def test_relabel():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs.columns = ['x', 'y', 'z', 'new_label', 't']
trajs.relabel(inplace=True)
new_values = [
[1.933058243735795, -14.581064591435775, 11.603556633147544, 0.0],
[-12.862215173899491, -2.8611502446443238, -2.2738941196781424, 0.0],
[9.100887851132633, 2.837252570763561, 2.875753940450461, 0.0],
[-9.253860446235523, 11.345550876585719, 22.118203258275745, 0.0]
]
assert trajs.iloc[:4].values.tolist() == new_values
trajs = Trajectories(data.brownian_trajs_df())
trajs.columns = ['x', 'y', 'z', 'new_label', 't']
trajs = trajs.relabel(inplace=False)
new_values = [
[1.933058243735795, -14.581064591435775, 11.603556633147544, 0.0],
[-12.862215173899491, -2.8611502446443238, -2.2738941196781424, 0.0],
[9.100887851132633, 2.837252570763561, 2.875753940450461, 0.0],
[-9.253860446235523, 11.345550876585719, 22.118203258275745, 0.0]
]
assert trajs.iloc[:4].values.tolist() == new_values
def test_merge_segments():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs.reset_index(inplace=True)
trajs.loc[15, ['label']] = 88
trajs.loc[20, ['label']] = 88
trajs.set_index(['t_stamp', 'label'], inplace=True)
new_trajs = trajs.merge_segments([0, 88], inplace=False)
assert_array_almost_equal(trajs.values, new_trajs.values)
trajs = Trajectories(data.brownian_trajs_df())
good_trajs = trajs.copy()
trajs.reset_index(inplace=True)
trajs.loc[15, ['label']] = 88
trajs.loc[20, ['label']] = 88
trajs.set_index(['t_stamp', 'label'], inplace=True)
trajs.merge_segments([0, 88], inplace=True)
assert_array_almost_equal(trajs.values, good_trajs.values)
def test_merge():
"""
"""
trajs1 = Trajectories(data.brownian_trajs_df())
trajs2 = Trajectories(data.brownian_trajs_df())
new = trajs1.merge(trajs2)
assert len(trajs1.labels) + len(trajs2.labels) == len(new.labels)
def test_reverse():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
assert trajs.reverse().shape == (25, 5)
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
trajs.reverse(inplace=True)
assert trajs.shape == (25, 5)
def test_get_colors():
"""
"""
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
colors = trajs.get_colors()
assert colors == {
0: '#FF0000',
1: '#ADFF00',
2: '#00FFA9',
3: '#0408FF',
4: '#FF00AC'
}
colors = trajs.get_colors(alpha=0.5)
assert colors == {
0: '#FF000080',
1: '#ADFF0080',
2: '#00FFA980',
3: '#0408FF80',
4: '#FF00AC80'
}
colors = trajs.get_colors(rgba=True)
good_colors = {
0: (1.0, 0.0, 0.0, 1.0),
1: (0.67977809154279767, 1.0, 0.0, 1.0),
2: (0.0, 1.0, 0.66360181783683614, 1.0),
3: (0.015440535661123769, 0.031618928677752463, 1.0, 1.0),
4: (1.0, 0.0, 0.67279469669175529, 1.0)
}
assert colors == good_colors
def test_attributes():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
assert_array_equal(trajs.t_stamps, np.array([0, 1, 2, 3, 4]))
assert_array_equal(trajs.labels, np.array([0, 1, 2, 3, 4]))
segments = {
0: [(0, 0), (1, 0), (2, 0), (3, 0), (4, 0)],
1: [(0, 1), (1, 1), (2, 1), (3, 1), (4, 1)],
2: [(0, 2), (1, 2), (2, 2), (3, 2), (4, 2)],
3: [(0, 3), (1, 3), (2, 3), (3, 3), (4, 3)],
4: [(0, 4), (1, 4), (2, 4), (3, 4), (4, 4)]
}
assert_dict_equal(trajs.segment_idxs, segments)
traj = np.array([[-9.25386045, 11.34555088, 22.11820326, 3., 0.],
[11.05321776, 3.23738477, 2.62790435, 2., 1.],
[16.6824928, 14.602054, -12.1218683, 4., 2.],
[17.22410516, 14.8068125, -11.87642753, 4., 3.],
[2.80222495, -13.13783042, 8.56406878, 0., 4.]])
t_stamp, traj_to_test = list(trajs.iter_segments)[0]
assert_array_almost_equal(traj, traj_to_test)
assert list(trajs.get_segments().keys()) == [0, 1, 2, 3, 4]
def test_project():
trajs = Trajectories(data.directed_motion_trajs_df())
trajs.rename(columns={'true_label': 'new_label'}, inplace=True)
trajs.relabel()
trajs.project([0, 1],
coords=['x', 'y'],
keep_first_time=False,
reference=None,
inplace=True,
progress=False)
excepted = np.array([[0.27027431, 0.], [-0.27027431, 0.],
[-0.25306519, 0.69683713], [0.04633664, 0.31722648]])
assert_array_almost_equal(excepted,
trajs.loc[:, ['x_proj', 'y_proj']].values[:4])
trajs = trajs.project([0, 1],
coords=['x', 'y'],
keep_first_time=False,
reference=None,
inplace=False,
progress=False)
assert_array_almost_equal(excepted,
trajs.loc[:, ['x_proj', 'y_proj']].values[:4])
assert_raises(ValueError,
trajs.project, [0, 1],
coords=['x', 'y', 'z', 't'])
def test_write_hdf():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
tmp_store = tempfile.NamedTemporaryFile(suffix='h5')
with pd.get_store(tmp_store.name) as store:
store['trajs'] = trajs
def get_get_diff():
trajs = Trajectories(data.brownian_trajs_df())
diffs = trajs.get_diff()
x_diffs = diffs.to_dict()['x']
real_x_diffs = {
(1, 2): 3.8452299074207819,
(3, 2): 4.7476193900872765,
(0, 0): np.nan,
(3, 0): 0.54161236467700746,
(0, 4): np.nan,
(1, 4): -5.6929349491048624,
(1, 3): -30.136494087633611,
(2, 3): 23.240228721514185,
(2, 1): -23.9264368052234,
(2, 4): 0.63465512968445115,
(4, 2): -4.5501817884252063,
(1, 0): 20.307078207040306,
(0, 3): np.nan,
(4, 0): -14.421880216023439,
(0, 1): np.nan,
(3, 3): -6.5845079821965991,
(4, 1): -19.329775838349192,
(3, 1): 18.084232469105203,
(4, 4): 24.644945052453025,
(0, 2): np.nan,
(2, 0): 5.6292750381105723,
(4, 3): 13.209596167161628,
(2, 2): -3.7469188310869228,
(3, 4): -17.381636024737336,
(1, 1): 13.827909766138866
}
assert_almost_equal(x_diffs, real_x_diffs)
def test_brownian_gap_close():
trajs = Trajectories(data.with_gaps_df())
max_speed = 10.
cost_func = BrownianGapCloseCostFunction(
parameters={'max_speed': max_speed})
in_idxs = [(3, 0), (3, 0), (5, 1), (13, 2), (13, 2), (16, 3)]
out_idxs = [(5, 3), (7, 4), (7, 4), (15, 5), (18, 6), (18, 6)]
cost_func.context['trajs'] = trajs
cost_func.context['idxs_in'] = in_idxs
cost_func.context['idxs_out'] = out_idxs
cost_func.get_block()
m = cost_func.mat
m_true = [[
np.nan, np.nan, np.nan, 2.0553543815820543, 45.74823093541038, np.nan,
np.nan
], [np.nan, np.nan, np.nan, np.nan, 1.1530875267623917, np.nan, np.nan],
[
np.nan, np.nan, np.nan, np.nan, np.nan, 0.8254178191605767,
23.84513283483594
],
[
np.nan, np.nan, np.nan, np.nan, np.nan, np.nan,
4.916599888362159
], [np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan],
[np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan],
[np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]]
assert_array_almost_equal(m, m_true)
def test_back_proj_pca():
trajs = Trajectories(data.with_gaps_df())
rotated, pca = transformation.do_pca(trajs, suffix='',
append=False, return_pca=True)
back_proj = transformation.back_proj_pca(rotated, pca, coords=['x', 'y', 'z'])
assert_almost_equal(back_proj.iloc[0].x, trajs.iloc[0].x)
def test_get_shortest_segments():
"""
"""
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
assert trajs.get_shortest_segments(1) == [0]
def test_radial_speed():
trajs = Trajectories(data.with_gaps_df())
trajs.reset_index(level='label', drop=True, inplace=True)
trajs.set_index('true_label', drop=True, append=True, inplace=True)
trajs.index.set_names(['t_stamp', 'label'], inplace=True)
rd_speeds = rotation.radial_speed(trajs,
in_coords=['x', 'y'],
from_polar=False)
def test_structure():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
assert_raises(ValueError, trajs.check_trajs_df_structure, ['t_idx'])
assert_raises(ValueError, trajs.check_trajs_df_structure,
['t_stamp', 'label'], ['dx'])
trajs.check_trajs_df_structure(['t_stamp', 'label'], ['x', 'y', 'z'])
def test_remove_segments():
"""
"""
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
trajs.remove_segments(1, inplace=True)
assert np.all(trajs.labels == [0, 2, 3, 4])
def test_get_t_stamps_correspondences():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs['t'] *= 33
data_values = [132, 33, 99, 66, 33, 33, 99., 99, 132]
t_stamps = trajs.get_t_stamps_correspondences(data_values, column='t')
assert_array_equal(t_stamps, [4, 1, 3, 2, 1, 1, 3, 3, 4])
def test_back_proj_interp():
trajs = Trajectories(data.with_gaps_df())
interpolated = transformation.time_interpolate(trajs,
sampling=3, s=0)
back_proj = transformation.back_proj_interp(interpolated, trajs, 3)
assert_almost_equal(back_proj.iloc[0].x, trajs.iloc[0].x)
interpolated = transformation.time_interpolate(trajs,
sampling=3, s=2)
back_proj = transformation.back_proj_interp(interpolated, trajs, 3)
assert_almost_equal(back_proj.iloc[0].x, trajs.iloc[0].x)
def test_duplicate_segments():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs = trajs.duplicate_segments(2)
new_indexes = [(0, 0), (0, 1), (0, 2), (0, 3), (0, 4), (0, 5), (1, 0),
(1, 1), (1, 2), (1, 3), (1, 4), (1, 5), (2, 0), (2, 1),
(2, 2), (2, 3), (2, 4), (2, 5), (3, 0), (3, 1), (3, 2),
(3, 3), (3, 4), (3, 5), (4, 0), (4, 1), (4, 2), (4, 3),
(4, 4), (4, 5)]
assert trajs.index.tolist() == new_indexes
def test_cut_segments():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs.cut_segments((2, 3), inplace=True)
new_indexes = [(0, 0), (0, 1), (0, 2), (0, 3), (0, 4), (1, 0), (1, 1),
(1, 2), (1, 3), (1, 4), (2, 0), (2, 1), (2, 2), (2, 3),
(2, 4), (3, 0), (3, 1), (3, 2), (3, 4), (3, 5), (4, 0),
(4, 1), (4, 2), (4, 4), (4, 5)]
assert trajs.index.tolist() == new_indexes
trajs = Trajectories(data.brownian_trajs_df())
trajs = trajs.cut_segments((2, 3), inplace=False)
new_indexes = [(0, 0), (0, 1), (0, 2), (0, 3), (0, 4), (1, 0), (1, 1),
(1, 2), (1, 3), (1, 4), (2, 0), (2, 1), (2, 2), (2, 3),
(2, 4), (3, 0), (3, 1), (3, 2), (3, 4), (3, 5), (4, 0),
(4, 1), (4, 2), (4, 4), (4, 5)]
assert trajs.index.tolist() == new_indexes
def test_interpolate():
trajs = Trajectories(data.with_gaps_df())
trajs.set_index('true_label', inplace=True, append=True)
trajs.reset_index(level='label', drop=True, inplace=True)
trajs.index.set_names(['t_stamp', 'label'], inplace=True)
interpolated = trajs.time_interpolate(sampling=3, time_step=0.1, s=1)
# t_stamps_in = interpolated.index.get_level_values('t_stamp')
# indexer = t_stamps_in % 2 == 0
# interpolated.loc[indexer].shape, trajs.shape
# indexer = interpolated.t_stamps % 3 == 0
# assert interpolated.loc[indexer].shape[0] == trajs.shape[0]
dts = interpolated.get_segments()[0].t.diff().dropna()
# All time points should be equaly spaced
assert_almost_equal(dts.min(), dts.max())
def test_get_bounds():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
trajs['t'] *= 10
t_stamp_bounds = {0: (0, 4), 1: (0, 4), 2: (0, 4), 3: (0, 4), 4: (0, 4)}
t_bounds = {
0: (0.0, 40.0),
1: (0.0, 40.0),
2: (0.0, 40.0),
3: (0.0, 40.0),
4: (0.0, 40.0)
}
assert trajs.get_bounds() == t_stamp_bounds
assert trajs.get_bounds(column='t') == t_bounds
def test_get_speeds():
trajs = Trajectories(data.brownian_trajs_df())
speeds = trajs.get_speeds().tolist()
real_speeds = [
np.nan, np.nan, np.nan, np.nan, np.nan, 857.99153458573994,
1596.9530747771976, 873.15267834726137, 1282.3088174598233,
408.98588960526808, 378.40023709328955, 1809.9895146014187,
917.93227668556324, 592.31881736181106, 0.48325048326444919,
0.39551116881922965, 798.29858694043128, 1085.3214310682606,
405.49164945495221, 550.37555144616226, 1406.707586739079,
1031.9444945962532, 1077.6619763794718, 1445.7789239945778,
739.66839622816326
]
assert_almost_equal(speeds, real_speeds)
def test_relabel_fromzero():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
original_labels = trajs.labels
idx = pd.IndexSlice
trajs.loc[idx[:, 1], :] = 55
relabeled = trajs.relabel_fromzero('label', inplace=False)
assert np.all(relabeled.labels == original_labels)
trajs.loc[idx[:, 1], :] = 55
relabeled = trajs.relabel_fromzero('label', inplace=False)
assert np.all(relabeled.labels == original_labels)
def linear_trajs():
data = np.array([
np.linspace(0, 30, 61),
np.zeros(61),
np.zeros(61),
np.linspace(0, 120, 61)
])
index = pd.MultiIndex.from_tuples([(t, i) for i in range(2)
for t in range(61)],
names=('t_stamp', 'label'))
trajs = pd.DataFrame(np.vstack([data.T, data.T]),
index=index,
columns=['x', 'y', 'z',
't']).sortlevel(['t_stamp', 'label'])
return Trajectories(trajs)
def test_scale():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
scaled = trajs.scale(factors=[2., 2., 2.],
coords=['x', 'y', 'z'],
inplace=False)
assert_array_almost_equal(scaled[['x', 'y', 'z']] / 2.,
trajs[['x', 'y', 'z']])
trajs = trajs.scale(factors=[2., 2., 2.],
coords=['x', 'y', 'z'],
inplace=True)
assert_array_almost_equal(scaled[['x', 'y', 'z']], trajs[['x', 'y', 'z']])
assert_raises(ValueError,
trajs.scale,
factors=[2., 2., 2.],
coords=['x', 'y'],
inplace=False)
def test_remove_spots():
"""
"""
trajs = Trajectories(data.brownian_trajs_df())
new_trajs = trajs.remove_spots([(3, 2), (0, 0)], inplace=False)
new_indexes = [(0, 1), (0, 2), (0, 3), (0, 4), (1, 0), (1, 1), (1, 2),
(1, 3), (1, 4), (2, 0), (2, 1), (2, 2), (2, 3), (2, 4),
(3, 0), (3, 1), (3, 3), (3, 4), (4, 0), (4, 1), (4, 2),
(4, 3), (4, 4)]
assert new_trajs.index.tolist() == new_indexes
new_trajs = trajs.remove_spots((0, 0), inplace=False)
new_indexes = [(0, 1), (0, 2), (0, 3), (0, 4), (1, 0), (1, 1), (1, 2),
(1, 3), (1, 4), (2, 0), (2, 1), (2, 2), (2, 3), (2, 4),
(3, 0), (3, 1), (3, 2), (3, 3), (3, 4), (4, 0), (4, 1),
(4, 2), (4, 3), (4, 4)]
assert new_trajs.index.tolist() == new_indexes
def test_polar_coords():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
polar = rotation.get_polar_coords(trajs, get_dtheta=True)
assert_array_almost_equal(trajs.x, polar.rho * np.cos(polar.theta))
def test_copy():
trajs = data.brownian_trajs_df()
trajs = Trajectories(trajs)
assert isinstance(trajs.copy(), Trajectories)