-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
52 lines (39 loc) · 1.58 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
import torch
import dill
import random
import numpy as np
from torch.utils.data import Dataset
class DrugRepurposing(Dataset):
def __init__(self, dill_path, transform=None):
'''
:param dill_path: dill path of a list of dicts containing at least 2 keys: x, y
'''
self.data = dill.load(open(dill_path, 'rb')) # list of dict containing at least the keys x, y
if transform is not None:
self.data = transform(self.data)
print('Dataset Loaded!')
def __len__(self):
return len(self.data)
def shuffle(self, seed=9):
random.seed = seed
random.shuffle(self.data)
def __getitem__(self, idx):
if isinstance(idx, str):
return self.getitem_from_key(idx=range(len(self.data)), key=idx)
if torch.is_tensor(idx):
idx = idx.tolist()
if isinstance(idx, list) or isinstance(idx, np.ndarray) or isinstance(idx, range):
return [self.data[i] for i in idx]
elif isinstance(idx, int):
return self.data[idx]['x'], self.data[idx]['y']
else:
raise ValueError("idx must be int or list, not ", type(idx))
def getitem_from_key(self, idx, key):
if torch.is_tensor(idx):
idx = idx.tolist()
if isinstance(idx, list) or isinstance(idx, np.ndarray) or isinstance(idx, range):
return [self.data[i][key] for i in idx]
elif isinstance(idx, int):
return self.data[idx][key]
else:
raise ValueError("idx must be int or list, not ", type(idx))