TopicDet.py 文件源码

def get_frequency_table(titles, vocab):
    tokenizer = RegexpTokenizer(r'\w+')
    freqtable = np.ndarray(shape=(len(titles),len(vocab)), dtype=int, order='C')
    freqtable.fill(0)
    for i in range(0,len(titles)):
        raw = titles[i].lower()
        tokens = tokenizer.tokenize(raw)
        for token in tokens:
            index = vocab[token]
            freqtable[i][index] +=1
    return freqtable