def write_data(generation, population, out_data, csv_name):
# get the fitness data set
data_set = []
for i in population:
data_set.append(i.fitness)
# calculate population fitness data
fittest = max(data_set)
unfittest = min(data_set)
average = int(sum(data_set) / len(population))
# debug print population fitness data
if out_data is True:
print("GENERATION " + str(generation))
# # find fittest memeber of population
# fittest = population[0]
# for p in population:
# if fittest.fitness < p.fitness:
# fittest = p
# # print fittest member's rules
# for r in range(R_count):
# print(str(fittest.rules[r]))
# print population stats
print("Fittest:\t" + str(fittest))
print("Average:\t" + str(average))
print("Unfittest:\t" + str(unfittest))
print("-------------------------")
# write population fitness data to csv
if csv_name is not None:
GA_csv.write(csv_name, [generation, fittest, average, unfittest])
def main():
# run algorithm & write results to csv
for s in range(SAMPLE_SIZE):
GA.main(SAMPLE_NAME + str(s))
# init averages samples csv file
GA_csv.init(SAMPLE_NAME)
# create averaged results csv from samples
average_samples()
# cleanup sample files
for s in range(SAMPLE_SIZE):
os.remove(SAMPLE_NAME + str(s) + '.csv')
def average_samples():
# get all csv data for each sample
rows = []
for s in range(SAMPLE_SIZE):
rows.append(GA_csv.get_rows(SAMPLE_NAME+str(s)))
# average out & write data to csv
for r in range(1, len(rows[0])-1):
# get averaged data in row r
fittest = 0
average = 0
unfittest = 0
for s in range(SAMPLE_SIZE):
fittest += int(rows[s][r][1])
average += int(rows[s][r][2])
unfittest += int(rows[s][r][3])
fittest = int(fittest / SAMPLE_SIZE)
average = int(average / SAMPLE_SIZE)
unfittest = int(unfittest / SAMPLE_SIZE)
GA_csv.write(SAMPLE_NAME, [r, fittest, average, unfittest])
def write_data(generation, population):
# get the fitness data set
data_set = []
for i in population:
data_set.append(i.fitness)
# calculate population fitness data
fittest = max(data_set)
unfittest = min(data_set)
average = int(sum(data_set) / len(population))
# debug print population fitness data
if OUT_DATA is True:
print("GENERATION " + str(generation))
print("Fittest:\t" + str(fittest))
print("Average:\t" + str(average))
print("Unfittest:\t" + str(unfittest))
print("-------------------------")
# write population fitness data to csv
if CSV_NAME is not None:
GA_csv.write(CSV_NAME, [generation, fittest, average, unfittest])
def main(argv=''):
global CSV_NAME
if argv != '':
CSV_NAME = argv
GA_csv.init(CSV_NAME)
run(generation_limit, P_size, G_size, C_rate, M_rate)
if new_population[p].fitness < new_population[n_worst].fitness:
n_worst = p
# replace least fit in new_population with fittest in new_population
for p in range(len(old_population)):
if p == n_worst and old_population[o_best].fitness > new_population[n_worst].fitness:
genome = old_population[o_best].genome
else:
genome = new_population[p].genome
survivors.append(Solution(genome))
return survivors
# ===========
# entry point
# ===========
if __name__ == '__main__':
if (len(sys.argv) >= 2):
CSV_NAME = sys.argv[1]
GA_csv.init(CSV_NAME)
run(generation_limit, P_size, G_size, C_rate, M_rate)
def main(argv=''):
global CSV_NAME
if argv != '':
CSV_NAME = argv
GA_csv.init(CSV_NAME)
run(generation_limit, P_size, G_size, C_rate, M_rate)