class NiaPy(Algorithm):
"""
Runs several algorithms from the NiaPy package
The default values comes from the Jie-ShengWang & Shu-Xia Li paper
"""
def __init__(self, algorithm_name, objective, maxfeval, population=30):
super().__init__(algorithm_name, objective, maxfeval)
self.algorithm_name = algorithm_name
self.algo = None #this is the suggest function from hyperopt
if algorithm_name not in __all__:
raise Exception('NiaPy does not have algorithm :' +
str(algorithm_name))
elif self.algorithm_name == 'NiaPyABC':
self.algo = ArtificialBeeColonyAlgorithm(NP=population, Limit=100)
elif self.algorithm_name == 'NiaPyBat':
self.algo = BatAlgorithm(NP=population)
elif self.algorithm_name == 'NiaPyCuckooSearch':
self.algo = CuckooSearch(N=population, pa=0.2, alpha=0.5)
elif self.algorithm_name == 'NiaPyDifferentialEvolution':
self.algo = DifferentialEvolution(NP=population, F=1, CR=0.8)
elif self.algorithm_name == 'NiaPyFireflyAlgorithm':
self.algo = FireflyAlgorithm(NP=population,
alpha=0.5,
betamin=0.2,
gamma=1.0)
elif self.algorithm_name == 'NiaPyGeneticAlgorithm':
self.algo = FireflyAlgorithm(NP=population,
Crossover=UniformCrossover,
Mutation=UniformMutation,
Cr=0.45,
Mr=0.9)
elif self.algorithm_name == 'NiaPyGWO':
self.algo = GreyWolfOptimizer(NP=population)
# config
elif self.algorithm_name == 'NiaPyNelderMead':
self.algo = NelderMeadMethod()
# config
elif self.algorithm_name == 'NiaPyPSO':
self.algo = ParticleSwarmAlgorithm(NP=population,
C1=2,
C2=2,
w=0.9,
vMin=-1.5,
vMax=1.5)
# config
# config
# config
elif self.algorithm_name == 'NiaPySimulatedAnnealing':
self.algo = SimulatedAnnealing(coolingMethod=coolLinear)
def assemblespace(self):
"""
We need to create a variable self.space to store the space to be used by the algorithm
Changes between algorithms
"""
try:
obj_space = self.objective.functionProperties['searchSpace']
self.x0 = self.objective.functionProperties['x0']
obj_space_type = self.objective.functionProperties['spaceType']
lower = []
higher = []
for v in obj_space:
lower.append(v[0])
higher.append(v[1])
self.lowerBound = min(lower)
self.higherBound = max(higher)
except:
logger.error('Error assembling the space for ' +
str(self.algorithm_name) + ' with cost function: ' +
str(self.objective.GetCostFunctionName()))
#since it will fail later in the optimize function we pass here
pass
def optimize(self):
"""
The main optimization procedure
Should return the best arm as and array/ list
:return:
THe instance to be optimized is the self.objective value
"""
try:
class MyBenchmark(Benchmark):
def __init__(self,
benchmark_objective,
Lower=self.lowerBound,
Upper=self.higherBound):
self.benchmark_objective = benchmark_objective
Benchmark.__init__(self, Lower, Upper)
def function(self):
bench = self.benchmark_objective
def evaluate(D, sol):
val = bench(sol)
return val
return evaluate
# run the random search algorithm
task = StoppingTask(D=len(self.x0),
nFES=self.maxfeval,
benchmark=MyBenchmark(
benchmark_objective=self.objective,
Lower=self.lowerBound,
Upper=self.higherBound))
best = self.algo.run(task)
#Some classes in the NiaPy return different formats for the best result
xopt = []
for xi in best[0]:
xopt.append(xi)
success = True
best_arm = convertToArray(xopt)
except:
logger.warning('Optimization for ' + str(self.algorithm_name) +
' failed.')
best_arm = NAN
success = False
return best_arm, success, self.objective
