-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtrain_model.py
195 lines (169 loc) · 8.39 KB
/
train_model.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
from asi.model import AttentionSpatialInterpolationModel as asi
import json
class train:
def __init__(self, sigma, learning_rate, batch_size, num_neuron, num_layers, size_embedded,
num_nearest_geo, num_nearest_eucli, id_dataset, label, graph_label, num_nearest,
epochs, validation_split, early_stopping, optimier, **kwargs):
"""
:param sigma:
:param learning_rate:
:param batch_size:
:param num_neuron:
:param num_layers:
:param size_embedded:
:param num_nearest_geo:
:param num_nearest_eucli:
:param id_dataset:
:param label:
:param graph_label:
:param num_nearest:
:param epochs:
:param validation_split:
:param early_stopping:
:param optimier:
:param kwargs:
"""
self.NUM_NEAREST = num_nearest
self.SIGMA = sigma
self.LEARNING_RATE = learning_rate
self.BATCH_SIZE = batch_size
self.NUM_NEURON = num_neuron
self.NUM_LAYERS = num_layers
self.SIZE_EMBEDDED = size_embedded
self.NUM_NEAREST_GEO = num_nearest_geo
self.NUM_NEAREST_EUCLI = num_nearest_eucli
self.ID_DATASET = id_dataset
self.EPOCHS = epochs
self.OPTIMIZER = optimier
self.VALIDATION_SPLIT = validation_split
self.LABEL = label
self.EARLY_STOPPING = early_stopping
self.GRAPH_LABEL = graph_label
self.num_image_features = kwargs.get('num_image_features')
self.scale = kwargs.get('scale', False)
self.image_scale = kwargs.get('image_scale', True)
self.image_feature_extractor = kwargs.get('image_feature_extractor', 'VGG')
def __call__(self):
####################################### Model ##########################################
# build of the object
spatial = asi(id_dataset=self.ID_DATASET,
num_nearest=self.NUM_NEAREST,
early_stopping=self.EARLY_STOPPING,
num_image_features=self.num_image_features, scale=self.scale,
image_feature_extractor=self.image_feature_extractor, image_scale=self.image_scale)
# build of the model
model = spatial.build(sigma=[0, self.SIGMA],
optimizer=self.OPTIMIZER,
learning_rate=self.LEARNING_RATE,
num_layers=self.NUM_LAYERS,
num_neuron=self.NUM_NEURON,
size_embedded=self.SIZE_EMBEDDED,
graph_label=self.GRAPH_LABEL,
num_nearest_geo=self.NUM_NEAREST_GEO,
num_nearest_eucli=self.NUM_NEAREST_EUCLI,
num_image_features=self.num_image_features)
# save architecture image
spatial.architecture(model, 'architecture_'+self.LABEL)
# fitt of the model
weight, fit = spatial.train(model=model,
epochs=self.EPOCHS,
batch_size=self.BATCH_SIZE,
validation_split=self.VALIDATION_SPLIT,
label=self.LABEL,
num_nearest_geo=self.NUM_NEAREST_GEO,
num_nearest_eucli=self.NUM_NEAREST_EUCLI)
# prediction
result = spatial.predict_value(model=model,
weights=weight,
num_nearest_geo=self.NUM_NEAREST_GEO,
num_nearest_eucli=self.NUM_NEAREST_EUCLI)
############################ Feature Extraction ###########################################
DATA_TRAIN = ([spatial.X_train,
spatial.train_x_d[:, :self.NUM_NEAREST_GEO, :],
spatial.train_x_p[:, :self.NUM_NEAREST_EUCLI, :],
spatial.train_x_g[:, :self.NUM_NEAREST_GEO],
spatial.train_x_e[:, :self.NUM_NEAREST_EUCLI]
]
)
DATA_TEST = ([spatial.X_test,
spatial.test_x_d[:, :self.NUM_NEAREST_GEO, :],
spatial.test_x_p[:, :self.NUM_NEAREST_EUCLI, :],
spatial.test_x_g[:, :self.NUM_NEAREST_GEO],
spatial.test_x_e[:, :self.NUM_NEAREST_EUCLI]
]
)
if self.num_image_features is not None:
DATA_TRAIN.append(spatial.X_train_image)
DATA_TEST.append(spatial.X_test_image)
#Embedded
embedded_train = spatial.output_layer(model=model,
weight=weight,
layer='embedded',
data=DATA_TRAIN,
batch=self.BATCH_SIZE,
file_name=self.ID_DATASET + '_embedded_train')
embedded_test = spatial.output_layer(model=model,
weight=weight,
layer='embedded',
data=DATA_TEST,
batch=self.BATCH_SIZE,
file_name=self.ID_DATASET + '_embedded_test')
#Regression
predict_regression_train = spatial.output_layer(model=model,
weight=weight,
layer='main_output',
data=DATA_TRAIN,
batch=self.BATCH_SIZE,
file_name=self.ID_DATASET + '_predict_regression_train')
predict_regression_test = spatial.output_layer(model=model,
weight=weight,
layer='main_output',
data=DATA_TEST,
batch=self.BATCH_SIZE,
file_name=self.ID_DATASET + '_predict_regression_test')
return spatial, result, fit, embedded_train, embedded_test, predict_regression_train, predict_regression_test
if __name__ == "__main__":
# %matplotlib inline
# import sys
# sys.path.append("../../")
from matplotlib import rcParams
rcParams['figure.figsize'] = (8, 4)
rcParams['figure.dpi'] = 100
rcParams['font.size'] = 8
rcParams['font.family'] = 'sans-serif'
rcParams['axes.facecolor'] = '#ffffff'
rcParams['lines.linewidth'] = 2.0
hyperparameter={
"num_nearest":60,
"sigma":2,
"learning_rate":0.001,
"batch_size":250,
"num_neuron":60,
"num_layers":3,
"size_embedded":50,
"num_nearest_geo":45,
"num_nearest_eucli":45,
"id_dataset":'sp',
"epochs":450,
"optimier":'adam',
"validation_split": 0.1,
"label":'asi_sp_vgg_64',
"early_stopping": False,
"graph_label":'matrix',
"num_image_features": 64,
"scale": True,
"image_feature_extractor": 'vgg', # 'vgg'; num_image_features-512 , 'vit'; num_image_features-768, 'resnet101'; num_image_features-2048
"image_scale": True
}
spatial = train(**hyperparameter)
dataset, result, fit, embedded_train, embedded_test, predict_regression_train, predict_regression_test = spatial()
mae_test, rmse_test, mape_test, mae_train, rmse_train, mape_train = result
res = {
'mae_test': mae_test,
'rmse_test': rmse_test,
'mape_test': mape_test,
'mae_train': mae_train,
'rmse_train': rmse_train,
'mape_train': mape_train
}
print(json.dumps(res, indent=4))