def test_rosen_de_parallel():
popsize = 8
dim = 2
testfun = Rosen(dim)
max_eval = 10000
limit = 0.01
for _ in range(5):
wrapper = Wrapper(testfun.fun, dim)
ret = de.minimize(wrapper.eval,
dim,
testfun.bounds,
max_evaluations=max_eval,
popsize=popsize,
workers=popsize)
if limit > ret.fun:
break
assert (limit > ret.fun) # optimization target not reached
assert (max_eval + popsize >= ret.nfev) # too much function calls
assert (max_eval // popsize + 2 > ret.nit) # too much iterations
assert (ret.nfev == wrapper.get_count()
) # wrong number of function calls returned
# assert(almost_equal(ret.x, wrapper.get_best_x())) # wrong best X returned
assert (almost_equal(ret.fun, wrapper.get_best_y(),
eps=1E-1)) # wrong best y returned
def optimize():
fprob = single_objective(pg.problem(udp))
print('interferometer optimization')
# Python Differential Evolution implementation, uses ask/tell for parallel function evaluation.
ret = de.minimize(fprob.fun,
bounds=fprob.bounds,
workers=16,
popsize=32,
max_evaluations=50000)
# Python CMAES implementation, uses ask/tell for parallel function evaluation.
#ret = cmaes.minimize(fprob.fun, bounds=fprob.bounds, workers=16, popsize=32, max_evaluations=50000)
# Parallel retry using DE
#ret = retry.minimize(fprob.fun, bounds=fprob.bounds, optimizer=De_cpp(20000, popsize=32), workers=16, num_retries=64)
# Parallel retry using Bite
# ret = retry.minimize(fprob.fun, bounds=fprob.bounds, optimizer=Bite_cpp(20000, M=1), workers=16, num_retries=64)
# Parallel retry using CMA-ES
#ret = retry.minimize(udp.fitness, bounds=bounds, optimizer=Cma_cpp(20000, popsize=32), workers=16, num_retries=64)
# Smart retry using DE
#ret = advretry.minimize(fprob.fun, bounds=fprob.bounds, optimizer=De_cpp(1500, popsize=32), workers=16)
# Smart retry using CMA-ES
#ret = advretry.minimize(fprob.fun, bounds=fprob.bounds, optimizer=Cma_cpp(1500, popsize=32), workers=16)
# Smart retry using DE->CMA sequence
#ret = advretry.minimize(fprob.fun, bounds=fprob.bounds, optimizer=de_cma(1500, popsize=32), workers=16)
print("best result is " + str(ret.fun) + ' x = ' +
", ".join(str(x) for x in ret.x))
def minimize(self,
fun,
bounds,
guess=None,
sdevs=None,
rg=Generator(MT19937()),
store=None):
ret = de.minimize(fun,
None,
bounds,
self.popsize,
self.max_eval_num(store),
workers=None,
stop_fitness=self.stop_fitness,
keep=self.keep,
f=self.f,
cr=self.cr,
rg=rg)
return ret.x, ret.fun, ret.nfev