public static void AnalyzeTwitterWordDistribution(string inputPath, TokenizeConfig tokenConfig)
        {
            var indexReader = LuceneOperations.GetIndexReader(inputPath);
            var docNum      = indexReader.NumDocs();

            int[] docWordCnt     = new int[docNum];
            int[] docUniqWordCnt = new int[docNum];
            Dictionary <string, int> wordDocCntDict = new Dictionary <string, int>();
            Dictionary <string, int> wordOccCntDict = new Dictionary <string, int>();

            var fieldWeights = tokenConfig.TokenizerType == TokenizerType.FeatureVector
                ? BingNewsFields.FeatureVectorFieldWeights
                : BingNewsFields.NewsFieldWeights;

            ProgramProgress progress = new ProgramProgress(docNum);

            for (int iDoc = 0; iDoc < docNum; iDoc++)
            {
                var document = indexReader.Document(iDoc);
                var content  = LuceneOperations.GetContent(document, fieldWeights);

                var words       = NLPOperations.Tokenize(content, tokenConfig);
                var uniqueWords = new HashSet <string>(words);
                docWordCnt[iDoc]     = words.Count;
                docUniqWordCnt[iDoc] = uniqueWords.Count;

                foreach (var word in uniqueWords)
                {
                    if (!wordDocCntDict.ContainsKey(word))
                    {
                        wordDocCntDict.Add(word, 0);
                    }
                    wordDocCntDict[word]++;
                }

                foreach (var word in words)
                {
                    if (!wordOccCntDict.ContainsKey(word))
                    {
                        wordOccCntDict.Add(word, 0);
                    }
                    wordOccCntDict[word]++;
                }

                progress.PrintIncrementExperiment();
            }
            progress.PrintTotalTime();

            indexReader.Close();

            //Statistics
            DoubleStatistics statDocWordCnt     = new DoubleStatistics();
            DoubleStatistics statDocUniqWordCnt = new DoubleStatistics();
            DoubleStatistics statWordDocCnt     = new DoubleStatistics();
            DoubleStatistics statWordOccCnt     = new DoubleStatistics();

            for (int iDoc = 0; iDoc < docNum; iDoc++)
            {
                statDocWordCnt.AddNumber(docWordCnt[iDoc]);
                statDocUniqWordCnt.AddNumber(docUniqWordCnt[iDoc]);
            }

            foreach (var kvp in wordDocCntDict)
            {
                statWordDocCnt.AddNumber(kvp.Value);
            }

            foreach (var kvp in wordOccCntDict)
            {
                statWordOccCnt.AddNumber(kvp.Value);
            }


            Console.WriteLine(statDocWordCnt.ToString("statDocWordCnt"));
            Console.WriteLine(statDocUniqWordCnt.ToString("statDocUniqWordCnt"));
            Console.WriteLine(statWordDocCnt.ToString("statWordDocCnt"));
            Console.WriteLine(statWordOccCnt.ToString("wordOccCnt"));

            //Hist
            var docWordCntHist       = new DoubleHistogram(docWordCnt.Select(i => (double)i), (double)1);
            var docUniqueWordCntList = new DoubleHistogram(docUniqWordCnt.Select(i => (double)i), (double)1);
            var wordDocCntHist       = new DoubleHistogram(wordDocCntDict.Select(kvp => (double)kvp.Value), 1000);
            var wordDocCntHist2      = new DoubleHistogram(wordDocCntDict.Select(kvp => (double)kvp.Value), (double)1);

            docWordCntHist.PrintToFile(StringOperations.EnsureFolderEnd(inputPath) + "docWordCntHist.csv");
            docUniqueWordCntList.PrintToFile(StringOperations.EnsureFolderEnd(inputPath) + "docUniqueWordCntList.csv");
            wordDocCntHist.PrintToFile(StringOperations.EnsureFolderEnd(inputPath) + "wordDocCntHist.csv");
            wordDocCntHist2.PrintToFile(StringOperations.EnsureFolderEnd(inputPath) + "wordDocCntHist2.csv");

            Console.Read();
        }