Add files via upload

dataset_mimic.py

@@ -0,0 +1,47 @@
+import pickle
+import os
+from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.optim as optim
+import json
+import numpy as np
+import time
+import copy
+import torchvision.transforms as transforms
+import cv2
+from PIL import Image
+
+class Dataset(torch.utils.data.Dataset):
+    #'Characterizes a dataset for PyTorch'
+    def __init__(self, list_IDs, labels, groups, transform=None):
+        'Initialization'
+        self.groups = groups
+        self.labels = labels
+        self.list_IDs = list_IDs
+        self.transform = transform
+
+    def __len__(self):
+       # 'Denotes the total number of samples'
+        return len(self.list_IDs)
+
+    def __getitem__(self, index):
+       # 'Generates one sample of data'
+        # Select sample
+        ID = self.list_IDs[index]
+
+        X = Image.open(ID).convert('RGB')
+#         X = Image.open('/prj0129/mil4012/glaucoma/NIH-chest-x-ray/CXR8/images/images/' + ID).convert('RGB')
+        # X = cv2.imread('/prj0129/mil4012/glaucoma/NIH-chest-x-ray/CXR8/images/images/' + ID)
+        # X = cv2.resize(X,(224,224))
+        if self.transform:
+             X = self.transform(X)
+
+        y = self.labels[index]
+        group = self.groups[index]
+
+
+        # return X, torch.tensor(y),torch.tensor(group)
+        # return X, torch.tensor(y),group
+        return X, torch.FloatTensor(y),group