def nearest_neighbor(user, users_data, measure):
"""Finds a list of users that are nearest to a given user.
Args:
user: the reviewerID of the user that we want to recommend items to
users_data: the complete datastore of ratings
measure: the distance measure to use
if 1: Adjusted cosine similarity
if 2: Cosine similarity
if 3: Pearson correlation
if 4: Manhattan distance
if 5: Euclidean distance
Return:
A sorted list of users based on their distance to the input user
"""
distances = []
for u in users_data:
if u != user:
if measure == 1:
distance = measures.adjusted_cosine_similarity(
users_data[user], users_data[u])
elif measure == 2:
distance = measures.cosine_similarity(users_data[user],
users_data[u])
elif measure == 3:
distance = measures.pearson_correlation(
users_data[user], users_data[u])
elif measure == 4:
distance = measures.manhattan_distance(users_data[user],
users_data[u])
elif measure == 5:
distance = measures.euclidean_distance(users_data[user],
users_data[u])
else:
"ERROR! The similarity measure doesn't exist!"
distances.append((u, distance))
distances.sort(key=lambda itemTuple: itemTuple[1], reverse=True)
return distances
def nearest_neighbor(user, users_data):
"""Finds a list of users that are nearest to a given user.
Args:
user: the reviewerID of the user that we want to recommend items to
users_data: the complete datastore of ratings
Return:
A sorted list of users based on their distance to the input user
"""
distances = []
for u in users_data:
if u != user:
distance = measures.adjusted_cosine_similarity(users_data[user], users_data[u])
distances.append((u, distance))
distances.sort(key=lambda itemTuple: itemTuple[1], reverse=True)
return distances
def nearest_neighbor(user, users_data, measure):
"""Finds a list of users that are nearest to a given user.
Args:
user: the reviewerID of the user that we want to recommend items to
users_data: the complete datastore of ratings
measure: the distance measure to use
if 1: Adjusted cosine similarity
if 2: Cosine similarity
if 3: Pearson correlation
if 4: Manhattan distance
if 5: Euclidean distance
Return:
A sorted list of users based on their distance to the input user
"""
distances = []
for u in users_data:
if u != user:
if measure == 1:
distance = measures.adjusted_cosine_similarity(users_data[user], users_data[u])
elif measure == 2:
distance = measures.cosine_similarity(users_data[user], users_data[u])
elif measure == 3:
distance = measures.pearson_correlation(users_data[user], users_data[u])
elif measure == 4:
distance = measures.manhattan_distance(users_data[user], users_data[u])
elif measure == 5:
distance = measures.euclidean_distance(users_data[user], users_data[u])
else:
"ERROR! The similarity measure doesn't exist!"
distances.append((u, distance))
distances.sort(key=lambda itemTuple: itemTuple[1], reverse=True)
return distances
