from load import model
import torch
import numpy as np
new_user_input = torch.FloatTensor([0, 0, 0, 3, 1, 2])
output = model(new_user_input)
output = (output + 1)
print(output)
# 주어진 테이블
food_table = []
print("모델 Output")
print(output)
print("최고 별점 음식 추천")
print(np.sort(output.detach().numpy()))
print("최고 별점의 food id 리스트")
sort_food_id = np.argmax(output.detach().numpy())
print(sort_food_id)
def AIModel(include, exclude):
num_epochs = 150
batch_size = 50
learning_rate = 1e-3
t1_data = pd.read_csv('./11/08.csv')
nb_users = int(max(t1_data.iloc[:, 0])) + 1 # 사용자의 수 +1
nb_foods = int(max(t1_data.iloc[:, 1])) + 1 # 음식 종류의 수 만약 인덱스 0부터 주었다면 +1
t1_data = t1_data.values
def convert(data):
new_data = []
for id_users in range(0, nb_users): # 총 사용자 수많큼 반복해라
id_foods = data[:, 1][data[:, 0] == id_users] # user가 본 영화들
id_foods = id_foods.astype(int) # user가 본 영화들의 별점
id_ratings = data[:, 2][data[:, 0] == id_users]
ratings = np.zeros(nb_foods) # 영화 숫자만큼 zero 배열 만들어줌
ratings[
id_foods] = id_ratings # id_movies영화갯수 1부터 하려고 -1을 해줌/ id_movies - 1번째 영화 /ratings[id_movies - 1]: n번째 영화 별점이 몇점인지 쭉 나열
ratings = ratings.astype(float)
new_data.append(list(ratings)) # 전체영화 zero에 배열되있는것에 점수 넣어줌
print('epoch [{}/{}], loss:{:.4f}'
.format(epoch + 1, num_epochs, loss.item()))
return new_data
t2_data = convert(t1_data)
t2_data = np.asarray(t2_data)
tensor = torch.FloatTensor(t2_data)
num_train_dataset = int(len(tensor) * 0.8)
num_test_dataset = len(tensor) - num_train_dataset
train_dataset, test_dataset = torch.utils.data.random_split(tensor, [num_train_dataset, num_test_dataset])
dataloader = DataLoader(tensor, batch_size=batch_size, shuffle=True)
aaa = []
lists = include
print(lists)
arr = [0 for i in range(1, 310)]
count = 0
for i in lists:
if count < 4:
arr[i] = 5
elif count < 8:
arr[i] = 3
elif count < 12:
arr[i] = 1
count += 1
aaa.append(arr)
bb = torch.FloatTensor(aaa)
new_user_input = bb
output = model(new_user_input)
output = (output + 1)
sort_food_id = np.argsort(-output.detach().numpy())
sort_food_id_list = sort_food_id.tolist()
food_real_list = np.ravel(sort_food_id_list, order='C').tolist()
file = pd.read_excel('food_label.xlsx')
rm_list = set()
list_remove = exclude
for j in list_remove:
for i in range(309):
if file[j][i] == 1:
rm_list.add(file['f_num'][i])
rm_real_list = list(rm_list)
for i in food_real_list:
if i in rm_list or i in include:
food_real_list.remove(i)
top_10 = food_real_list[:10]
count1 = 2
sampleList1 = include
random_list1 = random.sample(sampleList1, count1)
count = 4
sampleList = top_10
random_list2 = random.sample(sampleList, count)
final_list = random_list1 + random_list2
return final_list
action='store',
default=0.001,
type=float)
hyper_param.add_argument('-u',
'--hidden_units',
default=3136,
dest='u',
type=int)
hyper_param.add_argument('-e', '--epochs', dest='e', default=8, type=int)
args = parser.parse_args()
data_d = args.inputDirectory
architec = args.arch
lr = args.l
hidden_units = args.u
epochs = args.e
gpu = args.gpu
path = args.save_dir
train_image_datasets, test_image_data_sets, valid_image_datasets, trainloaders, testloaders, validloaders = load.arg_inputdir(
args.inputDirectory)
pretrained_model, output_size, i = model.arch(architec)
pretrained_model, criterion, optimizer = model.classifier(
pretrained_model, output_size, i, lr, hidden_units)
load.training_model(pretrained_model, criterion, optimizer, epochs, gpu)
load.validation_test(pretrained_model, testloaders)
complete_path = load.set_checkpoint_path(path)
load.save_model(pretrained_model, complete_path, architec)
load.model(architec)
# pretrained_model = load.load_model(pretrained_model,'checkpoint.pth')
nn.ReLU(True),
nn.Linear(200, 309), nn.Tanh())
def forward(self, x):
x = self.encoder(x)
x = self.decoder(x)
return x
model = autoencoder()
criterion = nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate, weight_decay=1e-3)
for epoch in range(num_epochs):
for data in dataloader:
output = model(data)
loss = criterion(output, data)
optimizer.zero_grad()
loss.backward()
optimizer.step()
def AIModel(include, exclude):
num_epochs = 150
batch_size = 50
learning_rate = 1e-3
t1_data = pd.read_csv('./11/08.csv')
def AiModel():
body = request.get_json(force=True)
output = model(FloatTensor(body['foods'])).detach().numpy()
# result를 가공하는 작업
result = [1, 2, 3, 4]
return jsonify({'recommendation': result})