/
example.py
112 lines (93 loc) · 4.93 KB
/
example.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
import os
import pandas as pd
import pickle
from ordinal_tsf.model import MordredStrategy
from ordinal_tsf.dataset import Dataset, Quantiser, Standardiser, WhiteCorrupter, AttractorStacker, Selector, TestDefinition
from ordinal_tsf.session import Session
from ordinal_tsf.util import cartesian
os.environ["CUDA_VISIBLE_DEVICES"]="7"
import tensorflow as tf
from keras.backend.tensorflow_backend import set_session
config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = 0.45
set_session(tf.Session(config=config))
MAX_LENGTH = 30000
LOOKBACK = 100
HORIZON = 100
ATTRACTOR_LAG = 10
EFFECTIVE_LAG = 0
VALIDATION_PREDICTIVE_HORIZON = 500
TEST_PREDICTIVE_HORIZON = 1000
DECODER_SEED_LENGTH = 1
# open/create session (folder name) (includes raw time series loading)
# choose experiment (includes model preparation)
best_models = {}
mordred_search_space = {'lam': [1e-6, 1e-7, 1e-8],
'dropout_rate': [0.25, 0.5],
'units': [64, 128, 256, 320],
'lookback': [LOOKBACK],
'horizon': [HORIZON]}
train_spec = {'epochs': 50, 'batch_size': 256, 'validation_split': 0.15}
DS = 'mg'
sess = Session('{}'.format(DS))
VALIDATION_START_INDEX = LOOKBACK + 2 * ATTRACTOR_LAG + DECODER_SEED_LENGTH
TEST_START_INDEX = LOOKBACK + 2 * ATTRACTOR_LAG + DECODER_SEED_LENGTH
## Transformations to be applied to the dataset
x = pd.read_feather('WebTS/{}.feather'.format(DS)).values[:MAX_LENGTH]
stand = Standardiser()
quant = Quantiser()
white_noise = WhiteCorrupter()
dataset = Dataset(x, LOOKBACK + HORIZON + DECODER_SEED_LENGTH,
p_val=0.15, p_test=0.15, preprocessing_steps=[stand, quant])
mordred_search_space['ordinal_bins'] = [quant.n_bins]
mordred_search_space['n_channels'] = [dataset.optional_params.get('n_channels', 1)]
## Definition of validation tests to choose best hyperparameters
selector = Selector(VALIDATION_START_INDEX, VALIDATION_PREDICTIVE_HORIZON)
continuous_ground_truth = dataset.apply_partial_preprocessing('val', [selector, stand])
ordinal_ground_truth = dataset.apply_partial_preprocessing('val', [selector, stand, quant])
validation_tests = [TestDefinition('mse', continuous_ground_truth),
TestDefinition('nll', ordinal_ground_truth),
TestDefinition('cum_nll', ordinal_ground_truth),
TestDefinition('median_dtw_distance', continuous_ground_truth),
TestDefinition('median_attractor_distance', continuous_ground_truth)]
validation_plots = {'plot_median_2std': {'ground_truth':continuous_ground_truth},
'plot_cum_nll': {'binned_ground_truth': ordinal_ground_truth},
'plot_like': {}}
## Trains models and chooses best hyperparameters
experiment = sess.start_experiment(dataset, MordredStrategy)
best_models[DS] = experiment.choose_model(validation_tests,
mordred_search_space.keys(),
cartesian(mordred_search_space.values()),
VALIDATION_START_INDEX - EFFECTIVE_LAG,
VALIDATION_PREDICTIVE_HORIZON,
plots=validation_plots,
fit_kwargs=train_spec,
eval_kwargs={'mc_samples': 50},
mode='val')
## Tests on actual test data
selector = Selector(TEST_START_INDEX, TEST_PREDICTIVE_HORIZON)
continuous_ground_truth = dataset.apply_partial_preprocessing('test', [selector, stand])
ordinal_ground_truth = dataset.apply_partial_preprocessing('test', [selector, stand, quant])
final_tests = [TestDefinition('mse', continuous_ground_truth),
TestDefinition('nll', ordinal_ground_truth),
TestDefinition('cum_nll', ordinal_ground_truth),
TestDefinition('median_dtw_distance', continuous_ground_truth),
TestDefinition('median_attractor_distance', continuous_ground_truth)]
test_plots = {'plot_median_2std': {'ground_truth': continuous_ground_truth},
'plot_median_dtw_alignment': {'ground_truth': continuous_ground_truth},
'plot_cum_nll': {'binned_ground_truth': ordinal_ground_truth},
'plot_like': {}}
for metric, best_model in best_models[DS].items():
print 'Test result for model with best {} performance'.format(metric)
experiment.choose_model(final_tests,
best_model.keys(),
[best_model.values()],
TEST_START_INDEX - EFFECTIVE_LAG,
TEST_PREDICTIVE_HORIZON,
plots=test_plots,
fit_kwargs=train_spec,
eval_kwargs={'mc_samples': 100},
mode='test')
with open('best_models_mordred', 'wb') as f:
pickle.dump(best_models, f)
exit()