get_failed_steps_for_dataset#

coffea.dataset_tools.get_failed_steps_for_dataset(dataset: dict | DatasetSpec, report: awkward.Array) → dict | DatasetSpec[source]#

Modify the input dataset to only contain the files and row-ranges for failed processing jobs as specified in the supplied report.

Parameters:

  • dataset (DatasetSpec | dict) – The dataset to be reduced to only contain files and row-ranges that have previously encountered failed file access.

  • report (awkward.Array) – The computed file-access error report from dask-awkward.

Returns:

out – The reduced dataset with only the row-ranges and files that failed processing, according to the input report.

Return type:

DatasetSpec | dict