def read_interaction_data(item_id_mapping, positive_threshold=4.0):
interactions = Interactions(item_id_mapping)
with open(os.path.join(DATA_DIR, "ratings.dat"), "r") as ratingfile:
for line in ratingfile:
(uid, iid, rating, timestamp) = line.split(SEPARATOR)
value = 1.0 if float(rating) >= positive_threshold else 0.0
interactions.add(uid, iid, value)
return interactions
def read_interaction_data(item_id_mapping, positive_threshold=4.0):
interactions = Interactions(item_id_mapping)
with open(os.path.join(DATA_DIR, 'ratings.dat'), 'r') as ratingfile:
for line in ratingfile:
(uid, iid, rating, timestamp) = line.split(SEPARATOR)
value = 1.0 if float(rating) >= positive_threshold else 0.0
interactions.add(uid, iid, value)
return interactions
def read_interactions(item_id_mapping):
interactions = Interactions(item_id_mapping)
interactions_dropped = 0
for post_id, user_id, tags, parent_post_id, body in read_post_data():
assert user_id is not None
assert post_id is not None
# Only answers count as interactions
if parent_post_id is not None:
try:
interactions.add(user_id, parent_post_id, 1.0)
except KeyError:
interactions_dropped += 1
logger.info('Dropped %s interactions due to None user issues', interactions_dropped)
return interactions
def read_interactions(item_id_mapping):
interactions = Interactions(item_id_mapping)
interactions_dropped = 0
for post_id, user_id, tags, parent_post_id, body in read_post_data():
assert user_id is not None
assert post_id is not None
# Only answers count as interactions
if parent_post_id is not None:
try:
interactions.add(user_id, parent_post_id, 1.0)
except KeyError:
interactions_dropped += 1
logger.info('Dropped %s interactions due to None user issues',
interactions_dropped)
return interactions