How to use the ngram.ngram.NGram function in ngram

def __get_n_grams(self, text, n_type, n):
        #removes the stop words
        tokens = list(self.__tokenize_input(text))
        if n_type == "word":
            if len(tokens) > n:
                return ["".join(j) for j in zip(*[tokens[i:] for i in range(n)])]
            else:
                #returns the word directly if n is greater than number of words
                a = list()
                a.append(text)
                return a
        if n_type == "character":
            gram_object = ngram.NGram(N=n)
            gram_char_tokens = list(gram_object.split(text))
            if len(text) > n:
                return gram_char_tokens
            else:
                a = list()
                a.append(text)
                return a
        else:
            return list(self.__tokenize_input(text))