apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-09 12:02:38 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

[merge]

Browse source
This commit is contained in:
Max Zwiessele 2015-09-02 14:34:23 +01:00
commit 74a7c6398b
2 changed files with 6 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

4
GPy/likelihoods/gaussian.py
View file

@ -326,7 +326,7 @@ class Gaussian(Likelihood):
dF_dtheta = -0.5/lik_var + 0.5*(np.square(Y) + np.square(m) + v - 2*m*Y)/(lik_var**2)
return F, dF_dmu, dF_dv, dF_dtheta.reshape(1, Y.shape[0], Y.shape[1])
class Heteroscedastic_Gaussian(Gaussian):
class HeteroscedasticGaussian(Gaussian):
def __init__(self, Y_metadata, gp_link=None, variance=1., name='het_Gauss'):
if gp_link is None:
gp_link = link_functions.Identity()
@ -335,7 +335,7 @@ class Heteroscedastic_Gaussian(Gaussian):
print("Warning, Exact inference is not implemeted for non-identity link functions,\
if you are not already, ensure Laplace inference_method is used")
super(Heteroscedastic_Gaussian, self).__init__(gp_link, np.ones(Y_metadata['output_index'].shape[0])*variance, name)
super(HeteroscedasticGaussian, self).__init__(gp_link, np.ones(Y_metadata['output_index'].shape[0])*variance, name)
def exact_inference_gradients(self, dL_dKdiag,Y_metadata=None):
return dL_dKdiag[Y_metadata['output_index']][:,0]

7
GPy/models/gp_heteroscedastic_regression.py
View file

@ -16,6 +16,8 @@ class GPHeteroscedasticRegression(GP):
:param X: input observations
:param Y: observed values
:param kernel: a GPy kernel, defaults to rbf
NB: This model does not make inference on the noise outside the training set
"""
def __init__(self, X, Y, kernel=None, Y_metadata=None):
@ -30,10 +32,7 @@ class GPHeteroscedasticRegression(GP):
kernel = kern.RBF(X.shape[1])
#Likelihood
#likelihoods_list = [likelihoods.Gaussian(name="Gaussian_noise_%s" %j) for j in range(Ny)]
noise_terms = np.unique(Y_metadata['output_index'].flatten())
likelihoods_list = [likelihoods.Gaussian(name="Gaussian_noise_%s" %j) for j in noise_terms]
likelihood = likelihoods.MixedNoise(likelihoods_list=likelihoods_list)
likelihood = likelihoods.HeteroscedasticGaussian(Y_metadata)
super(GPHeteroscedasticRegression, self).__init__(X,Y,kernel,likelihood, Y_metadata=Y_metadata)