Question

I have a map of word frequencies Map<String, Integer>. I need to make a set of the least occurring words. Say the lowest occurring words all appeared twice, I need to make a set of all these twice-occurring words. So far I have:

public Set findRarest()
{
    int occurrence = 1000;  //high initial value for word length
    for (Map.Entry<String,Integer> item : wcMap.entrySet())
    {
        if (item.getValue() > occurrence);        //most likely for performance
        else if (item.getValue() == occurrence)
        {
            rarest.add(item.getKey());
        }
        else                                      //found new lowest count
        {
            rarest.clear();
            rarest.add(item.getKey());
        }
    }
    return rarest;
}

This seems a little convoluted to me. Are there native collection tools to get this done?

Answer 1

I don’t think your code even works as written. Two things:

1. Initialize occurrence with Integer.MAX_VALUE instead of just some arbitrary large value.

2. Update the value of occurrence whenever you find a word which occurs less often.

Other than that, your solution is fine. I’m not sure you could get anything much cleaner restricting yourself to Java Collections Framework classes.

Updated code:

public Set findRarest()
{
    Set<String> rarest = new HashSet<String>();

    int occurrence = Integer.MAX_VALUE;  //high initial value for word length
    for (Map.Entry<String,Integer> item : wcMap.entrySet())
    {
        if (item.getValue() == occurrence)
        {
            rarest.add(item.getKey());
        }
        else if ( item.getValue() < occurrence )
        {
            occurrence = item.getValue();
            rarest.clear();
            rarest.add(item.getKey());
        }
    }
    return rarest;
}