Mercurial > repos > bgruening > keras_model_config

comparison keras_deep_learning.py @ 1:0fd7d8e90e2a draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ba6a47bdf76bbf4cb276206ac1a8cbf61332fd16"
author bgruening
date Fri, 13 Sep 2019 12:19:45 -0400
parents 1046cf73236b
children c3813c64d678
comparison
equal deleted inserted replaced
0:1046cf73236b 1:0fd7d8e90e2a
6 import six 6 import six
7 import warnings 7 import warnings
8 8
9 from ast import literal_eval 9 from ast import literal_eval
10 from keras.models import Sequential, Model 10 from keras.models import Sequential, Model
11 from galaxy_ml.utils import try_get_attr, get_search_params 11 from galaxy_ml.utils import try_get_attr, get_search_params, SafeEval
12
13
14 safe_eval = SafeEval()
12 15
13 16
14 def _handle_shape(literal): 17 def _handle_shape(literal):
15 """Eval integer or list/tuple of integers from string 18 """Eval integer or list/tuple of integers from string
16 19
98 continue 101 continue
99 102
100 if key in ['input_shape', 'noise_shape', 'shape', 'batch_shape', 103 if key in ['input_shape', 'noise_shape', 'shape', 'batch_shape',
101 'target_shape', 'dims', 'kernel_size', 'strides', 104 'target_shape', 'dims', 'kernel_size', 'strides',
102 'dilation_rate', 'output_padding', 'cropping', 'size', 105 'dilation_rate', 'output_padding', 'cropping', 'size',
103 'padding', 'pool_size', 'axis', 'shared_axes']: 106 'padding', 'pool_size', 'axis', 'shared_axes'] \
107 and isinstance(value, str):
104 params[key] = _handle_shape(value) 108 params[key] = _handle_shape(value)
105 109
106 elif key.endswith('_regularizer'): 110 elif key.endswith('_regularizer') and isinstance(value, dict):
107 params[key] = _handle_regularizer(value) 111 params[key] = _handle_regularizer(value)
108 112
109 elif key.endswith('_constraint'): 113 elif key.endswith('_constraint') and isinstance(value, dict):
110 params[key] = _handle_constraint(value) 114 params[key] = _handle_constraint(value)
111 115
112 elif key == 'function': # No support for lambda/function eval 116 elif key == 'function': # No support for lambda/function eval
113 params.pop(key) 117 params.pop(key)
114 118
127 layers = config['layers'] 131 layers = config['layers']
128 for layer in layers: 132 for layer in layers:
129 options = layer['layer_selection'] 133 options = layer['layer_selection']
130 layer_type = options.pop('layer_type') 134 layer_type = options.pop('layer_type')
131 klass = getattr(keras.layers, layer_type) 135 klass = getattr(keras.layers, layer_type)
132 other_options = options.pop('layer_options', {}) 136 kwargs = options.pop('kwargs', '')
133 options.update(other_options)
134 137
135 # parameters needs special care 138 # parameters needs special care
136 options = _handle_layer_parameters(options) 139 options = _handle_layer_parameters(options)
140
141 if kwargs:
142 kwargs = safe_eval('dict(' + kwargs + ')')
143 options.update(kwargs)
137 144
138 # add input_shape to the first layer only 145 # add input_shape to the first layer only
139 if not getattr(model, '_layers') and input_shape is not None: 146 if not getattr(model, '_layers') and input_shape is not None:
140 options['input_shape'] = input_shape 147 options['input_shape'] = input_shape
141 148
156 for layer in layers: 163 for layer in layers:
157 options = layer['layer_selection'] 164 options = layer['layer_selection']
158 layer_type = options.pop('layer_type') 165 layer_type = options.pop('layer_type')
159 klass = getattr(keras.layers, layer_type) 166 klass = getattr(keras.layers, layer_type)
160 inbound_nodes = options.pop('inbound_nodes', None) 167 inbound_nodes = options.pop('inbound_nodes', None)
161 other_options = options.pop('layer_options', {}) 168 kwargs = options.pop('kwargs', '')
162 options.update(other_options)
163 169
164 # parameters needs special care 170 # parameters needs special care
165 options = _handle_layer_parameters(options) 171 options = _handle_layer_parameters(options)
172
173 if kwargs:
174 kwargs = safe_eval('dict(' + kwargs + ')')
175 options.update(kwargs)
176
166 # merge layers 177 # merge layers
167 if 'merging_layers' in options: 178 if 'merging_layers' in options:
168 idxs = literal_eval(options.pop('merging_layers')) 179 idxs = literal_eval(options.pop('merging_layers'))
169 merging_layers = [all_layers[i-1] for i in idxs] 180 merging_layers = [all_layers[i-1] for i in idxs]
170 new_layer = klass(**options)(merging_layers) 181 new_layer = klass(**options)(merging_layers)