torchio.transforms.preprocessing.label.one_hot.OneHot.apply_transform() - Code Metrics - Inspection of "Add OneHot transform (#420)" - fepegar/torchio - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( 847e1d...ba5be7 )

by Fernando

created 2021-01-17 23:30 UTC

OneHot.apply_transform() A

↳ Parent: torchio.transforms.preprocessing.label.one_hot

Complexity

Conditions

Size

Total Lines	8
Code Lines	8

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	3
eloc	8
nop	2
dl	0
loc	8
rs	10
c	0
b	0
f	0

import torch.nn.functional as F  # noqa: N812

from .label_transform import LabelTransform


class OneHot(LabelTransform):
    r"""Reencode label maps using one-hot encoding.

    Args:
        num_classes: See :func:`~torch.nn.functional.one_hot`.
        **kwargs: See :class:`~torchio.transforms.Transform` for additional
            keyword arguments.
    """
    def __init__(
            self,
            num_classes: int = -1,
            **kwargs
            ):
        super().__init__(**kwargs)
        self.num_classes = num_classes
        self.args_names = []

    def apply_transform(self, subject):
        for image in self.get_images(subject):
            assert image.data.ndim == 4 and image.data.shape[0] == 1
            data = image.data.squeeze()
            num_classes = -1 if self.num_classes is None else self.num_classes
            one_hot = F.one_hot(data.long(), num_classes=num_classes)
            image.set_data(one_hot.permute(3, 0, 1, 2).type(data.type()))
        return subject


1			import torch.nn.functional as F # noqa: N812
2
3			from .label_transform import LabelTransform
4
5
6			class OneHot(LabelTransform):
7			r"""Reencode label maps using one-hot encoding.
8
9			Args:
10			num_classes: See :func:`~torch.nn.functional.one_hot`.
11			**kwargs: See :class:`~torchio.transforms.Transform` for additional
12			keyword arguments.
13			"""
14			def __init__(
15			self,
16			num_classes: int = -1,
17			**kwargs
18			):
19			super().__init__(**kwargs)
20			self.num_classes = num_classes
21			self.args_names = []
22
23			def apply_transform(self, subject):
24			for image in self.get_images(subject):
25			assert image.data.ndim == 4 and image.data.shape[0] == 1
26			data = image.data.squeeze()
27			num_classes = -1 if self.num_classes is None else self.num_classes
28			one_hot = F.one_hot(data.long(), num_classes=num_classes)
29			image.set_data(one_hot.permute(3, 0, 1, 2).type(data.type()))
30			return subject
31

fepegar / torchio

Push — master ( 847e1d...ba5be7 )

OneHot.apply_transform() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like