python3-nibabel 2.2.1-1 / usr / lib / python3 / dist-packages / nibabel / funcs.py

# emacs: -*- mode: python-mode; py-indent-offset: 4; indent-tabs-mode: nil -*-
# vi: set ft=python sts=4 ts=4 sw=4 et:
### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ##
#
#   See COPYING file distributed along with the NiBabel package for the
#   copyright and license terms.
#
### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ##
''' Processor functions for images '''
import numpy as np

from .orientations import io_orientation, OrientationError
from .loadsave import load


def squeeze_image(img):
    ''' Return image, remove axes length 1 at end of image shape

    For example, an image may have shape (10,20,30,1,1).  In this case
    squeeze will result in an image with shape (10,20,30).  See doctests
    for further description of behavior.

    Parameters
    ----------
    img : ``SpatialImage``

    Returns
    -------
    squeezed_img : ``SpatialImage``
       Copy of img, such that data, and data shape have been squeezed,
       for dimensions > 3rd, and at the end of the shape list

    Examples
    --------
    >>> import nibabel as nf
    >>> shape = (10,20,30,1,1)
    >>> data = np.arange(np.prod(shape)).reshape(shape)
    >>> affine = np.eye(4)
    >>> img = nf.Nifti1Image(data, affine)
    >>> img.shape == (10, 20, 30, 1, 1)
    True
    >>> img2 = squeeze_image(img)
    >>> img2.shape == (10, 20, 30)
    True

    If the data are 3D then last dimensions of 1 are ignored

    >>> shape = (10,1,1)
    >>> data = np.arange(np.prod(shape)).reshape(shape)
    >>> img = nf.ni1.Nifti1Image(data, affine)
    >>> img.shape == (10, 1, 1)
    True
    >>> img2 = squeeze_image(img)
    >>> img2.shape == (10, 1, 1)
    True

    Only *final* dimensions of 1 are squeezed

    >>> shape = (1, 1, 5, 1, 2, 1, 1)
    >>> data = data.reshape(shape)
    >>> img = nf.ni1.Nifti1Image(data, affine)
    >>> img.shape == (1, 1, 5, 1, 2, 1, 1)
    True
    >>> img2 = squeeze_image(img)
    >>> img2.shape == (1, 1, 5, 1, 2)
    True
    '''
    klass = img.__class__
    shape = img.shape
    slen = len(shape)
    if slen < 4:
        return klass.from_image(img)
    for bdim in shape[3::][::-1]:
        if bdim == 1:
            slen -= 1
        else:
            break
    if slen == len(shape):
        return klass.from_image(img)
    shape = shape[:slen]
    data = img.get_data()
    data = data.reshape(shape)
    return klass(data,
                 img.affine,
                 img.header,
                 img.extra)


def concat_images(images, check_affines=True, axis=None):
    ''' Concatenate images in list to single image, along specified dimension

    Parameters
    ----------
    images : sequence
       sequence of ``SpatialImage`` or filenames of the same dimensionality\s
    check_affines : {True, False}, optional
       If True, then check that all the affines for `images` are nearly
       the same, raising a ``ValueError`` otherwise.  Default is True
    axis : None or int, optional
        If None, concatenates on a new dimension.  This requires all images to
        be the same shape.  If not None, concatenates on the specified
        dimension.  This requires all images to be the same shape, except on
        the specified dimension.
    Returns
    -------
    concat_img : ``SpatialImage``
       New image resulting from concatenating `images` across last
       dimension
    '''
    images = [load(img) if not hasattr(img, 'get_data')
              else img for img in images]
    n_imgs = len(images)
    if n_imgs == 0:
        raise ValueError("Cannot concatenate an empty list of images.")
    img0 = images[0]
    affine = img0.affine
    header = img0.header
    klass = img0.__class__
    shape0 = img0.shape
    n_dim = len(shape0)
    if axis is None:
        # collect images in output array for efficiency
        out_shape = (n_imgs, ) + shape0
        out_data = np.empty(out_shape)
    else:
        # collect images in list for use with np.concatenate
        out_data = [None] * n_imgs
    # Get part of shape we need to check inside loop
    idx_mask = np.ones((n_dim,), dtype=bool)
    if axis is not None:
        idx_mask[axis] = False
    masked_shape = np.array(shape0)[idx_mask]
    for i, img in enumerate(images):
        if len(img.shape) != n_dim:
            raise ValueError(
                'Image {0} has {1} dimensions, image 0 has {2}'.format(
                    i, len(img.shape), n_dim))
        if not np.all(np.array(img.shape)[idx_mask] == masked_shape):
            raise ValueError('shape {0} for image {1} not compatible with '
                             'first image shape {2} with axis == {0}'.format(
                                 img.shape, i, shape0, axis))
        if check_affines and not np.all(img.affine == affine):
            raise ValueError('Affine for image {0} does not match affine '
                             'for first image'.format(i))
        # Do not fill cache in image if it is empty
        out_data[i] = img.get_data(caching='unchanged')

    if axis is None:
        out_data = np.rollaxis(out_data, 0, out_data.ndim)
    else:
        out_data = np.concatenate(out_data, axis=axis)

    return klass(out_data, affine, header)


def four_to_three(img):
    ''' Create 3D images from 4D image by slicing over last axis

    Parameters
    ----------
    img :  image
       4D image instance of some class with methods ``get_data``,
       ``header`` and ``affine``, and a class constructor
       allowing klass(data, affine, header)

    Returns
    -------
    imgs : list
       list of 3D images
    '''
    arr = img.get_data()
    header = img.header
    affine = img.affine
    image_maker = img.__class__
    if arr.ndim != 4:
        raise ValueError('Expecting four dimensions')
    imgs = []
    for i in range(arr.shape[3]):
        arr3d = arr[..., i]
        img3d = image_maker(arr3d, affine, header)
        imgs.append(img3d)
    return imgs


def as_closest_canonical(img, enforce_diag=False):
    ''' Return `img` with data reordered to be closest to canonical

    Canonical order is the ordering of the output axes.

    Parameters
    ----------
    img : ``spatialimage``
    enforce_diag : {False, True}, optional
       If True, before transforming image, check if the resulting image
       affine will be close to diagonal, and if not, raise an error

    Returns
    -------
    canonical_img : ``spatialimage``
       Version of `img` where the underlying array may have been
       reordered and / or flipped so that axes 0,1,2 are those axes in
       the input data that are, respectively, closest to the output axis
       orientation.  We modify the affine accordingly.  If `img` is
       already has the correct data ordering, we just return `img`
       unmodified.
    '''
    # Get the image class to transform the data for us
    img = img.as_reoriented(io_orientation(img.affine))

    # however, the affine may not be diagonal
    if enforce_diag and not _aff_is_diag(img.affine):
        raise OrientationError('Transformed affine is not diagonal')

    return img


def _aff_is_diag(aff):
    ''' Utility function returning True if affine is nearly diagonal '''
    rzs_aff = aff[:3, :3]
    return np.allclose(rzs_aff, np.diag(np.diag(rzs_aff)))