Bases: ABC
Dataset to load, transform, shuffle the data before training
Functions
__getitem__
abstractmethod
Returns the data of the item/batch at index
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def __getitem__(self, index: int):
"""Returns the data of the item/batch at index"""
pass
|
__len__
abstractmethod
Returns the number of batches/items
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def __len__(self):
"""Returns the number of batches/items"""
pass
|
get_data_by_key
abstractmethod
get_data_by_key(data_key)
Returns the data by the key (identifier of the data)
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def get_data_by_key(self, data_key):
"""Returns the data by the key (identifier of the data)"""
|
get_datainfo
abstractmethod
get_datainfo(batch_index)
returns the datainfo for the batch at index
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def get_datainfo(self, batch_index: int) -> List[DataInfo]:
"""returns the datainfo for the batch at index"""
|
get_dataset_stats
Returns the dataset stats
Source code in niceml/data/datasets/dataset.py
| def get_dataset_stats(self) -> dict:
"""Returns the dataset stats"""
return dict(size=self.get_item_count())
|
get_item_count
abstractmethod
Returns the current count of items in the dataset
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def get_item_count(self) -> int:
"""Returns the current count of items in the dataset"""
|
get_items_per_epoch
abstractmethod
Returns the items per epoch
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def get_items_per_epoch(self) -> int:
"""Returns the items per epoch"""
|
get_set_name
abstractmethod
Returns the name of the set e.g. train
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def get_set_name(self) -> str:
"""Returns the name of the set e.g. train"""
|
initialize
abstractmethod
initialize(data_description, exp_context)
Initializes with the data description and context
Source code in niceml/data/datasets/dataset.py
| @abstractmethod
def initialize(
self, data_description: DataDescription, exp_context: ExperimentContext
):
"""Initializes with the data description and context"""
|
iter_with_info
Iterates over the dataset and adds the data_info to each data
Source code in niceml/data/datasets/dataset.py
| def iter_with_info(self) -> Iterable:
"""Iterates over the dataset and adds the data_info to each data"""
return DataIterator(self)
|