def n_runs(param_dict):
# vars
var_count = param_dict['variables count']
param_dict['history'] = False
run_count = param_dict['runs count']
calc_count = param_dict['calculations count']
f_name = param_dict['function name']
# output
# Initialize a workbook
book = xlwt.Workbook(encoding="utf-8")
# Add a sheet to the workbook
sheet = book.add_sheet("N Runs")
sheet.write(0, 0, 'Values')
sheet.write(0, 1, 'Variables')
# Run!
pop = None
if f_name == 'cassini2':
print('matlab init')
cassini2_problem.process_init()
print('matlab init end!')
for i in range(run_count):
pop = ga.Population(param_dict)
pop.evolve()
best = pop.archive[0]
print('Run №' + str(i + 1) + ' complete!')
sheet.write(i + 1, 0, best.value.__str__())
for j in range(var_count):
sheet.write(i + 1, j + 1, best.variables[j].__str__())
if f_name == 'cassini2':
cassini2_problem.process_term()
print('---_Done_---')
# info
now = datetime.datetime.now()
ns = now.strftime("%Y-%m-%d %H:%M:%S")
s_info = book.add_sheet("Info")
s_info = book.add_sheet("Info")
s_info.write(0, 0, "Date-time")
s_info.write(0, 1, ns)
i = 1
for x in param_dict.keys():
s_info.write(i, 0, str(x))
i += 1
i = 1
for x in param_dict.values():
s_info.write(i, 1, str(x))
i += 1
# alg_inf = "Shard Algorithm with selection and crossover, version 1.2, developed by Vasiliev Dmitriy, September 2019"
# s_info.write(0, 3, alg_inf)
book.save("output_n.xls")
def run_mult():
print('script begin')
param_dict = param_dict_init()
runs = param_dict['runs count']
var_count = param_dict['variables count']
calc_count = param_dict['calculations count']
now = datetime.datetime.now()
ns = now.strftime("%Y_%m_%d-%H_%M_%S")
book = xlwt.Workbook(encoding="utf-8")
# Add a sheet to the workbook
sheet = book.add_sheet("N Runs")
sheet.write(0, 0, 'Values')
sheet.write(0, 1, 'Variables')
for x in range(runs):
print('run № ' + str(x + 1))
print('matlab init begin')
cassini2_problem.process_init()
print('matlab init end')
print('GA begin')
pop = ga.Population(param_dict)
pop.evolve()
print('GA end')
cassini2_problem.process_term()
print('matlab terminate')
best = pop.archive[0]
print('Archive best:')
print(best)
print('calculations count: ' + str(param_dict['calculations count']))
print('Run №' + str(x + 1) + ' complete!')
sheet.write(x + 1, 0, best.value.__str__())
for j in range(var_count):
sheet.write(x + 1, j + 1, best.variables[j].__str__())
s_info = book.add_sheet("Info")
s_info.write(0, 0, "Date-time")
s_info.write(0, 1, ns)
i = 1
for x in param_dict.keys():
s_info.write(i, 0, str(x))
i += 1
i = 1
for x in param_dict.values():
s_info.write(i, 1, str(x))
i += 1
print('script end')
book.save(ns + "_mruns.xls")
def run_single():
print('script begin')
param_dict = param_dict_init()
print('matlab init begin')
cassini2_problem.process_init()
print('matlab init end')
print('GA begin')
pop = ga.Population(param_dict)
pop.evolve()
print('GA end')
cassini2_problem.process_term()
print('matlab terminate')
best = pop.archive[0]
print('Archive best:')
print(best)
print('calculations count: ' + str(param_dict['calculations count']))
print('script end')
def test_runs(param_dict):
# TEST TEST TEST
print('Test is on')
'''
param_dict['function name'] = f_name
param_dict['is maximization'] = False
param_dict['variables count'] = var_count
param_dict['function borders'] = bord_all
param_dict['calculations count'] = calc_count
param_dict['runs count'] = runs_count
param_dict['maximum individuals count'] = max_pop_count
param_dict['initial individuals count'] = init_pop_count
param_dict['archive individuals count'] = max_pop_count
param_dict['mutation probability'] = mut_p
param_dict['maximal life probability'] = max_lp
param_dict['decay step'] = decay_step
param_dict['archive resurrect probability'] = archive_p
param_dict['archive use in crossover probability'] = archive_p
param_dict['history'] = False
'''
run_count = 100
init_count = 100
max_count = 100
mut_ch = 0.05
param_dict['runs count'] = run_count
param_dict['maximum individuals count'] = max_count
param_dict['initial individuals count'] = init_count
param_dict['mutation probability'] = mut_ch
test_list = ['85 10']
var_count = 0
calc_count = 0
for ttt in test_list:
print(ttt)
tt = ttt.split(' ')
t1 = float(tt[0]) * 0.01
t2 = float(tt[1]) * 0.01
param_dict['maximal life probability'] = t1
param_dict['decay step'] = t2
for func_number in range(1, 9):
func_name = ''
if func_number == 1:
func_name = 'sphere'
elif func_number == 2:
func_name = 'cigar'
elif func_number == 3:
func_name = 'rosenbrock'
elif func_number == 4:
func_name = 'ackley'
elif func_number == 5:
func_name = 'griewank'
elif func_number == 6:
func_name = 'rastrigin'
elif func_number == 7:
func_name = 'schaffer'
elif func_number == 8:
func_name = 'schwefel'
for test_n in range(0, 3):
if test_n == 0:
var_count = 2
calc_count = 5000
elif test_n == 1:
var_count = 5
calc_count = 20000
else:
# var_count = 20
var_count = 10
calc_count = 50000
# calc_count = 50000
mess = func_name + '_v' + str(var_count)
file_name = 'tests/' + ttt + '/EE_' + mess + '.xls'
print(mess)
bord_all = define_borders(func_name, var_count)
# output
# Initialize a workbook
book = xlwt.Workbook(encoding="utf-8")
# Add a sheet to the workbook
sheet = book.add_sheet("N Runs")
sheet.write(0, 0, 'Values')
sheet.write(0, 1, 'Variables')
# Run!
param_dict['calculations count'] = calc_count
param_dict['variables count'] = var_count
param_dict['function name'] = func_name
param_dict['function borders'] = bord_all
pop = None
if func_name == 'cassini2':
print('matlab init')
cassini2_problem.process_init()
print('matlab init end!')
for i in range(run_count):
pop = ga.Population(param_dict)
pop.evolve()
best = pop.archive[0]
print('Run №' + str(i + 1) + ' complete!')
sheet.write(i + 1, 0, best.value.__str__())
for j in range(var_count):
sheet.write(i + 1, j + 1, best.variables[j].__str__())
if func_name == 'cassini2':
cassini2_problem.process_term()
print('---_Done_---')
# info
now = datetime.datetime.now()
ns = now.strftime("%Y-%m-%d %H:%M:%S")
s_info = book.add_sheet("Info")
s_info.write(0, 0, "Date-time")
s_info.write(0, 1, ns)
i = 1
for x in param_dict.keys():
s_info.write(i, 0, str(x))
i += 1
i = 1
for x in param_dict.values():
s_info.write(i, 1, str(x))
i += 1
book.save(file_name)
def single_run(param_dict):
# vars
var_count = param_dict['variables count']
param_dict['history'] = True
calc_count = param_dict['calculations count']
f_name = param_dict['function name']
# Run!
if f_name == 'cassini2':
print('matlab init')
cassini2_problem.process_init()
print('matlab init end!')
pop = ga.Population(param_dict)
pop.evolve()
if f_name == 'cassini2':
cassini2_problem.process_term()
best = pop.archive[0]
print('Archive best:')
print(best)
print('Pop best:')
print(pop.individuals[0])
print('Mean lifeteme = ' + str(pop.AV_mean_lifetime))
print('Archive used: ' + str(pop.AV_archive_use_count) + ' times')
print('Mutations count: ' + str(pop.AV_mutations_count))
print('----------end----------')
# output
# Initialize a workbook
book = xlwt.Workbook(encoding="utf-8")
# Add a sheet to the workbook
sheet_arch = book.add_sheet("Archive")
sheet_arch.write(0, 0, 'Values')
sheet_arch.write(0, 1, 'Variables')
for i in range(pop.archive.__len__()):
ind = pop.archive[i]
sheet_arch.write(i + 1, 0, ind.value.__str__())
for j in range(var_count):
sheet_arch.write(i + 1, j + 1, ind.variables[j])
sheet_pop = book.add_sheet("Population")
sheet_pop.write(0, 0, 'Values')
sheet_pop.write(0, 1, 'Variables')
for i in range(pop.individuals.__len__()):
ind = pop.individuals[i]
sheet_pop.write(i + 1, 0, ind.value.__str__())
for j in range(var_count):
sheet_pop.write(i + 1, j + 1, ind.variables[j])
# info
now = datetime.datetime.now()
ns = now.strftime("%Y-%m-%d %H:%M:%S")
s_info = book.add_sheet("Info")
s_info.write(0, 0, "Date-time")
s_info.write(0, 1, ns)
i = 1
for x in param_dict.keys():
s_info.write(i, 0, str(x))
i += 1
i = 1
for x in param_dict.values():
s_info.write(i, 1, str(x))
i += 1
book.save("output_single.xls")
# additional graph
fig, ax = plt.subplots(figsize=(12, 7))
# ax.cla()
colors = ("red", "green", "blue", 'black', 'yellow', 'purple')
# colors = plt.get_cmap('hsv', clus_l + 1) # + 1 because of color[0] ~ color[n-1]
# Create plot
y1 = np.zeros(shape=(pop.mmm_inds.__len__()))
y2 = np.zeros(shape=(pop.mmm_inds.__len__()))
y3 = np.zeros(shape=(pop.mmm_inds.__len__()))
y4 = np.zeros(shape=(pop.mmm_inds.__len__()))
y5 = np.zeros(shape=(pop.mmm_inds.__len__()))
y6 = np.zeros(shape=(pop.mmm_inds.__len__()))
for i in range(pop.mmm_inds.__len__()):
cur_inds = pop.mmm_inds[i]
y1[i] = cur_inds[0].value
y2[i] = cur_inds[1].value
y3[i] = cur_inds[2].value
y4[i] = cur_inds[3].value
y5[i] = cur_inds[4].value
y6[i] = cur_inds[5].value
ax.plot(y1, c=colors[0], label='лучший архивный')
ax.plot(y2, c=colors[1], label='средний архивный')
ax.plot(y3, c=colors[2], label='худший архивный')
ax.plot(y4, c=colors[3], label='лучший из популяции')
ax.plot(y5, c=colors[4], label='средний из популяции')
ax.plot(y6, c=colors[5], label='худший из популяции')
plt.plot()
title = 'Разнообразие популяции'
plt.title(title)
ax.set_xlabel('Поколение')
ax.set_ylabel('Значение функции')
plt.legend(loc=1)
plt.grid(linestyle='--')
plt.show()
www = visualization.vis_main(pop.gen_history)
global windows
windows.append(www)