def test_data_splitter_with_np_array_with_larger_last_slice():
x = np.ones([32, 32, 32, 3], dtype=np.float32)
res2 = data_splitter(x, 5)
assert all(
map(lambda x: shape_as_list(x) == [6, 32, 32, 3],
[res2['slice{}'.format(i)] for i in range(4)])) is True
assert shape_as_list(res2['slice4']) == [8, 32, 32, 3]
def test_data_splitter_with_tensor_with_larger_last_slice():
x = tf.ones([32, 64, 64, 3], dtype=tf.float32)
x = Tensor(x)
res2 = data_splitter(x, 5)
assert all(
map(lambda x: shape_as_list(x) == [6, 64, 64, 3],
[res2['slice{}'.format(i)] for i in range(4)])) is True
assert all(
map(lambda x: isinstance(x, Tensor),
[res2['slice{}'.format(i)] for i in range(5)])) is True
assert shape_as_list(res2['slice4']) == [8, 64, 64, 3]
def _(x, nb_split, name='data_splitter'):
shape_x = shape_as_list(x)
split_size = shape_x[0] // nb_split
res = {}
offset = 0
for i in range(nb_split - 1):
res['slice{}'.format(i)] = x[offset:offset + split_size]
offset = offset + split_size
res['slice{}'.format(nb_split - 1)] = x[offset:]
return res
def _(x, nb_split, name='data_splitter'):
shape_x = shape_as_list(x)
split_size = shape_x[0] // nb_split
res = {}
offset = [0 for i in shape_x]
slice_size = copy.copy(shape_x)
slice_size[0] = split_size
for i in range(nb_split - 1):
res['slice{}'.format(i)] = tf.slice(x, offset, slice_size)
offset[0] = offset[0] + split_size
slice_size[0] = shape_x[0] - offset[0]
res['slice{}'.format(nb_split - 1)] = tf.slice(x, offset, slice_size)
return res
def test_data_splitter_with_tf_tensor():
x = tf.ones([32, 64, 64, 3], dtype=tf.float32)
res1 = data_splitter(x, 4)
assert all(
map(lambda x: shape_as_list(x) == [8, 64, 64, 3],
[res1['slice{}'.format(i)] for i in range(4)])) is True
def test_data_splitter_with_np_array():
x = np.ones([32, 32, 32, 3], dtype=np.float32)
res1 = data_splitter(x, 4)
assert all(
map(lambda x: shape_as_list(x) == [8, 32, 32, 3],
[res1['slice{}'.format(i)] for i in range(4)])) is True