def build_net(cin, a, b):
xor = XorNetwork.build_net(a, b)[0]
sum = XorNetwork.build_net(xor, cin)[0]
carry = model.OrNeuron(
model.AndNeuron(xor, cin),
model.AndNeuron(a, b)
)
return sum, carry
def build_net(a, b):
"""
[1] Built as guided by the description provided.
For the equation of the network, see:
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/ietron/xor4.gif
"""
return model.AndNeuron(
model.OrNeuron(a, b),
model.NotNeuron(model.AndNeuron(a, b))
),
def test_simple_net(self):
"""
a -\
AND \
b -/ \
AND ->
c -\ /
AND /
d -/
"""
a = model.Input("a")
b = model.Input("b")
g1 = model.AndNeuron(a, b)
c = model.Input("c")
d = model.Input("d")
g2 = model.AndNeuron(c, d)
final = model.AndNeuron(g1, g2)
def try_states(_a, _b, _c, _d, expected):
a.state = _a
b.state = _b
g1.update()
c.state = _c
d.state = _d
g2.update()
final.update()
self.assertEquals(final.state, expected,
"%s should result in %s" %
((_a, _b, _c, _d), expected))
try_states(0, 0, 0, 0, 0)
try_states(0, 0, 0, 1, 0)
try_states(0, 0, 1, 0, 0)
try_states(0, 0, 1, 1, 0)
try_states(0, 1, 0, 0, 0)
try_states(0, 1, 0, 1, 0)
try_states(0, 1, 1, 0, 0)
try_states(0, 1, 1, 1, 0)
try_states(1, 0, 0, 0, 0)
try_states(1, 0, 0, 1, 0)
try_states(1, 0, 1, 0, 0)
try_states(1, 0, 1, 1, 0)
try_states(1, 1, 0, 0, 0)
try_states(1, 1, 0, 1, 0)
try_states(1, 1, 1, 0, 0)
try_states(1, 1, 1, 1, 1)
def test_update(self):
a = model.Input("a")
b = model.Input("b")
c = model.Input("c")
d = model.Input("d")
e = model.Input("e")
f = model.Input("f")
g = model.Input("g")
h = model.Input("h")
net = model.Network(
model.AndNeuron(
model.AndNeuron(a, b),
model.AndNeuron(c, d)
),
model.AndNeuron(
model.AndNeuron(e, f),
model.AndNeuron(g, h)
)
)
states = net.update(a=1, b=0, c=1, d=0, e=1, f=0, g=1, h=0)
self.assertEquals(net["a"].state, 1)
self.assertEquals(net["b"].state, 0)
self.assertEquals(net["c"].state, 1)
self.assertEquals(net["d"].state, 0)
self.assertEquals(net["e"].state, 1)
self.assertEquals(net["f"].state, 0)
self.assertEquals(net["g"].state, 1)
self.assertEquals(net["h"].state, 0)
self.assertEquals(states, (0, 0))
states = net.update(a=0, b=1, c=0, d=1, e=0, f=1, g=0, h=1)
self.assertEquals(net["a"].state, 0)
self.assertEquals(net["b"].state, 1)
self.assertEquals(net["c"].state, 0)
self.assertEquals(net["d"].state, 1)
self.assertEquals(net["e"].state, 0)
self.assertEquals(net["f"].state, 1)
self.assertEquals(net["g"].state, 0)
self.assertEquals(net["h"].state, 1)
self.assertEquals(states, (0, 0))
states = net.update(a=1, b=1, c=1, d=1, e=0, f=0, g=0, h=0)
self.assertEquals(net["a"].state, 1)
self.assertEquals(net["b"].state, 1)
self.assertEquals(net["c"].state, 1)
self.assertEquals(net["d"].state, 1)
self.assertEquals(net["e"].state, 0)
self.assertEquals(net["f"].state, 0)
self.assertEquals(net["g"].state, 0)
self.assertEquals(net["h"].state, 0)
self.assertEquals(states, (1, 0))
states = net.update(a=0, b=0, c=0, d=0, e=1, f=1, g=1, h=1)
self.assertEquals(net["a"].state, 0)
self.assertEquals(net["b"].state, 0)
self.assertEquals(net["c"].state, 0)
self.assertEquals(net["d"].state, 0)
self.assertEquals(net["e"].state, 1)
self.assertEquals(net["f"].state, 1)
self.assertEquals(net["g"].state, 1)
self.assertEquals(net["h"].state, 1)
self.assertEquals(states, (0, 1))
def test_getitem(self):
a = model.Input("a")
b = model.Input("b")
c = model.Input("c")
d = model.Input("d")
e = model.Input("e")
f = model.Input("f")
g = model.Input("g")
h = model.Input("h")
and_aa = model.AndNeuron(a, b)
and_ab = model.AndNeuron(c, d)
and_a = model.AndNeuron(and_aa, and_ab)
and_ba = model.AndNeuron(e, f)
and_bb = model.AndNeuron(g, h)
and_b = model.AndNeuron(and_ba, and_bb)
net = model.Network(and_a, and_b)
self.assertEquals(net["a"], a)
self.assertEquals(net["b"], b)
self.assertEquals(net["c"], c)
self.assertEquals(net["d"], d)
self.assertEquals(net["e"], e)
self.assertEquals(net["f"], f)
self.assertEquals(net["g"], g)
self.assertEquals(net["h"], h)
self.assertEquals(net[0], and_a)
self.assertEquals(net[1], and_b)
def test_gather_inputs_collects_inputs(self):
a = model.Input("a")
b = model.Input("b")
c = model.Input("c")
d = model.Input("d")
e = model.Input("e")
f = model.Input("f")
g = model.Input("g")
h = model.Input("h")
net = model.Network(
model.AndNeuron(
model.AndNeuron(a, b),
model.AndNeuron(c, d)
),
model.AndNeuron(
model.AndNeuron(e, f),
model.AndNeuron(g, h)
)
)
self.assertEquals(len(net.inputs), 8)
self.assertEquals(net.inputs["a"], a)
self.assertEquals(net.inputs["b"], b)
self.assertEquals(net.inputs["c"], c)
self.assertEquals(net.inputs["d"], d)
self.assertEquals(net.inputs["e"], e)
self.assertEquals(net.inputs["f"], f)
self.assertEquals(net.inputs["g"], g)
self.assertEquals(net.inputs["h"], h)
def test_and_neuron(self):
a = model.Input("")
b = model.Input("")
gate = model.AndNeuron(a, b)
a.state = 0
b.state = 0
gate.update()
self.assertEquals(gate.state, 0)
a.state = 1
b.state = 0
gate.update()
self.assertEquals(gate.state, 0)
a.state = 0
b.state = 1
gate.update()
self.assertEquals(gate.state, 0)
a.state = 1
b.state = 1
gate.update()
self.assertEquals(gate.state, 1)
def test_iter_order(self):
a = model.Input("a")
b = model.Input("b")
c = model.Input("c")
d = model.Input("d")
e = model.Input("e")
f = model.Input("f")
g = model.Input("g")
h = model.Input("h")
and_aa = model.AndNeuron(a, b)
and_ab = model.AndNeuron(c, d)
and_a = model.AndNeuron(and_aa, and_ab)
and_ba = model.AndNeuron(e, f)
and_bb = model.AndNeuron(g, h)
and_b = model.AndNeuron(and_ba, and_bb)
net = model.Network(and_a, and_b)
expected = [a, b, and_aa, c, d, and_ab, and_a, e, f, and_ba, g, h,
and_bb, and_b]
self.assertListEqual(list(net), expected)
def build_net(a, b):
sum = XorNetwork.build_net(a, b)[0]
carry = model.AndNeuron(a, b)
return sum, carry
