def test_dataset_shuffle_setter():
try:
new_dataset = dataset.Dataset(shuffle=1.5)
except:
new_dataset = dataset.Dataset()
assert new_dataset.shuffle is True
def test_dataset_normalize_setter():
try:
new_dataset = dataset.Dataset(normalize=1.5)
except:
new_dataset = dataset.Dataset()
assert new_dataset.normalize == (0, 1)
def test_dataset_input_shape_setter():
try:
new_dataset = dataset.Dataset(input_shape=1.5)
except:
new_dataset = dataset.Dataset()
assert new_dataset.input_shape is None
def test_dataset_batch_size_setter():
try:
new_dataset = dataset.Dataset(batch_size=1.5)
except:
new_dataset = dataset.Dataset()
try:
new_dataset = dataset.Dataset(batch_size=-1)
except:
new_dataset = dataset.Dataset()
assert new_dataset.batch_size == 1
def test_dataset_preprocess():
new_dataset = dataset.Dataset(input_shape=(2, 3))
data = np.asarray([[1, 1, 1], [2, 2, 2]])
proc_data = new_dataset.preprocess(data)
assert proc_data.numpy()[0][0] == 0
new_dataset = dataset.Dataset()
proc_data = new_dataset.preprocess(data)
assert proc_data.numpy()[0][0] == 0
def test_dataset_build():
new_dataset = dataset.Dataset()
with pytest.raises(NotImplementedError):
new_dataset._build()
def test_dataset_batch_size():
new_dataset = dataset.Dataset()
assert new_dataset.batch_size == 1
def test_dataset_shuffle():
new_dataset = dataset.Dataset()
assert new_dataset.shuffle is True
def test_dataset_normalize():
new_dataset = dataset.Dataset()
assert new_dataset.normalize == (0, 1)
def test_dataset_input_shape():
new_dataset = dataset.Dataset()
assert new_dataset.input_shape is None
