Commit fc2c81d7 authored by Bharath Ramsundar's avatar Bharath Ramsundar
Browse files

Changes

parent cd076a8a

deepchem/trans/tests/test_IRV.py

0 → 100644
+28 −0
Original line number Diff line number Diff line 
import deepchem as dc

import numpy as np





def test_IRV_transformer():

  n_features = 128

  n_samples = 20

  test_samples = 5

  n_tasks = 2

  X = np.random.randint(2, size=(n_samples, n_features))

  y = np.zeros((n_samples, n_tasks))

  w = np.ones((n_samples, n_tasks))

  dataset = dc.data.NumpyDataset(X, y, w, ids=None)

  X_test = np.random.randint(2, size=(test_samples, n_features))

  y_test = np.zeros((test_samples, n_tasks))

  w_test = np.ones((test_samples, n_tasks))

  test_dataset = dc.data.NumpyDataset(X_test, y_test, w_test, ids=None)

  sims = np.sum(

      X_test[0, :] * X, axis=1, dtype=float) / np.sum(

          np.sign(X_test[0, :] + X), axis=1, dtype=float)

  sims = sorted(sims, reverse=True)

  IRV_transformer = dc.trans.IRVTransformer(10, n_tasks, dataset)

  test_dataset_trans = IRV_transformer.transform(test_dataset)

  dataset_trans = IRV_transformer.transform(dataset)

  assert test_dataset_trans.X.shape == (test_samples, 20 * n_tasks)

  assert np.allclose(test_dataset_trans.X[0, :10], sims[:10])

  assert np.allclose(test_dataset_trans.X[0, 10:20], [0] * 10)

  assert not np.isclose(dataset_trans.X[0, 0], 1.)

deepchem/trans/tests/test_transformers.py

+6 −57
Original line number Diff line number Diff line
@@ -9,32 +9,20 @@ import unittest 
import numpy as np

import deepchem as dc

import scipy.ndimage

from deepchem.trans.transformers import DataTransforms





def load_solubility_data():

  """Loads solubility dataset"""

  current_dir = os.path.dirname(os.path.abspath(__file__))

  featurizer = dc.feat.CircularFingerprint(size=1024)

  tasks = ["log-solubility"]

  task_type = "regression"

  input_file = os.path.join(current_dir, "../../models/tests/example.csv")

  loader = dc.data.CSVLoader(

      tasks=tasks, smiles_field="smiles", featurizer=featurizer)



  return loader.create_dataset(input_file)





class TestTransformers(unittest.TestCase):

class TestDataTransforms(unittest.TestCase):

  """

  Test top-level API for transformer objects.

  Test DataTransforms for images 

  """



  def setUp(self):

    """

     init to load the MNIST data for DataTransforms Tests

    """

    super(TestTransformers, self).setUp()

    self.current_dir = os.path.dirname(os.path.abspath(__file__))

    '''

       init to load the MNIST data for DataTransforms Tests

      '''

    (x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()

    train = dc.data.NumpyDataset(x_train, y_train)

    # extract only the images (no need of the labels)
@@ -43,45 +31,6 @@ class TestTransformers(unittest.TestCase): 
    data = np.reshape(data, (28, 28))

    self.d = data



  def test_coulomb_fit_transformer(self):

    """Test coulomb fit transformer on singletask dataset."""

    n_samples = 10

    n_features = 3

    n_tasks = 1

    ids = np.arange(n_samples)

    X = np.random.rand(n_samples, n_features, n_features)

    y = np.zeros((n_samples, n_tasks))

    w = np.ones((n_samples, n_tasks))

    dataset = dc.data.NumpyDataset(X, y, w, ids)

    fit_transformer = dc.trans.CoulombFitTransformer(dataset)

    X_t = fit_transformer.X_transform(dataset.X)

    assert len(X_t.shape) == 2



  def test_IRV_transformer(self):

    n_features = 128

    n_samples = 20

    test_samples = 5

    n_tasks = 2

    X = np.random.randint(2, size=(n_samples, n_features))

    y = np.zeros((n_samples, n_tasks))

    w = np.ones((n_samples, n_tasks))

    dataset = dc.data.NumpyDataset(X, y, w, ids=None)

    X_test = np.random.randint(2, size=(test_samples, n_features))

    y_test = np.zeros((test_samples, n_tasks))

    w_test = np.ones((test_samples, n_tasks))

    test_dataset = dc.data.NumpyDataset(X_test, y_test, w_test, ids=None)

    sims = np.sum(

        X_test[0, :] * X, axis=1, dtype=float) / np.sum(

            np.sign(X_test[0, :] + X), axis=1, dtype=float)

    sims = sorted(sims, reverse=True)

    IRV_transformer = dc.trans.IRVTransformer(10, n_tasks, dataset)

    test_dataset_trans = IRV_transformer.transform(test_dataset)

    dataset_trans = IRV_transformer.transform(dataset)

    assert test_dataset_trans.X.shape == (test_samples, 20 * n_tasks)

    assert np.allclose(test_dataset_trans.X[0, :10], sims[:10])

    assert np.allclose(test_dataset_trans.X[0, 10:20], [0] * 10)

    assert not np.isclose(dataset_trans.X[0, 0], 1.)



  def test_blurring(self):

    # Check Blurring

    dt = DataTransforms(self.d)

deepchem/trans/transformers.py

+84 −30
Original line number Diff line number Diff line
@@ -889,6 +889,8 @@ class BalancingTransformer(Transformer): 
  See Also

  --------

  deepchem.trans.DuplicateBalancingTransformer: Balance by duplicating samples.





  Note

  ----

  This transformer is only meaningful for classification datasets where `y`
@@ -1385,19 +1387,51 @@ class CoulombFitTransformer(Transformer): 




class IRVTransformer(Transformer):

  """Performs transform from ECFP to IRV features(K nearest neighbors)."""

  """Performs transform from ECFP to IRV features(K nearest neighbors).



  This transformer is required by `MultitaskIRVClassifier` as a preprocessing

  step before training.



  Examples

  --------

  Let's start by defining the parameters of the dataset we're about to

  transform.



  >>> n_feat = 128

  >>> N = 20

  >>> n_tasks = 2



  Let's now make our dataset object



  >>> import numpy as np

  >>> import deepchem as dc

  >>> X = np.random.randint(2, size=(N, n_feat))

  >>> y = np.zeros((N, n_tasks))

  >>> w = np.ones((N, n_tasks))

  >>> dataset = dc.data.NumpyDataset(X, y, w)



  And let's apply our transformer with 10 nearest neighbors.



  >>> K = 10

  >>> trans = dc.trans.IRVTransformer(K, n_tasks, dataset)

  >>> dataset = trans.transform(dataset)



  Note

  ----

  This class requires TensorFlow to be installed.

  """



  def __init__(self, K, n_tasks, dataset, transform_y=False, transform_x=False):

  def __init__(self, K, n_tasks, dataset):

    """Initializes IRVTransformer.



    Parameters

    ----------

    dataset: dc.data.Dataset object

      train_dataset

    K: int

      number of nearest neighbours being count

    n_tasks: int

      number of tasks

    dataset: dc.data.Dataset object

      train_dataset

    """

    self.X = dataset.X

    self.n_tasks = n_tasks
@@ -1424,7 +1458,6 @@ class IRVTransformer(Transformer): 
    features: list

      n_samples * np.array of size (2*K,)

      each array includes K similarity values and corresponding labels



    """

    features = []

    similarity_xs = similarity * np.sign(w)
@@ -1478,13 +1511,13 @@ class IRVTransformer(Transformer): 
    """

    X_target2 = []

    n_features = X_target.shape[1]

    print('start similarity calculation')

    logger.info('start similarity calculation')

    time1 = time.time()

    similarity = IRVTransformer.matrix_mul(X_target, np.transpose(

        self.X)) / (n_features - IRVTransformer.matrix_mul(

            1 - X_target, np.transpose(1 - self.X)))

    time2 = time.time()

    print('similarity calculation takes %i s' % (time2 - time1))

    logger.info('similarity calculation takes %i s' % (time2 - time1))

    for i in range(self.n_tasks):

      X_target2.append(self.realize(similarity, self.y[:, i], self.w[:, i]))

    return np.concatenate([z for z in np.array(X_target2)], axis=1)
@@ -1526,6 +1559,21 @@ class IRVTransformer(Transformer): 
    return all_result



  def transform(self, dataset, parallel=False, out_dir=None, **kwargs):

    """Transforms a given dataset



    Parameters

    ----------

    dataset: Dataset

      Dataset to transform

    parallel: bool, optional, (default False)

      Whether to parallelize this transformation. Currently ignored.

    out_dir: str, optional (default None)

      Directory to write resulting dataset.



    Returns

    -------

    `Dataset` object that is transformed.

    """

    X_length = dataset.X.shape[0]

    X_trans = []

    for count in range(X_length // 5000 + 1):
@@ -1733,6 +1781,10 @@ class ImageTransformer(Transformer): 


class ANITransformer(Transformer):

  """Performs transform from 3D coordinates to ANI symmetry functions



  Note

  ----

  This class requires TensorFlow to be installed.

  """



  def __init__(self,
@@ -1777,7 +1829,7 @@ class ANITransformer(Transformer): 
            [self.outputs], feed_dict={self.inputs: X_batch})[0]

        X_out.append(output)

        num_transformed = num_transformed + X_batch.shape[0]

        print('%i samples transformed' % num_transformed)

        logger.info('%i samples transformed' % num_transformed)

        start += 1

        if end >= len(X):

          break
@@ -1794,7 +1846,8 @@ class ANITransformer(Transformer): 
    """ tensorflow computation graph for transform """

    graph = tf.Graph()

    with graph.as_default():

      self.inputs = tf.placeholder(tf.float32, shape=(None, self.max_atoms, 4))

      self.inputs = tf.keras.Input(

          dtype=tf.float32, shape=(None, self.max_atoms, 4))

      atom_numbers = tf.cast(self.inputs[:, :, 0], tf.int32)

      flags = tf.sign(atom_numbers)

      flags = tf.cast(
@@ -1833,7 +1886,8 @@ class ANITransformer(Transformer): 
    # Cutoff with threshold Rc

    d_flag = flags * tf.sign(cutoff - d)

    d_flag = tf.nn.relu(d_flag)

    d_flag = d_flag * tf.expand_dims((1 - tf.eye(self.max_atoms)), 0)

    d_flag = d_flag * tf.expand_dims(

        tf.expand_dims((1 - tf.eye(self.max_atoms)), 0), -1)

    d = 0.5 * (tf.cos(np.pi * d / cutoff) + 1)

    return d * d_flag
@@ -1936,7 +1990,7 @@ class FeaturizationTransformer(Transformer): 
  >>> X = np.array(smiles)

  >>> y = np.array([1, 0])

  >>> dataset = dc.data.NumpyDataset(X, y)

  >>> trans = FeaturizerTransformer(dataset, dc.feat.CircularFingerprint())

  >>> trans = dc.trans.FeaturizationTransformer(dataset, dc.feat.CircularFingerprint())

  >>> dataset = trans.transform(dataset)

  """

CRA Git | Maintained and supported by SUSTech CRA and CCSE