Question

I have the following procedure that counts the document frequency for each entry of doctionary (freq)

  foreach (var termIndex in freq.Select(entry => GetTermIndex(entry.Key)))
         {                                                              
            _docFreq[i][termIndex]++;
         }

And also the procedure for getting term index

rivate int GetTermIndex(string term)
    {
        int i;
        if (_wordsIndex.TryGetValue(term, out i))
            return i;
        else
            return 0;
    }

Where, all terms are indexed in another dictionary (_wordsIndex).

Based on above procedures, if the entry.Key in freq not exist, then the GetTermIndex will return 0 and that will count up the _docFreq (_docFreq[i][0]) and that is the problem.
So, how could i avoid counting when the entry.Key not exist ?? I tried to do something like

rivate int GetTermIndex(string term)
    {
        int i;
        if (_wordsIndex.TryGetValue(term, out i))
            return i;
        else
            return -1;
    }

But of course an error “Index was outside the bounds of the array” appeared, because there no index for _docFreq[i][-1].

Could any help please ?? many Thanks

Answer 1

Just don’t count it if termIndex is -1:

foreach (var termIndex in freq.Select(entry => GetTermIndex(entry.Key)))
{
    if (termIndex >= 0)
        _docFreq[i][termIndex]++;
}

or using LINQ Where:

foreach (var termIndex in freq.Select(entry => GetTermIndex(entry.Key)).Where(ti => ti >= 0))
{
    _docFreq[i][termIndex]++;
}