add targets FAA getBraTS_4

@@ -5,3 +5,4 @@ torch
 hyperopt
 hyperopt
 pillow==6.2.1
 pillow==6.2.1
 natsort
 natsort
+fire
\ No newline at end of file

@@ -22,9 +22,9 @@ from sklearn.model_selection import KFold
 
 
 from networks import basenet
 from networks import basenet
 
 
-
-TRAIN_DATASET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/train_frame'
-VAL_DATASET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/val_frame'
+DATASET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/train_frame/'
+TRAIN_DATASET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/train_frame/'
+VAL_DATASET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/val_frame/'
 
 
 TRAIN_TARGET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/train_frame.csv'
 TRAIN_TARGET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/train_frame.csv'
 VAL_TARGET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/val_frame.csv'
 VAL_TARGET_PATH = '/content/drive/My Drive/CD2 Project/data/BraTS_Training/val_frame.csv'
@@ -35,7 +35,10 @@ current_epoch = 0
 def split_dataset(args, dataset, k):
 def split_dataset(args, dataset, k):
     # load dataset
     # load dataset
     X = list(range(len(dataset)))
     X = list(range(len(dataset)))
-    #Y = dataset.targets
+    Y = dataset.targets
+    #Y = [0]* len(X)
+
+    #print("X:\n", type(X), np.shape(X), '\n', X, '\n')
 
 
     # split to k-fold
     # split to k-fold
     # assert len(X) == len(Y)
     # assert len(X) == len(Y)
@@ -43,26 +46,49 @@ def split_dataset(args, dataset, k):
     def _it_to_list(_it):
     def _it_to_list(_it):
         return list(zip(*list(_it)))
         return list(zip(*list(_it)))
 
 
-    # sss = StratifiedShuffleSplit(n_splits=k, random_state=args.seed, test_size=0.1)
-    # Dm_indexes, Da_indexes = _it_to_list(sss.split(X, Y))
+    sss = StratifiedShuffleSplit(n_splits=k, random_state=args.seed, test_size=0.1)
+    Dm_indexes, Da_indexes = _it_to_list(sss.split(X, Y))
+
+    # print(type(Dm_indexes), np.shape(Dm_indexes))
+    # print("DM\n", len(Dm_indexes), Dm_indexes, "\nDA\n", len(Da_indexes),Da_indexes)
+    
+
+    return Dm_indexes, Da_indexes
+
+def split_dataset2222(args, dataset, k):
+    # load dataset
+    X = list(range(len(dataset)))
+    
+    # split to k-fold
+    #assert len(X) == len(Y)
 
 
-    x_train = []
-    x_test = []
+    def _it_to_list(_it):
+        return list(zip(*list(_it)))
 
 
+    x_train = ()
+    x_test = ()
 
 
     for i in range(k):
     for i in range(k):
-        xtr, xte = train_test_split(X, random_state=args.seed, test_size=0.1)
-        x_train.append(xtr)
-        x_test.append(xte)
+        #xtr, xte = train_test_split(X, random_state=args.seed, test_size=0.1)
+        xtr, xte = train_test_split(X, random_state=None, test_size=0.1)
+        x_train.append(np.array(xtr))
+        x_test.append(np.array(xte))
 
 
-    #kf = KFold(n_splits=k, random_state=args.seed, test)
-    #kf.split(x_train)
+    y_train = np.array([0]* len(x_train))
+    y_test = np.array([0]* len(x_test))
 
 
-    Dm_indexes, Da_indexes = np.array(x_train), np.array(x_test)
+    x_train = tuple(x_train)
+    x_test = tuple(x_test)
 
 
+    trainset = (zip(x_train, y_train),)
+    testset = (zip(x_test, y_test),)
 
 
-    return Dm_indexes, Da_indexes
+    Dm_indexes, Da_indexes = trainset, testset
 
 
+    print(type(Dm_indexes), np.shape(Dm_indexes))
+    print("DM\n", np.shape(Dm_indexes), Dm_indexes, "\nDA\n", np.shape(Da_indexes), Da_indexes)
+
+    return Dm_indexes, Da_indexes
 
 
 def concat_image_features(image, features, max_features=3):
 def concat_image_features(image, features, max_features=3):
     _, h, w = image.shape
     _, h, w = image.shape
@@ -169,22 +195,24 @@ def select_scheduler(args, optimizer):
 
 
 
 
 class CustomDataset(Dataset): 
 class CustomDataset(Dataset): 
-  def __init__(self, path, target_path, transform = None):
+  def __init__(self, path, transform = None):
       self.path = path
       self.path = path
       self.transform = transform
       self.transform = transform
       #self.imgpath = glob.glob(path + '/*.png'
       #self.imgpath = glob.glob(path + '/*.png'
-      #self.img = np.expand_dims(np.load(glob.glob(path + '/*.png'), axis = 3)
       self.imgs = natsorted(os.listdir(path))
       self.imgs = natsorted(os.listdir(path))
       self.len = len(self.imgs)
       self.len = len(self.imgs)
       #self.len = self.img.shape[0]
       #self.len = self.img.shape[0]
-      self.targets = pd.read_csv(target_path, header = None)
+      self.targets = [0]* self.len
 
 
   def __len__(self):
   def __len__(self):
       return self.len
       return self.len
 
 
   def __getitem__(self, idx):
   def __getitem__(self, idx):
+    #    print("\n\nIDX: ", idx, '\n', type(idx), '\n') 
+    #    print("\n\nimgs[idx]: ", self.imgs[idx], '\n', type(self.imgs[idx]), '\n') 
        #img, targets = self.img[idx], self.targets[idx]
        #img, targets = self.img[idx], self.targets[idx]
        img_loc = os.path.join(self.path, self.imgs[idx])
        img_loc = os.path.join(self.path, self.imgs[idx])
+       targets = self.targets[idx]
        #img = self.img[idx]
        #img = self.img[idx]
        image = Image.open(img_loc)
        image = Image.open(img_loc)
        
        
@@ -192,7 +220,7 @@ class CustomDataset(Dataset):
             #img = self.transform(img)
             #img = self.transform(img)
             tensor_image = self.transform(image)
             tensor_image = self.transform(image)
        #return img, targets
        #return img, targets
-       return tensor_image
+       return tensor_image, targets
 
 
 def get_dataset(args, transform, split='train'):
 def get_dataset(args, transform, split='train'):
     assert split in ['train', 'val', 'test', 'trainval']
     assert split in ['train', 'val', 'test', 'trainval']
@@ -224,9 +252,9 @@ def get_dataset(args, transform, split='train'):
 
 
     elif args.dataset == 'BraTS':
     elif args.dataset == 'BraTS':
       if split in ['train']:
       if split in ['train']:
-        dataset = CustomDataset(TRAIN_DATASET_PATH, TRAIN_TARGET_PATH, transform=transform)
+        dataset = CustomDataset(TRAIN_DATASET_PATH, transform=transform)
       else:
       else:
-        dataset = CustomDataset(VAL_DATASET_PATH, VAL_TARGET_PATH, transform=transform)
+        dataset = CustomDataset(VAL_DATASET_PATH, transform=transform)
       
       
 
 
     else:
     else:
@@ -250,6 +278,7 @@ def get_inf_dataloader(args, dataset):
 
 
     while True:
     while True:
         try:
         try:
+            #print("batch=dataloader:\n", batch, '\n')
             batch = next(data_loader)
             batch = next(data_loader)
 
 
         except StopIteration:
         except StopIteration:
@@ -334,8 +363,7 @@ def get_valid_transform(args, model):
 
 
 def train_step(args, model, optimizer, scheduler, criterion, batch, step, writer, device=None):
 def train_step(args, model, optimizer, scheduler, criterion, batch, step, writer, device=None):
     model.train()
     model.train()
-    print('\nBatch\n', batch)
-    print('\nBatch size\n', batch.size())
+    #print('\nBatch\n', batch)
     images, target = batch
     images, target = batch
 
 
     if device:
     if device: