ewokstomo.tasks.online.nabu_pipeline.get_nabu_pipeline

ewokstomo.tasks.online.nabu_pipeline.get_nabu_pipeline(config_dict, dataset_info, projs_batch_size, pipeline_backend='cuda')

Create a reconstruction pipeline object, based on configuration and scan info.

Parameters

config_dict: dict

dict of dict, see ‘default_nabu_config’ for structure and keys.

dataset_info: nabu.resources.dataset.Dataset

Dataset structure, InMemoryDataset in this case

projs_batch_size: int

Number of projections processed at once ; i.e we trigger the pipeline as soon as we have ‘projs_batch_size’ in memory