forked from NoKi41/DMM-Net
-
Notifications
You must be signed in to change notification settings - Fork 0
/
UCRDataset.py
63 lines (46 loc) · 2.29 KB
/
UCRDataset.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
61
62
63
import pandas as pd
import torch
from torch.utils.data import Dataset
from pyts.image import GramianAngularField, MarkovTransitionField, RecurrencePlot
from Image_Transform import ImageTransformer
from torchvision import transforms
import numpy as np
import os
class UCRDataset(Dataset):
def __init__(self, root_path, file_path, lenth, image_size):
self.df = pd.read_csv(file_path, sep='\t', header=None)
self.data = self.df.iloc[:, 1:].values
self.labels = self.df.iloc[:, 0].values
self.lenth =lenth
self.image_size = image_size
self.image_transformer = ImageTransformer(lenth= self.lenth, image_size=image_size)
def __len__(self):
return len(self.data)
def __getitem__(self, idx):
data = self.data[idx]
label = self.labels[idx]-1
# 数据转换为图像表示
gasf_img, gadf_img, mtf_img, rp_img = self.image_transformer.fit_transform(data.reshape(1, -1))
#reshape
gasf_tensor = torch.from_numpy(gasf_img).float()
gadf_tensor = torch.from_numpy(gadf_img).float()
mtf_tensor = torch.from_numpy(mtf_img).float()
rp_tensor = torch.from_numpy(rp_img).float()
# print(data.shape,gasf_tensor.shape,gadf_tensor.shape,mtf_tensor.shape,rp_tensor.shape)
img_input = torch.cat((gasf_tensor, gadf_tensor, mtf_tensor, rp_tensor), dim=0)
return torch.tensor(data).float(), img_input.float(), torch.tensor(label).long()
class StandardScaler():
def __init__(self):
self.mean = 0.
self.std = 1.
def fit(self, data):
self.mean = data.mean(0)
self.std = data.std(0)
def transform(self, data):
mean = torch.from_numpy(self.mean).type_as(data).to(data.device) if torch.is_tensor(data) else self.mean
std = torch.from_numpy(self.std).type_as(data).to(data.device) if torch.is_tensor(data) else self.std
return (data - mean) / std
def inverse_transform(self, data):
mean = torch.from_numpy(self.mean).type_as(data).to(data.device) if torch.is_tensor(data) else self.mean
std = torch.from_numpy(self.std).type_as(data).to(data.device) if torch.is_tensor(data) else self.std
return (data * std) + mean