LSTM_gait_model/data_utils_modified.py at master · margokhokhlova/LSTM_gait_model · 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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
import numpy as np
import os
import glob
import random


def loadfromfolder():
    ROOT ='data/best'
# 	"Function goes through the data and returns file  names and parameters neeeded for the NN in the next step"
    names = []
    T= []
    Cov_dim = []
    for filename in os.listdir(ROOT):
        print(filename)  # name composition is like angles_cycles_frames_06t_8f_angles_22persons
        t =  int(filename[21:23])
        #print(T)
        mat_dim = int(filename[25:27])
        #print(mat_dim)
        T.append(t)
        Cov_dim.append(mat_dim)
        names.append(filename)
    return names, T, Cov_dim


def Random_Selection_train_test (X_all, N):
    " This fucntion takes return train and test samples for the LSTM: X dataset, N - number of training samples (i.e persons)"
    Persons = X_all[:,1]
    uniquePersons = np.unique(Persons)
    random.shuffle(uniquePersons)
    num_dim = np.shape(X_all)[1]
    X_train = np.array([], dtype=np.int64).reshape(0,num_dim)
    Y_train = np.array([])
    X_test = np.array([])
    Y_test = np.array([])
    Persons_train = np.array([])
    Persons_test = np.array([])
    #print(X_train.shape)
    for Person in range(0,N):
        indexes = np.where(X_all[:,1]==uniquePersons[Person])
        #print(X_all[[indexes], :][0,0,:,:].shape)
        Person_X_val = X_all[[indexes], :][0,0,:,:]
        X_train= np.append(X_train,Person_X_val)
        label =  np.full((indexes[0].size),uniquePersons[Person])
        Y_train = np.append(Y_train, X_all[[indexes], 0])
        Persons_train = np.append(Persons_train, label, axis = 0)
    X_train=X_train.reshape(Y_train.size, num_dim)
    for Person in range(N,uniquePersons.size):
        indexes = np.where(X_all[:,1]==uniquePersons[Person])
        X_test = np.append(X_test, X_all[[indexes], :][0,0,:,:])
        label =  np.full((indexes[0].size),uniquePersons[Person])
        Y_test =  np.append(Y_test, X_all[[indexes], 0])
        Persons_test = np.append(Persons_test, label, axis = 0)
    X_test=X_test.reshape(Y_test.size, num_dim)
    print('Persons for train: ')
    print(np.unique(Persons_train))
    print('Persons for test: ')
    print(np.unique(Persons_test))
    return X_train, Y_train, X_test, Y_test, Persons_train


def Random_Selection_train_val_test (X_all, N):
#    import random
    " This fucntion takes return train, val and test samples for the LSTM: X dataset, N - number of training samples (i.e persons)"
    Persons = X_all[:,1]
    uniquePersons = np.unique(Persons)
    random.shuffle(uniquePersons)
    num_dim = np.shape(X_all)[1]
    X_train = np.array([], dtype=np.int64).reshape(0,num_dim)
    Y_train = np.array([])
    X_test = np.array([])
    Y_test = np.array([])
    X_val = np.array([])
    Y_val = np.array([])
    Persons_train = np.array([])
    Persons_val = np.array([])
    Persons_test = np.array([])

    train_N =int(np.ceil(N*2/3))

    for Person in range(0,train_N):
        indexes = np.where(X_all[:,1]==uniquePersons[Person])
        #print(X_all[[indexes], :][0,0,:,:].shape)
        Person_X_val = X_all[[indexes], :][0,0,:,:]
        X_train= np.append(X_train,Person_X_val)
        label =  np.full((indexes[0].size),uniquePersons[Person])
        Y_train = np.append(Y_train, X_all[[indexes], 0])
        Persons_train = np.append(Persons_train, label, axis = 0)
    X_train=X_train.reshape(Y_train.size, num_dim)


    for Person in range(train_N,N):
        indexes = np.where(X_all[:,1]==uniquePersons[Person])
        #print(X_all[[indexes], :][0,0,:,:].shape)
        Person_X_val = X_all[[indexes], :][0,0,:,:]
        X_val= np.append(X_val,Person_X_val)
        label =  np.full((indexes[0].size),uniquePersons[Person])
        Y_val = np.append(Y_val, X_all[[indexes], 0])
        Persons_val = np.append(Persons_val, label, axis = 0)
    X_val=X_val.reshape(Y_val.size, num_dim)

    for Person in range(N,uniquePersons.size):
        indexes = np.where(X_all[:,1]==uniquePersons[Person])
        X_test = np.append(X_test, X_all[[indexes], :][0,0,:,:])
        label =  np.full((indexes[0].size),uniquePersons[Person])
        Y_test =  np.append(Y_test, X_all[[indexes], 0])
        Persons_test = np.append(Persons_test, label, axis = 0)
    X_test=X_test.reshape(Y_test.size, num_dim)
    print('Persons for train: ')
    print(np.unique(Persons_train))
    returned_train = np.unique(Persons_train)
    print('Persons for val: ')
    print(np.unique(Persons_val))
    returned_val = np.unique(Persons_val)
    print('Persons for test: ')
    print(np.unique(Persons_test))
    return X_train, Y_train, X_test, Y_test, X_val, Y_val, returned_train, returned_val


def Selection_train_test(X_all, Persons_test):
#    import random
    " This fucntion return train and set samples, taking the specified persons for testing"
    Persons = X_all[:,1]
    uniquePersons = np.unique(Persons)
    num_dim = np.shape(X_all)[1]
    X_train = np.array([], dtype=np.int64).reshape(0,num_dim)
    Y_train = np.array([])
    X_test = np.array([])
    Y_test = np.array([])
    Persons_for_training = np.array([]) # for display
    Persons_for_testing = np.array([])
    ind2remove = Persons_test-1
    mask = np.ones(len(uniquePersons), dtype=bool)
    mask[ind2remove] = False
    Persons_train = uniquePersons[mask] # actually, removes all test persons from the list
    train_N = Persons_train.size


    for Person in range(0,train_N):
        indexes = np.where(X_all[:,1]==Persons_train[Person])
        Person_X_val = X_all[[indexes], :][0,0,:,:]
        X_train= np.append(X_train,Person_X_val)
        label =  np.full((indexes[0].size),Persons_train[Person])
        Y_train = np.append(Y_train, X_all[[indexes], 0])
        Persons_for_training = np.append(Persons_train, label, axis = 0)
    X_train=X_train.reshape(Y_train.size, num_dim)


    for Person in range(0,Persons_test.size):
        indexes = np.where(X_all[:,1]==Persons_test[Person])
        X_test = np.append(X_test, X_all[[indexes], :][0,0,:,:])
        label =  np.full((indexes[0].size),Persons_test[Person])
        Y_test =  np.append(Y_test, X_all[[indexes], 0])
        Persons_for_testing = np.append(Persons_test, label, axis = 0)
    X_test=X_test.reshape(Y_test.size, num_dim)

    print('Persons for train: ')
    print(np.unique(Persons_for_training))
    print('Persons for test: ')
    print(np.unique(Persons_for_testing))
    return X_train, Y_train, X_test, Y_test