def _check_array(
array, expected_dtype, expected_shape, expected_data_list=None,
device=None):
expected_dtype = chainerx.dtype(expected_dtype)
assert isinstance(array.dtype, chainerx.dtype)
assert isinstance(array.shape, tuple)
assert array.dtype == expected_dtype
assert array.shape == expected_shape
assert array.itemsize == expected_dtype.itemsize
assert array.size == array_utils.total_size(expected_shape)
assert array.nbytes == expected_dtype.itemsize * \
array_utils.total_size(expected_shape)
if expected_data_list is not None:
assert array._debug_flat_data == expected_data_list
assert array.is_contiguous
array_utils.check_device(array, device)
def _check_array(
array, expected_dtype, expected_shape, expected_data_list=None,
device=None):
expected_dtype = chainerx.dtype(expected_dtype)
assert isinstance(array.dtype, chainerx.dtype)
assert isinstance(array.shape, tuple)
assert array.dtype == expected_dtype
assert array.shape == expected_shape
assert array.itemsize == expected_dtype.itemsize
assert array.size == array_utils.total_size(expected_shape)
assert array.nbytes == expected_dtype.itemsize * \
array_utils.total_size(expected_shape)
if expected_data_list is not None:
assert array._debug_flat_data == expected_data_list
assert array.is_contiguous
array_utils.check_device(array, device)
def _make_inputs(shapes, dtypes):
# Generates input ndarrays.
assert isinstance(shapes, (list, tuple))
assert isinstance(dtypes, (list, tuple))
assert len(shapes) == len(dtypes)
inputs = []
for i, (shape, dtype) in enumerate(zip(shapes, dtypes)):
size = array_utils.total_size(shape)
a = numpy.arange(i * 100, i * 100 + size)
a = a.reshape(shape)
a = a.astype(dtype)
inputs.append(a)
assert len(inputs) > 0
return tuple(inputs)
def _make_inputs(shapes, dtypes):
# Generates input ndarrays.
assert isinstance(shapes, (list, tuple))
assert isinstance(dtypes, (list, tuple))
assert len(shapes) == len(dtypes)
inputs = []
for i, (shape, dtype) in enumerate(zip(shapes, dtypes)):
size = array_utils.total_size(shape)
a = numpy.arange(i * 100, i * 100 + size)
a = a.reshape(shape)
a = a.astype(dtype)
inputs.append(a)
assert len(inputs) > 0
return tuple(inputs)
def test_diagflat_invalid_ndim(xp, k, shape, device):
v = xp.arange(array_utils.total_size(shape)).reshape(shape)
return xp.diagflat(v, k)
def test_diag(xp, k, shape, transpose, device):
v = xp.arange(array_utils.total_size(shape)).reshape(shape)
if transpose: # Test non-contiguous inputs for multi-dimensional shapes.
v = v.T
return xp.diag(v, k)
def test_diagflat_invalid_ndim(xp, k, shape, device):
v = xp.arange(array_utils.total_size(shape), dtype='int32').reshape(shape)
return xp.diagflat(v, k)
def test_diag(xp, k, shape, transpose, device):
v = xp.arange(array_utils.total_size(shape), dtype='int32').reshape(shape)
if transpose: # Test non-contiguous inputs for multi-dimensional shapes.
v = v.T
return xp.diag(v, k)
def test_total_size(expected, shape):
assert expected == array_utils.total_size(shape)
def test_diagflat(xp, k, shape, device):
v = xp.arange(array_utils.total_size(shape), dtype='int32').reshape(shape)
return xp.diagflat(v, k)
def test_diagflat(xp, k, shape, device):
v = xp.arange(array_utils.total_size(shape)).reshape(shape)
return xp.diagflat(v, k)
def test_diag_invalid_ndim(xp, k, shape, device):
v = xp.arange(array_utils.total_size(shape)).reshape(shape)
return xp.diag(v, k)
def test_total_size(expected, shape):
assert expected == array_utils.total_size(shape)
