Loading README.md +1 −0 Original line number Diff line number Diff line Loading @@ -59,6 +59,7 @@ DeepChem has a number of "soft" requirements. These are packages which are neede - [RDKit](http://www.rdkit.org/docs/Install.html) - [simdna](https://github.com/kundajelab/simdna) - [XGBoost](https://xgboost.readthedocs.io/en/latest/) - [Tensorflow Probability](https://www.tensorflow.org/probability) ## Installation Loading deepchem/models/layers.py +17 −1 Original line number Diff line number Diff line # -*- coding: utf-8 -*- import tensorflow as tf import tensorflow_probability as tfp import numpy as np import collections from tensorflow.keras import activations, initializers, backend Loading Loading @@ -2184,6 +2183,17 @@ class WeaveLayer(tf.keras.layers.Layer): class WeaveGather(tf.keras.layers.Layer): """Implements the weave-gathering section of weave convolutions. Implements the gathering layer from the following paper: Kearnes, Steven, et al. "Molecular graph convolutions: moving beyond fingerprints." Journal of computer-aided molecular design 30.8 (2016): 595-608. The weave gathering layer gathers per-atom features to create a molecule-level fingerprint in a weave convolutional network. This layer can also perform Gaussian histogram expansion as detailed in the original paper. """ def __init__(self, batch_size, Loading @@ -2208,6 +2218,11 @@ class WeaveGather(tf.keras.layers.Layer): activation: str, optional Activation function applied """ try: import tensorflow_probability as tfp except ModuleNotFoundError: raise ValueError( "This class requires tensorflow-probability to be installed.") super(WeaveGather, self).__init__(**kwargs) self.n_input = n_input self.batch_size = batch_size Loading Loading @@ -2252,6 +2267,7 @@ class WeaveGather(tf.keras.layers.Layer): return output_molecules def gaussian_histogram(self, x): import tensorflow_probability as tfp gaussian_memberships = [(-1.645, 0.283), (-1.080, 0.170), (-0.739, 0.134), (-0.468, 0.118), (-0.228, 0.114), (0., 0.114), (0.228, 0.114), (0.468, 0.118), (0.739, 0.134), Loading deepchem/rl/a2c.py +17 −2 Original line number Diff line number Diff line Loading @@ -4,7 +4,6 @@ from deepchem.models import KerasModel from deepchem.models.optimizers import Adam import numpy as np import tensorflow as tf import tensorflow_probability as tfp import collections import copy import multiprocessing Loading Loading @@ -40,10 +39,20 @@ class A2CLossDiscrete(object): class A2CLossContinuous(object): """This class computes the loss function for A2C with continuous action spaces.""" """This class computes the loss function for A2C with continuous action spaces. Note ---- This class requires tensorflow-probability to be installed. """ def __init__(self, value_weight, entropy_weight, mean_index, std_index, value_index): try: import tensorflow_probability as tfp except ModuleNotFoundError: raise ValueError( "This class requires tensorflow-probability to be installed.") self.value_weight = value_weight self.entropy_weight = entropy_weight self.mean_index = mean_index Loading @@ -51,6 +60,7 @@ class A2CLossContinuous(object): self.value_index = value_index def __call__(self, outputs, labels, weights): import tensorflow_probability as tfp mean = outputs[self.mean_index] std = outputs[self.std_index] value = outputs[self.value_index] Loading Loading @@ -112,6 +122,11 @@ class A2C(object): except specifying the new goal. It should return that list of states, and the rewards that would have been received for taking the specified actions from those states. The output arrays may be shorter than the input ones, if the modified rollout would have terminated sooner. Note ---- Using this class on continuous action spaces requires that `tensorflow_probability` be installed. """ def __init__(self, Loading Loading
README.md +1 −0 Original line number Diff line number Diff line Loading @@ -59,6 +59,7 @@ DeepChem has a number of "soft" requirements. These are packages which are neede - [RDKit](http://www.rdkit.org/docs/Install.html) - [simdna](https://github.com/kundajelab/simdna) - [XGBoost](https://xgboost.readthedocs.io/en/latest/) - [Tensorflow Probability](https://www.tensorflow.org/probability) ## Installation Loading
deepchem/models/layers.py +17 −1 Original line number Diff line number Diff line # -*- coding: utf-8 -*- import tensorflow as tf import tensorflow_probability as tfp import numpy as np import collections from tensorflow.keras import activations, initializers, backend Loading Loading @@ -2184,6 +2183,17 @@ class WeaveLayer(tf.keras.layers.Layer): class WeaveGather(tf.keras.layers.Layer): """Implements the weave-gathering section of weave convolutions. Implements the gathering layer from the following paper: Kearnes, Steven, et al. "Molecular graph convolutions: moving beyond fingerprints." Journal of computer-aided molecular design 30.8 (2016): 595-608. The weave gathering layer gathers per-atom features to create a molecule-level fingerprint in a weave convolutional network. This layer can also perform Gaussian histogram expansion as detailed in the original paper. """ def __init__(self, batch_size, Loading @@ -2208,6 +2218,11 @@ class WeaveGather(tf.keras.layers.Layer): activation: str, optional Activation function applied """ try: import tensorflow_probability as tfp except ModuleNotFoundError: raise ValueError( "This class requires tensorflow-probability to be installed.") super(WeaveGather, self).__init__(**kwargs) self.n_input = n_input self.batch_size = batch_size Loading Loading @@ -2252,6 +2267,7 @@ class WeaveGather(tf.keras.layers.Layer): return output_molecules def gaussian_histogram(self, x): import tensorflow_probability as tfp gaussian_memberships = [(-1.645, 0.283), (-1.080, 0.170), (-0.739, 0.134), (-0.468, 0.118), (-0.228, 0.114), (0., 0.114), (0.228, 0.114), (0.468, 0.118), (0.739, 0.134), Loading
deepchem/rl/a2c.py +17 −2 Original line number Diff line number Diff line Loading @@ -4,7 +4,6 @@ from deepchem.models import KerasModel from deepchem.models.optimizers import Adam import numpy as np import tensorflow as tf import tensorflow_probability as tfp import collections import copy import multiprocessing Loading Loading @@ -40,10 +39,20 @@ class A2CLossDiscrete(object): class A2CLossContinuous(object): """This class computes the loss function for A2C with continuous action spaces.""" """This class computes the loss function for A2C with continuous action spaces. Note ---- This class requires tensorflow-probability to be installed. """ def __init__(self, value_weight, entropy_weight, mean_index, std_index, value_index): try: import tensorflow_probability as tfp except ModuleNotFoundError: raise ValueError( "This class requires tensorflow-probability to be installed.") self.value_weight = value_weight self.entropy_weight = entropy_weight self.mean_index = mean_index Loading @@ -51,6 +60,7 @@ class A2CLossContinuous(object): self.value_index = value_index def __call__(self, outputs, labels, weights): import tensorflow_probability as tfp mean = outputs[self.mean_index] std = outputs[self.std_index] value = outputs[self.value_index] Loading Loading @@ -112,6 +122,11 @@ class A2C(object): except specifying the new goal. It should return that list of states, and the rewards that would have been received for taking the specified actions from those states. The output arrays may be shorter than the input ones, if the modified rollout would have terminated sooner. Note ---- Using this class on continuous action spaces requires that `tensorflow_probability` be installed. """ def __init__(self, Loading
