annotate learner.py @ 133:b4657441dd65

Corrected typos
author Yoshua Bengio <bengioy@iro.umontreal.ca>
date Fri, 09 May 2008 13:38:54 -0400
parents f6505ec32dc3
children 3f4e5c9bdc5e
rev   line source
1
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
1
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
2 from dataset import AttributesHolder,AbstractFunction
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
3 import compile
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
4 from theano import tensor as t
1
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
5
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
6 class Learner(AttributesHolder):
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
7 """
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
8 Base class for learning algorithms, provides an interface
1
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
9 that allows various algorithms to be applicable to generic learning
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
10 algorithms.
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
11
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
12 A L{Learner} can be seen as a learning algorithm, a function that when
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
13 applied to training data returns a learned function, an object that
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
14 can be applied to other data and return some output data.
1
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
15 """
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
16
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
17 def __init__(self):
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
18 pass
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
19
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
20 def forget(self):
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
21 """
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
22 Reset the state of the learner to a blank slate, before seeing
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
23 training data. The operation may be non-deterministic if the
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
24 learner has a random number generator that is set to use a
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
25 different seed each time it forget() is called.
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
26 """
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
27 raise NotImplementedError
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
28
14
5ede27026e05 Working on gradient_based_learner
bengioy@bengiomac.local
parents: 13
diff changeset
29 def update(self,training_set,train_stats_collector=None):
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
30 """
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
31 Continue training a learner, with the evidence provided by the given training set.
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
32 Hence update can be called multiple times. This is particularly useful in the
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
33 on-line setting or the sequential (Bayesian or not) settings.
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
34 The result is a function that can be applied on data, with the same
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
35 semantics of the Learner.use method.
75
90e4c0784d6e Added draft of LinearRegression learner
bengioy@bengiomac.local
parents: 20
diff changeset
36
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
37 The user may optionally provide a training L{StatsCollector} that is used to record
75
90e4c0784d6e Added draft of LinearRegression learner
bengioy@bengiomac.local
parents: 20
diff changeset
38 some statistics of the outputs computed during training. It is update(d) during
90e4c0784d6e Added draft of LinearRegression learner
bengioy@bengiomac.local
parents: 20
diff changeset
39 training.
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
40 """
13
633453635d51 Starting to work on gradient_based_learner.py
bengioy@bengiomac.local
parents: 10
diff changeset
41 return self.use # default behavior is 'non-adaptive', i.e. update does not do anything
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
42
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
43
20
266c68cb6136 Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents: 14
diff changeset
44 def __call__(self,training_set,train_stats_collector=None):
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
45 """
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
46 Train a learner from scratch using the provided training set,
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
47 and return the learned function.
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
48 """
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
49 self.forget()
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
50 return self.update(training_set,train_stats_collector)
1
2cd82666b9a7 Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents: 0
diff changeset
51
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
52 def use(self,input_dataset,output_fieldnames=None,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
53 test_stats_collector=None,copy_inputs=True,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
54 put_stats_in_output_dataset=True,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
55 output_attributes=[]):
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
56 """
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
57 Once a L{Learner} has been trained by one or more call to 'update', it can
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
58 be used with one or more calls to 'use'. The argument is an input L{DataSet} (possibly
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
59 containing a single example) and the result is an output L{DataSet} of the same length.
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
60 If output_fieldnames is specified, it may be use to indicate which fields should
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
61 be constructed in the output L{DataSet} (for example ['output','classification_error']).
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
62 Otherwise, self.defaultOutputFields is called to choose the output fields.
20
266c68cb6136 Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents: 14
diff changeset
63 Optionally, if copy_inputs, the input fields (of the input_dataset) can be made
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
64 visible in the output L{DataSet} returned by this method.
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
65 Optionally, attributes of the learner can be copied in the output dataset,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
66 and statistics computed by the stats collector also put in the output dataset.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
67 Note the distinction between fields (which are example-wise quantities, e.g. 'input')
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
68 and attributes (which are not, e.g. 'regularization_term').
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
69
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
70 We provide here a default implementation that does all this using
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
71 a sub-class defined method: minibatchwiseUseFunction.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
72
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
73 @todo check if some of the learner attributes are actually SPECIFIED
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
74 as attributes of the input_dataset, and if so use their values instead
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
75 of the ones in the learner.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
76
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
77 The learner tries to compute in the output dataset the output fields specified.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
78 If None is specified then self.defaultOutputFields(input_dataset.fieldNames())
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
79 is called to determine the output fields.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
80
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
81 Attributes of the learner can also optionally be copied into the output dataset.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
82 If output_attributes is None then all of the attributes in self.AttributeNames()
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
83 are copied in the output dataset, but if it is [] (the default), then none are copied.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
84 If a test_stats_collector is provided, then its attributes (test_stats_collector.AttributeNames())
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
85 are also copied into the output dataset attributes.
10
80bf5492e571 Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents: 1
diff changeset
86 """
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
87 minibatchwise_use_function = self.minibatchwiseUseFunction(input_dataset.fieldNames(),
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
88 output_fieldnames,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
89 test_stats_collector)
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
90 virtual_output_dataset = ApplyFunctionDataSet(input_dataset,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
91 minibatchwise_use_function,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
92 True,DataSet.numpy_vstack,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
93 DataSet.numpy_hstack)
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
94 # actually force the computation
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
95 output_dataset = CachedDataSet(virtual_output_dataset,True)
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
96 if copy_inputs:
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
97 output_dataset = input_dataset | output_dataset
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
98 # copy the wanted attributes in the dataset
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
99 if output_attributes is None:
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
100 output_attributes = self.attributeNames()
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
101 if output_attributes:
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
102 assert set(attribute_names) <= set(self.attributeNames())
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
103 output_dataset.setAttributes(output_attributes,
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
104 self.names2attributes(output_attributes,return_copy=True))
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
105 if test_stats_collector:
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
106 test_stats_collector.update(output_dataset)
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
107 if put_stats_in_output_dataset:
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
108 output_dataset.setAttributes(test_stats_collector.attributeNames(),
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
109 test_stats_collector.attributes())
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
110 return output_dataset
77
1e2bb5bad636 toying with different ways to implement learners
bengioy@bengiomac.local
parents: 75
diff changeset
111
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
112 def minibatchwiseUseFunction(self, input_fields, output_fields, stats_collector):
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
113 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
114 Returns a function that can map the given input fields to the given output fields
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
115 and to the attributes that the stats collector needs for its computation.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
116 That function is expected to operate on minibatches.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
117 The function returned makes use of the self.useInputAttributes() and
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
118 sets the attributes specified by self.useOutputAttributes().
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
119 """
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
120 def attributeNames(self):
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
121 """
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
122 A Learner may have attributes that it wishes to export to other objects. To automate
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
123 such export, sub-classes should define here the names (list of strings) of these attributes.
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
124
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
125 @todo By default, attributeNames looks for all dictionary entries whose name does not start with _.
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
126 """
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
127 return []
77
1e2bb5bad636 toying with different ways to implement learners
bengioy@bengiomac.local
parents: 75
diff changeset
128
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
129 def attributes(self,return_copy=False):
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
130 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
131 Return a list with the values of the learner's attributes (or optionally, a deep copy).
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
132 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
133 return self.names2attributes(self.attributeNames(),return_copy)
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
134
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
135 def names2attributes(self,names,return_copy=False):
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
136 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
137 Private helper function that maps a list of attribute names to a list
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
138 of (optionally copies) values of attributes.
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
139 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
140 if return_copy:
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
141 return [copy.deepcopy(self.__getattribute__(name).data) for name in names]
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
142 else:
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
143 return [self.__getattribute__(name).data for name in names]
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
144
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
145 def updateInputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
146 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
147 A subset of self.attributeNames() which are the names of attributes needed by update() in order
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
148 to do its work.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
149 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
150 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
151
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
152 def useInputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
153 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
154 A subset of self.attributeNames() which are the names of attributes needed by use() in order
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
155 to do its work.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
156 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
157 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
158
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
159 def updateOutputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
160 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
161 A subset of self.attributeNames() which are the names of attributes modified/created by update() in order
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
162 to do its work.
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
163
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
164 By default these are inferred from the various update output attributes:
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
165 """
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
166 return ["parameters"] + self.updateMinibatchOutputAttributes() + self.updateEndOutputAttributes()
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
167
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
168 def useOutputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
169 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
170 A subset of self.attributeNames() which are the names of attributes modified/created by use() in order
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
171 to do its work.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
172 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
173 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
174
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
175
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
176 class TLearner(Learner):
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
177 """
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
178 TLearner is a virtual class of Learners that attempts to factor out of the definition
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
179 of a learner the steps that are common to many implementations of learning algorithms,
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
180 so as to leave only 'the equations' to define in particular sub-classes, using Theano.
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
181
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
182 In the default implementations of use and update, it is assumed that the 'use' and 'update' methods
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
183 visit examples in the input dataset sequentially. In the 'use' method only one pass through the dataset is done,
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
184 whereas the sub-learner may wish to iterate over the examples multiple times. Subclasses where this
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
185 basic model is not appropriate can simply redefine update or use.
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
186
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
187 Sub-classes must provide the following functions and functionalities:
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
188 - attributeNames(): defines all the names of attributes which can be used as fields or
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
189 attributes in input/output datasets or in stats collectors.
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
190 All these attributes are expected to be theano.Result objects
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
191 (with a .data property and recognized by theano.Function for compilation).
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
192 The sub-class constructor defines the relations between
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
193 the Theano variables that may be used by 'use' and 'update'
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
194 or by a stats collector.
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
195 - defaultOutputFields(input_fields): return a list of default dataset output fields when
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
196 None are provided by the caller of use.
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
197 The following naming convention is assumed and important.
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
198 Attributes whose names are listed in attributeNames() can be of any type,
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
199 but those that can be referenced as input/output dataset fields or as
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
200 output attributes in 'use' or as input attributes in the stats collector
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
201 should be associated with a Theano Result variable. If the exported attribute
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
202 name is <name>, the corresponding Result name (an internal attribute of
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
203 the TLearner, created in the sub-class constructor) should be _<name>.
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
204 Typically <name> will be numpy ndarray and _<name> will be the corresponding
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
205 Theano Tensor (for symbolic manipulation).
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
206
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
207 @todo pousser dans Learner toute la poutine qui peut l'etre sans etre
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
208 dependant de Theano
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
209 """
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
210
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
211 def __init__(self):
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
212 Learner.__init__(self)
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
213
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
214 def defaultOutputFields(self, input_fields):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
215 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
216 Return a default list of output field names (to put in the output dataset).
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
217 This will be used when None are provided (as output_fields) by the caller of the 'use' method.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
218 This may involve looking at the input_fields (names) available in the
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
219 input_dataset.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
220 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
221 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
222
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
223 def minibatchwiseUseFunction(self, input_fields, output_fields, stats_collector):
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
224 """
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
225 Implement minibatchwiseUseFunction by exploiting Theano compilation
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
226 and the expression graph defined by a sub-class constructor.
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
227 """
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
228 if not output_fields:
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
229 output_fields = self.defaultOutputFields(input_fields)
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
230 if stats_collector:
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
231 stats_collector_inputs = stats_collector.input2UpdateAttributes()
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
232 for attribute in stats_collector_inputs:
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
233 if attribute not in input_fields:
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
234 output_fields.append(attribute)
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
235 key = (input_fields,output_fields)
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
236 if key not in self.use_functions_dictionary:
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
237 use_input_attributes = self.useInputAttributes()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
238 use_output_attributes = self.useOutputAttributes()
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
239 complete_f = compile.function(self.names2OpResults(input_fields+use_input_attributes),
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
240 self.names2OpResults(output_fields+use_output_attributes))
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
241 def f(*input_field_values):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
242 input_attribute_values = self.names2attributes(use_input_attributes)
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
243 results = complete_f(*(input_field_values + input_attribute_values))
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
244 output_field_values = results[0:len(output_fields)]
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
245 output_attribute_values = results[len(output_fields):len(results)]
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
246 if use_output_attributes:
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
247 self.setAttributes(use_output_attributes,output_attribute_values)
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
248 return output_field_values
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
249 self.use_functions_dictionary[key]=f
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
250 return self.use_functions_dictionary[key]
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
251
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
252 def names2OpResults(self,names):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
253 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
254 Private helper function that maps a list of attribute names to a list
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
255 of corresponding Op Results (with the same name but with a '_' prefix).
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
256 """
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
257 return [self.__getattribute__('_'+name) for name in names]
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
258
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
259
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
260 class MinibatchUpdatesTLearner(TLearner):
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
261 """
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
262 This adds to L{TLearner} a
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
263 - updateStart(), updateEnd(), updateMinibatch(minibatch), isLastEpoch():
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
264 functions executed at the beginning, the end, in the middle
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
265 (for each minibatch) of the update method, and at the end
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
266 of each epoch. This model only
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
267 works for 'online' or one-shot learning that requires
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
268 going only once through the training data. For more complicated
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
269 models, more specialized subclasses of TLearner should be used
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
270 or a learning-algorithm specific update method should be defined.
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
271
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
272 - a 'parameters' attribute which is a list of parameters (whose names are
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
273 specified by the user's subclass with the parameterAttributes() method)
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
274
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
275 """
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
276
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
277 def __init__(self):
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
278 TLearner.__init__(self)
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
279 self.update_minibatch_function = compile.function
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
280 (self.names2OpResults(self.updateMinibatchOutputAttributes()+
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
281 self.updateMinibatchInputFields()),
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
282 self.names2OpResults(self.updateMinibatchOutputAttributes()))
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
283 self.update_end_function = compile.function
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
284 (self.names2OpResults(self.updateEndInputAttributes()),
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
285 self.names2OpResults(self.updateEndOutputAttributes()))
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
286
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
287 def allocate(self, minibatch):
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
288 """
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
289 This function is called at the beginning of each L{updateMinibatch}
128
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
290 and should be used to check that all required attributes have been
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
291 allocated and initialized (usually this function calls forget()
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
292 when it has to do an initialization).
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
293 """
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
294 raise AbstractFunction()
ee5507af2c60 minor edits
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 126
diff changeset
295
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
296 def updateMinibatchInputFields(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
297 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
298
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
299 def updateMinibatchInputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
300 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
301
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
302 def updateMinibatchOutputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
303 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
304
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
305 def updateEndInputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
306 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
307
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
308 def updateEndOutputAttributes(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
309 raise AbstractFunction()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
310
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
311 def parameterAttributes(self):
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
312 raise AbstractFunction()
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
313
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
314 def updateStart(self,training_set):
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
315 pass
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
316
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
317 def updateEnd(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
318 self.setAttributes(self.updateEndOutputAttributes(),
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
319 self.update_end_function
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
320 (self.names2attributes(self.updateEndInputAttributes())))
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
321 self.parameters = self.names2attributes(self.parameterAttributes())
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
322
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
323 def updateMinibatch(self,minibatch):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
324 # make sure all required fields are allocated and initialized
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
325 self.allocate(minibatch)
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
326 self.setAttributes(self.updateMinibatchOutputAttributes(),
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
327 # concatenate the attribute values and field values and then apply update fn
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
328 self.update_minibatch_function(*(self.names2attributes
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
329 (self.updateMinibatchInputAttributes()))
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
330 + minibatch(self.updateMinibatchInputFields())))
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
331
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
332 def isLastEpoch(self):
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
333 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
334 This method is called at the end of each epoch (cycling over the training set).
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
335 It returns a boolean to indicate if this is the last epoch.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
336 By default just do one epoch.
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
337 """
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
338 return True
78
3499918faa9d In the middle of designing TLearner
bengioy@bengiomac.local
parents: 77
diff changeset
339
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
340 def update(self,training_set,train_stats_collector=None):
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
341 """
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
342 @todo check if some of the learner attributes are actually SPECIFIED
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
343 in as attributes of the training_set.
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
344 """
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
345 self.updateStart(training_set)
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
346 stop=False
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
347 if hasattr(self,'_minibatch_size') and self._minibatch_size:
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
348 minibatch_size=self._minibatch_size
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
349 else:
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
350 minibatch_size=min(100,len(training_set))
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
351 while not stop:
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
352 if train_stats_collector:
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
353 train_stats_collector.forget() # restart stats collectin at the beginning of each epoch
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
354 for minibatch in training_set.minibatches(minibatch_size=minibatch_size):
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
355 self.updateMinibatch(minibatch)
107
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
356 if train_stats_collector:
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
357 minibatch_set = minibatch.examples()
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
358 minibatch_set.setAttributes(self.attributeNames(),self.attributes())
c4916445e025 Comments from Pascal V.
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 92
diff changeset
359 train_stats_collector.update(minibatch_set)
110
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
360 stop = self.isLastEpoch()
8fa1ef2411a0 Worked on OneShotTLearner and implementation of LinearRegression
bengioy@bengiomac.local
parents: 109
diff changeset
361 self.updateEnd()
92
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
362 return self.use
c4726e19b8ec Finished first draft of TLearner
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 78
diff changeset
363
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
364 class OnlineGradientTLearner(MinibatchUpdatesTLearner):
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
365 """
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
366 Specialization of L{MinibatchUpdatesTLearner} in which the minibatch updates
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
367 are obtained by performing an online (minibatch-based) gradient step.
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
368
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
369 Sub-classes must define the following:
132
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
370 - self._learning_rate (may be changed by the sub-class between epochs or minibatches)
f6505ec32dc3 Updated documentation slightly
Joseph Turian <turian@gmail.com>
parents: 131
diff changeset
371 - self.lossAttribute() = name of the loss field
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
372 """
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
373 def __init__(self,truly_online=False):
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
374 """
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
375 If truly_online then only one pass is made through the training set passed to update().
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
376
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
377 SUBCLASSES SHOULD CALL THIS CONSTRUCTOR ONLY AFTER HAVING DEFINED ALL THEIR THEANO FORMULAS
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
378 """
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
379 self.truly_online=truly_online
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
380
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
381 # create the formulas for the gradient update
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
382 old_params = [self.__getattribute__("_"+name) for name in self.parameterAttributes()]
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
383 new_params_names = ["_new_"+name for name in self.parameterAttributes()]
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
384 loss = self.__getattribute__("_"+self.lossAttribute())
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
385 self.setAttributes(new_params_names,
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
386 [t.add_inplace(param,self._learning_rate*t.grad(loss,param))
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
387 for param in old_params])
129
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
388 MinibatchUpdatesTLearner.__init__(self)
4c2280edcaf5 Fixed typos in learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 128
diff changeset
389
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
390 def isLastEpoch(self):
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
391 return self.truly_online
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
392
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
393 def updateMinibatchInputAttributes(self):
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
394 return self.parameterAttributes()
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
395
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
396 def updateMinibatchOutputAttributes(self):
133
b4657441dd65 Corrected typos
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 132
diff changeset
397 return ["new_"+name for name in self.parameterAttributes()]
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
398
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
399 def updateEndInputAttributes(self):
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
400 return self.parameterAttributes()
111
88257dfedf8c Added another work in progress, for mlp's
bengioy@bengiomac.local
parents: 110
diff changeset
401
118
d0a1bd0378c6 Finished draft of OneHiddenLayerNNetClassifier to debut learner.py
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 111
diff changeset
402