vocab = [
word[:][0] for word in vocab
] # we get vocab in each word in 100 most_common features, [:] is for all row, [0] is for index 0
print(vocab)
# now we get 300 words as vocab and content_final (content that has been cleared)
#this can take some time, this is from sklearn tfidfVectorizer
tfidf = TfidfVectorizer(analyzer='word',
stop_words=nltk.corpus.stopwords.words('indonesian'),
ngram_range=(1, 1),
min_df=0.04,
vocabulary=vocab)
tfidf_hasil = tfidf.fit_transform(content_final)
features = tfidf.get_feature_name()
print(features)
print(tfidf_hasil.toarray())
# In[5]:
import numpy
numpy.savetxt('D:/SKRIPSI/percobaan/tfidf1332.csv',
tfidf_hasil.todense(),
delimiter=',')
# In[1]:
#df = pd.DataFrame(data = vocab)
#df
