create a document-term frequency matrix

tf.py

# create candidate sentense set
docs = [['a', 'b', 'c'],
        ['b', 'd']]
terms = ['a', 'b', 'c', 'd']
vlist = []
n = len(docs)
d = len(terms)
for doc in docs:
    tmp = []
    for term in doc:
        tmp.append(terms.index(term))
    v = np.bincount(tmp, minlength=d)
    vlist.append(v)
V = np.vstack(vlist)