Commit f4d76863 authored by Bharath Ramsundar's avatar Bharath Ramsundar
Browse files

Changes

parent bf289292

deepchem/molnet/load_function/hppb_datasets.py

+25 −3
Original line number Diff line number Diff line
@@ -19,9 +19,27 @@ def remove_missing_entries(dataset): 
  feature vectors. Get rid of them.

  """

  for i, (X, y, w, ids) in enumerate(dataset.itershards()):

    available_rows = X.any(axis=1)

    ###########################################

    print("X.shape")

    print(X.shape)

    print("X[:10]")

    print(X[:10])

    print("type(X)")

    print(type(X))

    ###########################################

    #available_rows = X.any(axis=1)

    available_rows = np.any(X)

    ###########################################

    print("type(available_rows)")

    print(type(available_rows))

    ###########################################

    logger.info("Shard %d has %d missing entries." %

                (i, np.count_nonzero(~available_rows)))

                #(i, np.count_nonzero(~available_rows)))

                (i, len(X) - len(available_rows)))

    ##################################################

    print("type(available_rows)")

    print(type(available_rows))

    ##################################################

    X = X[available_rows]

    y = y[available_rows]

    w = w[available_rows]
@@ -61,7 +79,11 @@ def load_hppb(featurizer="ECFP", 
  dataset_file = os.path.join(data_dir, "hppb.csv")

  if not os.path.exists(dataset_file):

    logger.info("{} does not exist. Downloading it.".format(dataset_file))

    deepchem.utils.download_url(url=hppb_URL, dest_dir=data_dir)

    ################################################

    print("data_dir")

    print(data_dir)

    ################################################

    deepchem.utils.download_url(url=HPPB_URL, dest_dir=data_dir)



  if featurizer == 'ECFP':

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

examples/adme/README.md

+6 −4
Original line number Diff line number Diff line
@@ -21,12 +21,14 @@ Note that PPB dataset labels are transformed using %bound -> log(1 - %bound). 


# Running Benchmark

```sh

$ python run_benchmark.py model split dataset

$ python run_benchmark.py

```



- models: {GraphConv, PDNN, RF, SVR}

- splits: {scaffold, random, index}

- dataset: {az_clearance.csv, az_hppb.csv, az_logd.csv}

You can manually edit variables within `run_benchmarks.py` to the following values and fun



- `MODEL`: {GraphConv, RF, SVR}

- `SPLIT`: {scaffold, random, index}

- `DATASET_NAME`: {clearance, hppb}



License

----

examples/adme/__init__.py

deleted100644 → 0
+0 −0

Empty file deleted.

examples/adme/run_benchmarks.py

+60 −66
Original line number Diff line number Diff line 
# DeepCrystal Technologies 2017 - Patrick Hop

# MIT License - have fun!!



from __future__ import print_function

from __future__ import division

from __future__ import unicode_literals



import os

import numpy as np

np.random.seed(123)

from sklearn.ensemble import RandomForestRegressor

from sklearn import svm



import tensorflow as tf

tf.random.set_seed(123)

import deepchem as dc

from deepchem.models.graph_models import GraphConvModel

from sklearn.ensemble import RandomForestRegressor

from deepchem.models import GraphConvModel



MODEL = "GraphConv"

SPLIT = "scaffold"

DATASET_NAME = "hppb"





BATCH_SIZE = 128

# Set to higher values to get better numbers

MAX_EPOCH = 1

LR = 1e-3

LMBDA = 1e-4





def retrieve_datasets():

  os.system(

      'wget -c %s' %

      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_logd.csv')

  os.system(

      'wget -c %s' %

      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_hppb.csv')

  os.system(

      'wget -c %s' %

      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_clearance.csv'

  )





def load_dataset(dataset_file, featurizer='ECFP', split='index'):

  tasks = ['exp']



  if featurizer == 'ECFP':

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

  elif featurizer == 'GraphConv':

    featurizer = dc.feat.ConvMolFeaturizer()



  loader = dc.data.CSVLoader(

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

  dataset = loader.featurize(dataset_file, shard_size=8192)



  transformers = [

      dc.trans.NormalizationTransformer(transform_y=True, dataset=dataset)

  ]

  for transformer in transformers:

    dataset = transformer.transform(dataset)



  splitters = {

      'index': dc.splits.IndexSplitter(),

      'random': dc.splits.RandomSplitter(),

      'scaffold': dc.splits.ScaffoldSplitter()

  }

  splitter = splitters[split]

  train, valid, test = splitter.train_valid_test_split(dataset)

  return tasks, (train, valid, test), transformers





def experiment(dataset_file, method='GraphConv', split='scaffold'):



#def retrieve_datasets():

#  os.system(

#      'wget -c %s' %

#      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_logd.csv')

#  os.system(

#      'wget -c %s' %

#      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_hppb.csv')

#  os.system(

#      'wget -c %s' %

#      'https://s3-us-west-1.amazonaws.com/deep-crystal-california/az_clearance.csv'

#  )





def load_dataset(dataset_name, featurizer='ECFP', split='index'):

  #tasks = ['exp']



  #if featurizer == 'ECFP':

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

  #elif featurizer == 'GraphConv':

  #  featurizer = dc.feat.ConvMolFeaturizer()



  #loader = dc.data.CSVLoader(

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

  #dataset = loader.featurize(dataset_file, shard_size=8192)

  if dataset_name.lower() == "clearance":

    dataset = dc.molnet.load_clearance(featurizer=featurizer, split=split)

  elif dataset_name.lower() == "hppb":

    dataset = dc.molnet.load_hppb(featurizer=featurizer, split=split)



  #transformers = [

  #    dc.trans.NormalizationTransformer(transform_y=True, dataset=dataset)

  #]

  #for transformer in transformers:

  #  dataset = transformer.transform(dataset)



  #splitters = {

  #    'index': dc.splits.IndexSplitter(),

  #    'random': dc.splits.RandomSplitter(),

  #    'scaffold': dc.splits.ScaffoldSplitter()

  #}

  #splitter = splitters[split]

  #train, valid, test = splitter.train_valid_test_split(dataset)

  #return tasks, (train, valid, test), transformers





def experiment(dataset_name, method='GraphConv', split='scaffold'):

  featurizer = 'ECFP'

  if method == 'GraphConv':

    featurizer = 'GraphConv'

  tasks, datasets, transformers = load_dataset(

      dataset_file, featurizer=featurizer, split=split)

      dataset_name, featurizer=featurizer, split=split)

  train, val, test = datasets



  model = None
@@ -96,18 +94,14 @@ def experiment(dataset_file, method='GraphConv', split='scaffold'): 
#======================================================================

# Run Benchmarks {GC-DNN, SVR, RF}

def main():

  print("About to retrieve datasets")

  retrieve_datasets()



  MODEL = "GraphConv"

  SPLIT = "scaffold"

  DATASET = "az_hppb.csv"

  #print("About to retrieve datasets")

  #retrieve_datasets()



  metric = dc.metrics.Metric(dc.metrics.pearson_r2_score, np.mean)



  print("About to build model")

  model, train, val, test, transformers = experiment(

      DATASET, method=MODEL, split=SPLIT)

      DATASET_NAME, method=MODEL, split=SPLIT)

  if MODEL == 'GraphConv':

    print("running GraphConv search")

    best_val_score = 0.0
@@ -122,14 +116,14 @@ def main(): 
      if latest_val_score > best_val_score:

        best_val_score = latest_val_score

        train_score = latest_train_score

    print((MODEL, SPLIT, DATASET, train_score, best_val_score))

    print((MODEL, SPLIT, DATASET_NAME, train_score, best_val_score))

  else:

    model.fit(train)

    train_score = model.evaluate(train, [metric],

                                 transformers)['mean-pearson_r2_score']

    val_score = model.evaluate(val, [metric],

                               transformers)['mean-pearson_r2_score']

    print((MODEL, SPLIT, DATASET, train_score, val_score))

    print((MODEL, SPLIT, DATASET_NAME, train_score, val_score))





if __name__ == "__main__":

CRA Git | Maintained and supported by SUSTech CRA and CCSE