core.dataset.DingoDataset

core.dataset.DingoDataset(
    file_name=None,
    dictionary=None,
    data_keys=None,
    leave_on_disk_keys=None,
)

This is a generic dataset class with save / load methods.

A common use case is to inherit multiply from DingoDataset and torch.utils.data.Dataset, in which case the subclass picks up these I/O methods, and DingoDataset is acting as a Mixin class.

Alternatively, if the torch Dataset is not needed, then DingoDataset can be subclassed directly.

For constructing, provide either file_name, or dictionary containing data and settings entries, or neither.

Parameters

Name Type Description Default
file_name str HDF5 file containing a dataset None
dictionary dict Contains settings and data entries. The data keys should be the same as save_keys None
data_keys list Variables that should be saved / loaded. This allows for class to store additional variables beyond those that are saved. Typically, this list would be provided by any subclass. None
leave_on_disk_keys Optional[list] Keys for which the values are not loaded into RAM when initializing the dataset. This reduces the memory footprint during training. Instead, the values are loaded from the HDF5 file during training. None

Attributes

Name Description
dataset_type
settings

Methods

Name Description
from_dictionary
from_file
to_dictionary
to_file

from_dictionary

core.dataset.DingoDataset.from_dictionary(dictionary)

from_file

core.dataset.DingoDataset.from_file(file_name)

to_dictionary

core.dataset.DingoDataset.to_dictionary()

to_file

core.dataset.DingoDataset.to_file(file_name, mode='w')