# this file is added for extra experiment - CT3
from collections import defaultdict
import lib
from lib.recommenders import *
import pandas as pd
BASELINES = [DebiasedModel]
OUTFILE = "results/3_different_size.json"
INFILE = "results/2a_find_best.json"
dataset = lib.ImplicitMovieLens("ml-25m-implicit")
metrics = [
lib.eval.SumOfRanks(),
lib.eval.RecallAtK(100),
lib.eval.RecallAtK(50),
lib.eval.RecallAtK(25)
]
supervised_testset = dataset.load_rankings(
"datasets/labeled/similarity_judgements.test.csv",
"movieId",
"neighborId",
"sim_bin",
verbose=True)
results = dict()
table = list()
prev_models = lib.from_json(INFILE)
for size in [20, 50, 70, 100, 200, 380]:
return None
def get_valid_int(max):
valid = [str(x) for x in range(max + 1)]
while True:
res = input("Input option (0-%d) [empty to exit]: " % max)
if not res.strip():
return None
elif res in valid:
return int(res)
else:
print("Invalid option. Try again.")
movielens = lib.ImplicitMovieLens("ml-25m-implicit")
similarity_judgements = movielens.load_rankings(
"datasets/labeled/similarity_judgements.train.csv",
"movieId",
"neighborId",
"sim_bin",
verbose=False)
models = load_models(movielens, similarity_judgements)
while True:
title = input("Input (partial) movie title [empty to quit]: ")
if not title.strip():
break
results = search(movielens, title)