def tfidf(data):
tfidf_vectorizer = TfidfVectorizer(
max_df=0.8, min_df=2,
stop_words='english') #max_features=n_features,stop_words='english')
tfidf_vectorizer.decode_error = 'ignore'
tfidf = tfidf_vectorizer.fit_transform(data)
return [tfidf, tfidf_vectorizer]
