-
Notifications
You must be signed in to change notification settings - Fork 0
/
004_yolanda.py
176 lines (162 loc) · 7.98 KB
/
004_yolanda.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
import argparse
import os
from joblib import Parallel, delayed
import numpy as np
import autosklearn
import autosklearn.data
import autosklearn.data.competition_data_manager
from autosklearn.pipeline.regression import SimpleRegressionPipeline
parser = argparse.ArgumentParser()
parser.add_argument('input')
parser.add_argument('output')
args = parser.parse_args()
input = args.input
dataset = 'yolanda'
output = args.output
path = os.path.join(input, dataset)
D = autosklearn.data.competition_data_manager.CompetitionDataManager(path)
X = D.data['X_train']
y = D.data['Y_train']
X_valid = D.data['X_valid']
X_test = D.data['X_test']
# Use this version of lasagne commit of the lasagne master branch:
# 24c9ed2ffc25504c3b0df4598afb1e63fdd59eee
# https://github.com/Lasagne/Lasagne/commit/24c9ed2ffc25504c3b0df4598afb1e63fdd59eee
# Copy the file RegDeepNet into autosklearn.pipeline.components.regression
# Copy the file FeedForwardNet into autosklearn.pipeline.implementations
choices = \
[(0.360000, SimpleRegressionPipeline(configuration={
'imputation:strategy': 'mean',
'one_hot_encoding:minimum_fraction': 0.049682918006307676,
'one_hot_encoding:use_minimum_fraction': 'True',
'preprocessor:__choice__': 'no_preprocessing',
'regressor:RegDeepNet:activation': 'tanh',
'regressor:RegDeepNet:batch_size': 1865,
'regressor:RegDeepNet:dropout_layer_1': 0.017462492577406473,
'regressor:RegDeepNet:dropout_layer_2': 0.048354205627225436,
'regressor:RegDeepNet:dropout_output': 0.00962149073006804,
'regressor:RegDeepNet:lambda2': 1.0282444549550921e-05,
'regressor:RegDeepNet:learning_rate': 0.001,
'regressor:RegDeepNet:num_layers': 'd',
'regressor:RegDeepNet:num_units_layer_1': 2615,
'regressor:RegDeepNet:num_units_layer_2': 252,
'regressor:RegDeepNet:number_updates': 3225,
'regressor:RegDeepNet:solver': 'smorm3s',
'regressor:RegDeepNet:std_layer_1': 0.006861129306844183,
'regressor:RegDeepNet:std_layer_2': 0.002395977520245193,
'regressor:__choice__': 'RegDeepNet',
'rescaling:__choice__': 'standardize'})),
(0.320000, SimpleRegressionPipeline(configuration={
'imputation:strategy': 'mean',
'one_hot_encoding:minimum_fraction': 0.05112532429613385,
'one_hot_encoding:use_minimum_fraction': 'True',
'preprocessor:__choice__': 'no_preprocessing',
'regressor:RegDeepNet:activation': 'sigmoid',
'regressor:RegDeepNet:batch_size': 1840,
'regressor:RegDeepNet:dropout_layer_1': 0.15186663743978646,
'regressor:RegDeepNet:dropout_layer_2': 0.11387781420379316,
'regressor:RegDeepNet:dropout_layer_3': 0.19220971946536616,
'regressor:RegDeepNet:dropout_output': 0.5509953660515314,
'regressor:RegDeepNet:lambda2': 2.3655442216865217e-06,
'regressor:RegDeepNet:learning_rate': 0.1,
'regressor:RegDeepNet:num_layers': 'e',
'regressor:RegDeepNet:num_units_layer_1': 173,
'regressor:RegDeepNet:num_units_layer_2': 690,
'regressor:RegDeepNet:num_units_layer_3': 2761,
'regressor:RegDeepNet:number_updates': 4173,
'regressor:RegDeepNet:solver': 'smorm3s',
'regressor:RegDeepNet:std_layer_1': 0.006483588902887654,
'regressor:RegDeepNet:std_layer_2': 0.006696161430555593,
'regressor:RegDeepNet:std_layer_3': 0.0030798462419321746,
'regressor:__choice__': 'RegDeepNet',
'rescaling:__choice__': 'standardize'})),
(0.160000, SimpleRegressionPipeline(configuration={
'imputation:strategy': 'mean',
'one_hot_encoding:minimum_fraction': 0.00044746581915706805,
'one_hot_encoding:use_minimum_fraction': 'True',
'preprocessor:__choice__': 'no_preprocessing',
'regressor:RegDeepNet:activation': 'tanh',
'regressor:RegDeepNet:batch_size': 1867,
'regressor:RegDeepNet:dropout_layer_1': 0.0044842379741719856,
'regressor:RegDeepNet:dropout_output': 0.029970881815609602,
'regressor:RegDeepNet:lambda2': 3.922344043854585e-05,
'regressor:RegDeepNet:learning_rate': 0.001,
'regressor:RegDeepNet:num_layers': 'c',
'regressor:RegDeepNet:num_units_layer_1': 2775,
'regressor:RegDeepNet:number_updates': 4672,
'regressor:RegDeepNet:solver': 'smorm3s',
'regressor:RegDeepNet:std_layer_1': 0.0011091871005401157,
'regressor:__choice__': 'RegDeepNet',
'rescaling:__choice__': 'standardize'})),
(0.100000, SimpleRegressionPipeline(configuration={
'imputation:strategy': 'mean',
'one_hot_encoding:minimum_fraction': 0.0006151267694526832,
'one_hot_encoding:use_minimum_fraction': 'True',
'preprocessor:__choice__': 'no_preprocessing',
'regressor:RegDeepNet:activation': 'tanh',
'regressor:RegDeepNet:batch_size': 1293,
'regressor:RegDeepNet:dropout_layer_1': 0.024322298790122678,
'regressor:RegDeepNet:dropout_layer_2': 0.4831886801640319,
'regressor:RegDeepNet:dropout_layer_3': 0.7303058944461246,
'regressor:RegDeepNet:dropout_output': 0.43112081941910074,
'regressor:RegDeepNet:lambda2': 4.561723820100022e-06,
'regressor:RegDeepNet:learning_rate': 0.001,
'regressor:RegDeepNet:num_layers': 'e',
'regressor:RegDeepNet:num_units_layer_1': 2999,
'regressor:RegDeepNet:num_units_layer_2': 1630,
'regressor:RegDeepNet:num_units_layer_3': 897,
'regressor:RegDeepNet:number_updates': 4471,
'regressor:RegDeepNet:solver': 'smorm3s',
'regressor:RegDeepNet:std_layer_1': 0.0013646791717249367,
'regressor:RegDeepNet:std_layer_2': 0.012431732856634247,
'regressor:RegDeepNet:std_layer_3': 0.002351992156794049,
'regressor:__choice__': 'RegDeepNet',
'rescaling:__choice__': 'standardize'})),
(0.060000, SimpleRegressionPipeline(configuration={
'imputation:strategy': 'mean',
'one_hot_encoding:minimum_fraction': 0.006283026157824821,
'one_hot_encoding:use_minimum_fraction': 'True',
'preprocessor:__choice__': 'no_preprocessing',
'regressor:RegDeepNet:activation': 'tanh',
'regressor:RegDeepNet:batch_size': 1802,
'regressor:RegDeepNet:dropout_layer_1': 0.01257793094940521,
'regressor:RegDeepNet:dropout_output': 0.023821950297696383,
'regressor:RegDeepNet:lambda2': 8.078248563082777e-05,
'regressor:RegDeepNet:learning_rate': 0.001,
'regressor:RegDeepNet:num_layers': 'c',
'regressor:RegDeepNet:num_units_layer_1': 3293,
'regressor:RegDeepNet:number_updates': 4842,
'regressor:RegDeepNet:solver': 'smorm3s',
'regressor:RegDeepNet:std_layer_1': 0.001130906938022124,
'regressor:__choice__': 'RegDeepNet',
'rescaling:__choice__': 'standardize'})),
]
targets = []
predictions = []
predictions_valid = []
predictions_test = []
def fit_and_predict(estimator, weight, X, y):
try:
estimator.fit(X.copy(), y.copy())
pv = estimator.predict(X_valid.copy()) * weight
pt = estimator.predict(X_test.copy()) * weight
except Exception as e:
print(e)
print(estimator.configuration)
pv = None
pt = None
return pv, pt
# Make predictions and weight them
all_predictions = Parallel(n_jobs=-1)(delayed(fit_and_predict) \
(estimator, weight, X, y) for
weight, estimator in choices)
for pv, pt in all_predictions:
predictions_valid.append(pv)
predictions_test.append(pt)
# Output the predictions
for name, predictions in [('valid', predictions_valid),
('test', predictions_test)]:
predictions = np.array(predictions)
predictions = np.sum(predictions, axis=0).astype(np.float32)
filepath = os.path.join(output, '%s_%s_000.predict' % (dataset, name))
np.savetxt(filepath, predictions, delimiter=' ', fmt='%.4e')