from models import AdaptiveGA
from optproblems import cec2005
from os import makedirs
import statistics
import csv
if __name__ == '__main__':
dims = 10
bounds = [ [-100 for i in range(dims)], [100 for i in range(dims)] ]
functions = [ cec2005.F1(dims), cec2005.F2(dims), cec2005.F3(dims), cec2005.F4(dims), cec2005.F5(dims)]
optimums = [-450, -450, -450, -450, -310]
FESThresholds = [0, 0.001, 0.01, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]
numRuns = 25
# Creating result path
pathName = "Results/AGA"
makedirs(pathName, exist_ok=True)
# Initializing result files
tableFileName = pathName + "/AGA_" + str(dims) + "D.csv"
with open(tableFileName, "w") as resultsFile:
writer = csv.writer(resultsFile, delimiter = ',')
writer.writerow( ["Fi_10D", "Best", "Worst", "Median", "Mean", "Std_Dev", "Success_Rate"] )
"avgFits": avgFits}
except KeyboardInterrupt:
return
if __name__ == '__main__':
# Test of the ABC's performance over CEC2005's F1 (shifted sphere)
import time
from optproblems import cec2005
# np.seterr("raise") # any calculation error immediately stops the execution
dims = 10
bounds = [ [-100 for i in range(dims)], [100 for i in range(dims)] ] # 10-dimensional sphere (optimum: 0)
start = time.time()
# Initialization
ABC = ArtificialBeeColony(cec2005.F1(dims), bounds, popSize=50, workerOnlookerSplit=0.5, limit=None, numScouts=1, optimum=-450) # F5: -310 / others: -450
ABC.execute()
results = ABC.results
print("ABC: for criterion = " + ABC.crit + ", reached optimum of " + str(results["bestFits"][-1]) +
" (error of " + str(results["errors"][-1]) + ") (points " + str(results["bestPoints"][-1]) + ") with " + str(results["generations"][-1]) + " generations" +
" and " + str(results["FESCounts"][-1]) + " fitness evaluations" )
end = time.time()
print("time:" + str(end - start))
newPop.extend(self.pop)
self.pop = newPop[:self.popSize] # cuts the worst individuals here
if __name__ == '__main__':
# Test of the GA's performance over CEC2005's F1 (shifted sphere)
import time
from optproblems import cec2005
bounds = [[-100 for i in range(10)],
[100 for i in range(10)]] # 10-dimensional sphere (optimum: 0)
func = cec2005.F1(10)
start = time.time()
# Initialization
GA = GeneticAlgorithm(func,
bounds,
crit="min",
optimum=-450,
tol=1e-08,
eliteSize=1,
matingPoolSize=100,
popSize=100) #F5 = -310
GA.setParentSelection(GA.tournamentSelection, (True, ))
GA.setCrossover(GA.blxAlphaCrossover, (0.5, 1)) # alpha, prob
# GA.setMutation(GA.creepMutation, (1, 0, 1)) # prob, mean, sigma
from models import ArtificialBeeColony
from optproblems import cec2005
from os import makedirs
import statistics
import csv
if __name__ == '__main__':
dims = 10
bounds = [[-100 for i in range(dims)], [100 for i in range(dims)]]
functions = [
cec2005.F1(dims),
cec2005.F2(dims),
cec2005.F3(dims),
cec2005.F4(dims),
cec2005.F5(dims)
]
optimums = [-450, -450, -450, -450, -310]
FESThresholds = [
0, 0.001, 0.01, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1
]
numRuns = 25
# Creating result path
pathName = "Results/ABC"
makedirs(pathName, exist_ok=True)
# Initializing result files
tableFileName = pathName + "/ABC_" + str(dims) + "D.csv"
if __name__ == '__main__':
# Test of the SSA's performance over CEC2005's F1 (shifted sphere)
import time
from optproblems import cec2005
# np.seterr("raise") # any calculation error immediately stops the execution
dims = 10
bounds = [[-100 for i in range(dims)],
[100 for i in range(dims)]] # 10-dimensional sphere (optimum: 0)
start = time.time()
# Initialization
SSA = SocialSpiderAlgorithm(cec2005.F1(dims),
bounds,
popSize=10,
vibrationConstant=-700,
attenuationRate=1,
maskChangeProb=0.7,
maskOneProb=0.1,
optimum=-450) # F5: -310 / others: -450
#compare normalizing and non-normalizing
#compare populations of 20, 30 and 50
SSA.execute()
results = SSA.results
print("SSA: for criterion = " + SSA.crit + ", reached optimum of " +
str(results["bestFits"][-1]) + " (error of " +
str(results["errors"][-1]) + ") (points " +
elitism_bool = True
elitism_number = 1
hiper_parameters_dict = {
"cross_method": "aritmetic",
"cross_ratio": crossover_ratio,
"cross_alpha": alpha,
"mutation_method": "gauss",
"mutation_ratio": mutation_ratio,
"elitism_bool": True,
"selection_method": "tournment"
}
savePathParameters = "ag_function1.csv"
global_min = -450.0
function1 = cec2005.F1(n_genes)
ag = Genetic_Algorithm(Func_obj1,
n_genes,
bounds,
pop_size,
n_epochs,
hiper_parameters_dict,
optimum=global_min,
max_obj=False)
n_runs = 25
error_list = []
success_count = 0
for n in range(1):