from scripts.ssc.wc_offline.config_libraries.global_register_definitions import \
(
PATH_GR_SWISSROLL_EULER, PATH_GR_SWISSROLL_NOISE_EULER)
from src.data_preprocessing.witness_complex_offline.config import ConfigWC_Grid
from src.datasets.datasets import SwissRoll
swissroll_nonoise = ConfigWC_Grid(
dataset=[SwissRoll()],
sampling_kwargs=[dict(n_samples=2560)],
batch_size=[64, 128, 256, 512],
wc_kwargs=[dict()],
eval_size=[0.2],
n_jobs=[2],
seed=[36, 3851, 2570, 4304, 1935, 7954, 5095, 5310, 1577, 3288],
global_register=PATH_GR_SWISSROLL_EULER,
root_path=
'/cluster/home/schsimo/MT/output/WitnessComplexes/SwissRoll/nonoise',
verbose=True)
swissroll_noise005 = ConfigWC_Grid(
dataset=[SwissRoll()],
sampling_kwargs=[dict(n_samples=2560, noise=0.05)],
batch_size=[64, 128, 256, 512],
wc_kwargs=[dict()],
eval_size=[0.2],
n_jobs=[2],
seed=[6973, 5305, 6233, 1503, 3947, 1425, 3391, 2941, 1218, 7946],
global_register=PATH_GR_SWISSROLL_NOISE_EULER,
root_path=
'/cluster/home/schsimo/MT/output/WitnessComplexes/SwissRoll/noise',
verbose=True)
if k not in tmp:
tmp[k] = {}
tmp = tmp[k]
tmp[kc[-1]] = kc_v
ret.append(Config_Competitors(**ret_i))
return ret
placeholder_config_competitors = Config_Competitors(
model_class=tSNE,
model_kwargs=dict(),
dataset=SwissRoll(),
sampling_kwargs={'n_samples': [2560]},
eval=[
ConfigEval(
active=True,
evaluate_on=None,
save_eval_latent=True,
save_train_latent=True,
online_visualization=False,
k_min=5,
k_max=20,
k_step=5,
)
],
uid='uid',
verbose=False,
import random
import torch
import numpy as np
from scripts.ssc.persistence_pairings_visualization.utils_definitions import make_plot
from src.datasets.datasets import SwissRoll
from src.topology.witness_complex import WitnessComplex
if __name__ == "__main__":
dataset_sampler = SwissRoll()
N_WITNESSES = 2048
n_samples = 128
path_to_save = '/Users/simons/PycharmProjects/MT-VAEs-TDA/output/visualisation_nnsys/wc{}_w{}/'.format(
n_samples, N_WITNESSES)
N_sim = 100
ks = [1, 2, 3, 4, 6, 8, 12, 16]
ntot = int(len(ks) * N_sim)
counter = 1
for seed in list(set(np.random.randint(1, 100000, N_sim))):
witnesses, color_ = dataset_sampler.sample(N_WITNESSES, seed=seed)
ind = random.sample(range(N_WITNESSES), n_samples)
landmarks, color = witnesses[ind, :], color_[ind]
#dataset = SwissRoll()
# small = [0.5,0.8]
# large = [1, 1.3]
#
# area_l = math.pi*(large[1]**2-large[0]**2)
# area_s = math.pi*(small[1]**2-small[0]**2)
#
# sample_ratio = area_l/area_s
#
# infty_sign = tadasets.infty_sign(n=3000, noise=0.1)
for seed in [11, 22, 33, 44]:
for n in [16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 112, 128]:
name_plot_vr = 'VR_SwissRoll_n{n}_seed{seed}.pdf'.format(n=n,
seed=seed)
computeVR(data=SwissRoll().sample(n, seed=seed)[0],
path_to_save=os.path.join(path, name_plot_vr))
# df_an1 = annulus(int(sample_ratio*n), large[0], large[1], seed=seed)
# # df_an2 = annulus(n, small[0], small[1], label=1, seed=(seed+12))
# # df_an = df_an1.append(df_an2, ignore_index=True)
# # make_scatter(df_an, name='/annulus_manifold_ns{ns}_nl{nl}_s{s}'.format(ns=n, nl = int(sample_ratio*n), s=seed),
# # base_path=path)
# # data = df_an[['x','y']].to_numpy()
# #
# # name_plot = 'VR_ns{ns}_nl{nl}_seed{s}.pdf'.format(ns=n, nl=int(sample_ratio*n), s=seed)
# data = tadasets.infty_sign(n=n, noise=noise)
# labels = np.ones(n)
# name_plot_m = 'manifold_infty_n{n}_noise{noise}.pdf'.format(n=n, noise=noise)
# plot_2Dscatter(data,labels, path_to_save = os.path.join(path, name_plot_m), show = True)
# name_plot_vr = 'VR_infty_n{n}_noise{noise}.pdf'.format(n=n, noise=noise)
for v in itertools.product(*values):
ret_i = dict(root_path=self.root_path,
global_register=self.global_register,
verbose=self.verbose)
for kc, kc_v in zip(key_chains, v):
tmp = ret_i
for k in kc[:-1]:
if k not in tmp:
tmp[k] = {}
tmp = tmp[k]
tmp[kc[-1]] = kc_v
ret.append(ConfigWC(**ret_i))
return ret
placeholder_config_wc = ConfigWC(dataset=SwissRoll(),
sampling_kwargs=dict(n_samples=2560),
batch_size=64,
wc_kwargs=dict(),
eval_size=0.2,
n_jobs=1,
seed=1,
global_register='',
root_path='',
verbose=False)
import numpy as np
from torch.utils.data import TensorDataset, DataLoader
from src.data_preprocessing.witness_complex_offline.config import ConfigWC
from src.datasets.datasets import SwissRoll
if __name__ == "__main__":
n_samples = 5
#
labels = torch.from_numpy(np.array(range(n_samples)))
#
# dataset_train = TensorDataset(labels)
# dataset_train_2 = TensorDataset(labels)
#
# train_loader = DataLoader(dataset_train, batch_size=1, shuffle=False,
# pin_memory=True, drop_last=True)
#
# path = '/Users/simons/PycharmProjects/MT-VAEs-TDA/output/tests'
# torch.save(labels, os.path.join(path,'labels_tensor.pt'))
path = '/Users/simons/PycharmProjects/MT-VAEs-TDA/output/tests'
os.path.join(path, 'labels_tensor.pt')
labels_loaded = torch.load(os.path.join(path, 'labels_tensor.pt'))
print(torch.eq(labels_loaded, labels))
config = ConfigWC(SwissRoll(), dict(), dict(), 1, 1, 'global_bla',
'root_bla')
print(config.uid)
wcae = '/Users/simons/MT_data/sync/euler_sync_scratch/schsimo/output/WCAE_swissroll_nonoise/SwissRoll-n_samples2560-seed1935-Autoencoder_MLP_topoae-32-32-lr1_100-bs256-nep1000-rlw1-tlw8192-mepush_active21_20-k3-rmax10-seed1935-1edc2a73/'
eval_models_dict = {
'TopoAE64': topoae_64,
'TopoAE128': topoae_128,
'TopoAE256': topoae_256,
'WCTopoAE64': wctopoae_64,
'WCTopoAE128': wctopoae_128,
'WCTopoAE256': wctopoae_256,
}
eval_models_dict = {'wcae': wcae}
# sample data
n_samples = 2560
manifold = SwissRoll()
data_manifold, data, labels = manifold.sample_manifold(n_samples=n_samples,
seed=1)
model_names = []
values = []
for model_name, path in eval_models_dict.items():
# load WC-AE
model_kwargs = dict(input_dim=3,
latent_dim=2,
size_hidden_layers=[32, 32])
autoencoder = Autoencoder_MLP_topoae(**model_kwargs)
model = WitnessComplexAutoencoder(autoencoder)
state_dict = torch.load(os.path.join(path, 'model_state.pth'))
model.load_state_dict(state_dict)