def runSearcher(self, emax, emin):
modelbasics = modelBasics(self.modelName)
modelFunction = self.modelName()
anz = anzeigen()
hi, lo, kooling, indepSize, thresh, iterations = modelFunction.eigenschaften()
# emax, emin = modelbasics.baselining(self.modelName)
sb = s = [randi(lo, hi) for z in xrange(indepSize)]
eb = e = modelbasics.energy(s, emax, emin)
enRec = dynamikliste() # Creates a growing list.
enRec[0] = 0
# Since iterations start from 1, lets initialize enRec[0] to 0
analyser = analyzer.analyser()
epochs = 5 if self.early else iterations
k = 1
while epochs and k < iterations:
sn = modelbasics.neighbour(s, hi, lo)
en = modelbasics.energy(sn, emax, emin)
t = k / iterations
if en < eb:
eb, sb, enRec[k] = en, sn, en
if self.disp:
modelbasics.say("!")
if en < e:
s, e, enRec[k] = sn, en, en
if self.disp:
modelbasics.say("+")
if modelbasics.do_a_randJump(en, e, t, kooling):
# The cooling factor needs to be really low for some reason!!
s, e, enRec[k] = sn, en, en
if self.disp:
modelbasics.say("?")
else:
enRec[k] = en
if self.disp:
modelbasics.say(".")
if k % 50 == 0 and k > 50:
# print enRec[:-10]
proceed = analyser.isItGettinBetter(enRec[k - 100 :])
if proceed:
epochs += 1
else:
epochs -= 1
# print enRec[k-40:] #
k = k + 1
if k % 40 == 0:
if self.disp:
modelbasics.say("\n") # sa.say(format(sb,'0.2f'))
if self.disp:
modelbasics.say("\n"),
# Print Energy and best value.
for i in xrange(k):
if self.disp:
if i % 50 == 0:
print anz.xtile(enRec[i - 50 :])
if self.disp:
modelbasics.say("\n")
return eb
def runSearcher(self):
modelbasics = modelBasics(self.modelName);
modelFunction = self.modelName()
anz = anzeigen();
hi, lo, kooling, indepSize, iterations = \
modelFunction.eigenschaften()
emax, emin = self.emax, self.emin
sb = s = [randi(lo, hi) for z in xrange(indepSize)];
eb = e = modelbasics.energy(s, self.emax, self.emin)
enRec = dynamikliste() # Creates a growing list.
enRec[0] = 0;
# Since iterations start from 1, lets initialize enRec[0] to 0
analyser = analyzer.analyser()
epochs = 3 if self.early else iterations;
k = 1;
while epochs and k < iterations:
sn = modelbasics.neighbour(s, hi, lo)
en = modelbasics.energy(sn, emax, emin)
t = k / iterations
if en < eb:
eb, sb, enRec[k] = en, sn, en;
#if self.disp:
#modelbasics.say('!')
if en < e:
s, e, enRec[k] = sn, en, en;
#if self.disp:
#modelbasics.say('+')
if modelbasics.do_a_randJump(en, e, t, kooling):
# The cooling factor needs to be really low for some reason!!
s, e, enRec[k] = sn, en, en;
#if self.disp:
#modelbasics.say('?')
else:
enRec[k] = en
#if self.disp:
# modelbasics.say('.')
if k % 50 == 0 and k > 50:
# print enRec[:-10]
proceed = analyser.isItGettinBetter(enRec[k - 100:])
if proceed:
epochs += 1;
else:
epochs -= 1;
# print enRec[k-40:] #
k = k + 1
era=1;
# Print Energy and best value.
for i in xrange(k):
if self.disp:
if i % 50 == 0:
print era, anz.xtile(enRec[i - 50:], show='%0.2E')
era+=1
if self.disp:
modelbasics.say('\n')
return [eb, modelbasics.energyIndv(sb, emax, emin)]
sys.dont_write_bytecode = True
rdivDemo=sk.rdivDemo
for x in [Schaffer, Kursawe,
Fonseca, ZDT1, ZDT3, Viennet3]:
early=True
E=[]
for y in [GA,SimulatedAnnealer, MaxWalkSat]:
eb=30*[0]
print 'Model: ', x.__name__
print 'Searcher: ', y.__name__
print strftime("%a, %d %b %Y %H:%M:%S ", gmtime()), '\n'
k=x()
reps=30
dspl=anzeigen();
hi, lo, kooling, indepSize, iterations = k.eigenschaften()
print 'Settings:'
print 'min=', lo, ', max=', hi, ', Cooling Factor=', kooling, '\n'
if early: print 'Early Termination!' , '\n'
for r in xrange(reps):
a=y(x,disp=False,early=early)
eb[r] = a.runSearcher()
eb.insert(0,y.__doc__)
E.append(eb)
#print dspl.xtile(eb[1:])
"""for r in xrange(reps):
print dspl.xtile(eb[r:r+50], lo=lo, hi=hi)"""
print 'Energy: ', "{:.3E}".format(Decimal(str(np.sum(eb[1:])/reps)))
def _rDiv():