pytorch_pfn_extras.dataset.tabular.DelegateDataset

class pytorch_pfn_extras.dataset.tabular.DelegateDataset(dataset)

Bases: TabularDataset

A helper class to implement a TabularDataset.

This class wraps an instance of TabularDataset and provides methods of TabularDataset. This class is useful to create a custom dataset class by inheriting it.

>>> import numpy as np
>>>
>>> from pytorch_pfn_extras.dataset import tabular
>>>
>>> class MyDataset(tabular.DelegateDataset):
...
...     def __init__(self):
...         super().__init__(tabular.from_data((
...             ('a', np.arange(10)),
...             ('b', self.get_b),
...             ('c', [3, 1, 4, 5, 9, 2, 6, 8, 7, 0]),
...             (('d', 'e'), self.get_de))))
...
...     def get_b(self, i):
...         return 'b[{}]'.format(i)
...
...     def get_de(self, i):
...         return {'d': 'd[{}]'.format(i), 'e': 'e[{}]'.format(i)}
...
>>> dataset = MyDataset()
>>> len(dataset)
10
>>> dataset.keys
('a', 'b', 'c', 'd', 'e')
>>> dataset[0]
(0, 'b[0]', 3, 'd[0]', 'e[0]')

Parameters:

dataset (pytorch_pfn_extras.dataset.TabularDataset) – An underlying dataset.

Methods

__init__(dataset)
asdict()	Return a view with dict mode.
astuple()	Return a view with tuple mode.
concat(*datasets)	Stack datasets along rows.
convert(data)	Convert fetched data.
fetch()	Fetch data.
get_example(i)
get_examples(indices, key_indices)	Return a part of data.
join(*datasets)	Stack datasets along columns.
transform(keys, transform)	Apply a transform to each example.
transform_batch(keys, transform_batch)	Apply a transform to examples.
with_converter(converter)	Override the behaviour of convert().

Attributes

keys	Names of columns.
mode	Mode of representation.
slice	Get a slice of dataset.

__init__(dataset)

get_examples(indices, key_indices)

Return a part of data.

Parameters:

• indices (list of ints or slice) – Indices of requested rows. If this argument is None, it indicates all rows.

• key_indices (tuple of ints) – Indices of requested columns. If this argument is None, it indicates all columns.

Returns:

tuple of lists/arrays

property keys

Names of columns.

A tuple of strings that indicate the names of columns.

property mode

Mode of representation.

This indicates the type of value returned by fetch() and __getitem__(). tuple, dict, and None are supported.