def test_meanvar():
# Pick acoustic features for testing
_, X = example_file_data_sources_for_acoustic_model()
X = FileSourceDataset(X)
lengths = [len(x) for x in X]
D = X[0].shape[-1]
X_mean, X_var = P.meanvar(X)
X_std = np.sqrt(X_var)
assert np.isfinite(X_mean).all()
assert np.isfinite(X_var).all()
assert X_mean.shape[-1] == D
assert X_var.shape[-1] == D
_, X_std_hat = P.meanstd(X)
assert np.allclose(X_std, X_std_hat)
x = X[0]
x_scaled = P.scale(x, X_mean, X_std)
assert np.isfinite(x_scaled).all()
# For padded dataset
_, X = example_file_data_sources_for_acoustic_model()
X = PaddedFileSourceDataset(X, 1000)
# Should get same results with padded features
X_mean_hat, X_var_hat = P.meanvar(X, lengths)
assert np.allclose(X_mean, X_mean_hat)
assert np.allclose(X_var, X_var_hat)
# Inverse transform
x = X[0]
x_hat = P.inv_scale(P.scale(x, X_mean, X_std), X_mean, X_std)
assert np.allclose(x, x_hat, atol=1e-5)
def test_minmax():
# Pick linguistic features for testing
X, _ = example_file_data_sources_for_acoustic_model()
X = FileSourceDataset(X)
lengths = [len(x) for x in X]
D = X[0].shape[-1]
X_min, X_max = P.minmax(X)
assert np.isfinite(X_min).all()
assert np.isfinite(X_max).all()
x = X[0]
x_scaled = P.minmax_scale(x, X_min, X_max, feature_range=(0, 0.99))
assert np.max(x_scaled) <= 1
assert np.min(x_scaled) >= 0
assert np.isfinite(x_scaled).all()
# Need to specify (min, max) or (scale_, min_)
@raises(ValueError)
def __test_raise1(x, X_min, X_max):
P.minmax_scale(x)
@raises(ValueError)
def __test_raise2(x, X_min, X_max):
P.inv_minmax_scale(x)
__test_raise1(x, X_min, X_max)
__test_raise2(x, X_min, X_max)
# Explicit scale_ and min_
min_, scale_ = P.minmax_scale_params(X_min, X_max, feature_range=(0, 0.99))
x_scaled_hat = P.minmax_scale(x, min_=min_, scale_=scale_)
assert np.allclose(x_scaled, x_scaled_hat)
# For padded dataset
X, _ = example_file_data_sources_for_acoustic_model()
X = PaddedFileSourceDataset(X, 1000)
# Should get same results with padded features
X_min_hat, X_max_hat = P.minmax(X, lengths)
assert np.allclose(X_min, X_min_hat)
assert np.allclose(X_max, X_max_hat)
# Inverse transform
x = X[0]
x_hat = P.inv_minmax_scale(P.minmax_scale(x, X_min, X_max), X_min, X_max)
assert np.allclose(x, x_hat)
x_hat = P.inv_minmax_scale(P.minmax_scale(x, scale_=scale_, min_=min_),
scale_=scale_,
min_=min_)
assert np.allclose(x, x_hat)
def test_real_metrics():
_, source = example_file_data_sources_for_acoustic_model()
X = FileSourceDataset(source)
lengths = [len(x) for x in X]
X = X.asarray()
mgc = X[:, :, :source.mgc_dim // 3]
lf0 = X[:, :, source.lf0_start_idx]
vuv = (X[:, :, source.vuv_start_idx] > 0).astype(np.int)
bap = X[:, :, source.bap_start_idx]
mgc_tgt = mgc + 0.01
lf0_tgt = lf0 + 0.01
vuv_tgt = vuv.copy()
bap_tgt = bap + 0.01
mcd = metrics.melcd(mgc, mgc_tgt, lengths)
bap_mcd = metrics.melcd(bap, bap_tgt, lengths)
lf0_mse = metrics.lf0_mean_squared_error(lf0, vuv, lf0_tgt, vuv_tgt,
lengths)
vuv_err = metrics.vuv_error(vuv, vuv_tgt)
assert mcd > 0
assert bap_mcd > 0
assert lf0_mse > 0
assert vuv_err == 0.0
def _get_small_datasets(padded=False, duration=False, padded_length=1000):
if duration:
X, Y = example_file_data_sources_for_duration_model()
else:
X, Y = example_file_data_sources_for_acoustic_model()
if padded:
X = PaddedFileSourceDataset(X, padded_length=padded_length)
Y = PaddedFileSourceDataset(Y, padded_length=padded_length)
else:
X = FileSourceDataset(X)
Y = FileSourceDataset(Y)
return X, Y
