def similiar():
data, ids =get_notes_data()
vector = TfidfVectorizer(tokenizer=tokenize)
X = vector.fit_transform(data)
nbrs = NearestNeighbors(n_neighbors=3, algorithm='kd_tree').fit(X)
distances, indices = nbrs.kneighbors(X)
print distances,indices
# [[id,1,2],[id,2,3]]
print type(indices)
for i in xrange(len(indices)):
for j in xrange(len(indices[0])):
indices[i][j] = ids[indices[i][j]]
client = MongoClient()
notes = client.rss.notes
print indices
for x in indices:
notes.update({'_id': str(x[0])},{'$set': {'related':[str(i) for i in x[1:]]}})
if __name__ == '__main__':
# update_notes_key_words()
# similiar()
for x in xrange(10):
a=10
print x
client = MongoClient()
client.id11 = 1
print client.id11
