def generate_wavelet(list_ecg, parameters, name):
freq = parameters[0]
epi_qrs = [parameters[1] for i in range(5)]
wavelet = fb.WaveletExtraction(epi_qrs=epi_qrs)
for sig in list_ecg:
wavelet.evaluete_signal(sig.input_signal, freq, label=sig.label)
utils.write_file('models/wavelet_model_' + name, wavelet.serialize())
def generate_wavelet(list_ecg, freq, name):
wavelet = fb.WaveletExtraction()
for sig in list_ecg:
wavelet.evaluete_signal(sig.input_signal, freq,
is_valey=sig.is_valey)
utils.write_file('models/wavelet_model_' + name, wavelet.serialize())
def generate_wavelet_db1(list_ecg, freq, name):
wavelet_db1 = fb.WaveletTransform()
for sig in list_ecg:
wavelet_db1.evaluete_signal(sig.input_signal, label=sig.label)
utils.write_file('models/wavelet_db1_model_' + name,
wavelet_db1.serialize())
def generate_morfology(list_ecg, freq, name):
morfology = fb.MorfologyDetector()
for sig in list_ecg:
morfology.evaluete_signal(sig.input_signal,
int(len(sig.input_signal)/2)+1,
is_valey=sig.is_valey)
utils.write_file('models/morfology_model_' + name, morfology.serialize())
def generate_mathematical(list_ecg, offset, name):
# Matemathical Models Evaluetion
list_of_templates = utils.get_default_templates(offset)
list_of_packets = utils.mp_signal_apply(utils.consume_packet_helper,
list_ecg,
list_of_templates,
workers=3)
m_model = fb.ECG.build_from_packets(list_of_packets)
utils.write_file('models/m_model_' + name, m_model.serialize())
def generate_morfology(list_ecg, parameters, name):
morfology = fb.MorfologyDetector(a=parameters[:2],
b=parameters[2:4],
c=parameters[4:6],
d=parameters[6:])
for sig in list_ecg:
morfology.evaluete_signal(sig.input_signal,
int(len(sig.input_signal) / 2) + 1,
is_valey=sig.is_valey,
label=sig.label)
utils.write_file('models/morfology_model_' + name, morfology.serialize())
def my_generate_random(list_ecg, offset, name):
random.results = []
for sig in list_ecg:
random.evaluete_signal(sig.input_signal)
utils.write_file('models/random_model_' + name, random.serialize())
my_ecg = fb.ECG.build_from_packets(list_of_packets)
len_packet = 35
list_ecg = my_ecg.build_packets(len_packet)
list_ecg = my_ecg.packets
return list_ecg
if __name__ == '__main__':
list_ecg = main()
rd = fb.RandomExtraction(offset=35)
for sig in list_ecg:
rd.evaluete_signal(sig.input_signal)
utils.write_file('teste', rd.serialize())
data = utils.read_file('teste')
rd = fb.deserialize(data)
table = rd.get_table_erros()
t = []
for p in list_ecg:
t.append(p.label)
t = ['qRs_4' if x == 2 else x for x in t]
t = ['QS_5' if x == 14 else x for x in t]
t = ['rRs_8' if x == 10 else x for x in t]
t = ['RS_9' if x == 5 else x for x in t]
t = ['artifact' if x in [12, 11, 1, 18, 6, 15, -1] else x for x in t]
X = table.values
heart_rate = np.mean(heart_rate)
labels = [name for x in rpeaks]
my_ecg = fb.ECG(input_signal=ecg_signal,
indicators=rpeaks,
labels=labels,
is_filtered=False)
if name == 'qRs_4':
offset = int(intervals / 7)
elif name == 'QS_5':
offset = int(intervals / 7)
elif name == 'rRs_8':
offset = int(intervals / 7)
elif name == 'RS_9':
offset = int(intervals / 7)
else:
offset = int(intervals / 7)
offset = 40
my_ecg.packets = my_ecg.build_packets(offset)
utils.write_file('data/signal_' + name + '_' + str(i),
my_ecg.serialize())
list_of_signals = []
for name in ListName:
for i in range(10):
data = utils.read_file('data/signal_' + name + '_' + str(i))
list_of_signals.append(fb.deserialize(data))
for i in range(len(list_of_signals)):
plt.plot(list_of_signals[i].packets[0].input_signal)
plt.show()
plt.close()
# list_ecg, list_of_templates)
my_ecg.packets = lst_of_packets
return my_ecg
if __name__ == '__main__':
__spec__ = "ModuleSpec(name='builtins', loader=<class '_frozen_importlib.BuiltinImporter'>)"
results = main()
error = results.get_errors_table()
param1 = results.get_param_table()
[signals, models, names, errors,
types] = results.get_signals_and_models(normalized=False, only_best=True)
data = results.serialize()
utils.write_file('outfile', data)
data = utils.read_file('outfile')
c = fb.deserialize(data)
# t = []
# for p in results.packets:
# t.append(p.label)
#
# error['types'] = [int(x) for x in types]
# X = error.values
#
#
# X_train, X_test, y_train, y_test = train_test_split(
# X, t, test_size=0.4, random_state=0)
#
def generate_hermite(list_ecg, freq, name):
hermite = fb.HermiteExtraction()
for sig in list_ecg:
hermite.evaluete_signal(sig.input_signal, label=sig.label)
utils.write_file('models/hermite_model_' + name, hermite.serialize())
def generate_hos(list_ecg, freq, name):
hos = fb.HOSExtraction()
for sig in list_ecg:
hos.evaluete_signal(sig.input_signal, label=sig.label)
utils.write_file('models/hos_model_' + name, hos.serialize())
def generate_ulbp(list_ecg, freq, name):
ulbp = fb.ULBP1DExtraction()
for sig in list_ecg:
ulbp.evaluete_signal(sig.input_signal, label=sig.label)
utils.write_file('models/ulbp_model_' + name, ulbp.serialize())