implement Mallet's get_sorted_words

[colinallen]

Please implement get_sorted_words as described here for Mallet: https://stackoverflow.com/questions/40478209/how-is-the-weight-of-a-word-in-a-topic-calculated-in-mallet

If we already have this functionality under a different name, please create an alias for it

[zhaoyang9425]

I have added the function get_sorted_words() successfully. It is in https://github.com/inpho/vsm/blob/master/vsm/viewer/ldacgsviewer.py
I test it with notebook and it can work well.

from corpus import *
from vsm import *
k = 100
v = lda_v[k]
test = v.get_sorted_words(topic=1)
test[:10]

[(u'\u7ecf', 22807),
(u'\u5377', 16443),
(u'\u8bd1', 8818),
(u'\u7b2c', 6823),
(u'\u5f55', 5781),
(u'\u672c', 4192),
(u'\u85cf', 3544),
(u'\u6cd5', 3448),
(u'\u7eb8', 2897),
(u'\u90e8', 2657)]