def test_correct_results(self, angle, keep_size, mode, padding_mode,
align_corners):
rotate_fn = Rotate(angle, keep_size, mode, padding_mode, align_corners)
rotated = rotate_fn(self.imt[0])
if keep_size:
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)
_order = 0 if mode == "nearest" else 1
if padding_mode == "border":
_mode = "nearest"
elif padding_mode == "reflection":
_mode = "reflect"
else:
_mode = "constant"
expected = []
for channel in self.imt[0]:
expected.append(
scipy.ndimage.rotate(
channel,
-np.rad2deg(angle),
(0, 1),
not keep_size,
order=_order,
mode=_mode,
prefilter=False,
))
expected = np.stack(expected).astype(np.float32)
np.testing.assert_allclose(expected, rotated, atol=1e-1)
def test_correct_shape(self, im_type, angle, mode, padding_mode, align_corners):
rotate_fn = Rotate(angle, True, align_corners=align_corners, dtype=np.float64)
im = im_type(self.imt[0])
set_track_meta(False)
rotated = rotate_fn(im, mode=mode, padding_mode=padding_mode)
self.assertNotIsInstance(rotated, MetaTensor)
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)
set_track_meta(True)
rotated = rotate_fn(im, mode=mode, padding_mode=padding_mode)
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)
test_local_inversion(rotate_fn, rotated, im)
def test_correct_results(self, angle, spatial_axes, reshape, order, mode, cval, prefilter):
rotate_fn = Rotate(angle, spatial_axes, reshape, order, mode, cval, prefilter)
rotated = rotate_fn(self.imt[0])
expected = list()
for channel in self.imt[0]:
expected.append(
scipy.ndimage.rotate(
channel, angle, spatial_axes, reshape, order=order, mode=mode, cval=cval, prefilter=prefilter
)
)
expected = np.stack(expected).astype(np.float32)
self.assertTrue(np.allclose(expected, rotated))
def test_correct_results(self, im_type, angle, keep_size, mode, padding_mode, align_corners):
rotate_fn = Rotate([angle, 0, 0], keep_size, mode, padding_mode, align_corners, dtype=np.float64)
rotated = rotate_fn(im_type(self.imt[0]))
if keep_size:
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)
_order = 0 if mode == "nearest" else 1
if padding_mode == "border":
_mode = "nearest"
elif padding_mode == "reflection":
_mode = "reflect"
else:
_mode = "constant"
expected = []
for channel in self.imt[0]:
expected.append(
scipy.ndimage.rotate(
channel, -np.rad2deg(angle), (1, 2), not keep_size, order=_order, mode=_mode, prefilter=False
)
)
expected = np.stack(expected).astype(np.float32)
rotated = rotated.cpu() if isinstance(rotated, torch.Tensor) else rotated
n_good = np.sum(np.isclose(expected, rotated, atol=1e-3))
self.assertLessEqual(expected.size - n_good, 5, "diff at most 5 pixels")
def test_correct_shape(self, angle, mode, padding_mode, align_corners):
rotate_fn = Rotate(angle, True, align_corners=align_corners)
rotated = rotate_fn(self.imt[0], mode=mode, padding_mode=padding_mode)
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)
def test_correct_shape(self, angle, order, mode, align_corners):
rotate_fn = Rotate(angle, True, align_corners)
rotated = rotate_fn(self.imt[0], interp_order=order, mode=mode)
np.testing.assert_allclose(self.imt[0].shape, rotated.shape)