eurocropsml.dataset.dataset.EuroCropsDataset

class EuroCropsDataset(file_dict: dict[str, list[Path]], encode: dict[int, int], mmap_store: MMapStore, config: EuroCropsDatasetConfig, preprocess_config: EuroCropsDatasetPreprocessConfig, pad_seq_to_366: bool = False, padding_value: float = 0.0)

Bases: Dataset[LabelledData]

PyTorch dataset class for the EuroCropsML dataset.

Parameters:

• file_dict – Dictionary of file paths (one file per parcel/time series) for each satellite.

• encode – Encoding used to encode the classes into integers.

• mmap_store – Instance of memory map store.

• config – EuroCropsDatasetConfig instance.

• preprocess_config – EuroCropsDatasetPreprocessConfig instance.

• pad_seq_to_366 – If sequence should be padded to 366 days.

• padding_value – Padding value used for padding data sequences.

Methods

__init__(file_dict, encode, mmap_store, ...)