def test_guess_multiscale():
data = np.random.random((10, 15))
assert not guess_multiscale(data)[0]
data = np.random.random((10, 15, 6))
assert not guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6))]
assert not guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6)), np.random.random((5, 7, 3))]
assert guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6)), np.random.random((10, 7, 3))]
assert guess_multiscale(data)[0]
data = tuple(data)
assert guess_multiscale(data)[0]
data = tuple(
pyramid_gaussian(np.random.random((10, 15)), multichannel=False)
)
assert guess_multiscale(data)[0]
data = np.asarray(
tuple(pyramid_gaussian(np.random.random((10, 15)), multichannel=False))
)
assert guess_multiscale(data)[0]
# Check for integer overflow with big data
s = 8192
data = [da.ones((s,) * 3), da.ones((s // 2,) * 3), da.ones((s // 4,) * 3)]
assert guess_multiscale(data)[0]
def test_guess_multiscale_non_array_list():
"""Check that non-decreasing list input raises ValueError"""
data = [
np.empty((10, 15, 6)),
] * 2 # noqa: E231
with pytest.raises(ValueError):
_, _ = guess_multiscale(data)
def test_guess_multiscale():
data = np.random.random((10, 15))
assert not guess_multiscale(data)[0]
data = np.random.random((10, 15, 6))
assert not guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6))]
assert not guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6)), np.random.random((5, 7, 3))]
assert guess_multiscale(data)[0]
data = [np.random.random((10, 15, 6)), np.random.random((10, 7, 3))]
assert guess_multiscale(data)[0]
data = tuple(data)
assert guess_multiscale(data)[0]
if skimage.__version__ > '0.19':
pyramid_kwargs = {'channel_axis': None}
else:
pyramid_kwargs = {'multichannel': False}
data = tuple(pyramid_gaussian(np.random.random((10, 15)),
**pyramid_kwargs))
assert guess_multiscale(data)[0]
data = np.asarray(
tuple(pyramid_gaussian(np.random.random((10, 15)), **pyramid_kwargs)),
dtype=object,
)
assert guess_multiscale(data)[0]
# Check for integer overflow with big data
s = 8192
data = [
da.ones((s, ) * 3),
da.ones((s // 2, ) * 3),
da.ones((s // 4, ) * 3)
]
assert guess_multiscale(data)[0]
def test_timing_multiscale_big():
assert not guess_multiscale(data_dask)[0]
def test_guess_multiscale_incorrect_order():
data = [np.empty((10, 15)), np.empty((5, 6)), np.empty((20, 15))]
with pytest.raises(ValueError):
_, _ = guess_multiscale(data)
def test_guess_multiscale_strip_single_scale():
data = [np.empty((10, 10))]
guess, data_out = guess_multiscale(data)
assert data_out is data[0]
assert guess is False
def test_timing_multiscale_big():
now = time.monotonic()
assert not guess_multiscale(data_dask)[0]
elapsed = time.monotonic() - now
assert elapsed < 2, "test was too slow, computation was likely not lazy"
