def test_output_of_transformer():
X = _make_nested_from_array(np.array([4, 6, 10, 12, 8, 6, 5, 5]),
n_instances=1,
n_columns=1)
s = SlopeTransformer(num_intervals=2).fit(X)
res = s.transform(X)
orig = convert_list_to_dataframe([[(5 + math.sqrt(41)) / 4,
(1 + math.sqrt(101)) / -10]])
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
X = _make_nested_from_array(np.array(
[-5, 2.5, 1, 3, 10, -1.5, 6, 12, -3, 0.2]),
n_instances=1,
n_columns=1)
s = s.fit(X)
res = s.transform(X)
orig = convert_list_to_dataframe([[
(104.8 + math.sqrt(14704.04)) / 61,
(143.752 + math.sqrt(20790.0775)) / -11.2,
]])
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
def test_output_dimensions():
# test with univariate
X = _make_nested_from_array(np.ones(12), n_instances=10, n_columns=1)
p = PAA(num_intervals=5).fit(X)
res = p.transform(X)
# get the dimension of the generated dataframe.
corr_time_series_length = res.iloc[0, 0].shape[0]
num_rows = res.shape[0]
num_cols = res.shape[1]
assert corr_time_series_length == 5
assert num_rows == 10
assert num_cols == 1
# test with multivariate
X = _make_nested_from_array(np.ones(12), n_instances=10, n_columns=5)
p = PAA(num_intervals=5).fit(X)
res = p.transform(X)
# get the dimension of the generated dataframe.
corr_time_series_length = res.iloc[0, 0].shape[0]
num_rows = res.shape[0]
num_cols = res.shape[1]
assert corr_time_series_length == 5
assert num_rows == 10
assert num_cols == 5
def test_output_of_transformer():
X = _make_nested_from_array(
np.array([4, 6, 10, 12, 8, 6, 5, 5]), n_instances=1, n_columns=1
)
d = DWTTransformer(num_levels=2).fit(X)
res = d.transform(X)
orig = convert_list_to_dataframe(
[[16, 12, -6, 2, -math.sqrt(2), -math.sqrt(2), math.sqrt(2), 0]]
)
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
X = _make_nested_from_array(
np.array([-5, 2.5, 1, 3, 10, -1.5, 6, 12, -3]), n_instances=1, n_columns=1
)
d = d.fit(X)
res = d.transform(X)
orig = convert_list_to_dataframe(
[
[
0.75000,
13.25000,
-3.25000,
-4.75000,
-5.303301,
-1.414214,
8.131728,
-4.242641,
]
]
)
def test_no_levels_does_no_change():
X = _make_nested_from_array(
np.array([1, 2, 3, 4, 5, 56]), n_instances=1, n_columns=1
)
d = DWTTransformer(num_levels=0).fit(X)
res = d.transform(X)
assert check_if_dataframes_are_equal(res, X)
def test_bad_input_args(bad_num_intervals):
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
if not isinstance(bad_num_intervals, int):
with pytest.raises(TypeError):
PAA(num_intervals=bad_num_intervals).fit(X).transform(X)
else:
with pytest.raises(ValueError):
PAA(num_intervals=bad_num_intervals).fit(X).transform(X)
def test_bad_input_args(bad_components):
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
if isinstance(bad_components, str):
with pytest.raises(TypeError):
PCATransformer(n_components=bad_components).fit(X)
else:
with pytest.raises(ValueError):
PCATransformer(n_components=bad_components).fit(X)
def test_output_of_transformer():
X = _make_nested_from_array(np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]),
n_instances=1,
n_columns=1)
p = PAA(num_intervals=3).fit(X)
res = p.transform(X)
orig = convert_list_to_dataframe([[2.2, 5.5, 8.8]])
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
def test_paa_performs_correcly_along_each_dim():
X = _make_nested_from_array(np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]),
n_instances=1,
n_columns=2)
p = PAA(num_intervals=3).fit(X)
res = p.transform(X)
orig = convert_list_to_dataframe([[2.2, 5.5, 8.8], [2.2, 5.5, 8.8]])
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
def test_shape_descriptor_function(bad_sdf):
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
y = np.zeros(10)
if not isinstance(bad_sdf, str):
with pytest.raises(TypeError):
ShapeDTW(shape_descriptor_function=bad_sdf).fit(X, y)
else:
with pytest.raises(ValueError):
ShapeDTW(shape_descriptor_function=bad_sdf).fit(X, y)
def test_subsequence_length(bad_subsequence_length):
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
y = np.zeros(10)
if not isinstance(bad_subsequence_length, int):
with pytest.raises(TypeError):
ShapeDTW(subsequence_length=bad_subsequence_length).fit(X, y)
else:
with pytest.raises(ValueError):
ShapeDTW(subsequence_length=bad_subsequence_length).fit(X, y)
def test_output_format_dim(n_instances, n_timepoints, n_intervals, features):
X = _make_nested_from_array(np.ones(n_timepoints),
n_instances=n_instances,
n_columns=1)
n_rows, n_cols = X.shape
trans = RandomIntervalFeatureExtractor(n_intervals=n_intervals,
features=features)
Xt = trans.fit_transform(X)
assert isinstance(Xt, pd.DataFrame)
assert Xt.shape[0] == n_rows
assert np.array_equal(Xt.values, np.ones(Xt.shape))
def test_slope_performs_correcly_along_each_dim():
X = _make_nested_from_array(np.array([4, 6, 10, 12, 8, 6, 5, 5]),
n_instances=1,
n_columns=2)
s = SlopeTransformer(num_intervals=2).fit(X)
res = s.transform(X)
orig = convert_list_to_dataframe([
[(5 + math.sqrt(41)) / 4, (1 + math.sqrt(101)) / -10],
[(5 + math.sqrt(41)) / 4, (1 + math.sqrt(101)) / -10],
])
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
def test_shape_descriptor_functions(bad_sdfs):
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
y = np.zeros(10)
if not len(bad_sdfs) == 2:
with pytest.raises(ValueError):
ShapeDTW(
shape_descriptor_function="compound",
shape_descriptor_functions=bad_sdfs,
).fit(X, y)
else:
ShapeDTW(
shape_descriptor_function="compound", shape_descriptor_functions=bad_sdfs
).fit(X, y)
def test_output_dimensions(num_intervals, corr_series_length):
X = _make_nested_from_array(np.ones(13), n_instances=10, n_columns=1)
s = SlopeTransformer(num_intervals=num_intervals).fit(X)
res = s.transform(X)
# get the dimension of the generated dataframe.
act_time_series_length = res.iloc[0, 0].shape[0]
num_rows = res.shape[0]
num_cols = res.shape[1]
assert act_time_series_length == corr_series_length
assert num_rows == 10
assert num_cols == 1
def test_row_transformer_sklearn_transfomer():
mu = 10
X = _make_nested_from_array(
np.random.normal(loc=mu, scale=5, size=(100,)), n_instances=10, n_columns=1
)
t = StandardScaler(with_mean=True, with_std=True)
r = RowTransformer(t)
Xt = r.fit_transform(X)
assert Xt.shape == X.shape
assert isinstance(
Xt.iloc[0, 0], (pd.Series, np.ndarray)
) # check series-to-series transform
np.testing.assert_almost_equal(Xt.iloc[0, 0].mean(), 0) # check standardisation
np.testing.assert_almost_equal(Xt.iloc[0, 0].std(), 1, decimal=2)
def test_results(n_instances, n_timepoints, n_intervals):
x = np.random.normal(size=n_timepoints)
X = _make_nested_from_array(x, n_instances=n_instances, n_columns=1)
t = RandomIntervalFeatureExtractor(
n_intervals=n_intervals, features=[np.mean, np.std, time_series_slope])
Xt = t.fit_transform(X)
# Check results
intervals = t.intervals_
for start, end in intervals:
expected_mean = np.mean(x[start:end])
expected_std = np.std(x[start:end])
expected_slope = time_series_slope(x[start:end])
actual_means = Xt.filter(like=f"*_{start}_{end}_mean").values
actual_stds = Xt.filter(like=f"_{start}_{end}_std").values
actual_slopes = Xt.filter(
like=f"_{start}_{end}_time_series_slope").values
assert np.all(actual_means == expected_mean)
assert np.all(actual_stds == expected_std)
assert np.all(actual_slopes == expected_slope)
def test_dwt_performs_correcly_along_each_dim():
X = _make_nested_from_array(
np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]), n_instances=1, n_columns=2
)
d = DWTTransformer(num_levels=3).fit(X)
res = d.transform(X)
orig = convert_list_to_dataframe(
[
[
9 * math.sqrt(2),
-4 * math.sqrt(2),
-2,
-2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
],
[
9 * math.sqrt(2),
-4 * math.sqrt(2),
-2,
-2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
-math.sqrt(2) / 2,
],
]
)
orig.columns = X.columns
assert check_if_dataframes_are_equal(res, orig)
def test_metric_params():
X = _make_nested_from_array(np.ones(10), n_instances=10, n_columns=1)
y = np.zeros(10)
# test the raw shape descriptor
shp = ShapeDTW()
assert shp._get_transformer("rAw") is None
# test the paa shape descriptor
shp = ShapeDTW(metric_params={"num_intERvals_paa": 3})
assert shp._get_transformer("pAA").num_intervals == 3
shp = ShapeDTW()
assert shp._get_transformer("pAA").num_intervals == 8
assert isinstance(shp._get_transformer("paa"), PAA)
# test the dwt shape descriptor
assert shp._get_transformer("dWt").num_levels == 3
shp = ShapeDTW(metric_params={"num_LEvEls_dwt": 5})
assert shp._get_transformer("Dwt").num_levels == 5
assert isinstance(shp._get_transformer("dwt"), DWTTransformer)
# test the slope shape descriptor
shp = ShapeDTW()
assert shp._get_transformer("sLoPe").num_intervals == 8
shp = ShapeDTW(metric_params={"num_inTErvals_slope": 2})
assert shp._get_transformer("slope").num_intervals == 2
assert isinstance(shp._get_transformer("slope"), SlopeTransformer)
# test the derivative shape descriptor
shp = ShapeDTW()
assert isinstance(shp._get_transformer("derivative"), DerivativeSlopeTransformer)
# test the hog1d shape descriptor
assert (
shp._get_transformer("hOG1d").num_intervals == 2
and shp._get_transformer("hOG1d").num_bins == 8
and shp._get_transformer("hog1d").scaling_factor == 0.1
)
# test hog1d with only 1 custom parameter
shp = ShapeDTW(metric_params={"NUM_intervals_hog1d": 5})
assert (
shp._get_transformer("hoG1d").num_intervals == 5
and shp._get_transformer("hOG1d").num_bins == 8
and shp._get_transformer("hog1d").scaling_factor == 0.1
)
shp = ShapeDTW(metric_params={"nUM_BinS_hog1d": 63})
assert (
shp._get_transformer("hoG1d").num_intervals == 2
and shp._get_transformer("hOG1d").num_bins == 63
and shp._get_transformer("hog1d").scaling_factor == 0.1
)
shp = ShapeDTW(metric_params={"scaling_factor_hog1d": 0.5})
assert (
shp._get_transformer("hoG1d").num_intervals == 2
and shp._get_transformer("hOG1d").num_bins == 8
and shp._get_transformer("hog1d").scaling_factor == 0.5
)
# test hog1d with 2 custom parameters
shp = ShapeDTW(metric_params={"NUM_intervals_hog1d": 5, "nUM_BinS_hog1d": 63})
assert (
shp._get_transformer("hoG1d").num_intervals == 5
and shp._get_transformer("hOG1d").num_bins == 63
and shp._get_transformer("hog1d").scaling_factor == 0.1
)
shp = ShapeDTW(metric_params={"NUM_bins_hog1d": 63, "scaling_factor_hog1d": 0.5})
assert (
shp._get_transformer("hoG1d").num_intervals == 2
and shp._get_transformer("hOG1d").num_bins == 63
and shp._get_transformer("hog1d").scaling_factor == 0.5
)
shp = ShapeDTW(
metric_params={"scaling_factor_hog1d": 0.5, "nUM_intervals_hog1d": 5}
)
assert (
shp._get_transformer("hoG1d").num_intervals == 5
and shp._get_transformer("hOG1d").num_bins == 8
and shp._get_transformer("hog1d").scaling_factor == 0.5
)
# test hog1d with all 3 custom parameters
shp = ShapeDTW(
metric_params={
"scaling_factor_hog1d": 0.5,
"nUM_intervals_hog1d": 5,
"num_bins_hog1d": 63,
}
)
assert (
shp._get_transformer("hoG1d").num_intervals == 5
and shp._get_transformer("hOG1d").num_bins == 63
and shp._get_transformer("hog1d").scaling_factor == 0.5
)
shp = ShapeDTW()
assert isinstance(shp._get_transformer("hog1d"), HOG1DTransformer)
# test compound shape descriptor (mix upper and lower cases)
shp = ShapeDTW(
shape_descriptor_function="compound",
shape_descriptor_functions=["raw", "derivative"],
metric_params={"weighting_FACtor": 20},
)
shp.fit(X, y)
assert shp.fit(X, y).weighting_factor == 20
with pytest.raises(ValueError):
ShapeDTW(
shape_descriptor_function="paa", metric_params={"num_intervals": 8}
).fit(X, y)
def test_early_trans_fail():
X = _make_nested_from_array(np.ones(10), n_instances=1, n_columns=1)
pca = PCATransformer(n_components=1)
with pytest.raises(NotFittedError):
pca.transform(X)