
check 1 == 3. no, doing nothing.
check 2 == 1. no, doing nothing.
check 2 == 2. yes, there is duplicate, assigning duplicate to true.
check 2 == 3. no, doing nothing.
check 3 == 1. no, doing nothing.
check 3 == 2. no, doing nothing.
check 3 == 3. yes, there is duplicate, assigning duplicate to true.

a better algorithm:

for (j=0;j<zipcodeList.length;j++) {
    for (k=j+1;k<zipcodeList.length;k++) {
        if (zipcodeList[k]==zipcodeList[j]){ // or use .equals()
            return true;
        }
    }
}
return false;

Answer 4

Don't use == use .equals.

try this instead (IIRC, ZipCode needs to implement Comparable for this to work.

boolean unique;
Set<ZipCode> s = new TreeSet<ZipCode>();
for( ZipCode zc : zipcodelist )
    unique||=s.add(zc);
duplicates = !unique;

Answer 5

Dear God Above, I'm Going to Bed Now.

On the nose answer..

Ugh.

duplicates=false;
for (j=0;j<zipcodeList.length;j++)
  for (k=j+1;k<zipcodeList.length;k++)
    if (k!=j && zipcodeList[k] == zipcodeList[j])
      duplicates=true;

Edited to switch .equals() back to == since I read somewhere you're using int, which wasn't clear in the initial question. Also to set k=j+1, to halve execution time, but it's still O(n^2).

A faster (in the limit) way

Here's a hash based approach. You gotta pay for the autoboxing, but it's O(ln(n)) instead of O(n²). An enterprising soul would go find a primitive int-based hash set (Apache or Google Collections has such a thing, methinks.)

boolean duplicates(final int[] zipcodelist)
{
  Set<Integer> lump = new HashSet<Integer>();
  for (int i : zipcodelist)
  {
    if (lump.contains(i)) return true;
    lump.add(i);
  }
  return false;
}

Bow to HuyLe

See HuyLe's answer for a more or less O(n) solution, which I think needs a couple of add'l steps:

static boolean duplicates(final int[] zipcodelist)
{
   final int MAXZIP = 99999;
   boolean[] bitmap = new boolean[MAXZIP+1];
   java.util.Arrays.fill(bitmap, false);
   for (int item : zipcodeList)
     if (!bitmap[item]) bitmap[item] = true;
     else return true;
   }
   return false;
}

Or Just to be Compact

static boolean duplicates(final int[] zipcodelist)
{
   final int MAXZIP = 99999;
   boolean[] bitmap = new boolean[MAXZIP+1];  // Java guarantees init to false
   for (int item : zipcodeList)
     if (!(bitmap[item] ^= true)) return true;
   return false;
}

Does it Matter?

Well, so I ran a little benchmark, which is iffy all over the place, but here's the code:

import java.util.BitSet;

class Yuk
{
  static boolean duplicatesZero(final int[] zipcodelist)
  {
    boolean duplicates=false;
    for (int j=0;j<zipcodelist.length;j++)
      for (int k=j+1;k<zipcodelist.length;k++)
        if (k!=j && zipcodelist[k] == zipcodelist[j])
          duplicates=true;

    return duplicates;
  }


  static boolean duplicatesOne(final int[] zipcodelist)
  {
    final int MAXZIP = 99999;
    boolean[] bitmap = new boolean[MAXZIP + 1];
    java.util.Arrays.fill(bitmap, false);
    for (int item : zipcodelist) {
      if (!(bitmap[item] ^= true))
        return true;
    }
    return false;
  }

  static boolean duplicatesTwo(final int[] zipcodelist)
  {
    final int MAXZIP = 99999;

    BitSet b = new BitSet(MAXZIP + 1);
    b.set(0, MAXZIP, false);
    for (int item : zipcodelist) {
      if (!b.get(item)) {
        b.set(item, true);
      } else
        return true;
    }
    return false;
  }

  enum ApproachT { NSQUARED, HASHSET, BITSET};

  /**
   * @param args
   */
  public static void main(String[] args)
  {
    ApproachT approach = ApproachT.BITSET;

    final int REPS = 100;
    final int MAXZIP = 99999;

    int[] sizes = new int[] { 10, 1000, 10000, 100000, 1000000 };
    long[][] times = new long[sizes.length][REPS];

    boolean tossme = false;

    for (int sizei = 0; sizei < sizes.length; sizei++) {
      System.err.println("Trial for zipcodelist size= "+sizes[sizei]);
      for (int rep = 0; rep < REPS; rep++) {
        int[] zipcodelist = new int[sizes[sizei]];
        for (int i = 0; i < zipcodelist.length; i++) {
          zipcodelist[i] = (int) (Math.random() * (MAXZIP + 1));
        }
        long begin = System.currentTimeMillis();
        switch (approach) {
        case NSQUARED :
          tossme ^= (duplicatesZero(zipcodelist));
          break;
        case HASHSET :
          tossme ^= (duplicatesOne(zipcodelist));
          break;
        case BITSET :
          tossme ^= (duplicatesTwo(zipcodelist));
          break;

        }
        long end = System.currentTimeMillis();
        times[sizei][rep] = end - begin;


      }
      long avg = 0;
      for (int rep = 0; rep < REPS; rep++) {
        avg += times[sizei][rep];
      }
      System.err.println("Size=" + sizes[sizei] + ", avg time = "
            + avg / (double)REPS + "ms");
    }
  }

}

With NSQUARED:

Trial for size= 10
Size=10, avg time = 0.0ms
Trial for size= 1000
Size=1000, avg time = 0.0ms
Trial for size= 10000
Size=10000, avg time = 100.0ms
Trial for size= 100000
Size=100000, avg time = 9923.3ms

With HashSet

Trial for zipcodelist size= 10
Size=10, avg time = 0.16ms
Trial for zipcodelist size= 1000
Size=1000, avg time = 0.15ms
Trial for zipcodelist size= 10000
Size=10000, avg time = 0.0ms
Trial for zipcodelist size= 100000
Size=100000, avg time = 0.16ms
Trial for zipcodelist size= 1000000
Size=1000000, avg time = 0.0ms

With BitSet

Trial for zipcodelist size= 10
Size=10, avg time = 0.0ms
Trial for zipcodelist size= 1000
Size=1000, avg time = 0.0ms
Trial for zipcodelist size= 10000
Size=10000, avg time = 0.0ms
Trial for zipcodelist size= 100000
Size=100000, avg time = 0.0ms
Trial for zipcodelist size= 1000000
Size=1000000, avg time = 0.0ms

BITSET Wins!

But only by a hair... .15ms is within the error for currentTimeMillis(), and there are some gaping holes in my benchmark. Note that for any list longer than 100000, you can simply return true because there will be a duplicate. In fact, if the list is anything like random, you can return true WHP for a much shorter list. What's the moral? In the limit, the most efficient implementation is:

 return true;

And you won't be wrong very often.

Answer 6

Initialize k = j+1. You won't compare elements to themselves and you'll also not duplicate comparisons. For example, j = 0, k = 1 and k = 0, j = 1 compare the same set of elements. This would remove the k = 0, j = 1 comparison.

Answer 7

You can use bitmap for better performance with large array.

java.util.Arrays.fill(bitmap, false);
for (int item : zipcodeList)
   if (!bitmap[item]) bitmap[item] = true;
   else {
      duplicate = true;
      break;
   }