help='proposal_type')
parser.add_argument('--scale',
type=str,
default='scale_64',
help='scale_size')
parser.add_argument('--noise', type=str, default='0.05', help='scale_size')
parser.add_argument('--exp',
type=str,
default='inversion_16_64',
help='scale_size')
args = parser.parse_args()
device = torch.device("cuda:2" if torch.cuda.is_available() else "cpu")
latent_c = latent_c(device)
mean_c = np.loadtxt(os.getcwd() +
f'/Gaussian/prediction/inversion_16/mean.dat')
init_c = latent_c.coarse_latent(args.exp, mean_c).reshape(16, )
# load experiment data #
obs = np.loadtxt(os.getcwd() + f'/Gaussian/test_data/obs_{args.noise}.dat')
sigma = np.loadtxt(os.getcwd() +
f'/Gaussian/test_data/sigma_{args.noise}.dat')
true_permeability = np.loadtxt(
os.getcwd() +
f'/Gaussian/test_data/true_permeability_{args.noise}.dat')
true_pressure = np.loadtxt(
os.getcwd() + f'/Gaussian/test_data/true_pressure_{args.noise}.dat')
obs_position = np.loadtxt(
os.getcwd() + f'/Gaussian/test_data/obs_position_{args.noise}.dat')
inference = inference(args.steps, args.burn_in, args.dim, obs, sigma,
true_permeability, true_pressure, obs_position,
args.scale, device)
np.random.seed(64)
'''
plot MDGM with 2 scales(16-64),16 is single scale(vanilla VAE)
'''
# torch.random.manual_seed(68)
torch.random.manual_seed(7)
device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")
latent_c = latent_c(device)
prediction = Prediction('scale_64', device)
nz_x, nz_y = 4, 4
z = torch.randn(4, 1, nz_x, nz_y)
model = prediction.model('inversion_16')
gen_imgs = prediction.permeability('inversion_16', model, z)
samples = np.squeeze(gen_imgs.data.cpu().numpy())
img = np.loadtxt(os.getcwd() + f'/Channel/MDGM_plot/channel_16.dat')
zc = latent_c.coarse_latent('inversion_16_64', img).reshape(16, )
zc = (np.ones([3, 1, 4, 4])) * zc.reshape(4, 4)
z64 = torch.randn(3, 1, 16, 16)
model = prediction.model('inversion_16_64')
gen_imgs1 = prediction.permeability('inversion_16_64', model, zc, zf=z64)
samples1 = np.squeeze(gen_imgs1.data.cpu().numpy())
samples2 = [
samples[0], samples[1], samples[2], samples[3], img, samples1[0],
samples1[1], samples1[2]
]
filename = '/channel_MDGM_16_64.pdf'
prediction.plot_ite(samples2, filename)
'''
plot MDGM with 3 scales(16-32-64), 16 is single scale(vanilla VAE) present above
'''
bbox_inches='tight',
dpi=1000,
pad_inches=0.0)
plt.close()
if __name__ == "__main__":
torch.random.manual_seed(10)
device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")
latent_c = latent_c(device)
prediction = Prediction('scale_64', device)
nz_x, nz_y = 4, 4
z = torch.randn(4, 1, nz_x, nz_y)
model = prediction.model('inversion_16')
gen_imgs = prediction.permeability('inversion_16', model, z)
samples = np.squeeze(gen_imgs.data.cpu().numpy())
condition = np.loadtxt(os.getcwd() +
f'/Gaussian/MDGM_plot/gaussian_16.dat')
zc = latent_c.coarse_latent('inversion_16_64', condition).reshape(16, )
zc = (np.ones([3, 1, 4, 4])) * zc.reshape(4, 4)
z64 = torch.randn(3, 1, 16, 16)
model = prediction.model('inversion_16_64')
gen_imgs1 = prediction.permeability('inversion_16_64', model, zc, zf=z64)
samples1 = np.squeeze(gen_imgs1.data.cpu().numpy())
samples2 = [
samples[0], samples[1], samples[2], samples[3], condition, samples1[0],
samples1[1], samples1[2]
]
filename = '/gaussian_MDGM_16_64.pdf'
prediction.plot_ite(samples2, filename)
if __name__ == "__main__":
'''
plot MDGM with 2 scales(16-64),16 is single scale(vanilla VAE)
'''
torch.random.manual_seed(9)
device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")
latent_c = latent_c(device)
prediction = Prediction('scale_64', device)
nz_x, nz_y = 4, 4
z = torch.randn(4, 1, nz_x, nz_y)
model = prediction.model('inversion_16')
gen_imgs = prediction.permeability('inversion_16', model, z)
samples = np.squeeze(gen_imgs.data.cpu().numpy())
img = np.loadtxt(os.getcwd() + f'/Channel/MDGM_plot/channel_16.dat')
zc = latent_c.coarse_latent('inversion_16_64', img).reshape(16, )
zc = (np.ones([3, 1, 4, 4])) * zc.reshape(4, 4)
z64 = torch.randn(3, 1, 16, 16)
model = prediction.model('inversion_16_64')
gen_imgs1 = prediction.permeability('inversion_16_64', model, zc, zf=z64)
samples1 = np.squeeze(gen_imgs1.data.cpu().numpy())
samples2 = [
samples[0], samples[1], samples[2], samples[3], img, samples1[0],
samples1[1], samples1[2]
]
filename = '/channel_MDGM_16_64.pdf'
prediction.plot_ite(samples2, filename)
'''
plot MDGM with 3 scales(16-32-64), 16 is single scale(vanilla VAE) present above
'''