class Tsp2Opt:
def __init__(self, t, G):
self.G = G
self.t = t
self.w = G.m
self.nt = Trip(G)
def execute(self):
while True:
n = len(self.t.trip)-1
maxVariation = 0
best = None
for a in range(n-3):
b = a+1
if a != 0:
lim = n
else:
lim = n-1
for c in range(a+2, lim):
if c != n:
d = c+1
else:
d = 0
variation = self.w[self.t.trip[a]-1][self.t.trip[b]-1] + self.w[self.t.trip[c]-1][self.t.trip[d]-1] \
- self.w[self.t.trip[a]-1][self.t.trip[c]-1] - self.w[self.t.trip[b]-1][self.t.trip[d]-1]
if variation > maxVariation:
maxVariation = variation
best = a, b, c, d
if maxVariation != 0:
a, b, c, d = best
aux = self.t.trip[b]
self.t.trip[b] = self.t.trip[c]
self.t.trip[c] = aux
else:
self.calculateNewTrip()
break
def calculateNewTrip(self):
for v in self.t.trip:
self.nt.addVertex(self.G.V[v-1])
class TspMst:
def __init__(self, G):
self.G = G
self.r = self.G.V[0]
self.mstPrim = MstPrim(self.G, self.r)
self.trip = Trip(G)
def execute(self):
self.mstPrim.execute()
self.preOrder(self.r)
self.trip.addVertex(self.r)
def preOrder(self, u):
self.trip.addVertex(u)
for v in self.G.V:
if v.pi == u:
self.preOrder(v)
class TspNn:
def __init__(self, G):
self.G = G
self.r = self.G.V[0]
self.trip = Trip(G)
self.V = self.G.V
def execute(self):
self.trip.addVertex(self.r)
self.V.remove(self.r)
v = self.r
while self.V:
u = self.nearest(v)
self.trip.addVertex(u)
self.V.remove(u)
v = u
self.trip.addVertex(self.r)
def nearest(self, u):
Min = maxint
for v in self.G.V:
if self.G.m[u.n-1][v.n-1] < Min:
nearest = v
Min = self.G.m[u.n-1][v.n-1]
return nearest
class Tsp3Opt:
def __init__(self, t, G):
self.G = G
self.t = t
self.w = G.m
self.nt = Trip(G)
def execute(self):
while True:
n = len(self.t.trip)-1
maxVariation = 0
best = None
variation = [0, 0, 0, 0]
for a in range(n-5):
b = a + 1
for c in range(a+2, n-3):
d = c + 1
if a != 0:
lim = n
else:
lim = n-1
for e in range(c+2, lim):
if e != n:
f = e+1
else:
f = 0
oldWeight = self.w[self.t.trip[a]-1][self.t.trip[b]-1] \
+ self.w[self.t.trip[c]-1][self.t.trip[d]-1] \
+ self.w[self.t.trip[e]-1][self.t.trip[f]-1]
newWeight1 = self.w[self.t.trip[a]-1][self.t.trip[e]-1] \
+ self.w[self.t.trip[d]-1][self.t.trip[b]-1] \
+ self.w[self.t.trip[c]-1][self.t.trip[f]-1]
newWeight2 = self.w[self.t.trip[a]-1][self.t.trip[d]-1] \
+ self.w[self.t.trip[e]-1][self.t.trip[b]-1] \
+ self.w[self.t.trip[c]-1][self.t.trip[f]-1] \
newWeight3 = self.w[self.t.trip[a]-1][self.t.trip[d]-1] \
+ self.w[self.t.trip[e]-1][self.t.trip[c]-1] \
+ self.w[self.t.trip[b]-1][self.t.trip[f]-1] \
newWeight4 = self.w[self.t.trip[a]-1][self.t.trip[c]-1] \
+ self.w[self.t.trip[b]-1][self.t.trip[e]-1] \
+ self.w[self.t.trip[d]-1][self.t.trip[f]-1] \
variation[0] = oldWeight - newWeight1
variation[1] = oldWeight - newWeight2
variation[2] = oldWeight - newWeight3
variation[3] = oldWeight - newWeight4
for i in range(4):
var = variation[i]
if var > maxVariation:
maxVariation = var
best = (a, b, c, d, e, f)
nVariation = i
if maxVariation != 0:
a, b, c, d, e, f = best
if nVariation == 0:
self.swap1(a, b, c, d, e, f)
elif nVariation == 1:
self.swap2(a, b, c, d, e, f)
elif nVariation == 2:
self.swap3(a, b, c, d, e, f)
elif nVariation == 3:
self.swap4(a, b, c, d, e, f)
else:
self.calculateNewTrip()
break
def swap1(self, a, b, c, d, e, f):
aux = self.t.trip[b]
self.t.trip[b] = self.t.trip[e]
self.t.trip[e] = self.t.trip[c]
self.t.trip[c] = self.t.trip[d]
self.t.trip[d] = aux
def swap2(self, a, b, c, d, e, f):
aux = self.t.trip[b]
self.t.trip[b] = self.t.trip[d]
self.t.trip[d] = aux
aux = self.t.trip[e]
self.t.trip[e] = self.t.trip[c]
self.t.trip[c] = aux
def swap3(self, a, b, c, d, e, f):
aux = self.t.trip[b]
self.t.trip[b] = self.t.trip[d]
self.t.trip[d] = self.t.trip[c]
self.t.trip[c] = self.t.trip[e]
self.t.trip[e] = aux
def swap4(self, a, b, c, d, e, f):
aux = self.t.trip[b]
self.t.trip[b] = self.t.trip[c]
self.t.trip[c] = aux
aux = self.t.trip[e]
self.t.trip[e] = self.t.trip[d]
self.t.trip[d] = aux
def calculateNewTrip(self):
for v in self.t.trip:
self.nt.addVertex(self.G.V[v-1])
