torchio.transforms.transform.Transform.inverse() - Code Metrics - Inspection of "Add label remapping transforms (#402)" - fepegar/torchio - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( b9ac52...6aebda )

by Fernando

created 2020-12-29 14:50 UTC

torchio.transforms.transform.Transform.inverse() A

↳ Parent: torchio.transforms.transform

Complexity

Conditions

Size

Total Lines	6
Code Lines	6

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	2
eloc	6
nop	1
dl	0
loc	6
rs	10
c	0
b	0
f	0

import copy
import numbers
import warnings
from typing import Union, Tuple
from abc import ABC, abstractmethod
from contextlib import contextmanager

import torch
import numpy as np
import SimpleITK as sitk

from ..utils import to_tuple
from ..data.subject import Subject
from ..data.io import nib_to_sitk, sitk_to_nib
from ..data.image import LabelMap
from ..typing import TypeData, TypeNumber, TypeKeys, TypeCallable, TypeTripletInt
from .interpolation import Interpolation, get_sitk_interpolator
from .data_parser import DataParser, TypeTransformInput

TypeSixBounds = Tuple[int, int, int, int, int, int]
TypeBounds = Union[
    int,
    TypeTripletInt,
    TypeSixBounds,
]
TypeMaskingMethod = Union[str, TypeCallable, TypeBounds, None]


class Transform(ABC):
    """Abstract class for all TorchIO transforms.

    All subclasses must overwrite
    :meth:`Transform.apply_transform`,
    which takes data, applies some transformation and returns the result.

    The input can be an instance of
    :class:`torchio.Subject`,
    :class:`torchio.Image`,
    :class:`numpy.ndarray`,
    :class:`torch.Tensor`,
    :class:`SimpleITK.Image`,
    or :class:`dict`.

    Args:
        p: Probability that this transform will be applied.
        copy: Make a shallow copy of the input before applying the transform.
        include: Sequence of strings with the names of the only images to which
            the transform will be applied.
            Mandatory if the input is a :class:`dict`.
        exclude: Sequence of strings with the names of the images to which the
            the transform will not be applied, apart from the ones that are
            excluded because of the transform type.
            For example, if a subject includes an MRI, a CT and a label map, and
            the CT is added to the list of exclusions of an intensity transform
            such as :class:`~torchio.transforms.RandomBlur`,
            the transform will be only applied to the MRI, as the label map is
            excluded by default by spatial transforms.
    """
    def __init__(
            self,
            p: float = 1,
            copy: bool = True,
            include: TypeKeys = None,
            exclude: TypeKeys = None,
            keys: TypeKeys = None,
            ):
        self.probability = self.parse_probability(p)
        self.copy = copy
        if keys is not None:
            message = (
                'The "keys" argument is deprecated and will be removed in the'
                ' future. Use "include" instead.'
            )
            warnings.warn(message, DeprecationWarning)
            include = keys
        self.include, self.exclude = self.parse_include_and_exclude(
            include, exclude)

    def __call__(
            self,
            data: TypeTransformInput,
            ) -> TypeTransformInput:
        """Transform data and return a result of the same type.

        Args:
            data: Instance of 1) :class:`~torchio.Subject`, 4D
                :class:`torch.Tensor` or :class:`numpy.ndarray` with dimensions
                :math:`(C, W, H, D)`, where :math:`C` is the number of channels
                and :math:`W, H, D` are the spatial dimensions. If the input is
                a tensor, the affine matrix will be set to identity. Other
                valid input types are a SimpleITK image, a
                :class:`torchio.Image`, a NiBabel Nifti1 image or a
                :class:`dict`. The output type is the same as the input type.
        """
        if torch.rand(1).item() > self.probability:
            return data
        data_parser = DataParser(data, keys=self.include)
        subject = data_parser.get_subject()
        if self.copy:
            subject = copy.copy(subject)
        with np.errstate(all='raise'):
            transformed = self.apply_transform(subject)
        self.add_transform_to_subject_history(transformed)
        for image in transformed.get_images(intensity_only=False):
            ndim = image.data.ndim
            assert ndim == 4, f'Output of {self.name} is {ndim}D'

        output = data_parser.get_output(transformed)
        return output

    def __repr__(self):
        if hasattr(self, 'args_names'):
            names = self.args_names
            args_strings = [f'{arg}={getattr(self, arg)}' for arg in names]
            if hasattr(self, 'invert_transform') and self.invert_transform:
                args_strings.append('invert=True')
            args_string = ', '.join(args_strings)
            return f'{self.name}({args_string})'
        else:
            return super().__repr__()

    @property
    def name(self):
        return self.__class__.__name__

    @abstractmethod
    def apply_transform(self, subject: Subject):
        raise NotImplementedError

    def add_transform_to_subject_history(self, subject):
        from .augmentation import RandomTransform
        from . import Compose, OneOf, CropOrPad
        from .preprocessing.label import SequentialLabels
        call_others = (
            RandomTransform,
            Compose,
            OneOf,
            CropOrPad,
            SequentialLabels,
        )
        if not isinstance(self, call_others):
            subject.add_transform(self, self._get_reproducing_arguments())

    @staticmethod
    def to_range(n, around):
        if around is None:
            return 0, n
        else:
            return around - n, around + n

    def parse_params(self, params, around, name, make_ranges=True, **kwargs):
        params = to_tuple(params)
        if len(params) == 1 or (len(params) == 2 and make_ranges):  # d or (a, b)
            params *= 3  # (d, d, d) or (a, b, a, b, a, b)
        if len(params) == 3 and make_ranges:  # (a, b, c)
            items = [self.to_range(n, around) for n in params]
            # (-a, a, -b, b, -c, c) or (1-a, 1+a, 1-b, 1+b, 1-c, 1+c)
            params = [n for prange in items for n in prange]
        if make_ranges:
            if len(params) != 6:
                message = (
                    f'If "{name}" is a sequence, it must have length 2, 3 or 6,'
                    f' not {len(params)}'
                )
                raise ValueError(message)
            for param_range in zip(params[::2], params[1::2]):
                self._parse_range(param_range, name, **kwargs)
        return tuple(params)

    @staticmethod
    def _parse_range(
            nums_range: Union[TypeNumber, Tuple[TypeNumber, TypeNumber]],
            name: str,
            min_constraint: TypeNumber = None,
            max_constraint: TypeNumber = None,
            type_constraint: type = None,
            ) -> Tuple[TypeNumber, TypeNumber]:
        r"""Adapted from :class:`torchvision.transforms.RandomRotation`.

        Args:
            nums_range: Tuple of two numbers :math:`(n_{min}, n_{max})`,
                where :math:`n_{min} \leq n_{max}`.
                If a single positive number :math:`n` is provided,
                :math:`n_{min} = -n` and :math:`n_{max} = n`.
            name: Name of the parameter, so that an informative error message
                can be printed.
            min_constraint: Minimal value that :math:`n_{min}` can take,
                default is None, i.e. there is no minimal value.
            max_constraint: Maximal value that :math:`n_{max}` can take,
                default is None, i.e. there is no maximal value.
            type_constraint: Precise type that :math:`n_{max}` and
                :math:`n_{min}` must take.

        Returns:
            A tuple of two numbers :math:`(n_{min}, n_{max})`.

        Raises:
            ValueError: if :attr:`nums_range` is negative
            ValueError: if :math:`n_{max}` or :math:`n_{min}` is not a number
            ValueError: if :math:`n_{max} \lt n_{min}`
            ValueError: if :attr:`min_constraint` is not None and
                :math:`n_{min}` is smaller than :attr:`min_constraint`
            ValueError: if :attr:`max_constraint` is not None and
                :math:`n_{max}` is greater than :attr:`max_constraint`
            ValueError: if :attr:`type_constraint` is not None and
                :math:`n_{max}` and :math:`n_{max}` are not of type
                :attr:`type_constraint`.
        """
        if isinstance(nums_range, numbers.Number):  # single number given
            if nums_range < 0:
                raise ValueError(
                    f'If {name} is a single number,'
                    f' it must be positive, not {nums_range}')
            if min_constraint is not None and nums_range < min_constraint:
                raise ValueError(
                    f'If {name} is a single number, it must be greater'
                    f' than {min_constraint}, not {nums_range}'
                )
            if max_constraint is not None and nums_range > max_constraint:
                raise ValueError(
                    f'If {name} is a single number, it must be smaller'
                    f' than {max_constraint}, not {nums_range}'
                )
            if type_constraint is not None:
                if not isinstance(nums_range, type_constraint):
                    raise ValueError(
                        f'If {name} is a single number, it must be of'
                        f' type {type_constraint}, not {nums_range}'
                    )
            min_range = -nums_range if min_constraint is None else nums_range
            return (min_range, nums_range)

        try:
            min_value, max_value = nums_range
        except (TypeError, ValueError):
            raise ValueError(
                f'If {name} is not a single number, it must be'
                f' a sequence of len 2, not {nums_range}'
            )

        min_is_number = isinstance(min_value, numbers.Number)
        max_is_number = isinstance(max_value, numbers.Number)
        if not min_is_number or not max_is_number:
            message = (
                f'{name} values must be numbers, not {nums_range}')
            raise ValueError(message)

        if min_value > max_value:
            raise ValueError(
                f'If {name} is a sequence, the second value must be'
                f' equal or greater than the first, but it is {nums_range}')

        if min_constraint is not None and min_value < min_constraint:
            raise ValueError(
                f'If {name} is a sequence, the first value must be greater'
                f' than {min_constraint}, but it is {min_value}'
            )

        if max_constraint is not None and max_value > max_constraint:
            raise ValueError(
                f'If {name} is a sequence, the second value must be smaller'
                f' than {max_constraint}, but it is {max_value}'
            )

        if type_constraint is not None:
            min_type_ok = isinstance(min_value, type_constraint)
            max_type_ok = isinstance(max_value, type_constraint)
            if not min_type_ok or not max_type_ok:
                raise ValueError(
                    f'If "{name}" is a sequence, its values must be of'
                    f' type "{type_constraint}", not "{type(nums_range)}"'
                )
        return nums_range

    @staticmethod
    def parse_interpolation(interpolation: str) -> str:
        if not isinstance(interpolation, str):
            itype = type(interpolation)
            raise TypeError(f'Interpolation must be a string, not {itype}')
        interpolation = interpolation.lower()
        is_string = isinstance(interpolation, str)
        supported_values = [key.name.lower() for key in Interpolation]
        is_supported = interpolation.lower() in supported_values
        if is_string and is_supported:
            return interpolation
        message = (
            f'Interpolation "{interpolation}" of type {type(interpolation)}'
            f' must be a string among the supported values: {supported_values}'
        )
        raise ValueError(message)

    @staticmethod
    def parse_probability(probability: float) -> float:
        is_number = isinstance(probability, numbers.Number)
        if not (is_number and 0 <= probability <= 1):
            message = (
                'Probability must be a number in [0, 1],'
                f' not {probability}'
            )
            raise ValueError(message)
        return probability

    @staticmethod
    def parse_include_and_exclude(
            include: TypeKeys = None,
            exclude: TypeKeys = None,
            ) -> Tuple[TypeKeys, TypeKeys]:
        if include is not None and exclude is not None:
            raise ValueError('Include and exclude cannot both be specified')
        return include, exclude

    @staticmethod
    def nib_to_sitk(data: TypeData, affine: TypeData) -> sitk.Image:
        return nib_to_sitk(data, affine)

    @staticmethod
    def sitk_to_nib(image: sitk.Image) -> Tuple[torch.Tensor, np.ndarray]:
        return sitk_to_nib(image)

    def _get_reproducing_arguments(self):
        """
        Return a dictionary with the arguments that would be necessary to
        reproduce the transform exactly.
        """
        reproducing_arguments = {'include': self.include, 'exclude': self.exclude, 'copy': self.copy}
        reproducing_arguments.update({name: getattr(self, name) for name in self.args_names})
        return reproducing_arguments

    def is_invertible(self):
        return hasattr(self, 'invert_transform')

    def inverse(self):
        if not self.is_invertible():
            raise RuntimeError(f'{self.name} is not invertible')
        new = copy.deepcopy(self)
        new.invert_transform = not self.invert_transform
        return new

    @staticmethod
    @contextmanager
    def _use_seed(seed):
        """Perform an operation using a specific seed for the PyTorch RNG"""
        torch_rng_state = torch.random.get_rng_state()
        torch.manual_seed(seed)
        yield
        torch.random.set_rng_state(torch_rng_state)

    @staticmethod
    def get_sitk_interpolator(interpolation: str) -> int:
        return get_sitk_interpolator(interpolation)

    @staticmethod
    def parse_bounds(bounds_parameters: TypeBounds) -> TypeSixBounds:
        try:
            bounds_parameters = tuple(bounds_parameters)
        except TypeError:
            bounds_parameters = (bounds_parameters,)

        # Check that numbers are integers
        for number in bounds_parameters:
            if not isinstance(number, (int, np.integer)) or number < 0:
                message = (
                    'Bounds values must be integers greater or equal to zero,'
                    f' not "{bounds_parameters}" of type {type(number)}'
                )
                raise ValueError(message)
        bounds_parameters = tuple(int(n) for n in bounds_parameters)
        bounds_parameters_length = len(bounds_parameters)
        if bounds_parameters_length == 6:
            return bounds_parameters
        if bounds_parameters_length == 1:
            return 6 * bounds_parameters
        if bounds_parameters_length == 3:
            return tuple(np.repeat(bounds_parameters, 2).tolist())
        message = (
            'Bounds parameter must be an integer or a tuple of'
            f' 3 or 6 integers, not {bounds_parameters}'
        )
        raise ValueError(message)

    @staticmethod
    def ones(tensor: torch.Tensor) -> torch.Tensor:
        return torch.ones_like(tensor, dtype=torch.bool)

    @staticmethod
    def mean(tensor: torch.Tensor) -> torch.Tensor:
        mask = tensor > tensor.mean()
        return mask

    @staticmethod
    def get_mask(masking_method: TypeMaskingMethod, subject: Subject, tensor: torch.Tensor) -> torch.Tensor:
        if masking_method is None:
            return Transform.ones(tensor)
        elif callable(masking_method):
            return masking_method(tensor)
        elif type(masking_method) is str:
            if masking_method in subject and isinstance(subject[masking_method], LabelMap):
                return subject[masking_method].data.bool()
            if masking_method.title() in ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'):
                return Transform.get_mask_from_anatomical_label(masking_method.title(), tensor)
        elif type(masking_method) in (tuple, list, int):
            return Transform.get_mask_from_bounds(masking_method, tensor)
        message = (
            'Masking method parameter must be a function, a label map name,'
            " an anatomical label: ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'),"
            ' or a bounds parameter: (an int, tuple of 3 ints, or tuple of 6 ints)'
            f' not {masking_method} of type {type(masking_method)}'
        )
        raise ValueError(message)

    @staticmethod
    def get_mask_from_anatomical_label(anatomical_label: str, tensor: torch.Tensor) -> torch.Tensor:
        anatomical_label = anatomical_label.title()
        if anatomical_label.title() not in ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'):
            message = (
                "Anatomical label must be one of ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior')"
                f' not {anatomical_label}'
            )
            raise ValueError(message)
        mask = torch.zeros_like(tensor, dtype=torch.bool)
        _, width, height, depth = tensor.shape
        if anatomical_label == 'Right':
            mask[:, width // 2:] = True
        elif anatomical_label == 'Left':
            mask[:, :width // 2] = True
        elif anatomical_label == 'Anterior':
            mask[:, :, height // 2:] = True
        elif anatomical_label == 'Posterior':
            mask[:, :, :height // 2] = True
        elif anatomical_label == 'Superior':
            mask[:, :, :, depth // 2:] = True
        elif anatomical_label == 'Inferior':
            mask[:, :, :, :depth // 2] = True
        return mask

    @staticmethod
    def get_mask_from_bounds(bounds_parameters: TypeBounds, tensor: torch.Tensor) -> torch.Tensor:
        bounds_parameters = Transform.parse_bounds(bounds_parameters)
        low = bounds_parameters[::2]
        high = bounds_parameters[1::2]
        i0, j0, k0 = low
        i1, j1, k1 = np.array(tensor.shape[1:]) - high
        mask = torch.zeros_like(tensor, dtype=torch.bool)
        mask[:, i0:i1, j0:j1, k0:k1] = True
        return mask


1			import copy
2			import numbers
3			import warnings
4			from typing import Union, Tuple
5			from abc import ABC, abstractmethod
6			from contextlib import contextmanager
7
8			import torch
9			import numpy as np
10			import SimpleITK as sitk
11
12			from ..utils import to_tuple
13			from ..data.subject import Subject
14			from ..data.io import nib_to_sitk, sitk_to_nib
15			from ..data.image import LabelMap
16			from ..typing import TypeData, TypeNumber, TypeKeys, TypeCallable, TypeTripletInt
17			from .interpolation import Interpolation, get_sitk_interpolator
18			from .data_parser import DataParser, TypeTransformInput
19
20			TypeSixBounds = Tuple[int, int, int, int, int, int]
21			TypeBounds = Union[
22			int,
23			TypeTripletInt,
24			TypeSixBounds,
25			]
26			TypeMaskingMethod = Union[str, TypeCallable, TypeBounds, None]
27
28
29			class Transform(ABC):
30			"""Abstract class for all TorchIO transforms.
31
32			All subclasses must overwrite
33			:meth:`Transform.apply_transform`,
34			which takes data, applies some transformation and returns the result.
35
36			The input can be an instance of
37			:class:`torchio.Subject`,
38			:class:`torchio.Image`,
39			:class:`numpy.ndarray`,
40			:class:`torch.Tensor`,
41			:class:`SimpleITK.Image`,
42			or :class:`dict`.
43
44			Args:
45			p: Probability that this transform will be applied.
46			copy: Make a shallow copy of the input before applying the transform.
47			include: Sequence of strings with the names of the only images to which
48			the transform will be applied.
49			Mandatory if the input is a :class:`dict`.
50			exclude: Sequence of strings with the names of the images to which the
51			the transform will not be applied, apart from the ones that are
52			excluded because of the transform type.
53			For example, if a subject includes an MRI, a CT and a label map, and
54			the CT is added to the list of exclusions of an intensity transform
55			such as :class:`~torchio.transforms.RandomBlur`,
56			the transform will be only applied to the MRI, as the label map is
57			excluded by default by spatial transforms.
58			"""
59			def __init__(
60			self,
61			p: float = 1,
62			copy: bool = True,
63			include: TypeKeys = None,
64			exclude: TypeKeys = None,
65			keys: TypeKeys = None,
66			):
67			self.probability = self.parse_probability(p)
68			self.copy = copy
69			if keys is not None:
70			message = (
71			'The "keys" argument is deprecated and will be removed in the'
72			' future. Use "include" instead.'
73			)
74			warnings.warn(message, DeprecationWarning)
75			include = keys
76			self.include, self.exclude = self.parse_include_and_exclude(
77			include, exclude)
78
79			def __call__(
80			self,
81			data: TypeTransformInput,
82			) -> TypeTransformInput:
83			"""Transform data and return a result of the same type.
84
85			Args:
86			data: Instance of 1) :class:`~torchio.Subject`, 4D
87			:class:`torch.Tensor` or :class:`numpy.ndarray` with dimensions
88			:math:`(C, W, H, D)`, where :math:`C` is the number of channels
89			and :math:`W, H, D` are the spatial dimensions. If the input is
90			a tensor, the affine matrix will be set to identity. Other
91			valid input types are a SimpleITK image, a
92			:class:`torchio.Image`, a NiBabel Nifti1 image or a
93			:class:`dict`. The output type is the same as the input type.
94			"""
95			if torch.rand(1).item() > self.probability:
96			return data
97			data_parser = DataParser(data, keys=self.include)
98			subject = data_parser.get_subject()
99			if self.copy:
100			subject = copy.copy(subject)
101			with np.errstate(all='raise'):
102			transformed = self.apply_transform(subject)
103			self.add_transform_to_subject_history(transformed)
104			for image in transformed.get_images(intensity_only=False):
105			ndim = image.data.ndim
106			assert ndim == 4, f'Output of {self.name} is {ndim}D'
107
108			output = data_parser.get_output(transformed)
109			return output
110
111			def __repr__(self):
112			if hasattr(self, 'args_names'):
113			names = self.args_names
114			args_strings = [f'{arg}={getattr(self, arg)}' for arg in names]
115			if hasattr(self, 'invert_transform') and self.invert_transform:
116			args_strings.append('invert=True')
117			args_string = ', '.join(args_strings)
118			return f'{self.name}({args_string})'
119			else:
120			return super().__repr__()
121
122			@property
123			def name(self):
124			return self.__class__.__name__
125
126			@abstractmethod
127			def apply_transform(self, subject: Subject):
128			raise NotImplementedError
129
130			def add_transform_to_subject_history(self, subject):
131			from .augmentation import RandomTransform
132			from . import Compose, OneOf, CropOrPad
133			from .preprocessing.label import SequentialLabels
134			call_others = (
135			RandomTransform,
136			Compose,
137			OneOf,
138			CropOrPad,
139			SequentialLabels,
140			)
141			if not isinstance(self, call_others):
142			subject.add_transform(self, self._get_reproducing_arguments())
143
144			@staticmethod
145			def to_range(n, around):
146			if around is None:
147			return 0, n
148			else:
149			return around - n, around + n
150
151			def parse_params(self, params, around, name, make_ranges=True, **kwargs):
152			params = to_tuple(params)
153			if len(params) == 1 or (len(params) == 2 and make_ranges): # d or (a, b)
154			params *= 3 # (d, d, d) or (a, b, a, b, a, b)
155			if len(params) == 3 and make_ranges: # (a, b, c)
156			items = [self.to_range(n, around) for n in params]
157			# (-a, a, -b, b, -c, c) or (1-a, 1+a, 1-b, 1+b, 1-c, 1+c)
158			params = [n for prange in items for n in prange]
159			if make_ranges:
160			if len(params) != 6:
161			message = (
162			f'If "{name}" is a sequence, it must have length 2, 3 or 6,'
163			f' not {len(params)}'
164			)
165			raise ValueError(message)
166			for param_range in zip(params[::2], params[1::2]):
167			self._parse_range(param_range, name, **kwargs)
168			return tuple(params)
169
170			@staticmethod
171			def _parse_range(
172			nums_range: Union[TypeNumber, Tuple[TypeNumber, TypeNumber]],
173			name: str,
174			min_constraint: TypeNumber = None,
175			max_constraint: TypeNumber = None,
176			type_constraint: type = None,
177			) -> Tuple[TypeNumber, TypeNumber]:
178			r"""Adapted from :class:`torchvision.transforms.RandomRotation`.
179
180			Args:
181			nums_range: Tuple of two numbers :math:`(n_{min}, n_{max})`,
182			where :math:`n_{min} \leq n_{max}`.
183			If a single positive number :math:`n` is provided,
184			:math:`n_{min} = -n` and :math:`n_{max} = n`.
185			name: Name of the parameter, so that an informative error message
186			can be printed.
187			min_constraint: Minimal value that :math:`n_{min}` can take,
188			default is None, i.e. there is no minimal value.
189			max_constraint: Maximal value that :math:`n_{max}` can take,
190			default is None, i.e. there is no maximal value.
191			type_constraint: Precise type that :math:`n_{max}` and
192			:math:`n_{min}` must take.
193
194			Returns:
195			A tuple of two numbers :math:`(n_{min}, n_{max})`.
196
197			Raises:
198			ValueError: if :attr:`nums_range` is negative
199			ValueError: if :math:`n_{max}` or :math:`n_{min}` is not a number
200			ValueError: if :math:`n_{max} \lt n_{min}`
201			ValueError: if :attr:`min_constraint` is not None and
202			:math:`n_{min}` is smaller than :attr:`min_constraint`
203			ValueError: if :attr:`max_constraint` is not None and
204			:math:`n_{max}` is greater than :attr:`max_constraint`
205			ValueError: if :attr:`type_constraint` is not None and
206			:math:`n_{max}` and :math:`n_{max}` are not of type
207			:attr:`type_constraint`.
208			"""
209			if isinstance(nums_range, numbers.Number): # single number given
210			if nums_range < 0:
211			raise ValueError(
212			f'If {name} is a single number,'
213			f' it must be positive, not {nums_range}')
214			if min_constraint is not None and nums_range < min_constraint:
215			raise ValueError(
216			f'If {name} is a single number, it must be greater'
217			f' than {min_constraint}, not {nums_range}'
218			)
219			if max_constraint is not None and nums_range > max_constraint:
220			raise ValueError(
221			f'If {name} is a single number, it must be smaller'
222			f' than {max_constraint}, not {nums_range}'
223			)
224			if type_constraint is not None:
225			if not isinstance(nums_range, type_constraint):
226			raise ValueError(
227			f'If {name} is a single number, it must be of'
228			f' type {type_constraint}, not {nums_range}'
229			)
230			min_range = -nums_range if min_constraint is None else nums_range
231			return (min_range, nums_range)
232
233			try:
234			min_value, max_value = nums_range
235			except (TypeError, ValueError):
236			raise ValueError(
237			f'If {name} is not a single number, it must be'
238			f' a sequence of len 2, not {nums_range}'
239			)
240
241			min_is_number = isinstance(min_value, numbers.Number)
242			max_is_number = isinstance(max_value, numbers.Number)
243			if not min_is_number or not max_is_number:
244			message = (
245			f'{name} values must be numbers, not {nums_range}')
246			raise ValueError(message)
247
248			if min_value > max_value:
249			raise ValueError(
250			f'If {name} is a sequence, the second value must be'
251			f' equal or greater than the first, but it is {nums_range}')
252
253			if min_constraint is not None and min_value < min_constraint:
254			raise ValueError(
255			f'If {name} is a sequence, the first value must be greater'
256			f' than {min_constraint}, but it is {min_value}'
257			)
258
259			if max_constraint is not None and max_value > max_constraint:
260			raise ValueError(
261			f'If {name} is a sequence, the second value must be smaller'
262			f' than {max_constraint}, but it is {max_value}'
263			)
264
265			if type_constraint is not None:
266			min_type_ok = isinstance(min_value, type_constraint)
267			max_type_ok = isinstance(max_value, type_constraint)
268			if not min_type_ok or not max_type_ok:
269			raise ValueError(
270			f'If "{name}" is a sequence, its values must be of'
271			f' type "{type_constraint}", not "{type(nums_range)}"'
272			)
273			return nums_range
274
275			@staticmethod
276			def parse_interpolation(interpolation: str) -> str:
277			if not isinstance(interpolation, str):
278			itype = type(interpolation)
279			raise TypeError(f'Interpolation must be a string, not {itype}')
280			interpolation = interpolation.lower()
281			is_string = isinstance(interpolation, str)
282			supported_values = [key.name.lower() for key in Interpolation]
283			is_supported = interpolation.lower() in supported_values
284			if is_string and is_supported:
285			return interpolation
286			message = (
287			f'Interpolation "{interpolation}" of type {type(interpolation)}'
288			f' must be a string among the supported values: {supported_values}'
289			)
290			raise ValueError(message)
291
292			@staticmethod
293			def parse_probability(probability: float) -> float:
294			is_number = isinstance(probability, numbers.Number)
295			if not (is_number and 0 <= probability <= 1):
296			message = (
297			'Probability must be a number in [0, 1],'
298			f' not {probability}'
299			)
300			raise ValueError(message)
301			return probability
302
303			@staticmethod
304			def parse_include_and_exclude(
305			include: TypeKeys = None,
306			exclude: TypeKeys = None,
307			) -> Tuple[TypeKeys, TypeKeys]:
308			if include is not None and exclude is not None:
309			raise ValueError('Include and exclude cannot both be specified')
310			return include, exclude
311
312			@staticmethod
313			def nib_to_sitk(data: TypeData, affine: TypeData) -> sitk.Image:
314			return nib_to_sitk(data, affine)
315
316			@staticmethod
317			def sitk_to_nib(image: sitk.Image) -> Tuple[torch.Tensor, np.ndarray]:
318			return sitk_to_nib(image)
319
320			def _get_reproducing_arguments(self):
321			"""
322			Return a dictionary with the arguments that would be necessary to
323			reproduce the transform exactly.
324			"""
325			reproducing_arguments = {'include': self.include, 'exclude': self.exclude, 'copy': self.copy}
326			reproducing_arguments.update({name: getattr(self, name) for name in self.args_names})
327			return reproducing_arguments
328
329			def is_invertible(self):
330			return hasattr(self, 'invert_transform')
331
332			def inverse(self):
333			if not self.is_invertible():
334			raise RuntimeError(f'{self.name} is not invertible')
335			new = copy.deepcopy(self)
336			new.invert_transform = not self.invert_transform
337			return new
338
339			@staticmethod
340			@contextmanager
341			def _use_seed(seed):
342			"""Perform an operation using a specific seed for the PyTorch RNG"""
343			torch_rng_state = torch.random.get_rng_state()
344			torch.manual_seed(seed)
345			yield
346			torch.random.set_rng_state(torch_rng_state)
347
348			@staticmethod
349			def get_sitk_interpolator(interpolation: str) -> int:
350			return get_sitk_interpolator(interpolation)
351
352			@staticmethod
353			def parse_bounds(bounds_parameters: TypeBounds) -> TypeSixBounds:
354			try:
355			bounds_parameters = tuple(bounds_parameters)
356			except TypeError:
357			bounds_parameters = (bounds_parameters,)
358
359			# Check that numbers are integers
360			for number in bounds_parameters:
361			if not isinstance(number, (int, np.integer)) or number < 0:
362			message = (
363			'Bounds values must be integers greater or equal to zero,'
364			f' not "{bounds_parameters}" of type {type(number)}'
365			)
366			raise ValueError(message)
367			bounds_parameters = tuple(int(n) for n in bounds_parameters)
368			bounds_parameters_length = len(bounds_parameters)
369			if bounds_parameters_length == 6:
370			return bounds_parameters
371			if bounds_parameters_length == 1:
372			return 6 * bounds_parameters
373			if bounds_parameters_length == 3:
374			return tuple(np.repeat(bounds_parameters, 2).tolist())
375			message = (
376			'Bounds parameter must be an integer or a tuple of'
377			f' 3 or 6 integers, not {bounds_parameters}'
378			)
379			raise ValueError(message)
380
381			@staticmethod
382			def ones(tensor: torch.Tensor) -> torch.Tensor:
383			return torch.ones_like(tensor, dtype=torch.bool)
384
385			@staticmethod
386			def mean(tensor: torch.Tensor) -> torch.Tensor:
387			mask = tensor > tensor.mean()
388			return mask
389
390			@staticmethod
391			def get_mask(masking_method: TypeMaskingMethod, subject: Subject, tensor: torch.Tensor) -> torch.Tensor:
392			if masking_method is None:
393			return Transform.ones(tensor)
394			elif callable(masking_method):
395			return masking_method(tensor)
396			elif type(masking_method) is str:
397			if masking_method in subject and isinstance(subject[masking_method], LabelMap):
398			return subject[masking_method].data.bool()
399			if masking_method.title() in ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'):
400			return Transform.get_mask_from_anatomical_label(masking_method.title(), tensor)
401			elif type(masking_method) in (tuple, list, int):
402			return Transform.get_mask_from_bounds(masking_method, tensor)
403			message = (
404			'Masking method parameter must be a function, a label map name,'
405			" an anatomical label: ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'),"
406			' or a bounds parameter: (an int, tuple of 3 ints, or tuple of 6 ints)'
407			f' not {masking_method} of type {type(masking_method)}'
408			)
409			raise ValueError(message)
410
411			@staticmethod
412			def get_mask_from_anatomical_label(anatomical_label: str, tensor: torch.Tensor) -> torch.Tensor:
413			anatomical_label = anatomical_label.title()
414			if anatomical_label.title() not in ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior'):
415			message = (
416			"Anatomical label must be one of ('Left', 'Right', 'Anterior', 'Posterior', 'Inferior', 'Superior')"
417			f' not {anatomical_label}'
418			)
419			raise ValueError(message)
420			mask = torch.zeros_like(tensor, dtype=torch.bool)
421			_, width, height, depth = tensor.shape
422			if anatomical_label == 'Right':
423			mask[:, width // 2:] = True
424			elif anatomical_label == 'Left':
425			mask[:, :width // 2] = True
426			elif anatomical_label == 'Anterior':
427			mask[:, :, height // 2:] = True
428			elif anatomical_label == 'Posterior':
429			mask[:, :, :height // 2] = True
430			elif anatomical_label == 'Superior':
431			mask[:, :, :, depth // 2:] = True
432			elif anatomical_label == 'Inferior':
433			mask[:, :, :, :depth // 2] = True
434			return mask
435
436			@staticmethod
437			def get_mask_from_bounds(bounds_parameters: TypeBounds, tensor: torch.Tensor) -> torch.Tensor:
438			bounds_parameters = Transform.parse_bounds(bounds_parameters)
439			low = bounds_parameters[::2]
440			high = bounds_parameters[1::2]
441			i0, j0, k0 = low
442			i1, j1, k1 = np.array(tensor.shape[1:]) - high
443			mask = torch.zeros_like(tensor, dtype=torch.bool)
444			mask[:, i0:i1, j0:j1, k0:k1] = True
445			return mask
446

fepegar / torchio

Push — master ( b9ac52...6aebda )

torchio.transforms.transform.Transform.inverse() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like