def pdf_points():
topology = 'topos/eon.simpltopo.txt'
wavelengths = 16
iterations = 100000
departures = 1
curves = (
("FF", {}),
("FF - Adaptive", {}),
("FF/LF - Adaptive", {}),
("FF/LF/2 - Adaptive", {}),
("MinCost - Adaptive", {})
)
# Sulle x la distance
# Sulle y interference
x = []
with open("pdf-all.txt", "w+") as f:
for load in xrange(80, 151, 10):
for idx, method in enumerate(curves):
sim = Simulator(topology, wavelengths, iterations, load, departures, idx, None)
distances, interferences = sim.run(wantpdf=True)
for d, i in zip(distances, interferences):
f.write("%d %d %.20f\n" % (idx, d, i))
def interference():
topology = 'topos/eon.simpltopo.txt'
wavelengths = 16
iterations = 200000
departures = 1
curves = (
("FF", []),
("FF - Adaptive", []),
("FF/LF - Adaptive", []),
("FF/LF/2 - Adaptive", []),
("MinCost - Adaptive", [])
)
x = []
#for i in xrange(64, 151):
for i in xrange(80, 151, 10):
x.append(i)
for method, (label, points) in enumerate(curves):
sim = Simulator(topology, wavelengths, iterations, i, departures, method, 0)
stats = sim.run()
points.append((stats[0], stats[-1]))
f = open('graph.py', 'w+')
f.write("from pylab import *\nfrom matplotlib import rc\n")
f.write("rc('text', usetex=True)\nrc('font', family='serif')\nfigure()\n")
f.write("x = " + repr(x) + "\n")
colors = ('k', 'r', 'g', 'b', 'm')
labels = map(lambda x: x[0], curves)
for idx, (label, points) in enumerate(curves):
y = map(lambda x: x[0], points)
e = map(lambda x: x[1], points)
f.write(("y%d = " % idx) + repr(y) + "\n")
f.write(("e%d = " % idx) + repr(e) + "\n")
f.write("errorbar(x, y%d, yerr=e%d, ls='-', fmt='%s.')\n" % (idx, idx, colors[idx]))
f.write("grid(True)\n")
f.write("xlabel(r'\\textit{Traffic load (Erlang)}')\n")
f.write("ylabel(r'\\textit{Average interference level}')\n")
f.write("legend(" + repr(labels) + ", loc=4)\n")
f.write("savefig('graph.eps')\n")
f.write("show()\n")
f.close()
def capture(topology, wavelengths, iterations, arrivals, departures, reference):
# Use a fixed seed if you want to see the real difference
sim = Simulator(topology, wavelengths, iterations, arrivals, departures, reference, None)
# snapshot = sim.simple_run(3000)
# samples = random.sample(snapshot, min(3000, len(snapshot)))
samples = sim.simple_run(3000)
if len(samples) > 10:
statistics = stats(map(lambda x: float(x[0]) / float(x[1]), samples))
average, confidence = statistics[0], statistics[-1]
print statistics
print "Collected:", average, confidence
return average, confidence
else:
return 0, 0
