"""
This file serves as an example of how to
a) select a problem to be solved
b) select a network type
c) train the network to minimize recovery MSE
"""
import numpy as np
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # BE QUIET!!!!
# os.environ['CUDA_VISIBLE_DEVICES'] = '7'
import tensorflow as tf
np.random.seed(1) # numpy is good about making repeatable output
tf.set_random_seed(
1) # on the other hand, this is basically useless (see issue 9171)
from tools import problems, networks, train
A = np.load('A_250_500.npy')
prob = problems.bernoulli_gaussian_trial(
A=A, M=250, N=500, L=10000, is_train=False
) #a Bernoulli-Gaussian x, noisily observed through a random matrix
T = 14
layers = networks.build_LAMP(prob, T=T, shrink='soft', untied=False)
train.do_testing_another(layers, prob, T,
'LAMP-soft T=' + str(T) + ' trainrate=0.001.npz')
import numpy as np
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # BE QUIET!!!!
import tensorflow as tf
np.random.seed(1) # numpy is good about making repeatable output
tf.set_random_seed(
1) # on the other hand, this is basically useless (see issue 9171)
# import our problems, networks and training modules
from tools import problems, networks, train
# Create the basic problem structure.
prob = problems.bernoulli_gaussian_trial(
kappa=None, M=250, N=500, L=1000, pnz=.1,
SNR=40) #a Bernoulli-Gaussian x, noisily observed through a random matrix
#prob = problems.random_access_problem(2) # 1 or 2 for compressive random access or massive MIMO
# build a LISTA network to solve the problem and get the intermediate results so we can greedily extend and then refine(fine-tune)
layers = networks.build_LISTA(prob, T=6, initial_lambda=.1, untied=False)
# plan the learning
training_stages = train.setup_training(layers,
prob,
trinit=1e-3,
refinements=(.5, .1, .01))
# do the learning (takes a while)
sess = train.do_training(training_stages, prob, 'LISTA_bg_giid.npz')
xs = []
ys_mean = []
ys_var = []
ys_std = []
T = 16
count = 1
M = 250
N = 500
loss_total = np.zeros([len(eps), 10000])
for i in range(len(eps)):
print(eps[i])
prob = problems.bernoulli_gaussian_trial(A=A,
M=M,
N=N,
L=1000,
is_train=False)
layers = networks.build_LAMP_act(prob,
tao,
T=T,
eps=eps[i],
shrink='soft',
untied=False)
savefile = 'LAMP_T_' + str(T) + '_tao_' + str(tao) + '_count_' + str(
count) + '.npz'
num, num_min, num_max, loss_mean, loss_var, loss_std, loss = train.do_testing_another(
layers, prob, tao,
'T_10_tao_' + str(tao) + '_thresh_' + str(eps[i]) + '_test.txt',
savefile)
print(num_min)
import tensorflow as tf
np.random.seed(1) # numpy is good about making repeatable output
tf.set_random_seed(
1) # on the other hand, this is basically useless (see issue 9171)
# import our problems, networks and training modules
from tools import problems, networks, train
A = np.load('A_250_500.npy')
tao = 10
eps = 0.01 #this value will not influence the traning results, it's used to see the layers that different signals use, the average layer is reduced if eps increase
for i in range(16, 17):
prob = problems.bernoulli_gaussian_trial(A=A,
M=250,
N=500,
L=1000,
is_train=True)
layers = networks.build_LAMP_act(prob,
tao,
T=i,
eps=eps,
shrink='soft',
untied=False)
training_stages = train.setup_training(layers,
i,
prob,
tao=tao,
trinit=1e-4,
type='LAMP')
sess = train.do_training(training_stages,
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # BE QUIET!!!!
import tensorflow as tf
# np.random.seed(1) # numpy is good about making repeatable output
# tf.set_random_seed(1) # on the other hand, this is basically useless (see issue 9171)
# import our problems, networks and training modules
from tools import problems,networks,train
from scipy.io import savemat
if __name__ == "__main__":
if not os.path.exists("probs"):
os.mkdir("probs")
start = 4.0
end = 15.0
step = 0.5
for kappa in np.arange(start, end+step, step):
for i in range(5):
base = "probs/" + "problem_k{0:04.1f}_{1}".format(kappa, i)
problems.save_problem(base,problems.bernoulli_gaussian_trial(M=50,N=100,L=500,pnz=.1,kappa=kappa,SNR=40))
# save_problem('problem_k5',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=5,SNR=40))
# save_problem('problem_k15',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=15,SNR=40))
# save_problem('problem_k100',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=100,SNR=40))
# save_problem('problem_k1000',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=1000,SNR=40))
# save_problem('problem_rap1',problems.random_access_problem(1))
# save_problem('problem_rap2',problems.random_access_problem(2))
# save_problem('problem_k100',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=100,SNR=40))
def save_problem(base, prob):
print('saving {b}.mat,{b}.npz norm(x)={x:.7f} norm(y)={y:.7f}'.format(
b=base, x=la.norm(prob.xval), y=la.norm(prob.yval)))
D = dict(A=prob.A, x=prob.xval, y=prob.yval)
np.savez(base + '.npz', **D)
savemat(base + '.mat', D, oned_as='column')
#save_problem('problem_Giid',problems.bernoulli_gaussian_trial(M=250,N=500,L=1000,pnz=.1,kappa=None,SNR=40))
save_problem(
'problem_Giid',
problems.bernoulli_gaussian_trial(M=250,
N=500,
L=1000,
pnz=.1,
kappa=0,
SNR=40))
save_problem(
'problem_k15',
problems.bernoulli_gaussian_trial(M=250,
N=500,
L=1000,
pnz=.1,
kappa=15,
SNR=40))
save_problem(
'problem_k100',
problems.bernoulli_gaussian_trial(M=250,
N=500,
L=1000,
parser.add_argument('--shrink',
type=str,
default='bg',
help='choose your shrinkage function')
if __name__ == "__main__":
parsed, unparsed = parser.parse_known_args()
if not os.path.exists(parsed.dest_models):
os.mkdir(parsed.dest_models)
if not os.path.exists(parsed.dest_probs):
os.mkdir(parsed.dest_probs)
prob = problems.bernoulli_gaussian_trial(M=parsed.M,
N=parsed.N,
L=parsed.L,
pnz=parsed.pnz,
kappa=parsed.kappa,
SNR=parsed.SNR)
# save problems
base = parsed.dest_probs + "/problem_k{0:04.1f}_{1}".format(
parsed.kappa, parsed.id)
problems.save_problem(base, prob)
if parsed.model == 'LVAMP':
layers = networks.build_LVAMP(prob, T=parsed.T, shrink=parsed.shrink)
elif parsed.model == 'LAMP':
layers = networks.build_LAMP(prob,
T=parsed.T,
shrink=parsed.shrink,
untied=False)
elif parsed.model == 'LISTA':
