from GP.covariance_functions import SquaredExponential, Matern, GammaExponential
data_params = np.array([1.0, 0.5, 0.1])
data_covariance_obj = SquaredExponential(data_params)
model_params = np.array([0.7, 0.3, 0.2])
model_covariance_obj = SquaredExponential(model_params)
gp = GPR(data_covariance_obj)
num = 500
test_num = 100
input_nums = [5, 10, 15, 20, 30, 50]
dim = 1
seed = 10
file_name = 'd1_n500'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr, y_tr, x_test, y_test, model_params, input_nums, file_name, title, True)
data_params = np.array([1.0, 0.5, 0.1])
data_covariance_obj = SquaredExponential(data_params)
model_params = np.array([0.7, 0.3, 0.2])
model_covariance_obj = SquaredExponential(model_params)
gp = GPR(data_covariance_obj)
num = 500
test_num = 100
input_nums = [5, 10, 30, 50, 60]
# input_nums = [40, 50]#, 60]
dim = 4
seed = 10
file_name = 'd5_n500'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr, y_tr, x_test, y_test, model_params, input_nums, file_name, title, vi=True, full=True, show=True)
from GP.covariance_functions import SquaredExponential, Matern, GammaExponential
data_params = np.array([1.0, 0.5, 0.1])
data_covariance_obj = SquaredExponential(data_params)
model_params = np.array([0.7, 0.3, 0.2])
model_covariance_obj = SquaredExponential(model_params)
gp = GPR(data_covariance_obj)
num = 500
test_num = 100
input_nums = [5, 10, 15, 20, 30, 50]
dim = 1
seed = 10
file_name = 'd1_n500'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr, y_tr, x_test, y_test, model_params, input_nums, file_name,
title, True)
from GP.covariance_functions import SquaredExponential, Matern, GammaExponential
data_params = np.array([1.0, 0.5, 0.1])
data_covariance_obj = SquaredExponential(data_params)
model_params = np.array([0.7, 0.3, 0.2])
model_covariance_obj = SquaredExponential(model_params)
gp = GPR(data_covariance_obj)
num = 4000
test_num = 100
input_nums = [10, 20, 50, 100, 200, 300, 400, 500, 600]
dim = 10
seed = 10
file_name = 'd10_n4000'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr, y_tr, x_test, y_test, model_params, input_nums, file_name, title, True, full=False, vi=False)
test_num = 100
input_nums = [5, 10, 30, 50, 60]
# input_nums = [40, 50]#, 60]
dim = 4
seed = 10
file_name = 'd5_n500'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr,
y_tr,
x_test,
y_test,
model_params,
input_nums,
file_name,
title,
vi=True,
full=True,
show=True)
num = 4000
test_num = 100
input_nums = [10, 20, 50, 100, 200, 300, 400, 500, 600]
dim = 10
seed = 10
file_name = 'd10_n4000'
title = 'Generated dataset, n = ' + str(num) + ', d = ' + str(dim)
# Generating data points
np.random.seed(seed)
x_tr = np.random.rand(dim, num)
if dim == 1:
x_test = np.linspace(0, 1, test_num)
x_test = x_test.reshape(1, test_num)
else:
x_test = np.random.rand(dim, test_num)
y_tr, y_test = gp.generate_data(x_tr, x_test, seed=seed)
run_methods(x_tr,
y_tr,
x_test,
y_test,
model_params,
input_nums,
file_name,
title,
True,
full=False,
vi=False)