def test1(self):
red = Red(cantidad=[1])
red.capas[0][0].bias = 0.5
red.capas[0][0].peso[0] = 0.4
red.capas[0][0].peso[1] = 0.3
red.capas[1][0].bias = 0.4
red.capas[1][0].peso[0] = 0.3
red.entrenar([[1, 1]], [[1]], [], [], epocas=1)
self.assertEqual(0.502101508999489, red.capas[0][0].bias)
self.assertEqual(0.40210150899948904, red.capas[0][0].peso[0])
self.assertEqual(0.302101508999489, red.capas[0][0].peso[1])
self.assertEqual(0.43937745312797394, red.capas[1][0].bias)
self.assertEqual(0.33026254863991883, red.capas[1][0].peso[0])
total = [0,0,0,0,0]
for lineas in contents:
dic = {3:0,4:0,5:1,6:2,7:3,8:4,9:4}
x = lineas.split(";")
if(primera):
primera = False
else:
entradas.append([normalizar(float(x[0]),14.2,3.8),normalizar(float(x[1]),1.1,0.08),normalizar(float(x[2]),1.66,0),normalizar(float(x[3]),65.8,0.6),
normalizar(float(x[4]),0.346,.009),normalizar(float(x[5]),289.0,2.0),normalizar(float(x[6]),440.0,9.0),
normalizar(float(x[7]),1.04,0.99),normalizar(float(x[8]),3.82,2.72),normalizar(float(x[9]),1.08,0.22),normalizar(float(x[10]),14.2,8)])
aux = [0,0,0,0,0,0,0]
aux[dic[int(x[11])]] = 1
total[dic[int(x[11])]] = total[dic[int(x[11])]] + 1
salidas.append(aux)
entradas_train = entradas[:4500]
entradas_test = entradas[4500:]
salidas_train = salidas[:4500]
salidas_test = salidas[4500:]
neuronas = 5
epocas = 5000
red = Red(cantidad=[neuronas], activacion=["sigmoid"], dim_input=11, dim_output=5)
red.entrenar(entradas_train,salidas_train,entradas_test,salidas_test,epocas = epocas,guardar = True)
