def get_user_profile(user_id):
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
ratings = client.get_data_table(session, keyspace, table)
if ratings.empty:
return empty_json
ratings = ratings.drop(columns="rating_id")
avg_ratings = calculate_avg_ratings(ratings, converted_genres)
avg_user_ratings = pd.read_json(get_user_avg_rating(user_id),
orient='records')
appended_avg_ratings = avg_ratings.append(avg_user_ratings,
ignore_index=True,
sort=False)
profile = pd.DataFrame(columns=converted_genres)
# FIXME: Change profile
profile.loc[0] = [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
]
for genre in converted_genres:
profile.iloc[0, profile.columns.get_loc(genre)] = appended_avg_ratings.iloc[0][genre] - \
appended_avg_ratings.iloc[1][genre]
return profile.to_json(orient='records')
def delete_rating(json_rating):
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
rating = pd.read_json(json_rating, orient=records)
client.delete_rows(session, keyspace, table, rating)
return json_rating
def add_rating(json_rating):
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
new_rating = pd.read_json(json_rating, orient=records)
client.push_data_table(session, keyspace, table, new_rating)
return json_rating
def get_ratings():
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
df = client.get_data_table(session, keyspace, table)
if df.empty:
return empty_json
df = df.drop(columns="rating_id")
return df.to_json(orient='records')
def get_user_avg_rating(user_id):
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
ratings = client.get_data_table(session, keyspace, table)
ratings = ratings.sort_values(by=['userid'])
ratings.set_index('userid')
ratings = ratings[int(user_id) == ratings['userid']]
return calculate_avg_ratings(ratings,
converted_genres).to_json(orient=records)
def get_avg_ratings():
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
ratings = client.get_data_table(session, keyspace, table)
if ratings.empty:
return empty_json
ratings = ratings.drop(columns="rating_id")
return calculate_avg_ratings(ratings,
converted_genres).to_json(orient=records)
def __init__(self):
self.genres = wtiproj03_ETL.get_genres_list()
self.keyspace = "user_ratings"
self.rating_table = "rating"
self.user_avg_table = "user_avg"
self.all_avg_table = "all_avg"
self.cluster = Cluster(['127.0.0.1'], port=9042)
self.session = self.cluster.connect()
wtiproj06_cassandra_client.create_keyspace(self.session, self.keyspace)
wtiproj06_cassandra_client.create_rating_table(self.session,
self.keyspace,
self.rating_table)
wtiproj06_cassandra_client.create_user_avg_table(
self.session, self.keyspace, self.user_avg_table)
wtiproj06_cassandra_client.create_all_avg_table(
self.session, self.keyspace, self.all_avg_table)
self.session.set_keyspace(self.keyspace)
self.session.row_factory = dict_factory
def init():
np.set_printoptions(threshold=sys.maxsize)
users = pd.read_csv("user_ratedmovies.dat",
header=0,
delimiter="\t",
usecols=['userID', 'movieID', 'rating'])
movies = pd.read_csv("movie_genres.dat",
header=0,
delimiter="\t",
usecols=['movieID', 'genre'])
joinedTable = join_tables(users, movies, 'movieID')
mergedTable, genres = build_dataframe(joinedTable, movies)
session = client.create_session()
client.create_keyspace(session, keyspace)
session.set_keyspace(keyspace)
session.row_factory = dict_factory
client.push_data_table(session, keyspace, table, mergedTable)