def test_transformation_identity():
data_shape = np.random.randint(1, 77, 3, dtype=int)
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = data_shape[1:]
m = image_transformation_matrix(source_shape=source_shape,
target_shape=target_shape,
affine_transformation=lambda x: x)
assert np.allclose(data, apply_matrix(data, m, target_shape))
def test_transformation_transpose():
data_shape = np.random.randint(1, 77, 3, dtype=int)
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = tuple(reversed(data_shape[1:]))
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=lambda x: np.flip(x, axis=-1))
assert np.allclose(np.swapaxes(data, 1, 2),
apply_matrix(data, m, target_shape))
def test_fftshift_matrix():
data_shape = np.random.randint(1, 77, 3, dtype=int)
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = data_shape[1:]
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=lambda x: x,
pre_transform=fftshift_coords(target_shape))
assert np.allclose(np.fft.fftshift(data, axes=(1, 2)),
apply_matrix(data, m, target_shape))
def test_difftodect_com_flip_rot_scale(dim):
lt_ctx = lt.Context(InlineJobExecutor())
data_shape = (2, 2, dim, dim)
data = np.zeros(data_shape)
data[0, 0, 7, 7] = 1
data[0, 1, 7, 8] = 1
data[1, 1, 8, 8] = 1
data[1, 0, 8, 7] = 1
source_shape = data_shape[2:]
target_shape = data_shape[2:]
f = diffraction_to_detector(lamb=1,
diffraction_shape=target_shape,
pixel_size_real=1,
pixel_size_detector=1 /
(np.array(target_shape)) * 4,
cy=source_shape[0] / 2,
cx=source_shape[1] / 2,
flip_y=True,
scan_rotation=-90.)
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=f,
)
transformed_data = apply_matrix(data, m, target_shape)
ds = lt_ctx.load('memory', data=data, sig_dims=2)
transformed_ds = lt_ctx.load('memory', data=transformed_data, sig_dims=2)
com_a = lt_ctx.create_com_analysis(dataset=ds,
mask_radius=np.inf,
flip_y=True,
scan_rotation=-90.,
cy=target_shape[0] / 2,
cx=target_shape[1] / 2)
com_res = lt_ctx.run(com_a)
trans_com_a = lt_ctx.create_com_analysis(dataset=transformed_ds,
mask_radius=np.inf,
flip_y=False,
scan_rotation=0.,
cy=target_shape[0] / 2,
cx=target_shape[1] / 2)
trans_com_res = lt_ctx.run(trans_com_a)
print(com_res.field.raw_data)
print(trans_com_res.field.raw_data)
assert np.allclose(com_res.field.raw_data,
np.array(trans_com_res.field.raw_data) / 4)
def test_transformation_scale():
scale = np.random.randint(1, 7)
data_shape = np.random.randint(1, 17, 3, dtype=int) * scale
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = tuple(np.array(source_shape) // scale)
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=lambda x: x * scale,
)
print(data_shape, source_shape, target_shape)
res = apply_matrix(data, m, target_shape)
# Binning, not accumulating intensity but keeping same absolute values
ref = data.reshape((data.shape[0], target_shape[0], scale, target_shape[1],
scale)).mean(axis=(2, 4))
assert np.allclose(ref, res)
def test_transformation_rot90():
data_shape = np.random.randint(1, 77, 3, dtype=int)
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = tuple(reversed(data_shape[1:]))
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=lambda x: x @ rotate_deg(90),
pre_transform=lambda x: x - np.array(target_shape) / 2 + 0.5,
post_transform=lambda x: np.round(x + np.array(source_shape) / 2 - 0.5
).astype(int))
res = apply_matrix(data, m, target_shape)
# positive rotations are clockwise, upper right corner
# is now lower right corner
assert np.allclose(data[:, 0, -1], res[:, -1, -1])
# alternative rotation: transpose and flip
assert np.allclose(np.flip(np.transpose(data, axes=(0, 2, 1)), axis=2),
res)
def test_difftodect_identity():
data_shape = np.random.randint(1, 77, 3, dtype=int)
data = np.random.random(data_shape)
source_shape = data_shape[1:]
target_shape = data_shape[1:]
f = diffraction_to_detector(lamb=1,
diffraction_shape=target_shape,
pixel_size_real=1,
pixel_size_detector=1 /
(np.array(target_shape)),
cy=source_shape[0] / 2,
cx=source_shape[1] / 2,
flip_y=False,
scan_rotation=0.)
m = image_transformation_matrix(
source_shape=source_shape,
target_shape=target_shape,
affine_transformation=f,
)
assert np.allclose(data, apply_matrix(data, m, target_shape))