def choose_random_user(minGames=10):
user_ids = list(userIds("steamData/game_purchase.dat"))
randomID = int(random.choice(user_ids))
usr = userVectorData(str(randomID), game_Ids)
while len(usr.gamePurchasedIds) < 10:
randomID = int(random.choice(user_ids))
usr = userVectorData(str(randomID), game_Ids)
removedGameID_local = choose_random_game(usr)
return usr, removedGameID_local
def recommend_similar_user():
updateMatrix, current_user, remove_game = create_new_user_vector_matrix()
results = create_recommend_pairs_eval(updateMatrix)
matched_user_id = int(results[float(current_user)])
usr = userVectorData(str(matched_user_id), game_Ids)
if str(remove_game) in usr.gameNames.keys():
print('Similar user: game removed Found!')
return 1
else:
print('Similar user: game not Found!')
return 0
def recommend(userId, top=5):
usr = userVectorData(str(userId), game_Ids)
owned_games = usr.print_Owned_games()
usr.loadUsrVector('Vectors_data/user_vector.csv')
results = usr.calculate(VectorDict)
results_list = list()
for i in range(top):
results_list.append(
str(i + 1) + ' ' + game_Ids[str(results[i][0])] + ' Score:' +
str(results[i][1]))
return owned_games, results_list
def run_feature_testing(numberOfTests=5):
final_report = list()
local_sum = 0
for _ in range(numberOfTests):
randomUser, removedGameID = choose_random_user()
localUser = create_vector_user(randomUser.userId, removedGameID)
usr = userVectorData(str(randomUser.userId), game_Ids)
usr.usrVector = copy.deepcopy(localUser.vector)
res = float(recommend_eval(usr, removedGameID))
local_sum += res
print('Test Num: {0} | Rec Score: {1}'.format(_, res))
final_report.append((_, res))
print('==== Final Report Feature Evaluation')
print(final_report)
result = local_sum / numberOfTests
print(result)
def owned_game_list(UserID):
usr = userVectorData(str(UserID), game_Ids)
owned_games = usr.print_Owned_games()
return owned_games