# training
for epoch in range(0,epoch):
for v,d in zip(train_in,train_out) :
o1 = l1.inout(v)
o2 = l2.inout(o1)
#back propagation
# output layer
print('output layer')
print('============')
for n in l2.get_neurons() :
e_out = n.update_weight(desired=d[0])
print('\n')
# output layer
print('Input Layer')
print('===========')
for n,e in zip(l1.get_neurons(),e_out) :
e_in = n.update_weight(above_errors=[e])
print('\n')
