def test_simple(self):
length = 128
frames = 3
x, y = _get_dummy_tracking_data(length, frames=frames)
num_objects = len(np.unique(y)) - 1
model = DummyModel()
_ = tracking.CellTracker(x, y, model=model)
# test data with bad rank
with pytest.raises(ValueError):
tracking.CellTracker(np.random.random((32, 32, 1)),
np.random.randint(num_objects,
size=(32, 32, 1)),
model=model)
# test mismatched x and y shape
with pytest.raises(ValueError):
tracking.CellTracker(np.random.random((3, 32, 32, 1)),
np.random.randint(num_objects,
size=(2, 32, 32, 1)),
model=model)
# test bad features
with pytest.raises(ValueError):
tracking.CellTracker(x, y, model=model, features=None)
# test bad data_format
with pytest.raises(ValueError):
tracking.CellTracker(x, y, model=model, data_format='invalid')
def test__sub_area(self):
length = 128
frames = 3
model = DummyModel()
# TODO: Fix for channels_first
for data_format in ('channels_last', ): # 'channels_first'):
x, y = _get_dummy_tracking_data(length,
frames=frames,
data_format=data_format)
tracker = tracking.CellTracker(x,
y,
model=model,
data_format=data_format)
for f in range(frames):
if data_format == 'channels_first':
xf = x[:, f]
yf = y[:, f]
else:
xf = x[f]
yf = y[f]
sub = tracker._sub_area(xf, yf, 1)
expected_shape = tracker.get_feature_shape('neighborhood')
assert sub.shape == expected_shape
def test_simple(self):
data_format = 'channels_last'
frames = 3
labels_per_frame = 5
y = get_annotated_movie(img_size=256,
labels_per_frame=labels_per_frame,
frames=frames,
mov_type='sequential',
seed=0,
data_format=data_format)
x = np.random.random(y.shape)
num_objects = len(np.unique(y)) - 1
model = DummyModel()
encoder = DummyEncoder()
_ = tracking.CellTracker(x,
y,
tracking_model=model,
neighborhood_encoder=encoder)
# test data with bad rank
with pytest.raises(ValueError):
tracking.CellTracker(np.random.random((32, 32, 1)),
np.random.randint(num_objects,
size=(32, 32, 1)),
tracking_model=model,
neighborhood_encoder=encoder)
# test mismatched x and y shape
with pytest.raises(ValueError):
tracking.CellTracker(np.random.random((3, 32, 32, 1)),
np.random.randint(num_objects,
size=(2, 32, 32, 1)),
tracking_model=model,
neighborhood_encoder=encoder)
# test bad data_format
with pytest.raises(ValueError):
tracking.CellTracker(x,
y,
tracking_model=model,
neighborhood_encoder=encoder,
data_format='invalid')
def test_get_feature_shape(self):
length = 128
frames = 3
x, y = _get_dummy_tracking_data(length, frames=frames)
model = DummyModel()
for data_format in ('channels_first', 'channels_last'):
tracker = tracking.CellTracker(x,
y,
model=model,
data_format=data_format)
for f in tracker.features:
shape = tracker.get_feature_shape(f)
assert isinstance(shape, tuple)
with pytest.raises(ValueError):
tracker.get_feature_shape('bad feature name')
def test_fetch_tracked_features(self):
length = 128
frames = 5
# TODO: Fix for channels_first
for data_format in ('channels_last', ): # 'channels_first'):
x, y = _get_dummy_tracking_data(length,
frames=frames,
data_format=data_format)
for track_length in (1, frames // 2 + 1, frames + 1):
tracker = tracking.CellTracker(x,
y,
model=DummyModel(),
track_length=track_length,
data_format=data_format)
tracker._initialize_tracks()
axis = tracker.channel_axis
tracked_features = tracker.fetch_tracked_features()
for feature_name in tracker.features:
feature_shape = tracker.get_feature_shape(feature_name)
feature = tracked_features[feature_name]
axis = tracker.channel_axis
assert feature.shape[axis] == feature_shape[axis]
assert feature.shape[0] == len(tracker.tracks)
assert feature.shape[tracker.time_axis + 1] == track_length
tracked_features = tracker.fetch_tracked_features(
before_frame=frames // 2 + 1)
for feature_name in tracker.features:
feature_shape = tracker.get_feature_shape(feature_name)
feature = tracked_features[feature_name]
axis = tracker.channel_axis
assert feature.shape[axis] == feature_shape[axis]
assert feature.shape[0] == len(tracker.tracks)
assert feature.shape[tracker.time_axis + 1] == track_length
def test_track_cells(self):
frames = 10
track_length = 3
labels_per_frame = 3
# TODO: test detected divisions
# TODO: test creating new track
# TODO: Fix for channels_first
for data_format in ('channels_last', ): # 'channels_first'):
y1 = get_annotated_movie(img_size=256,
labels_per_frame=labels_per_frame,
frames=frames,
mov_type='sequential',
seed=1,
data_format=data_format)
y2 = get_annotated_movie(img_size=256,
labels_per_frame=labels_per_frame * 2,
frames=frames,
mov_type='sequential',
seed=2,
data_format=data_format)
y3 = get_annotated_movie(img_size=256,
labels_per_frame=labels_per_frame,
frames=frames,
mov_type='sequential',
seed=3,
data_format=data_format)
y = np.concatenate((y1, y2, y3))
x = np.random.random(y.shape)
tracker = tracking.CellTracker(x,
y,
tracking_model=DummyModel(),
neighborhood_encoder=DummyEncoder(),
track_length=track_length,
data_format=data_format)
tracker.track_cells()
# test tracker.dataframe
df = tracker.dataframe(cell_type='test-value')
assert isinstance(df, pd.DataFrame)
assert 'cell_type' in df.columns # pylint: disable=E1135
# test incorrect values in tracker.dataframe
with pytest.raises(ValueError):
tracker.dataframe(bad_value=-1)
try:
# test tracker.postprocess
tempdir = tempfile.mkdtemp() # create dir
path = os.path.join(tempdir, 'postprocess.xyz')
tracker.postprocess(filename=path)
post_saved_path = os.path.join(tempdir, 'postprocess.trk')
assert os.path.isfile(post_saved_path)
# test tracker.dump
path = os.path.join(tempdir, 'test.xyz')
tracker.dump(path)
dump_saved_path = os.path.join(tempdir, 'test.trk')
assert os.path.isfile(dump_saved_path)
# utility tests for loading trk files
# TODO: move utility tests into utils_test.py
# test trk_folder_to_trks
utils.trk_folder_to_trks(tempdir,
os.path.join(tempdir, 'all.trks'))
assert os.path.isfile(os.path.join(tempdir, 'all.trks'))
# test load_trks
data = utils.load_trks(post_saved_path)
assert isinstance(data['lineages'], list)
assert all(isinstance(d, dict) for d in data['lineages'])
np.testing.assert_equal(data['X'], tracker.X)
np.testing.assert_equal(data['y'], tracker.y_tracked)
# load trks instead of trk
data = utils.load_trks(os.path.join(tempdir, 'all.trks'))
# test trks_stats
utils.trks_stats(os.path.join(tempdir, 'test.trk'))
finally:
try:
shutil.rmtree(tempdir) # delete directory
except OSError as exc:
if exc.errno != errno.ENOENT: # no such file or directory
raise # re-raise exception
def test_track_cells(self):
length = 128
frames = 5
track_length = 2
features = ['appearance', 'neighborhood', 'regionprop', 'distance']
# TODO: Fix for channels_first
for data_format in ('channels_last', ): # 'channels_first'):
x, y = _get_dummy_tracking_data(length,
frames=frames,
data_format=data_format)
tracker = tracking.CellTracker(x,
y,
model=DummyModel(),
track_length=track_length,
data_format=data_format,
features=features)
tracker.track_cells()
# test tracker.dataframe
df = tracker.dataframe(cell_type='test-value')
assert isinstance(df, pd.DataFrame)
assert 'cell_type' in df.columns # pylint: disable=E1135
# test incorrect values in tracker.dataframe
with pytest.raises(ValueError):
tracker.dataframe(bad_value=-1)
try:
# test tracker.postprocess
tempdir = tempfile.mkdtemp() # create dir
path = os.path.join(tempdir, 'postprocess.xyz')
tracker.postprocess(filename=path)
post_saved_path = os.path.join(tempdir, 'postprocess.trk')
assert os.path.isfile(post_saved_path)
# test tracker.dump
path = os.path.join(tempdir, 'test.xyz')
tracker.dump(path)
dump_saved_path = os.path.join(tempdir, 'test.trk')
assert os.path.isfile(dump_saved_path)
# utility tests for loading trk files
# TODO: move utility tests into utils_test.py
# test trk_folder_to_trks
utils.trk_folder_to_trks(tempdir,
os.path.join(tempdir, 'all.trks'))
assert os.path.isfile(os.path.join(tempdir, 'all.trks'))
# test load_trks
data = utils.load_trks(post_saved_path)
assert isinstance(data['lineages'], list)
assert all(isinstance(d, dict) for d in data['lineages'])
np.testing.assert_equal(data['X'], tracker.x)
np.testing.assert_equal(data['y'], tracker.y_tracked)
# load trks instead of trk
data = utils.load_trks(os.path.join(tempdir, 'all.trks'))
# test trks_stats
utils.trks_stats(os.path.join(tempdir, 'test.trk'))
finally:
try:
shutil.rmtree(tempdir) # delete directory
except OSError as exc:
if exc.errno != errno.ENOENT: # no such file or directory
raise # re-raise exception