apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-18 13:55:14 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

[mean_func] added parameters in additive mean func and tests for mean functions

Browse source
This commit is contained in:
mzwiessele 2016-08-03 08:32:46 +01:00
parent 02457660a2
commit e0c70a52a0
3 changed files with 43 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

3
GPy/mappings/additive.py
View file

@ -23,9 +23,10 @@ class Additive(Mapping):
assert(mapping1.input_dim==mapping2.input_dim) assert(mapping1.input_dim==mapping2.input_dim)
assert(mapping1.output_dim==mapping2.output_dim) assert(mapping1.output_dim==mapping2.output_dim)
input_dim, output_dim = mapping1.input_dim, mapping1.output_dim input_dim, output_dim = mapping1.input_dim, mapping1.output_dim
Mapping.__init__(self, input_dim=input_dim, output_dim=output_dim) super(Additive, self).__init__(input_dim=input_dim, output_dim=output_dim)
self.mapping1 = mapping1 self.mapping1 = mapping1
self.mapping2 = mapping2 self.mapping2 = mapping2
self.link_parameters(self.mapping1, self.mapping2)
def f(self, X): def f(self, X):
return self.mapping1.f(X) + self.mapping2.f(X) return self.mapping1.f(X) + self.mapping2.f(X)

2
GPy/mappings/linear.py
View file

@ -33,7 +33,7 @@ class Linear(Mapping):
return np.dot(X, self.A) return np.dot(X, self.A)
def update_gradients(self, dL_dF, X): def update_gradients(self, dL_dF, X):
self.A.gradient = np.dot( X.T, dL_dF) self.A.gradient = np.dot(X.T, dL_dF)
def gradients_X(self, dL_dF, X): def gradients_X(self, dL_dF, X):
return np.dot(dL_dF, self.A.T) return np.dot(dL_dF, self.A.T)

41
GPy/testing/meanfunc_tests.py
View file

@ -28,10 +28,49 @@ class MFtests(unittest.TestCase):
A linear mean function with parameters that we'll learn alongside the kernel A linear mean function with parameters that we'll learn alongside the kernel
""" """
X = np.linspace(-1,10,50).reshape(-1,1)
Y = 3-np.abs((X-6))
Y += .5*np.cos(3*X) + 0.3*np.random.randn(*X.shape)
mf = GPy.mappings.PiecewiseLinear(1, 1, [-1,1], [9,2])
k =GPy.kern.RBF(1)
lik = GPy.likelihoods.Gaussian()
m = GPy.core.GP(X, Y, kernel=k, likelihood=lik, mean_function=mf)
self.assertTrue(m.checkgrad())
def test_parametric_mean_function_composition(self):
"""
A linear mean function with parameters that we'll learn alongside the kernel
"""
X = np.linspace(0,10,50).reshape(-1,1) X = np.linspace(0,10,50).reshape(-1,1)
Y = np.sin(X) + 0.5*np.cos(3*X) + 0.1*np.random.randn(*X.shape) + 3*X Y = np.sin(X) + 0.5*np.cos(3*X) + 0.1*np.random.randn(*X.shape) + 3*X
mf = GPy.mappings.Linear(1,1) mf = GPy.mappings.Compound(GPy.mappings.Linear(1,1),
GPy.mappings.Kernel(1, 1, np.random.normal(0,1,(1,1)),
GPy.kern.RBF(1))
)
k =GPy.kern.RBF(1)
lik = GPy.likelihoods.Gaussian()
m = GPy.core.GP(X, Y, kernel=k, likelihood=lik, mean_function=mf)
self.assertTrue(m.checkgrad())
def test_parametric_mean_function_additive(self):
"""
A linear mean function with parameters that we'll learn alongside the kernel
"""
X = np.linspace(0,10,50).reshape(-1,1)
Y = np.sin(X) + 0.5*np.cos(3*X) + 0.1*np.random.randn(*X.shape) + 3*X
mf = GPy.mappings.Additive(GPy.mappings.Constant(1,1,3),
GPy.mappings.Additive(GPy.mappings.MLP(1,1),
GPy.mappings.Identity(1,1)
)
)
k =GPy.kern.RBF(1) k =GPy.kern.RBF(1)
lik = GPy.likelihoods.Gaussian() lik = GPy.likelihoods.Gaussian()