def read_ratings(self, file_name):
if self.raw_ratings is None:
Dataset.read_ratings(self, file_name)
else:
return self.raw_ratings
[stdout.write(' %.4f' % p) for p in propensities]
stdout.write('\n')
with open(propensity_file) as fin:
line = fin.readline()
propensities = np.asarray([float(f) for f in line.split()])
propensities /= propensities.sum()
# [stdout.write('%.4f ' % p) for p in propensities]
# stdout.write('\n')
weights = 1.0 / propensities
# [stdout.write('%.4f ' % w) for w in weights]
# stdout.write('\n')
reader = Reader(line_format='user item rating', sep='\t')
data = Dataset(reader=reader, rating_scale=rating_scale)
raw_trainset = data.read_ratings(train_file)
raw_testset = data.read_ratings(test_file)
trainset = data.construct_trainset(raw_trainset)
testset = data.construct_testset(raw_testset)
#### default
n_factors_opt = [
100,
]
n_epochs_opt = [
20,
]
biased_opt = [
True,
]
reg_all_opt = [
