-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathkubernetes_script.py
More file actions
214 lines (165 loc) · 11.3 KB
/
kubernetes_script.py
File metadata and controls
214 lines (165 loc) · 11.3 KB
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
208
209
210
211
212
213
214
"""Kubernetes script"""
# Training initial models
# from brainreader import train
# from brainreader import params
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 1, 'data_params': 1}, 'training_params <=6',
# reserve_jobs=True) # MSE + none
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params <=6', reserve_jobs=True) # MSE + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >6 AND training_params <=12',
# reserve_jobs=True) # MSE + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 2, 'data_params': 3},
# 'training_params >12 AND training_params <=18',
# reserve_jobs=True) # poisson + exp
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >18 AND training_params <=24',
# reserve_jobs=True) # poisson + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >18 AND training_params <=24',
# reserve_jobs=True) # poisson + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 2, 'data_params': 3},
# 'training_params >24 AND training_params <=30',
# reserve_jobs=True) # exp + exp
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >24 AND training_params <=30',
# reserve_jobs=True) # exp + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >30 AND training_params <=36',
# reserve_jobs=True) # exp + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 1, 'data_params': 1},
# 'training_params >36 AND training_params <=42',
# reserve_jobs=True) # mse + none
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 2, 'data_params': 3},
# 'training_params >36 AND training_params <=42',
# reserve_jobs=True) # mse + exp
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >42 AND training_params <=48',
# reserve_jobs=True) # mse + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >36 AND training_params <=42',
# reserve_jobs=True) # mse + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 2, 'data_params': 3},
# 'training_params >48 AND training_params <=54',
# reserve_jobs=True) # poisson + exp
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >54 AND training_params <=60',
# reserve_jobs=True) # poisson + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >54 AND training_params <=60',
# reserve_jobs=True) # poisson + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 2, 'data_params': 3},
# 'training_params >60 AND training_params <=66',
# reserve_jobs=True) # exp + exp
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >60 AND training_params <=66',
# reserve_jobs=True) # exp + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >66 AND training_params <=72',
# reserve_jobs=True) # exp + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 3, 'data_params': 3},
# 'training_params >72 AND training_params <=96',
# reserve_jobs=True) # poissson + elu
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >72 AND training_params <=96',
# reserve_jobs=True) # poissson + expscaled
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 1, 'data_params': 2},
# 'training_params <=6', reserve_jobs=True) # zscore-resps
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 1, 'data_params': 4},
# 'training_params <=6', reserve_jobs=True) # df/f
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 1, 'data_params': 5},
# 'training_params <=6', reserve_jobs=True) # df/std(df)
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 6},
# 'training_params >72 AND training_params <=84', reserve_jobs=True) # stddev-resps
# train.TrainedModel.populate({'dset_id': 1, 'model_params': 4, 'data_params': 3},
# 'training_params >96 AND training_params <=132',
# reserve_jobs=True) # different batch size
# train.TrainedModel.populate({'dset_id': 1, 'data_params': 3}, 'model_params in (5, 6, 7)',
# 'training_params >72 AND training_params <=84',
# reserve_jobs=True) # different mlps
# train.TrainedModel.populate({'dset_id': 1, 'data_params': 3, 'model_params': 4},
# 'training_params >132 AND training_params <=144',
# reserve_jobs=True) # weighted poisson
# train.TrainedModel.populate({'dset_id': 1, 'data_params': 3}, 'model_params in (29, 30)',
# 'training_params >72 AND training_params <=84',
# reserve_jobs=True) # konstinets
# train.TrainedModel.populate({'dset_id': 1, 'data_params': 3, 'training_params': 145},
# 'model_params in (29, 30)', reserve_jobs=True) # konstinets
# from brainreader.encoding import train
# train.TrainedModel.populate('dset_id in (1, 5)', reserve_jobs=True)
# train.Evaluation.populate(reserve_jobs=True)
# train.Ensemble.populate(reserve_jobs=True)
# train.EnsembleEvaluation.populate(reserve_jobs=True)
# train Gabor
# from brainreader import decoding
# decoding.GaborModel.populate('dset_id in (1, 5)', reserve_jobs=True)
# train AHP
# from brainreader import reconstructions
# reconstructions.AHPValEvaluation.populate(reserve_jobs=True)
# from brainreader import reconstructions
# reconstructions.GradientOneReconstruction.fill_recons('ensemble_dset=5 AND gradient_params > 200', split='val')
# Populate all models for dataset 4
# from brainreader import decoding
# decoding.LinearModel.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.LinearValEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.LinearReconstructions.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.LinearEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.MLPModel.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.MLPValEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.MLPReconstructions.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.MLPEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.GaborModel.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.GaborValEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.GaborReconstructions.populate({'dset_id': 4}, reserve_jobs=True)
# decoding.GaborEvaluation.populate({'dset_id': 4}, reserve_jobs=True)
# from brainreader.encoding import train
# train.TrainedModel.populate({'dset_id': 4}, reserve_jobs=True)
# train.Evaluation.populate({'dset_id': 4}, reserve_jobs=True)
# train.Ensemble.populate({'ensemble_dset': 4}, reserve_jobs=True)
# train.EnsembleEvaluation.populate({'ensemble_dset': 4}, reserve_jobs=True)
# from brainreader import reconstructions
# Use val_corr in EnsembleEvaluation to add an entry in reconstructions.BestEnsemble
# reconstructions.ModelResponses.populate({'ensemble_dset': 4}, reserve_jobs=True) # needs to be done only once per dset
# reconstructions.AHPValEvaluation.populate({'ensemble_dset': 4}, reserve_jobs=True)
# reconstructions.AHPReconstructions.populate({'ensemble_dset': 4}, reserve_jobs=True)
# reconstructions.AHPEvaluation.populate({'ensemble_dset': 4}, reserve_jobs=True)
# reconstructions.GradientOneReconstruction.fill_recons({'ensemble_dset': 4}, split='test')
# reconstructions.GradientEvaluation.populate({'ensemble_dset': 4}, reserve_jobs=True)
# reconstructions.GradientOneReconstruction.fill_recons({'ensemble_dset': 4}, split='val')
# reconstructions.GradientValEvaluation.populate({'ensemble_dset': 4}, reserve_jobs=True)
# # Populate models for all scans
# from brainreader.encoding import train
# train.TrainedModel.populate(reserve_jobs=True)
# train.Evaluation.populate(reserve_jobs=True)
# train.Ensemble.populate(reserve_jobs=True)
# train.EnsembleEvaluation.populate(reserve_jobs=True)
# from brainreader import decoding
# decoding.LinearModel.populate(reserve_jobs=True)
# decoding.LinearValEvaluation.populate(reserve_jobs=True)
# decoding.MLPModel.populate(reserve_jobs=True)
# decoding.MLPValEvaluation.populate(reserve_jobs=True)
# decoding.DeconvModel.populate(reserve_jobs=True)
# decoding.DeconvValEvaluation.populate(reserve_jobs=True)
# decoding.GaborModel.populate(reserve_jobs=True)
# decoding.GaborValEvaluation.populate(reserve_jobs=True)
from brainreader import reconstructions
# # Use val_corr in EnsembleEvaluation to add an entry in reconstructions.BestEnsemble
# # reconstructions.ModelResponses.populate(reserve_jobs=True) # needs to be done only once per dset
# reconstructions.AHPValEvaluation.populate(reserve_jobs=True)
reconstructions.GradientOneReconstruction.fill_recons({'dset_id': 4, 'ensemble_dset': 4}, split='test')
reconstructions.GradientOneReconstruction.fill_recons({'dset_id': 5, 'ensemble_dset': 5}, split='test')
reconstructions.GradientValEvaluation.populate(reserve_jobs=True)
reconstructions.GradientEvaluation.populate(reserve_jobs=True)
# # Populate all test set evaluations for relevant scans
# for dset_id in [21, 22]:#[5, 6, 7, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]:
# key = {'dset_id': dset_id, 'ensemble_dset': dset_id}
# decoding.LinearReconstructions.populate(key, reserve_jobs=True)
# decoding.LinearEvaluation.populate(key, reserve_jobs=True)
# decoding.MLPReconstructions.populate(key, reserve_jobs=True)
# decoding.MLPEvaluation.populate(key, reserve_jobs=True)
# decoding.DeconvReconstructions.populate(key, reserve_jobs=True)
# decoding.DeconvEvaluation.populate(key, reserve_jobs=True)
# decoding.GaborReconstructions.populate(key, reserve_jobs=True)
# decoding.GaborEvaluation.populate(key, reserve_jobs=True)
# reconstructions.AHPReconstructions.populate(key, reserve_jobs=True)
# reconstructions.AHPEvaluation.populate(key, reserve_jobs=True)