rxn.onmt_models.training_files.RxnPreprocessingFiles

class rxn.onmt_models.training_files.RxnPreprocessingFiles(processed_data_dir)

Bases: object

Class to make it easy to get the names/paths of the files generated during data preprocessing.

This assumes that the default paths were used when calling rxn-data-pipeline.

Parameters

processed_data_dir (Union[str, PathLike]) –

__init__(processed_data_dir)

Parameters

processed_data_dir (Union[str, PathLike]) –

Methods

__init__(processed_data_dir)	param processed_data_dir
augmented(data_path)	Get the path for the augmented version of a data file.
get_context_src_for_split(split)	param split
get_context_tags_for_split(split)	param split
get_context_tgt_for_split(split)	param split
get_precursors_for_split(split)	param split
get_processed_csv_for_split(split)	param split
get_products_for_split(split)	param split
get_src_file(split, model_task)	Get the source file for the given task.
get_tgt_file(split, model_task)	Get the target file for the given task.

Attributes

FILENAME_ROOT
processed_csv	rtype Path
processed_test_csv	rtype Path
processed_train_csv	rtype Path
processed_validation_csv	rtype Path
standardized_csv	rtype Path
test_precursors	rtype Path
test_products	rtype Path
train_precursors	rtype Path
train_products	rtype Path
validation_precursors	rtype Path
validation_products	rtype Path

static augmented(data_path)

Get the path for the augmented version of a data file.

Parameters

data_path (Path) –

Return type

Path

get_src_file(split, model_task)

Get the source file for the given task.

Note: the file is tokenized for the forward and retro tasks, but not for the context task.

Parameters

• split (str) –

• model_task (str) –

Return type

Path

get_tgt_file(split, model_task)

Get the target file for the given task.

Note: the file is tokenized for the forward and retro tasks, but not for the context task.

Parameters

• split (str) –

• model_task (str) –

Return type

Path