获取名称错误名称'sentence_stream'未定义

Question

我想使用三元组和双字母，因为我不想只使用unigrams。

bigramer = gensim.models.Phrases(sentences)
model = Word2Vec(bigramer[sentences], workers=num_workers, \
            size=num_features, min_count = min_word_count, \
            window = context, sample = downsampling)
from nltk import bigrams
from nltk import trigrams
from gensim.models import Phrases
from gensim.models.phrases import Phraser
trigrams = Phrases(bigrams[sentence_stream])

但是，我遇到了这个错误。

NameErrorTraceback (most recent call last)
<ipython-input-161-15b0101c13b1> in <module>()
----> 1 trigrams = Phrases(bigrams[sentence_stream])

NameError: name 'sentence_stream' is not defined

Answer 1

我通过将代码重写为：

解决了这个问题

bigram = Phrases(sentences, min_count=1, threshold=1)
print list(bigram[sentences])

trigram = Phrases(bigram[sentences],min_count=1, threshold=1)
print list(trigram[bigram[sentences]])