def test_func_args(self):
X = np.linspace(0, 1, 10)[:, None]
Y = np.random.randn(10, 1)
with Model() as model:
# make a Gaussian model
with pytest.raises(ValueError):
random_test = gp.GP('random_test', cov_func=gp.mean.Zero(), observed={'X':X, 'Y':Y})
with pytest.raises(ValueError):
random_test = gp.GP('random_test', mean_func=gp.cov.Matern32(1, 1),
cov_func=gp.cov.Matern32(1, 1), observed={'X':X, 'Y':Y})
def test_sample(self):
X = np.linspace(0,1,100)[:,None]
Y = np.random.randn(100,1)
with Model() as model:
M = gp.mean.Zero()
l = Uniform('l', 0, 5)
K = gp.cov.Matern32(1, l)
sigma = Uniform('sigma', 0, 10)
# make a Gaussian model
random_test = gp.GP('random_test', mean_func=M, cov_func=K, sigma=sigma, observed={'X':X, 'Y':Y})
tr = sample(500, init=None, progressbar=False, random_seed=self.random_seed)
def gaussian(x, y):
X = x.reshape(-1, 1)
Z = linspace(-6, 6, 100).reshape(-1, 1)
with Model() as gp_fit:
ρ = Gamma('ρ', 1, 1)
nu = Gamma('nu', 1, 1)
K = nu * gp.cov.Matern32(1, ρ)
with gp_fit:
M = gp.mean.Zero()
σ = HalfCauchy('σ', 2.5)
with gp_fit:
y_obs = gp.GP('y_obs', mean_func=M, cov_func=K, sigma=σ, observed={'X': X, 'Y': y})
with gp_fit:
gp_samples = pm.gp.sample_gp(trace[1000:], y_obs, Z, samples=50)
def test_sample(self):
X = np.linspace(0, 1, 10)[:, None]
Y = np.random.randn(10)
with Model() as model:
M = gp.mean.Zero()
l = Uniform('l', 0, 5)
K = gp.cov.Matern32(1, l)
sigma = Uniform('sigma', 0, 10)
# make a Gaussian model
random_test = gp.GP('random_test', mean_func=M, cov_func=K, sigma=sigma, observed={'X':X, 'Y':Y})
tr = sample(20, init=None, progressbar=False, random_seed=self.random_seed)
# test prediction
Z = np.linspace(0, 1, 5)[:, None]
with model:
out = gp.sample_gp(tr[-3:], gp=random_test, X_values=Z, obs_noise=False,
random_seed=self.random_seed, progressbar=False, chol_const=True)