How to calculate cosine similarity between two columns? - Python

Question

tfvect = TfidfVectorizer(use_idf=True, stop_words = 'english')
wholeword = df_all['search_term']+" "+df_all['product_title']
vocab = tfvect.fit_transform(wholeword)
st = tfvect.transform(df_all['search_term'])
pt = tfvect.transform(df_all['product_title'])

I want to get the cosine similarity between each row of st and pt and store it in df_all['similarity'].

Answer 1

对每对行使用scipy.distance.cosine（v1，v2） http://docs.scipy.org/doc/scipy-0.14.0/reference/generation/scipy.spatial.distance.cosine.html