def testZeros(self):
for s in self.all_shapes:
actual = array_ops.zeros(s)
expected = np.zeros(s)
msg = 'shape: {}'.format(s)
self.match(actual, expected, msg)
for s, t in itertools.product(self.all_shapes, self.all_types):
actual = array_ops.zeros(s, t)
expected = np.zeros(s, t)
msg = 'shape: {}, dtype: {}'.format(s, t)
self.match(actual, expected, msg)
def _testEvalOnShapes(self, transformer):
def f(a, b):
return array_ops.sum(math_ops.sqrt(math_ops.exp(a)) + b)
f_prime = transformer(f)
shape = [10]
dtype = np.float16
a = array_ops.zeros(shape=shape, dtype=dtype)
b = array_ops.zeros(shape=shape, dtype=dtype)
expected = f(a, b)
got = f_prime(a, b)
self.assertAllEqual(expected.shape, got.shape)
self.assertAllEqual(expected.dtype, got.dtype)
# Call again since the code path is different on second call
got = f_prime(a, b)
self.assertAllEqual(expected.shape, got.shape)
self.assertAllEqual(expected.dtype, got.dtype)
def testAsAnyArray(self):
for a, dtype in itertools.product(self.all_arrays, self.all_types):
self.match(array_ops.asanyarray(a, dtype=dtype),
np.asanyarray(a, dtype=dtype))
zeros_list = array_ops.zeros(5)
# Same instance is returned if no dtype is specified and input is ndarray.
self.assertIs(array_ops.asanyarray(zeros_list), zeros_list)
# Different instance is returned if dtype is specified and input is ndarray.
self.assertIsNot(array_ops.asanyarray(zeros_list, dtype=int),
zeros_list)
def testArray(self):
ndmins = [0, 1, 2, 5]
for a, dtype, ndmin, copy in itertools.product(self.all_arrays,
self.all_types, ndmins,
[True, False]):
self.match(array_ops.array(a, dtype=dtype, ndmin=ndmin, copy=copy),
np.array(a, dtype=dtype, ndmin=ndmin, copy=copy))
zeros_list = array_ops.zeros(5)
# TODO(srbs): Test that copy=True when context.device is different from
# tensor device copies the tensor.
# Backing tensor is the same if copy=False, other attributes being None.
self.assertIs(
array_ops.array(zeros_list, copy=False).data, zeros_list.data)
self.assertIs(
array_ops.array(zeros_list.data, copy=False).data, zeros_list.data)
# Backing tensor is different if ndmin is not satisfied.
self.assertIsNot(
array_ops.array(zeros_list, copy=False, ndmin=2).data,
zeros_list.data)
self.assertIsNot(
array_ops.array(zeros_list.data, copy=False, ndmin=2).data,
zeros_list.data)
self.assertIs(
array_ops.array(zeros_list, copy=False, ndmin=1).data,
zeros_list.data)
self.assertIs(
array_ops.array(zeros_list.data, copy=False, ndmin=1).data,
zeros_list.data)
# Backing tensor is different if dtype is not satisfied.
self.assertIsNot(
array_ops.array(zeros_list, copy=False, dtype=int).data,
zeros_list.data)
self.assertIsNot(
array_ops.array(zeros_list.data, copy=False, dtype=int).data,
zeros_list.data)
self.assertIs(
array_ops.array(zeros_list, copy=False, dtype=float).data,
zeros_list.data)
self.assertIs(
array_ops.array(zeros_list.data, copy=False, dtype=float).data,
zeros_list.data)
def f_prime(a, b): shape_dtype = extensions.eval_on_shapes(f)(a, b) return array_ops.zeros(shape=shape_dtype.shape, dtype=shape_dtype.dtype)
def zeros(x): return tf.nest.map_structure(lambda _: array_ops.zeros([], np.float32), x)