Commit 6441cb2f authored by Milosz Grabski's avatar Milosz Grabski
Browse files

classes name change 

Changed names of classes to better match purpose of those layers
parent a626d8f5

deepchem/models/layers.py

+31 −31
Original line number Diff line number Diff line
@@ -370,11 +370,11 @@ class GraphGather(tf.keras.layers.Layer): 
    return mol_features





class GraphConvolutionLayer(tf.keras.layers.Layer):

class MolGANConvolutionLayer(tf.keras.layers.Layer):

  """

  Graph convolution layer used in MolGAN model.

  MolGAN is a WGAN type model for generation of small molecules.

  Not used directly, higher level layers like MultiGraphConvolutionLayer use it.

  Not used directly, higher level layers like MolGANMultiConvolutionLayer use it.

  This layer performs basic convolution on one-hot encoded matrices containing

  atom and bond information. This layer also accepts three inputs for the case

  when convolution is performed more than once and results of previous convolution
@@ -385,14 +385,14 @@ class GraphConvolutionLayer(tf.keras.layers.Layer): 


  Example

  --------

  See: MultiGraphConvolutionLayer for using in layers.

  See: MolGANMultiConvolutionLayer for using in layers.



  vertices = 9

  nodes = 5

  edges = 5

  units = 128



  layer = GraphConvolutionLayer(units=units,edges=edges)

  layer = MolGANConvolutionLayer(units=units,edges=edges)

  adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

  node_tensor = layers.Input(shape=(vertices,nodes))

  hidden1 = layer([adjacency_tensor,node_tensor])
@@ -430,7 +430,7 @@ class GraphConvolutionLayer(tf.keras.layers.Layer): 
      Name of the layer

    """



    super(GraphConvolutionLayer, self).__init__(name=name, **kwargs)

    super(MolGANConvolutionLayer, self).__init__(name=name, **kwargs)

    self.activation = activation

    self.dropout_rate = dropout_rate

    self.units = units
@@ -462,7 +462,7 @@ class GraphConvolutionLayer(tf.keras.layers.Layer): 
    """



    ic = len(inputs)

    assert ic > 1, "GraphConvolutionLayer requires at least two inputs: [adjacency_tensor, node_features_tensor]"

    assert ic > 1, "MolGANConvolutionLayer requires at least two inputs: [adjacency_tensor, node_features_tensor]"



    adjacency_tensor = inputs[0]

    node_tensor = inputs[1]
@@ -489,7 +489,7 @@ class GraphConvolutionLayer(tf.keras.layers.Layer): 
    Returns config dictionary for this layer.

    """



    config = super(GraphConvolutionLayer, self).get_config()

    config = super(MolGANConvolutionLayer, self).get_config()

    config["activation"] = self.activation

    config["dropout_rate"] = self.dropout_rate

    config["units"] = self.units
@@ -497,13 +497,13 @@ class GraphConvolutionLayer(tf.keras.layers.Layer): 
    return config





class GraphAggregationLayer(tf.keras.layers.Layer):

class MolGANAggregationLayer(tf.keras.layers.Layer):

  """

  Graph Aggregation layer used in MolGAN model.

  MolGAN is a WGAN type model for generation of small molecules.

  Performs aggregation on tensor resulting from convolution layers.

  Given its simple nature it might be removed in future and to

  GraphEncoderLayer.

  Given its simple nature it might be removed in future and moved to

  MolGANEncoderLayer.





  Example
@@ -513,9 +513,9 @@ class GraphAggregationLayer(tf.keras.layers.Layer): 
  edges = 5

  units = 128



  layer_1 = GraphConvolutionLayer(units=units,edges=edges)

  layer_2 = GraphConvolutionLayer(units=units,edges=edges)

  layer_3 = GraphAggregationLayer(units=128)

  layer_1 = MolGANConvolutionLayer(units=units,edges=edges)

  layer_2 = MolGANConvolutionLayer(units=units,edges=edges)

  layer_3 = MolGANAggregationLayer(units=128)

  adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

  node_tensor = layers.Input(shape=(vertices,nodes))

  hidden_1 = layer_1([adjacency_tensor,node_tensor])
@@ -550,7 +550,7 @@ class GraphAggregationLayer(tf.keras.layers.Layer): 
      Name of the layer

    """



    super(GraphAggregationLayer, self).__init__(name=name, **kwargs)

    super(MolGANAggregationLayer, self).__init__(name=name, **kwargs)

    self.units = units

    self.activation = activation

    self.dropout_rate = dropout_rate
@@ -590,7 +590,7 @@ class GraphAggregationLayer(tf.keras.layers.Layer): 
    Returns config dictionary for this layer.

    """



    config = super(GraphAggregationLayer, self).get_config()

    config = super(MolGANAggregationLayer, self).get_config()

    config["units"] = self.units

    config["activation"] = self.activation

    config["dropout_rate"] = self.dropout_rate
@@ -598,14 +598,14 @@ class GraphAggregationLayer(tf.keras.layers.Layer): 
    return config





class MultiGraphConvolutionLayer(tf.keras.layers.Layer):

class MolGANMultiConvolutionLayer(tf.keras.layers.Layer):

  """

  Multiple pass convolution layer used in MolGAN model.

  MolGAN is a WGAN type model for generation of small molecules.

  It takes outputs of previous convolution layer and uses

  them as inputs for the next one.

  It simplifies the overall framework, but might be moved to

  GraphEncoderLayer in the future in order to reduce number of layers.

  MolGANEncoderLayer in the future in order to reduce number of layers.



  Example

  --------
@@ -614,8 +614,8 @@ class MultiGraphConvolutionLayer(tf.keras.layers.Layer): 
  edges = 5

  units = 128



  layer_1 = MultiGraphConvolutionLayer(units=(128,64))

  layer_2 = GraphAggregationLayer(units=128)

  layer_1 = MolGANMultiConvolutionLayer(units=(128,64))

  layer_2 = MolGANAggregationLayer(units=128)

  adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

  node_tensor = layers.Input(shape=(vertices,nodes))

  hidden = layer_1([adjacency_tensor,node_tensor])
@@ -654,7 +654,7 @@ class MultiGraphConvolutionLayer(tf.keras.layers.Layer): 
      Name of the layer

    """



    super(MultiGraphConvolutionLayer, self).__init__(name=name, **kwargs)

    super(MolGANMultiConvolutionLayer, self).__init__(name=name, **kwargs)

    assert len(units) > 1, "Layer requires at least two values"



    self.units = units
@@ -662,11 +662,11 @@ class MultiGraphConvolutionLayer(tf.keras.layers.Layer): 
    self.dropout_rate = dropout_rate

    self.edges = edges



    self.first_convolution = GraphConvolutionLayer(

    self.first_convolution = MolGANConvolutionLayer(

        self.units[0], self.activation, self.dropout_rate, self.edges)

    self.gcl = [

        GraphConvolutionLayer(u, self.activation, self.dropout_rate, self.edges)

        for u in self.units[1:]

        MolGANConvolutionLayer(u, self.activation, self.dropout_rate,

                               self.edges) for u in self.units[1:]

    ]



  def call(self, inputs, training=False):
@@ -705,7 +705,7 @@ class MultiGraphConvolutionLayer(tf.keras.layers.Layer): 
    Returns config dictionary for this layer.

    """



    config = super(MultiGraphConvolutionLayer, self).get_config()

    config = super(MolGANMultiConvolutionLayer, self).get_config()

    config["units"] = self.units

    config["activation"] = self.activation

    config["dropout_rate"] = self.dropout_rate
@@ -713,7 +713,7 @@ class MultiGraphConvolutionLayer(tf.keras.layers.Layer): 
    return config





class GraphEncoderLayer(tf.keras.layers.Layer):

class MolGANEncoderLayer(tf.keras.layers.Layer):

  """

  Main learning layer used by MolGAN model.

  MolGAN is a WGAN type model for generation of small molecules.
@@ -730,7 +730,7 @@ class GraphEncoderLayer(tf.keras.layers.Layer): 
    adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

    node_tensor = layers.Input(shape=(vertices, nodes))



    graph = GraphEncoderLayer(units = [(128,64),128],

    graph = MolGANEncoderLayer(units = [(128,64),128],

                              dropout_rate= dropout_rate,

                              edges=edges)([adjacency_tensor,node_tensor])

    dense = layers.Dense(units=128, activation='tanh')(graph)
@@ -760,7 +760,7 @@ class GraphEncoderLayer(tf.keras.layers.Layer): 
    Parameters

    ---------

    units: List, optional (default=[(128, 64), 128])

      List of units for MultiGraphConvolutionLayer and GraphAggregationLayer

      List of units for MolGANMultiConvolutionLayer and GraphAggregationLayer

      i.e. [(128,64),128] means two convolution layers dims = [128,64]

      followed by aggregation layer dims=128

    activation: function, optional (default=Tanh)
@@ -774,17 +774,17 @@ class GraphEncoderLayer(tf.keras.layers.Layer): 
      Name of the layer

    """



    super(GraphEncoderLayer, self).__init__(name=name, **kwargs)

    super(MolGANEncoderLayer, self).__init__(name=name, **kwargs)

    assert len(units) == 2

    self.graph_convolution_units, self.auxiliary_units = units

    self.activation = activation

    self.dropout_rate = dropout_rate

    self.edges = edges



    self.multi_graph_convolution_layer = MultiGraphConvolutionLayer(

    self.multi_graph_convolution_layer = MolGANMultiConvolutionLayer(

        self.graph_convolution_units, self.activation, self.dropout_rate,

        self.edges)

    self.graph_aggregation_layer = GraphAggregationLayer(

    self.graph_aggregation_layer = MolGANAggregationLayer(

        self.auxiliary_units, self.activation, self.dropout_rate)



  def call(self, inputs, training=False):
@@ -826,7 +826,7 @@ class GraphEncoderLayer(tf.keras.layers.Layer): 
    Returns config dictionary for this layer.

    """



    config = super(GraphEncoderLayer, self).get_config()

    config = super(MolGANEncoderLayer, self).get_config()

    config["graph_convolution_units"] = self.graph_convolution_units

    config["auxiliary_units"] = self.auxiliary_units

    config["activation"] = self.activation

deepchem/models/tests/test_molgan_layers.py

+5 −5
Original line number Diff line number Diff line
@@ -3,7 +3,7 @@ import unittest 
from tensorflow import keras

from tensorflow.keras.layers import Input

from tensorflow.keras import activations

from deepchem.models.layers import GraphConvolutionLayer, MultiGraphConvolutionLayer, GraphAggregationLayer, GraphEncoderLayer

from deepchem.models.layers import MolGANConvolutionLayer, MolGANMultiConvolutionLayer, MolGANAggregationLayer, MolGANEncoderLayer





class test_molgan_layers(unittest.TestCase):
@@ -17,7 +17,7 @@ class test_molgan_layers(unittest.TestCase): 
    edges = 5

    units = 128



    layer = GraphConvolutionLayer(units=units, edges=edges)

    layer = MolGANConvolutionLayer(units=units, edges=edges)

    adjacency_tensor = Input(shape=(vertices, vertices, edges))

    node_tensor = Input(shape=(vertices, nodes))

    output = layer([adjacency_tensor, node_tensor])
@@ -36,7 +36,7 @@ class test_molgan_layers(unittest.TestCase): 
    vertices = 9

    units = 128



    layer = GraphAggregationLayer(units=units)

    layer = MolGANAggregationLayer(units=units)

    hidden_tensor = Input(shape=(vertices, units))

    output = layer(hidden_tensor)

    model = keras.Model(inputs=[hidden_tensor], outputs=[output])
@@ -54,7 +54,7 @@ class test_molgan_layers(unittest.TestCase): 
    second_convolution_unit = 64

    units = [first_convolution_unit, second_convolution_unit]



    layer = MultiGraphConvolutionLayer(units=units, edges=edges)

    layer = MolGANMultiConvolutionLayer(units=units, edges=edges)

    adjacency_tensor = Input(shape=(vertices, vertices, edges))

    node_tensor = Input(shape=(vertices, nodes))

    hidden_tensor = layer([adjacency_tensor, node_tensor])
@@ -77,7 +77,7 @@ class test_molgan_layers(unittest.TestCase): 
    units = [(first_convolution_unit, second_convolution_unit),

             aggregation_unit]



    layer = GraphEncoderLayer(units=units, edges=edges)

    layer = MolGANEncoderLayer(units=units, edges=edges)

    adjacency_tensor = Input(shape=(vertices, vertices, edges))

    node_tensor = Input(shape=(vertices, nodes))

    output = layer([adjacency_tensor, node_tensor])

docs/source/api_reference/layers.rst

+4 −4
Original line number Diff line number Diff line
@@ -16,16 +16,16 @@ another tensor. DeepChem maintains an extensive collection of layers which perfo 
.. autoclass:: deepchem.models.layers.GraphGather

  :members:



.. autoclass:: deepchem.models.layers.GraphConvolutionLayer

.. autoclass:: deepchem.models.layers.MolGANConvolutionLayer

  :members:



.. autoclass:: deepchem.models.layers.GraphAggregationLayer

.. autoclass:: deepchem.models.layers.MolGANAggregationLayer

  :members:



.. autoclass:: deepchem.models.layers.MultiGraphConvolutionLayer

.. autoclass:: deepchem.models.layers.MolGANMultiConvolutionLayer

  :members:



.. autoclass:: deepchem.models.layers.GraphEncoderLayer

.. autoclass:: deepchem.models.layers.MolGANEncoderLayer

  :members:



.. autoclass:: deepchem.models.layers.LSTMStep

CRA Git | Maintained and supported by SUSTech CRA and CCSE