Question

Below is the problem description and algorithm that I have written. Is there anything to be done to improve this algorithm?

Given an integer array of unknown size, containing only numbers between 0 and 30, write a function to return an integer array containing all of the duplicates.

int[] findDupes(int[] array) {
    int[] found = new int[30];
    int[] dupes = new int[30];
    int dupesCount = 0;
    for (int i = 0; i < array.length; i++) {
        if (found[array[i]] <= 1) {
            found[array[i]]++;              
        }else{
            continue;
        }
        if(found[array[i]] > 1){
            dupes[dupesCount++] = array[i];
            if (dupesCount == 30)
                break;
        }
    }
    if (dupesCount == 0)
        return new int[0];
    return dupes;
}

Am assuming that the best case for running this algorithm would n or 30 whichever is lower
and the worst case for running this algorithm is n, since I have to scan the entire array to find duplicates. Any comments?

Answer 1

here is the modified version with comments embedded.

int[] found = new int[3];
    int[] dupes = new int[3];
    int dupesCount = 0;
    for (int i = 0; i < array.length; i++) {
        if (found[array[i]] <= 1) {
            found[array[i]]++;              
        }
        if(found[array[i]] > 1){ //duplicate found
            dupes[dupesCount++] = array[i];

            // if 30 duplicate are found don't scan the array any more
            // since the distinct numbers are only 30
            if (dupesCount == 30) 
                break;
        }
    }
    if (dupesCount == 0)
        return null;
    return dupes;