def getNN(self, input, hlayers, hnodes, output):
NN = []
if hlayers == 0:
first_layer = Layer.getLayer(self, output, input)
else:
first_layer = Layer.getLayer(self, hnodes, input)
NN.append(first_layer)
for layer in range(hlayers - 1):
hlayer = Layer.getLayer(self, hnodes, len(NN[-1]))
NN.append(hlayer)
if hlayers != 0:
outputL = Layer.getLayer(self, output, len(NN[-1]))
NN.append(outputL)
return (NN)
def getNN(self, input, hlayers, hnodes, output):
"""Kieran Ringel
The NN is made up of layers that contain nodes that contain weights
for this reason the NN just has layers for the hidden layers and output layer
to know how many weights each node has, it must know the size of the previous layer"""
NN = []
if hlayers == 0: #if there are no hidden layers, the only layer is the output layer, with its previous layer being the input
first_layer = Layer.getLayer(self, output, input)
else: #otherwise the first layer is the first hidden layer
first_layer = Layer.getLayer(self, hnodes, input)
NN.append(first_layer)
for layer in range(hlayers - 1): #adds on remaining hidden layers
hlayer = Layer.getLayer(self, hnodes, len(NN[-1]))
NN.append(hlayer)
if hlayers != 0: #if the output layer wasn't added as the only layer, add the output layer
outputL = Layer.getLayer(self, output, len(NN[-1]))
NN.append(outputL)
return (NN)
