Add MoleculeNet wrappers for Factors/Kinase/UV datasets

I recently ran python benchmark.py from examples/ using 1 K80 on a Amazon EC2 p2.xlarge instance. The benchmark failed after 18.5 hours with the following error:

Traceback (most recent call last):
  File "benchmark.py", line 1097, in <module>
    test=test)
  File "benchmark.py", line 179, in benchmark_loading_datasets
    featurizer=featurizer)
  File "/home/ubuntu/deepchem/examples/kaggle/kaggle_datasets.py", line 139, in load_kaggle
    shard_size=shard_size)
  File "/home/ubuntu/deepchem/examples/kaggle/kaggle_datasets.py", line 74, in gen_kaggle
    train_dataset = loader.featurize(train_files, shard_size=shard_size)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/deepchem/data/data_loader.py", line 197, in featurize
    shard_generator(), data_dir, self.tasks, verbose=self.verbose)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/deepchem/data/datasets.py", line 440, in create_dataset
    for shard_num, (X, y, w, ids) in enumerate(shard_generator):
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/deepchem/data/data_loader.py", line 174, in shard_generator
    self.get_shards(input_files, shard_size)):
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/deepchem/utils/save.py", line 98, in load_csv_files
    for df in pd.read_csv(filename, chunksize=shard_size):
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/pandas/io/parsers.py", line 646, in parser_f
    return _read(filepath_or_buffer, kwds)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/pandas/io/parsers.py", line 389, in _read
    parser = TextFileReader(filepath_or_buffer, **kwds)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/pandas/io/parsers.py", line 730, in __init__
    self._make_engine(self.engine)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/pandas/io/parsers.py", line 923, in _make_engine
    self._engine = CParserWrapper(self.f, **self.options)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/site-packages/pandas/io/parsers.py", line 1390, in __init__
    self._reader = _parser.TextReader(src, **kwds)
  File "pandas/parser.pyx", line 373, in pandas.parser.TextReader.__cinit__ (pandas/parser.c:4184)
  File "pandas/parser.pyx", line 609, in pandas.parser.TextReader._setup_parser_source (pandas/parser.c:7348)
  File "/home/ubuntu/anaconda3/envs/deepchem/lib/python3.5/gzip.py", line 163, in __init__
    fileobj = self.myfileobj = builtins.open(filename, mode or 'rb')
FileNotFoundError: [Errno 2] No such file or directory: '/home/ubuntu/deepchem/examples/kaggle/KAGGLE_training_disguised_combined_full.csv.gz'

The time elapsed was: real 1125m29.127s user 1723m4.204s sys 170m16.504s

This issue is to ensure that the /home/ubuntu/deepchem/examples/kaggle/KAGGLE_training_disguised_combined_full.csv.gz file is either made present before benchmarking, and if not to handle more gracefully.

[rbharath]

Haven’t had a chance to write a MoleculeNet load function, but the datasets are online at:

https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/FACTORS_test1_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/FACTORS_test2_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/FACTORS_training_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/KINASE_test1_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/KINASE_test2_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/KINASE_training_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/UV_test1_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/UV_test2_disguised_combined_full.csv.gz https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/UV_training_disguised_combined_full.csv.gz

Contributions welcome for adding simple dc.molnet wrappers. Will rename the issue to reflect that.

[rbharath]

This one’s on me. Ping me in a few weeks if I haven’t taken care of it yet.