think-openly/selector.py

18 lines
495 B
Python

import nlp
import random
# get user preference from database (i.e. how many times they clicked on some certain type of article)
# prob = [1/nlp.NUM_TOPICS for i in range(nlp.NUM_TOPICS)]
# manipulate prob based on user preference
def get_topics(weights, num_reccomendations):
"""
Takes in weights as list/tuple, ex: (0.1, 0.2, 0.3)
Returns a list of topics
"""
return random.choices([*range(nlp.NUM_TOPICS)], weights, k=num_reccomendations)
# print(get_topics(prob, 4))