def test_spectral_entropy(self):
spectral_entropy(RANDOM_TS, SF_TS, method='fft')
spectral_entropy(RANDOM_TS, SF_TS, method='welch')
spectral_entropy(RANDOM_TS, SF_TS, method='welch', nperseg=400)
self.assertEqual(np.round(spectral_entropy(RANDOM_TS, SF_TS,
normalize=True), 1), 0.9)
self.assertEqual(np.round(spectral_entropy(PURE_SINE, 100), 2), 0.0)
# 2D data
params = dict(sf=SF_TS, normalize=True, method='welch', nperseg=100)
assert_equal(aal(spectral_entropy, axis=1, arr=data, **params),
spectral_entropy(data, **params))
def compute_entropy_phase(score, ver_ratio=0.2, aw_size=4):
arr1 = produce_pitch_class_matrix_from_filename(score, aw_size=aw_size)
utm = np.round(np.angle(apply_dft_to_pitch_class_matrix(arr1)), 2)
vec = []
coeffs = []
height = int(utm.shape[0] * ver_ratio) - 1
sel = np.array(utm[height, :, :]).T
entr = ent.spectral_entropy(sel, 1, method='fft')
return entr[1:]
def compute_magnitude_entropy(score,
ver_ratio=0.2,
hor_ratio=(0, 1),
aw_size=4):
arr1 = produce_pitch_class_matrix_from_filename(score, aw_size=aw_size)
utm = np.abs(apply_dft_to_pitch_class_matrix(arr1))
vec = []
for i in range(7):
vec.append(utm[int(utm.shape[0] * ver_ratio) - 1, :,
i][utm[int(utm.shape[0] * ver_ratio) - 1, :, i] != 0])
sel = np.array([
ve[int(utm.shape[1] * hor_ratio[0]):int(utm.shape[1] * hor_ratio[1])]
for ve in vec
])
entr = ent.spectral_entropy(sel, 1, method='fft')
return entr[1:]
def create_database(
path_data="C:/Users/kb39309/Documents/Studia/Vovel_sounds/src/vowels_dataset/*",
path_csv="data.csv"):
""" Funkcja pozyskująca cechy z próbek dźwiękowych do ramki danych (Pandas).
:param path_data: Ścieżka do plików dźwiękowych (glob)
:param path_csv: Ścieżka do zapisu ramki danych z rozszerzeniem .csv
:return: Ramka danych (pandas.DataFrame)
"""
filenames = glob.glob(path_data)
df = pd.DataFrame(columns=[
"vowel", "sex", "f0", "form1", "form2", "form3", "form4", "spc1",
"spc2", "spc3", "entrspc"
])
for fname in filenames:
input, fs = get_signal(fname, fs=10000)
f0 = get_base_freq(input, fs)
formants = get_formants(input, fs)
spc = get_spectral_centroids(input, fs)
spcent = ant.spectral_entropy(input, sf=fs, method='welch')
df = df.append(
{
'vowel': os.path.basename(fname)[2],
"sex": 1 if 'f' in fname else 0,
"f0": f0,
"form1": formants[0],
"form2": formants[1],
"form3": formants[2],
"form4": formants[3],
"spc1": spc[0],
"spc2": spc[1],
"spc3": spc[2],
"entrspc": spcent
},
ignore_index=True)
df.to_csv(path_csv)
return df