torchio.data.sampler.uniform.UniformSampler.__call__() - Code Metrics - Inspection of "Fix flake8 length" - fepegar/torchio - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Pull Request — master (#403)

by Fernando

created 2021-01-01 13:42 UTC

UniformSampler.call() B

↳ Parent: torchio.data.sampler.uniform

Complexity

Conditions

Size

Total Lines	25
Code Lines	19

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	5
eloc	19
nop	3
dl	0
loc	25
rs	8.9833
c	0
b	0
f	0

import torch
from ...data.subject import Subject
from ...typing import TypePatchSize
from .sampler import RandomSampler
from typing import Generator
import numpy as np


class UniformSampler(RandomSampler):
    """Randomly extract patches from a volume with uniform probability.

    Args:
        patch_size: See :class:`~torchio.data.PatchSampler`.
    """
    def __init__(self, patch_size: TypePatchSize):
        super().__init__(patch_size)

    def get_probability_map(self, subject: Subject) -> torch.Tensor:
        return torch.ones(1, *subject.spatial_shape)

    def __call__(
            self,
            subject: Subject,
            num_patches: int = None,
            ) -> Generator[Subject, None, None]:
        subject.check_consistent_spatial_shape()

        if np.any(self.patch_size > subject.spatial_shape):
            message = (
                f'Patch size {tuple(self.patch_size)} cannot be'
                f' larger than image size {tuple(subject.spatial_shape)}'
            )
            raise RuntimeError(message)

        valid_range = subject.spatial_shape - self.patch_size
        patches_left = num_patches if num_patches is not None else True
        while patches_left:
            index_ini = [
                torch.randint(x + 1, (1,)).item()
                for x in valid_range
            ]
            index_ini_array = np.asarray(index_ini)
            yield self.extract_patch(subject, index_ini_array)
            if num_patches is not None:
                patches_left -= 1


1			import torch
2			from ...data.subject import Subject
3			from ...typing import TypePatchSize
4			from .sampler import RandomSampler
5			from typing import Generator
6			import numpy as np
7
8
9			class UniformSampler(RandomSampler):
10			"""Randomly extract patches from a volume with uniform probability.
11
12			Args:
13			patch_size: See :class:`~torchio.data.PatchSampler`.
14			"""
15			def __init__(self, patch_size: TypePatchSize):
16			super().__init__(patch_size)
17
18			def get_probability_map(self, subject: Subject) -> torch.Tensor:
19			return torch.ones(1, *subject.spatial_shape)
20
21			def __call__(
22			self,
23			subject: Subject,
24			num_patches: int = None,
25			) -> Generator[Subject, None, None]:
26			subject.check_consistent_spatial_shape()
27
28			if np.any(self.patch_size > subject.spatial_shape):
29			message = (
30			f'Patch size {tuple(self.patch_size)} cannot be'
31			f' larger than image size {tuple(subject.spatial_shape)}'
32			)
33			raise RuntimeError(message)
34
35			valid_range = subject.spatial_shape - self.patch_size
36			patches_left = num_patches if num_patches is not None else True
37			while patches_left:
38			index_ini = [
39			torch.randint(x + 1, (1,)).item()
40			for x in valid_range
41			]
42			index_ini_array = np.asarray(index_ini)
43			yield self.extract_patch(subject, index_ini_array)
44			if num_patches is not None:
45			patches_left -= 1
46

fepegar / torchio

Pull Request — master (#403)

UniformSampler.__call__() B

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like

UniformSampler.call() B