forked from hycis/Mozi
-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
295 lines (219 loc) · 9.2 KB
/
dataset.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
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
import mozi.datasets.iterator as iterators
import numpy as np
import theano
floatX = theano.config.floatX
import logging
internal_logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.DEBUG)
from mozi.log import Log
class IterMatrix(object):
def __init__(self, X, y, iter_class='SequentialSubsetIterator',
batch_size=100, num_batches=None, rng=None):
self.X = X
self.y = y
self.batch_size = batch_size
self.num_batches = num_batches
self.iter_class = iter_class
self.rng = rng
self.iterator = getattr(iterators, self.iter_class)
def __iter__(self):
return self.iterator(dataset_size=self.dataset_size,
batch_size=self.batch_size,
num_batches=self.num_batches,
rng=self.rng)
def set_iterator(self, iterator):
self.iterator = iterator
def __getitem__(self, key):
return self.X[key], self.y[key]
@property
def dataset_size(self):
return self.X.shape[0] if self.X is not None else -1
class IterDatasets(object):
def __init__(self, datasets, labels, iter_class='SequentialSubsetIterator',
batch_size=100, num_batches=None, rng=None):
self.datasets = datasets
self.labels = labels
self.batch_size = batch_size
self.num_batches = num_batches
self.iter_class = iter_class
self.rng = rng
self.iterator = getattr(iterators, self.iter_class)
def __iter__(self):
return self.iterator(dataset_size=self.dataset_size,
batch_size=self.batch_size,
num_batches=self.num_batches,
rng=self.rng)
def set_iterator(self, iterator):
self.iterator = iterator
def __getitem__(self, key):
Xslice = []
yslice = []
for dataset in self.datasets:
Xslice.append(dataset[key])
for label in self.labels:
yslice.append(label[key])
return Xslice + yslice
@property
def dataset_size(self):
return len(self.datasets[0]) if self.datasets is not None else -1
class Dataset(object):
def __init__(self, train_valid_test_ratio=[8,1,1], log=None, batch_size=100,
num_batches=None, iter_class='SequentialSubsetIterator', rng=None):
assert len(train_valid_test_ratio) == 3, 'the size of list is not 3'
self.ratio = train_valid_test_ratio
self.iter_class = iter_class
self.batch_size = batch_size
self.num_batches = num_batches
self.rng = rng
self.log = log
if self.log is None:
# use default Log setting, using the internal logger
self.log = Log(logger=internal_logger)
def __iter__(self):
raise NotImplementedError(str(type(self))+" does not implement the __iter__ method.")
def next(self):
raise NotImplementedError(str(type(self))+" does not implement the next method.")
@property
def nblocks(self):
raise NotImplementedError(str(type(self))+" does not implement the nblocks method.")
class SingleBlock(Dataset):
def __init__(self, X=None, y=None, train_valid_test_ratio=[8,1,1], log=None, **kwargs):
'''
All the data is loaded into memory for one go training
'''
super(SingleBlock, self).__init__(train_valid_test_ratio, log, **kwargs)
self.train = IterMatrix(X=None, y=None, **kwargs)
self.valid = IterMatrix(X=None, y=None, **kwargs)
self.test = IterMatrix(X=None, y=None, **kwargs)
assert len(self.ratio) == 3, 'the size of list is not 3'
if X is not None and y is not None:
self.set_Xy(X, y)
def __iter__(self):
self.iter = True
return self
def next(self):
if self.iter:
self.iter = False
return self
else:
raise StopIteration
@property
def nblocks(self):
return 1
def set_Xy(self, X, y):
num_examples = len(X)
total_ratio = sum(self.ratio)
num_train = int(self.ratio[0] * 1.0 * num_examples / total_ratio)
num_valid = int(self.ratio[1] * 1.0 * num_examples / total_ratio)
train_X = X[:num_train]
train_y = y[:num_train]
valid_X = X[num_train:num_train+num_valid]
valid_y = y[num_train:num_train+num_valid]
test_X = X[num_train+num_valid:]
test_y = y[num_train+num_valid:]
self.train.X = train_X
self.train.y = train_y
if self.ratio[1] == 0:
self.log.info('Valid set is empty! It is needed for early stopping and saving best model')
self.valid.X = valid_X
self.valid.y = valid_y
if self.ratio[2] == 0:
self.log.info('Test set is empty! It is needed for testing the best model')
self.test.X = test_X
self.test.y = test_y
def get_train(self):
return self.train
def get_valid(self):
return self.valid
def get_test(self):
return self.test
def set_train(self, X, y):
self.train.X = X
self.train.y = y
def set_valid(self, X, y):
self.valid.X = X
self.valid.y = y
def set_test(self, X, y):
self.test.X = X
self.test.y = y
class DataBlocks(Dataset):
def __init__(self, data_paths, train_valid_test_ratio=[8,1,1], log=None, **kwargs):
"""
DESCRIPTION:
This is class for processing blocks of data, whereby dataset is loaded
and unloaded into memory one block at a time.
PARAM:
data_paths(list): contains the paths to the numpy data files. It's a
list of tuples whereby the first element of the tuple
is the X path, and the second is the y path.
example [(X_path1, y_path1),(X_path2, y_path2)]
"""
super(DataBlocks, self).__init__(train_valid_test_ratio, log, **kwargs)
assert isinstance(data_paths, list), "data_paths is not a list"
self.data_paths = data_paths
self.single_block = SingleBlock(None, None, train_valid_test_ratio, log, **kwargs)
def __iter__(self):
self.files = iter(self.data_paths)
return self
def next(self):
file = next(self.files)
assert isinstance(file, tuple) or isintance(file, list), str(type(file)) + "is not a tuple or list"
with open(file[0], 'rb') as X_fin, open(file[1], 'rb') as y_fin:
X = np.load(X_fin)
y = np.load(y_fin)
self.single_block.set_Xy(X=X, y=y)
return self.single_block
@property
def nblocks(self):
return len(self.data_paths)
class MultiInputsData(SingleBlock):
def __init__(self, datasets, labels, train_valid_test_ratio=[8,1,1], log=None, **kwargs):
"""
DESCRIPTION:
This class is used for multitask learning where we have multiple data
inputs and one output.
PARAM:
datasets (tuple of arrays): If our input is X1 and X2, both with same number
of rows, then X = (X1, X2)
labels (tuple of arrays): label of same number of rows as input data
"""
assert isinstance(datasets, tuple), "dataset is not a tuple of arrays"
assert isinstance(labels, tuple), "labels is not a tuple of arrays"
self.num_examples = len(datasets[0])
for dataset in datasets:
assert len(dataset) == self.num_examples, 'number of rows for different datasets is not the same'
for label in labels:
assert len(label) == self.num_examples, 'number of rows for different labels is not the same'
super(MultiInputsData, self).__init__(train_valid_test_ratio, log, **kwargs)
self.train = IterDatasets(None, None, **kwargs)
self.valid = IterDatasets(None, None, **kwargs)
self.test = IterDatasets(None, None, **kwargs)
self.set(datasets, labels)
def set(self, datasets, labels):
total_ratio = sum(self.ratio)
num_train = int(float(self.ratio[0]) * self.num_examples / total_ratio)
num_valid = int(float(self.ratio[1]) * self.num_examples / total_ratio)
trainset = []
validset = []
testset = []
for dataset in datasets:
trainset.append(dataset[:num_train])
validset.append(dataset[num_train:num_train+num_valid])
testset.append(dataset[num_train+num_valid:])
trainlbl = []
validlbl = []
testlbl = []
for label in labels:
trainlbl.append(label[:num_train])
validlbl.append(label[num_train:num_train+num_valid])
testlbl.append(label[num_train+num_valid:])
self.train.datasets = trainset
self.train.labels = trainlbl
if self.ratio[1] == 0:
self.log.info('Valid set is empty! It is needed for early stopping and saving best model')
self.valid.datasets = validset
self.valid.labels = validlbl
if self.ratio[2] == 0:
self.log.info('Test set is empty! It is needed for testing the best model')
self.test.datasets = testset
self.test.labels = testlbl