-
Notifications
You must be signed in to change notification settings - Fork 2
/
dataset.py
60 lines (47 loc) · 2.03 KB
/
dataset.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
import pickle
import os
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image
from torch.utils.data import Dataset
class CIFAR10(Dataset):
def __init__(self,root,train=True,transforms=None):
self.root = root
self.transforms = transforms
self.split = train
self.data = []
self.targets = []
self.train_data = [file for file in os.listdir(root) if "data_batch" in file]
self.test_data = [file for file in os.listdir(root) if "test_batch" in file]
data_split = self.train_data if self.split else self.test_data
for files in data_split:
entry = self.extract(os.path.join(root,files))
self.data.append(entry["data"])
self.targets.extend(entry["labels"])
self.data = np.vstack(self.data).reshape(-1, 3, 32, 32)
self.data = self.data.transpose((0, 2, 3, 1))
self.load_meta()
def extract(self,filename):
with open(filename,"rb") as f:
batch_data = pickle.load(f,encoding="latin1")
return batch_data
def load_meta(self):
path = os.path.join(self.root,"batches.meta")
with open(path,"rb") as infile:
data = pickle.load(infile,encoding="latin1")
self.classes = data["label_names"]
self.classes_to_idx = {_class:i for i,_class in enumerate(self.classes)}
def plot(self,image,target=None):
if target is not None:
print(f"Target :{target} class :{self.classes[target]}")
plt.figure(figsize=(2,2))
plt.imshow(image.permute(1,2,0))
plt.show()
def __len__(self):
return len(self.data)
def __getitem__(self,idx):
image,target = self.data[idx],self.targets[idx]
image = Image.fromarray(image)
if self.transforms:
image = self.transforms(image)
return image,target