def sigmoidXOR():
expected_output_xor = (([1, 1], [False]),
([1, 0], [True]),
([0, 1], [True]),
([0, 0], [False]))
print(f"XOR Gate expectation = {expected_output_xor}")
table = makeTruthTable(2)
OR1 = im.Neuron(inputWeight=[24, 24], bias=-12, idPerceptron='OR 1')
NAND1 = im.Neuron(inputWeight=[-12, -12], bias=18, idPerceptron='NAND 1')
AND1 = im.Neuron(inputWeight=[12, 12], bias=-18, idPerceptron='AND 1')
layerOneXOR = ptl.NeuronLayer([NAND1, OR1], idLayer='FirstLayer')
layerTwoXOR = ptl.NeuronLayer([AND1], idLayer='SecondLayer')
networkOneXOR = ptn.NeuronNetwork([layerOneXOR, layerTwoXOR])
for index, i in enumerate(table):
output = networkOneXOR.feed_forward(list(i))
print(f'input = {list(i)} and output {output} expected output = {expected_output_xor[index][1]} difference == {expected_output_xor[index][1][0] - output[0]}')
def sigmoidINV():
expected_output_inv = (([1], [False]),
([0], [True]))
print(f"Inv Gate expectation = {expected_output_inv}")
INV1 = im.Neuron(inputWeight=[-12], bias=6, idPerceptron='INV 1')
table = makeTruthTable(1)
for index, i in enumerate(table):
output = INV1.activate(list(i))
print(f'input = {list(i)} and output {output} expected output = {expected_output_inv[index][1]} difference == {expected_output_inv[index][1][0] - output}')
print(f"The error = {INV1.error(expected_output_inv)} \n")
def sigmoidNAND():
expected_output_nand = (([1, 1], [False]),
([1, 0], [True]),
([0, 1], [True]),
([0, 0], [True]))
print(f"NAND Gate expectation = {expected_output_nand}")
table = makeTruthTable(2)
NAND1 = im.Neuron(inputWeight=[-12, -12], bias=18, idPerceptron='NAND 1')
for index, i in enumerate(table):
output = NAND1.activate(list(i))
print(f'input = {list(i)} and output {output} expected output = {expected_output_nand[index][1]} difference == {expected_output_nand[index][1][0] - output}')
print(f"The error = {NAND1.error(expected_output_nand)} \n")
def sigmoidOR():
expected_output_or = (([1, 1], [True]),
([1, 0], [True]),
([0, 1], [True]),
([0, 0], [False]))
print(f"OR Gate expectation = {expected_output_or}")
table = makeTruthTable(2)
OR1 = im.Neuron(inputWeight=[24, 24], bias=-12, idPerceptron='OR 1')
for index, i in enumerate(table):
output = OR1.activate(list(i))
print(f'input = {list(i)} and output {output} expected output = {expected_output_or[index][1]} difference == {expected_output_or[index][1][0] - output}')
print(f"The error = {OR1.error(expected_output_or)} \n")
def sigmoidHalfAdder():
expected_output_ha = (([1, 1], [False, True]),
([1, 0], [True, False]),
([0, 1], [True, False]),
([0, 0], [False, False]))
print(f"Half adder expectation = {expected_output_ha}")
table = makeTruthTable(2)
OR1 = im.Neuron(inputWeight=[24, 24], bias=-12, idPerceptron='OR 1')
NAND1 = im.Neuron(inputWeight=[-12, -12], bias=18, idPerceptron='NAND 1')
AND1 = im.Neuron(inputWeight=[12, 12], bias=-18, idPerceptron='AND 1')
AND2 = im.Neuron(inputWeight=[12, 12, 0], bias=-18, idPerceptron='AND 2')
REP1 = im.Neuron(inputWeight=[0, 0, 24], bias=-12, idPerceptron='REP 1')
layer1 = ptl.NeuronLayer([OR1, NAND1, AND1], idLayer='FirstLayer')
layer2 = ptl.NeuronLayer([AND2, REP1], idLayer='SecondLayer')
network = ptn.NeuronNetwork([layer1, layer2])
for index, i in enumerate(table):
output = network.feed_forward(list(i))
print(f'input = {list(i)} and output {output} expected output = {expected_output_ha[index][1]} difference == {expected_output_ha[index][1][0] - output[0]}')