datasets.QM7X

class datasets.QM7X(*args: Any, **kwargs: Any)[source]

QM7-X a comprehensive dataset of > 40 physicochemical properties for ~4.2 M equilibrium and non-equilibrium structure of small organic molecules with up to seven non-hydrogen (C, N, O, S, Cl) atoms. This class adds convenient functions to download QM7-X and load the data into pytorch.

References

Parameters:

  • datapath – path to dataset

  • batch_size – (train) batch size

  • raw_data_path – path to raw data. If None use tmp dir otherwise persist data and not remove it.

  • remove_duplicates – remove duplicated equilibrium structures with different non-equilibrium structures

  • only_equilibrium – only use equilibrium structures

  • num_train – number of training examples

  • num_val – number of validation examples

  • num_test – number of test examples

  • split_file – path to npz file with data partitions

  • format – dataset format

  • load_properties – subset of properties to load

  • val_batch_size – validation batch size. If None, use test_batch_size, then batch_size.

  • test_batch_size – test batch size. If None, use val_batch_size, then batch_size.

  • transforms – Transform applied to each system separately before batching.

  • train_transforms – Overrides transform_fn for training.

  • val_transforms – Overrides transform_fn for validation.

  • test_transforms – Overrides transform_fn for testing.

  • num_workers – Number of data loader workers.

  • num_val_workers – Number of validation data loader workers (overrides num_workers).

  • num_test_workers – Number of test data loader workers (overrides num_workers).

  • property_units – Dictionary from property to corresponding unit as a string (eV, kcal/mol, …).

  • distance_unit – Unit of the atom positions and cell as a string (Ang, Bohr, …).

  • data_workdir – Copy data here as part of setup, e.g. cluster scratch for faster performance.

  • splitting – Method to generate train/validation/test partitions (default: GroupSplit(splitting_key=”smiles_id”))