apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-06 18:42:39 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'newGP' of github.com:SheffieldML/GPy into newGP

Browse source
This commit is contained in:
Ricardo Andrade 2013-02-01 15:17:13 +00:00
commit 5593d53828
2 changed files with 12 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

2
GPy/likelihoods/EP.py
View file

@ -50,7 +50,7 @@ class EP(likelihood):
mu_diff_2 = (self.v_/self.tau_ - mu_tilde)**2
self.Z = np.sum(np.log(self.Z_hat)) + 0.5*np.sum(np.log(sigma_sum)) + 0.5*np.sum(mu_diff_2/sigma_sum) #Normalization constant, aka Z_ep
self.Y = mu_tilde[:,None]
self.Y = mu_tilde[:,None]
self.YYT = np.dot(self.Y,self.Y.T)
self.precision = self.tau_tilde
self.covariance_matrix = np.diag(1./self.precision)

15
GPy/models/sparse_GP.py
View file

@ -70,16 +70,23 @@ class sparse_GP(GP):
self.psi0 = self.kern.psi0(self.Z,self.X, self.X_uncertainty).sum()
self.psi1 = self.kern.psi1(self.Z,self.X, self.X_uncertainty).T
self.psi2 = self.kern.psi2(self.Z,self.X, self.X_uncertainty)
self.psi2_beta_scaled = (self.psi2*(self.beta/sf2)).sum(0)
if self.likelihood.is_heteroscedastic:
self.psi2_beta_scaled = (self.psi2*(self.likelihood.precision.reshape(self.N,1,1)/sf2)).sum(0)
#TODO: what is the likelihood is heterscedatic and there are multiple independent outputs?
else:
self.psi2_beta_scaled = (self.psi2*(self.likelihood.precision/sf2)).sum(0)
else:
self.psi0 = self.kern.Kdiag(self.X,slices=self.Xslices).sum()
self.psi1 = self.kern.K(self.Z,self.X)
tmp = self.psi1*(np.sqrt(self.likelihood.beta)/sf)
if self.likelihood.is_heteroscedastic:
tmp = self.psi1*(np.sqrt(self.likelihood.precision.reshape(self.N,1))/sf)
else:
tmp = self.psi1*(np.sqrt(self.likelihood.precision)/sf)
self.psi2_beta_scaled = np.dot(tmp,tmp.T)
self.Kmmi, self.Lm, self.Lmi, self.Kmm_logdet = pdinv(self.Kmm)#+np.eye(self.M)*1e-3)
self.Kmmi, self.Lm, self.Lmi, self.Kmm_logdet = pdinv(self.Kmm)
self.V = (self.likelihood.beta/self.scale_factor)*self.Y
self.V = (self.likelihood.precision/self.scale_factor)*self.Y
self.A = mdot(self.Lmi, self.psi2_beta_scaled, self.Lmi.T)
self.B = np.eye(self.M)/sf2 + self.A