def ga_result1(out='result'):
''' アニメーション表示
'''
file = ut.fsort(glob.glob(os.path.join(out, f'epoch*.pickle')))[-1]
optimizer = MOEAD.load(file=file)
print('epoch:', len(optimizer))
elite = optimizer.get_elite()
print('elite:', len(elite))
imax = len(optimizer) - 1
for epoch in range(0, imax, 20):
print(epoch, end='\n')
# plt.cla()
population = optimizer[epoch]
# population = optimizer.calc_rank(population)
front = [x for x in population]
x, y = np.array([x.data.value for x in front]).T
plt.scatter(x,
y,
color='blue',
alpha=epoch / imax,
label=f'epoch{epoch}')
# plt.legend()
plt.xlim((0, 50))
# plt.ylim((0, 50))
plt.pause(0.2)
plt.show()
def get_model():
# モデル読み込み
# model_cls = {'nsga2':NSGA2, 'moead':MOEAD}[model]
files = ut.fsort(glob.glob(os.path.join(out, f'*epoch*.pkl')))
for i, file in enumerate(files):
print(f'[{i}]', file)
print('select file')
n = int(input())
if n < 0:
return
file = files[n]
print('file:', file)
env, optimizer, history = ut.load(file)
return env, optimizer, history
def ga_result14(model, out='result', show=False):
''' エリート個体数を数える
'''
# モデル読み込み
model_cls = {'nsga2': NSGA2, 'moead': MOEAD}[model]
files = ut.fsort(glob.glob(os.path.join(out, f'*epoch*.pickle')))
for i, file in enumerate(files):
print(f'[{i}]', file)
print('select file')
n = int(input())
if n < 0:
return
file = files[n]
print('file:', file)
optimizer = model_cls.load(file=file)
imax = len(optimizer)
# imax = 100
idx = 0
cache = file.replace('.pickle', '_ndom.npy')
if os.path.isfile(cache) and not force:
data = np.load(cache)
else:
data = []
for epoch in range(0, imax, 1):
print(epoch, end='\r')
# plt.cla()
population = optimizer[epoch]
population = optimizer.calc_rank(population)
front = [x for x in population if x.rank == 1]
indivs = [fit.get_indiv() for fit in population]
if hasattr(population[0], 'id'):
idx = max(*map(attrgetter('id'), population), idx)
else:
idx = max(*map(attrgetter('id'), indivs), idx)
data.append([idx, len(front)])
np.save(cache, data)
def ga_result2(out='result'):
file = ut.fsort(glob.glob(os.path.join(out, f'epoch*.pickle')))[-1]
optimizer = MOEAD.load(file=file)
population = optimizer[-1]
front = [x for x in population if x.rank == 1]
front.sort(key=attrgetter('data.value'))
for fit in front:
# print(ind.value, ind.data.value)
print(fit.get_indiv().get_gene().get_ivalue(), fit.value)
crowdings = [ind.value[1] for ind in front]
fig, axes = plt.subplots(2)
axes[0].plot(crowdings)
x, y = np.array([x.data.value for x in front]).T
im = axes[1].scatter(x, y, c=crowdings, cmap='jet')
# plt.xlim((0, 1))
# plt.ylim((0, 1))
fig.colorbar(im)
plt.show()
def ga_result13(model, out='result', show=False):
''' HVプロット
'''
# モデル読み込み
model_cls = {'nsga2': NSGA2, 'moead': MOEAD}[model]
files = ut.fsort(glob.glob(os.path.join(out, f'*epoch*.pickle')))
for i, file in enumerate(files):
print(f'[{i}]', file)
print('select file')
n = int(input())
if n < 0:
return
file = files[n]
print('file:', file)
optimizer = model_cls.load(file=file)
ref_point = [100, 0]
hypervolume = HyperVolume(ref=ref_point)
# モデルパラメータ表示
print('[PARAMS]')
print('popsize:', optimizer.popsize)
print('epoch:', len(optimizer))
print('alternation:', optimizer.alternation)
print('Indiv:', optimizer.indiv_type.__name__)
optimizer.calc_rank(optimizer.population)
elite = optimizer.get_elite()
print('elite:', len(elite))
imax = len(optimizer)
# imax = 100
idx = 0
force = False
cache = file.replace('.pickle', '_hv.npy')
if os.path.isfile(cache) and not force:
hvs = np.load(cache)
else:
hvs = []
population = []
for epoch in range(0, imax, 1):
# print(epoch, end='\r')
# plt.cla()
population = optimizer[epoch] + population
optimizer.calc_rank(population)
population = [fit for fit in population if fit.rank == 1]
hv = hypervolume(population)
print('epoch:', epoch, 'HV=', hv)
if hasattr(population[0], 'id'):
idx = max(*map(attrgetter('id'), population), idx)
else:
idx = max(*map(attrgetter('id'), indivs), idx)
hvs.append([idx, hv])
np.save(cache, hvs)
i, v = np.array(hvs).T
fig, ax = plt.subplots()
im = ax.plot(i, v, label=model)
ax.legend()
ax.set_xlabel('Evaluation Step')
ax.set_ylabel('Hypervolume')
if show:
plt.show()
else:
basename = os.path.splitext(os.path.basename(file))[0]
plt.savefig(os.path.join('result', f'front_{model}_{basename}_hv.png'))
def ga_result12(model, out='result', show=False):
''' 最適値プロット
'''
# モデル読み込み
model_cls = {'nsga2': NSGA2, 'moead': MOEAD}[model]
files = ut.fsort(glob.glob(os.path.join(out, f'*epoch*.pickle')))
for i, file in enumerate(files):
print(f'[{i}]', file)
print('select file')
n = int(input())
if n < 0:
return
file = files[n]
print('file:', file)
optimizer = model_cls.load(file=file)
# モデルパラメータ表示
print('[PARAMS]')
print('popsize:', optimizer.popsize)
print('epoch:', len(optimizer))
print('alternation:', optimizer.alternation)
print('Indiv:', optimizer.indiv_type.__name__)
optimizer.calc_rank(optimizer.population)
elite = optimizer.get_elite()
print('elite:', len(elite))
imax = len(optimizer)
# imax = 100
idx = 0
cache = file.replace('.pickle', '_data.npy')
if os.path.isfile(cache) and not force:
data = np.load(cache)
else:
data = []
for epoch in range(0, imax, 1):
print(epoch, end='\r')
# plt.cla()
population = optimizer[epoch]
indivs = [fit.get_indiv() for fit in population]
dv = min(map(itemgetter(0), indivs))
rcs = max(map(itemgetter(1), indivs))
if hasattr(population[0], 'id'):
idx = max(*map(attrgetter('id'), population), idx)
else:
idx = max(*map(attrgetter('id'), indivs), idx)
data.append([idx, dv, rcs])
np.save(cache, data)
i, x, y = np.array(data).T
fig, ax = plt.subplots()
# color_list = [(0, '#AADDFF'), (0.33, '#002288'), (0.66, '#EEDD22'), (1, '#CC0000')]
# color_list = [(0, '#0022AA'), (0.3, '#11AA00'), (0.5, '#EEDD22'), (1, '#CC0000')]
# color_list = [(0, 'blue'), (0.3, 'green'), (0.5, 'yellow'), (1, 'red')]
# cmap = plc.LinearSegmentedColormap.from_list('custom_cmap', color_list, N=25)
im = ax.plot(i, x)
im = ax.plot(i, y * 100)
print('min dV=', x.min())
print('max RCS=', y.max())
# plt.legend()
# ax.set_xlabel('$\\Delta\\it{V}$, $km/s$')
# ax.set_ylabel('$RCS$, $m^2$')
# ax.set_xlim((0, 100))
# ax.set_ylim((0, 0.45))
# cbar = fig.colorbar(im, label='generation')
if show:
plt.show()
else:
basename = os.path.splitext(os.path.basename(file))[0]
plt.savefig(os.path.join('result', f'front_{model}_{basename}.png'))
def ga_result11(model, out='result', show=False):
''' 一括表示
'''
# モデル読み込み
model_cls = {'nsga2': NSGA2, 'moead': MOEAD}[model]
files = ut.fsort(glob.glob(os.path.join(out, f'*epoch*.pickle')))
for i, file in enumerate(files):
print(f'[{i}]', file)
print('select file')
n = int(input())
if n < 0:
return
file = files[n]
print('file:', file)
optimizer = model_cls.load(file=file)
# モデルパラメータ表示
print('[PARAMS]')
print('popsize:', optimizer.popsize)
print('epoch:', len(optimizer))
print('alternation:', optimizer.alternation)
if hasattr(optimizer, 'ksize'):
print('ksize:', optimizer.ksize)
print('Indiv:', optimizer.indiv_type.__name__)
optimizer.calc_rank(optimizer.population)
elite = optimizer.get_elite()
print('elite:', len(elite))
imax = len(optimizer)
# imax = 51
data = []
for epoch in range(0, imax, 1):
print(epoch, end='\r')
# plt.cla()
population = optimizer[epoch]
front = [x for x in population]
data.extend([[*x.data.value, epoch] for x in front])
x, y, c = np.array(data).T
fig, ax = plt.subplots()
# color_list = [(0, '#AADDFF'), (0.33, '#002288'), (0.66, '#EEDD22'), (1, '#CC0000')]
color_list = [(0, '#0022AA'), (0.3, '#11AA00'), (0.5, '#EEDD22'),
(1, '#CC0000')]
color_list = [(0, 'blue'), (0.3, 'green'), (0.5, 'yellow'), (1, 'red')]
cmap = plc.LinearSegmentedColormap.from_list('custom_cmap',
color_list,
N=25)
im = ax.scatter(x, y, c=c, cmap=cmap, marker='o', alpha=1)
print('min dV=', x.min())
print('max RCS=', y.max())
# plt.legend()
ax.set_xlabel('$\\Delta\\it{V}$, $km/s$')
ax.set_ylabel('$RCS$, $m^2$')
ax.set_xlim((0, 100))
ax.set_ylim((0, 0.45))
# plt.pause(0.2)
cbar = fig.colorbar(im, label='Generation')
cbar.set_label('Generation', size=24)
plt.tight_layout()
# show or save
if show:
plt.show()
else:
basename = os.path.splitext(os.path.basename(file))[0]
plt.savefig(os.path.join('result', f'front_{model}_{basename}.png'))