[keras] Networks: now all LinearNetworks come from one class

keras.py

@@ -1117,28 +1117,24 @@ class DenseNetBlock(tf.keras.layers.Layer):
         return {"block_size": self.block_size, "sub_kwargs": self.sub_kwargs}
 
 
-class FullyConnected(tf.keras.layers.Layer):
-    """
-    The FullyConnected object is an implementation of a fully connected DNN.
-
-    Parameters
-    ----------
-    layers : int
-        The number of layers.
-    kwargs :
-        Arguments for DenseLayer.
+class LinearNetwork(tf.keras.layers.Layer):
+    @property
+    def name(self):
+        raise NotImplementedError
 
-    """
+    @property
+    def substructure(self):
+        raise NotImplementedError
 
     def __init__(self, layers=0, sub_kwargs=None, **kwargs):
-        super().__init__(name="FullyConnected")
+        super().__init__(name=self.name)
         self.layers = layers
         self.sub_kwargs = kwargs if sub_kwargs is None else sub_kwargs
 
     def build(self, input_shape):
         network_layers = []
         for layer in range(self.layers):
-            network_layers.append(DenseLayer(**self.sub_kwargs))
+            network_layers.append(self.substructure(**self.sub_kwargs))
         self.network_layers = network_layers
 
     def call(self, input_tensor, training=False):
@@ -1151,41 +1147,41 @@ class FullyConnected(tf.keras.layers.Layer):
         return {"layers": self.layers, "sub_kwargs": self.sub_kwargs}
 
 
-class ResNet(tf.keras.layers.Layer):
+class FullyConnected(LinearNetwork):
     """
-    The ResNet object is an implementation of a Residual Neural Network.
+    The FullyConnected object is an implementation of a fully connected DNN.
 
     Parameters
     ----------
     layers : int
-        The number of residual blocks.
+        The number of layers.
     kwargs :
-        Arguments for ResNetBlock.
+        Arguments for DenseLayer.
 
     """
 
-    def __init__(self, layers=1, sub_kwargs=None, **kwargs):
-        super().__init__(name="ResNet")
-        self.layers = layers
-        self.sub_kwargs = kwargs if sub_kwargs is None else sub_kwargs
+    name = "FullyConnected"
+    substructure = DenseLayer
 
-    def build(self, input_shape):
-        _layers = []
-        for i in range(self.layers):
-            _layers.append(ResNetBlock(**self.sub_kwargs))
-        self._layers = _layers
 
-    def call(self, input_tensor, training=False):
-        x = input_tensor
-        for layer in self._layers:
-            x = layer(x, training=training)
-        return x
+class ResNet(LinearNetwork):
+    """
+    The ResNet object is an implementation of a Residual Neural Network.
 
-    def get_config(self):
-        return {"layers": self.layers, "sub_kwargs": self.sub_kwargs}
+    Parameters
+    ----------
+    layers : int
+        The number of residual blocks.
+    kwargs :
+        Arguments for ResNetBlock.
+
+    """
+
+    name = "ResNet"
+    substructure = ResNetBlock
 
 
-class DenseNet(tf.keras.layers.Layer):
+class DenseNet(LinearNetwork):
     """
     The DenseNet object is an implementation of a DenseNet Neural Network.
 
@@ -1198,25 +1194,8 @@ class DenseNet(tf.keras.layers.Layer):
 
     """
 
-    def __init__(self, layers=1, sub_kwargs=None, **kwargs):
-        super().__init__(name="DenseNet")
-        self.layers = layers
-        self.sub_kwargs = kwargs if sub_kwargs is None else sub_kwargs
-
-    def build(self, input_shape):
-        _layers = []
-        for i in range(self.layers):
-            _layers.append(DenseNetBlock(**self.sub_kwargs))
-        self._layers = _layers
-
-    def call(self, input_tensor, training=False):
-        x = input_tensor
-        for layer in self._layers:
-            x = layer(x, training=training)
-        return x
-
-    def get_config(self):
-        return {"layers": self.layers, "sub_kwargs": self.sub_kwargs}
+    name = "DenseNet"
+    substructure = DenseNetBlock
 
 
 class Xception1D(tf.keras.layers.Layer):