-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathencoding_utils.py
207 lines (196 loc) · 10.1 KB
/
encoding_utils.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
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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
import numpy as np
def ord2ohe(X_ord, dataset):
continuous_availability = dataset['continuous_availability']
discrete_availability = dataset['discrete_availability']
ohe_feature_encoder = dataset['ohe_feature_encoder']
len_continuous_ord = dataset['len_continuous_ord']
len_discrete_ord = dataset['len_discrete_ord']
if X_ord.shape.__len__() == 1:
if continuous_availability and discrete_availability:
X_continuous = X_ord[len_continuous_ord[0]:len_continuous_ord[1]]
X_discrete = X_ord[len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ohe_feature_encoder.transform(X_discrete.reshape(1,-1)).ravel()
X_ohe = np.r_[X_continuous, X_discrete]
return X_ohe
elif continuous_availability:
X_continuous = X_ord[len_continuous_ord[0]:len_continuous_ord[1]]
X_ohe = X_continuous.copy()
return X_ohe
elif discrete_availability:
X_discrete = X_ord[len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ohe_feature_encoder.transform(X_discrete.reshape(1, -1)).ravel()
X_ohe = X_discrete.copy()
return X_ohe
else:
if continuous_availability and discrete_availability:
X_continuous = X_ord[:,len_continuous_ord[0]:len_continuous_ord[1]]
X_discrete = X_ord[:,len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ohe_feature_encoder.transform(X_discrete)
X_ohe = np.c_[X_continuous,X_discrete]
return X_ohe
elif continuous_availability:
X_continuous = X_ord[:,len_continuous_ord[0]:len_continuous_ord[1]]
X_ohe = X_continuous.copy()
return X_ohe
elif discrete_availability:
X_discrete = X_ord[:,len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ohe_feature_encoder.transform(X_discrete)
X_ohe = X_discrete.copy()
return X_ohe
def ohe2ord(X_ohe, dataset):
continuous_availability = dataset['continuous_availability']
discrete_availability = dataset['discrete_availability']
ohe_feature_encoder = dataset['ohe_feature_encoder']
len_continuous_ohe = dataset['len_continuous_ohe']
len_discrete_ohe = dataset['len_discrete_ohe']
if X_ohe.shape.__len__() == 1:
if continuous_availability and discrete_availability:
X_continuous = X_ohe[len_continuous_ohe[0]:len_continuous_ohe[1]]
X_discrete = X_ohe[len_discrete_ohe[0]:len_discrete_ohe[1]]
X_discrete = ohe_feature_encoder.inverse_transform(X_discrete.reshape(1,-1)).ravel()
X_ord = np.r_[X_continuous, X_discrete]
return X_ord
elif continuous_availability:
X_continuous = X_ohe[len_continuous_ohe[0]:len_continuous_ohe[1]]
X_ord = X_continuous.copy()
return X_ord
elif discrete_availability:
X_discrete = X_ohe[len_discrete_ohe[0]:len_discrete_ohe[1]]
X_discrete = ohe_feature_encoder.inverse_transform(X_discrete.reshape(1,-1)).ravel()
X_ord = X_discrete.copy()
return X_ord
else:
if continuous_availability and discrete_availability:
X_continuous = X_ohe[:,len_continuous_ohe[0]:len_continuous_ohe[1]]
X_discrete = X_ohe[:,len_discrete_ohe[0]:len_discrete_ohe[1]]
X_discrete = ohe_feature_encoder.inverse_transform(X_discrete)
X_ord = np.c_[X_continuous,X_discrete]
return X_ord
elif continuous_availability:
X_continuous = X_ohe[:,len_continuous_ohe[0]:len_continuous_ohe[1]]
X_ord = X_continuous.copy()
return X_ord
elif discrete_availability:
X_discrete = X_ohe[:,len_discrete_ohe[0]:len_discrete_ohe[1]]
X_discrete = ohe_feature_encoder.inverse_transform(X_discrete)
X_ord = X_discrete.copy()
return X_ord
def org2ord(X_org, dataset):
continuous_availability = dataset['continuous_availability']
discrete_availability = dataset['discrete_availability']
num_feature_scaler = dataset['num_feature_scaler']
ord_feature_encoder = dataset['ord_feature_encoder']
len_continuous_org = dataset['len_continuous_org']
len_discrete_org = dataset['len_discrete_org']
if X_org.shape.__len__() == 1:
if continuous_availability and discrete_availability:
X_continuous = X_org[len_continuous_org[0]:len_continuous_org[1]]
X_continuous = num_feature_scaler.transform(X_continuous.reshape(1, -1)).ravel()
X_discrete = X_org[len_discrete_org[0]:len_discrete_org[1]]
X_discrete = ord_feature_encoder.transform(X_discrete.reshape(1,-1)).ravel()
X_ord = np.r_[X_continuous, X_discrete]
return X_ord
elif continuous_availability:
X_continuous = X_org[len_continuous_org[0]:len_continuous_org[1]]
X_continuous = num_feature_scaler.transform(X_continuous.reshape(1, -1)).ravel()
X_ord = X_continuous.copy()
return X_ord
elif discrete_availability:
X_discrete = X_org[len_discrete_org[0]:len_discrete_org[1]]
X_discrete = ord_feature_encoder.transform(X_discrete.reshape(1,-1)).ravel()
X_ord = X_discrete.copy()
return X_ord
else:
if continuous_availability and discrete_availability:
X_continuous = X_org[:,len_continuous_org[0]:len_continuous_org[1]]
X_continuous = num_feature_scaler.transform(X_continuous)
X_discrete = X_org[:,len_discrete_org[0]:len_discrete_org[1]]
X_discrete = ord_feature_encoder.transform(X_discrete)
X_ord = np.c_[X_continuous,X_discrete]
return X_ord
elif continuous_availability:
X_continuous = X_org[:,len_continuous_org[0]:len_continuous_org[1]]
X_continuous = num_feature_scaler.transform(X_continuous)
X_ord = X_continuous.copy()
return X_ord
elif discrete_availability:
X_discrete = X_org[:,len_discrete_org[0]:len_discrete_org[1]]
X_discrete = ord_feature_encoder.transform(X_discrete)
X_ord = X_discrete.copy()
return X_ord
def ord2org(X_ord, dataset):
continuous_availability = dataset['continuous_availability']
discrete_availability = dataset['discrete_availability']
num_feature_scaler = dataset['num_feature_scaler']
ord_feature_encoder = dataset['ord_feature_encoder']
len_continuous_ord = dataset['len_continuous_ord']
len_discrete_ord = dataset['len_discrete_ord']
continuous_precision = dataset['continuous_precision']
if X_ord.shape.__len__() == 1:
if continuous_availability and discrete_availability:
X_continuous = X_ord[len_continuous_ord[0]:len_continuous_ord[1]]
X_continuous = num_feature_scaler.inverse_transform(X_continuous.reshape(1, -1)).ravel()
for f, dec in enumerate(continuous_precision):
X_continuous[f] = np.around(X_continuous[f], decimals=dec)
X_discrete = X_ord[len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ord_feature_encoder.inverse_transform(X_discrete.reshape(1,-1)).ravel()
X_org = np.r_[X_continuous, X_discrete]
return X_org
elif continuous_availability:
X_continuous = X_ord[len_continuous_ord[0]:len_continuous_ord[1]]
X_continuous = num_feature_scaler.inverse_transform(X_continuous.reshape(1, -1)).ravel()
for f, dec in enumerate(continuous_precision):
X_continuous[f] = np.around(X_continuous[f], decimals=dec)
X_org = X_continuous.copy()
return X_org
elif discrete_availability:
X_discrete = X_ord[len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ord_feature_encoder.inverse_transform(X_discrete.reshape(1,-1)).ravel()
X_org = X_discrete.copy()
return X_org
else:
if continuous_availability and discrete_availability:
X_continuous = X_ord[:,len_continuous_ord[0]:len_continuous_ord[1]]
X_continuous = num_feature_scaler.inverse_transform(X_continuous)
for f, dec in enumerate(continuous_precision):
X_continuous[:,f] = np.around(X_continuous[:,f], decimals=dec)
X_discrete = X_ord[:,len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ord_feature_encoder.inverse_transform(X_discrete)
X_org = np.c_[X_continuous,X_discrete]
return X_org
elif continuous_availability:
X_continuous = X_ord[:,len_continuous_ord[0]:len_continuous_ord[1]]
X_continuous = num_feature_scaler.inverse_transform(X_continuous)
for f, dec in enumerate(continuous_precision):
X_continuous[:,f] = np.around(X_continuous[:,f], decimals=dec)
X_org = X_continuous.copy()
return X_org
elif discrete_availability:
X_discrete = X_ord[:,len_discrete_ord[0]:len_discrete_ord[1]]
X_discrete = ord_feature_encoder.inverse_transform(X_discrete)
X_org = X_discrete.copy()
return X_org
def ord2theta(X_ord, featureScaler):
if X_ord.shape.__len__() == 1:
X_theta = featureScaler.transform(X_ord.reshape(1,-1)).ravel()
return X_theta
else:
X_theta = featureScaler.transform(X_ord)
return X_theta
def theta2ord(X_theta, featureScaler, dataset):
discrete_availability = dataset['discrete_availability']
discrete_indices = dataset['discrete_indices']
if X_theta.shape.__len__() == 1:
X_ord = featureScaler.inverse_transform(X_theta.reshape(1,-1)).ravel()
if discrete_availability:
X_ord[discrete_indices] = np.rint(X_ord[discrete_indices])
return X_ord
else:
X_ord = featureScaler.inverse_transform(X_theta)
if discrete_availability:
X_ord[:, discrete_indices] = np.rint(X_ord[:, discrete_indices])
return X_ord
def theta2org(X_theta, featureScaler, dataset):
X_ord = theta2ord(X_theta, featureScaler, dataset)
X_org = ord2org(X_ord, dataset)
return X_org