def test_directory_parsing_label_dataset(self):
dataset = DirectoryParsingLabelDataset(
self.tmp_dir, color=self.color)
if self.depth == 1:
expected_legnth = self.n_img_per_class * self.n_class
elif self.depth == 2:
expected_legnth =\
self.n_img_per_class * self.n_sub_directory * self.n_class
self.assertEqual(len(dataset), expected_legnth)
assert_is_label_dataset(dataset, self.n_class, color=self.color)
label_names = directory_parsing_label_names(self.tmp_dir)
self.assertEqual(
label_names, ['class_{}'.format(i) for i in range(self.n_class)])
if self.depth == 1:
self.assertEqual(
dataset.img_paths,
['{}/class_{}/img{}.{}'.format(self.tmp_dir, i, j, self.suffix)
for i in range(self.n_class)
for j in range(self.n_img_per_class)])
elif self.depth == 2:
self.assertEqual(
dataset.img_paths,
['{}/class_{}/nested_{}/img{}.{}'.format(
self.tmp_dir, i, j, k, self.suffix)
for i in range(self.n_class)
for j in range(self.n_sub_directory)
for k in range(self.n_img_per_class)])
def test_directory_parsing_label_dataset(self):
dataset = DirectoryParsingLabelDataset(
self.tmp_dir, color=self.color)
if self.depth == 1:
expected_legnth = self.n_img_per_class * self.n_class
elif self.depth == 2:
expected_legnth =\
self.n_img_per_class * self.n_sub_directory * self.n_class
self.assertEqual(len(dataset), expected_legnth)
assert_is_label_dataset(dataset, self.n_class, color=self.color)
label_names = directory_parsing_label_names(self.tmp_dir)
self.assertEqual(
label_names, ['class_{}'.format(i) for i in range(self.n_class)])
if self.depth == 1:
self.assertEqual(
dataset.img_paths,
['{}/class_{}/img{}.{}'.format(self.tmp_dir, i, j, self.suffix)
for i in range(self.n_class)
for j in range(self.n_img_per_class)])
elif self.depth == 2:
self.assertEqual(
dataset.img_paths,
['{}/class_{}/nested_{}/img{}.{}'.format(
self.tmp_dir, i, j, k, self.suffix)
for i in range(self.n_class)
for j in range(self.n_sub_directory)
for k in range(self.n_img_per_class)])
def test_numerical_sort(self):
dataset = DirectoryParsingLabelDataset(
self.tmp_dir, numerical_sort=True)
assert_is_label_dataset(dataset, self.n_class)
label_names = directory_parsing_label_names(
self.tmp_dir, numerical_sort=True)
self.assertEqual(
label_names, ['{}'.format(i) for i in range(self.n_class)])
def test_numerical_sort(self):
dataset = DirectoryParsingLabelDataset(
self.tmp_dir, numerical_sort=True)
assert_is_label_dataset(dataset, self.n_class)
label_names = directory_parsing_label_names(
self.tmp_dir, numerical_sort=True)
self.assertEqual(
label_names, ['{}'.format(i) for i in range(self.n_class)])
def test_assert_is_label_dataset(self):
if hasattr(self, 'option'):
dataset = self.dataset(self.color, self.option)
else:
dataset = self.dataset(self.color)
if self.valid:
assert_is_label_dataset(dataset, 20, color=self.color)
else:
with self.assertRaises(AssertionError):
assert_is_label_dataset(dataset, 20, color=self.color)
def test_assert_is_label_dataset(self):
if hasattr(self, 'option'):
dataset = self.dataset(self.color, self.option)
else:
dataset = self.dataset(self.color)
if self.valid:
assert_is_label_dataset(dataset, 20, color=self.color)
else:
with self.assertRaises(AssertionError):
assert_is_label_dataset(dataset, 20, color=self.color)
def test_cub_label_dataset(self):
assert_is_label_dataset(
self.dataset, len(cub_label_names), n_example=10)
idx = np.random.choice(np.arange(10))
if self.return_bb:
bb = self.dataset[idx][2]
assert_is_bbox(bb[np.newaxis])
if self.return_prob_map:
img = self.dataset[idx][0]
prob_map = self.dataset[idx][-1]
self.assertEqual(prob_map.dtype, np.float32)
self.assertEqual(prob_map.shape, img.shape[1:])
self.assertTrue(np.min(prob_map) >= 0)
self.assertTrue(np.max(prob_map) <= 1)
def test_online_products_dataset(self):
assert_is_label_dataset(
self.dataset, 22634, n_example=10)
for _ in range(10):
i = np.random.randint(0, len(self.dataset))
_, _, super_label = self.dataset[i]
assert isinstance(super_label, np.int32), \
'label must be a numpy.int32.'
assert super_label.ndim == 0, 'The ndim of label must be 0'
assert (super_label >= 0 and
super_label < len(online_products_super_label_names)), \
'The value of label must be in [0, n_class - 1].'
def test_cub_label_dataset(self):
assert_is_label_dataset(self.dataset,
len(cub_label_names),
n_example=10)
idx = np.random.choice(np.arange(10))
if self.return_bb:
bb = self.dataset[idx][2]
assert_is_bbox(bb[np.newaxis])
if self.return_prob_map:
img = self.dataset[idx][0]
prob_map = self.dataset[idx][-1]
self.assertEqual(prob_map.dtype, np.float32)
self.assertEqual(prob_map.shape, img.shape[1:])
self.assertTrue(np.min(prob_map) >= 0)
self.assertTrue(np.max(prob_map) <= 1)