create_custom_splits

create_split_file_qmugs_bin(dataset: str, override: bool)

Create a split file containing the first bin of QMUGS.

create_split_file_qmugs_bin0_qm9(dataset: str, override: bool)

Create a split file containing qm9 train set + the first bin of QMUGS.

create_split_file_qmugs_large_bins(dataset: str, override: bool)

Create a split file containing only validation molecules for the QMUGSLargeBins dataset.

Train and test splits are empty.

get_qmugs_paths_and_group_ids(qmugs_path: Path, csv_file: Path) → tuple[list[Path], list[int]]

Determine paths of QMUGS label files based on the id given in the csv file.