MHANet/data_process.py at main · SunshineGreeny/MHANet · GitHub

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
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
import math
import pandas as pd
from sklearn.preprocessing import scale
from scipy.interpolate import griddata
from scipy.io import loadmat
import numpy as np
import random

def read_prepared_data(args):
    data = []

    for l in range(len(args.ConType)):
        label = pd.read_csv(args.data_document_path + "/csv/" + args.name + args.ConType[l] + ".csv")
        target = []
        for k in range(args.trail_number):
            filename = args.data_document_path + "/" + args.ConType[l] + "/" + args.name + "Tra" + str(k + 1) + ".csv"
            data_pf = pd.read_csv(filename, header=None)
            eeg_data = data_pf.iloc[:, 2:] #KUL,DTU

            data.append(eeg_data)
            target.append(label.iloc[k, args.label_col])

    return data, target

def get_data_from_mat(mat_path):

        mat_eeg_data = []
        mat_event_data = []
        matstruct_contents = loadmat(mat_path)
        matstruct_contents = matstruct_contents['data']
        mat_event = matstruct_contents[0, 0]['event']['eeg'].item()
        mat_event_value = mat_event[0]['value']  # 1*60 1=male, 2=female
        mat_eeg = matstruct_contents[0, 0]['eeg']  # 60 trials 3200*66
        for i in range(mat_eeg.shape[1]):
            mat_eeg_data.append(mat_eeg[0, i])
            mat_event_data.append(mat_event_value[i][0][0])

        return mat_eeg_data, mat_event_data


def sliding_window(eeg_datas, labels, args, out_channels):
    window_size = args.window_length
    stride = int(window_size * (1 - args.overlap))

    train_eeg = []
    test_eeg = []
    train_label = []
    test_label = []

    for m in range(len(labels)):
        eeg = eeg_datas[m]
        label = labels[m]
        windows = []
        new_label = []
        for i in range(0, eeg.shape[0] - window_size + 1, stride):
            window = eeg[i:i+window_size, :]
            windows.append(window)
            new_label.append(label)
        train_eeg.append(np.array(windows)[:int(len(windows)*0.9)])
        test_eeg.append(np.array(windows)[int(len(windows)*0.9):])
        train_label.append(np.array(new_label)[:int(len(windows)*0.9)])
        test_label.append(np.array(new_label)[int(len(windows)*0.9):])

    train_eeg = np.stack(train_eeg, axis=0).reshape(-1, window_size, out_channels)
    test_eeg = np.stack(test_eeg, axis=0).reshape(-1, window_size, out_channels)
    train_label = np.stack(train_label, axis=0).reshape(-1, 1)
    test_label = np.stack(test_label, axis=0).reshape(-1, 1)

    return train_eeg, test_eeg, train_label, test_label

def new_sliding_window(eeg_datas, labels, args, out_channels):
    window_size = args.window_length
    stride = int(128 * (1 - args.overlap))

    train_eeg = []
    test_eeg = []
    train_label = []
    test_label = []

    for m in range(len(labels)):
        eeg = eeg_datas[m]
        label = labels[m]
        windows = []
        new_label = []
        for i in range(0, eeg.shape[0] - window_size + 1, stride):
            window = eeg[i:i+window_size, :]
            windows.append(window)
            new_label.append(label)
        train_eeg.append(np.array(windows)[:int(len(windows)*0.9)])
        test_eeg.append(np.array(windows)[int(len(windows)*0.9):])
        train_label.append(np.array(new_label)[:int(len(windows)*0.9)])
        test_label.append(np.array(new_label)[int(len(windows)*0.9):])

    train_eeg = np.stack(train_eeg, axis=0).reshape(-1, window_size, out_channels)
    test_eeg = np.stack(test_eeg, axis=0).reshape(-1, window_size, out_channels)
    train_label = np.stack(train_label, axis=0).reshape(-1, 1)
    test_label = np.stack(test_label, axis=0).reshape(-1, 1)

    return train_eeg, test_eeg, train_label, test_label

def sliding_window_csp(eeg_datas, labels, args, out_channels):
    window_size = args.window_length
    stride = int(window_size * (1 - args.overlap))

    eeg_set = []
    label_set = []

    for m in range(len(labels)): #labels 0-19
        eeg = eeg_datas[m]
        label = labels[m]
        windows = []
        new_label = []
        for i in range(0, eeg.shape[0] - window_size + 1, stride):
            window = eeg[i:i+window_size, :]
            windows.append(window)
            new_label.append(label)

        eeg_set.append(np.array(windows))
        label_set.append(np.array(new_label))

    eeg_set = np.stack(eeg_set, axis=0).reshape(-1, window_size, out_channels)
    label_set = np.stack(label_set, axis=0).reshape(-1, 1)

    return eeg_set, label_set

def within_data(eeg_datas, labels):
        train_datas = []
        test_datas = []
        train_labels = []
        test_labels = []

        for m in range(len(labels)): #labels 0-19
            eeg = eeg_datas[m]
            label = labels[m]

            train_datas.append(np.array(eeg)[:, :int(eeg.shape[1]*0.9)])
            test_datas.append(np.array(eeg)[:, int(eeg.shape[1]*0.9):])
            train_labels.append(np.array(label))
            test_labels.append(np.array(label))

        train_datas = np.stack(train_datas, axis=0)
        test_datas = np.stack(test_datas, axis=0)
        train_labels = np.stack(train_labels, axis=0)
        test_labels = np.stack(test_labels, axis=0)

        return train_datas, test_datas, train_labels, test_labels