From 5b433a3f73504604240b4425b699987d53a9408e Mon Sep 17 00:00:00 2001
From: Ricardo <acq11ra@sheffield.ac.uk>
Date: Mon, 15 Jul 2013 18:09:26 +0100
Subject: [PATCH] random changes

---
 GPy/core/gp.py                                      |  7 ++++++-
 GPy/likelihoods/ep.py                               |  3 ++-
 GPy/likelihoods/noise_models/gaussian_noise.py      | 11 ++++++++---
 GPy/likelihoods/noise_models/noise_distributions.py |  3 +++
 4 files changed, 19 insertions(+), 5 deletions(-)

diff --git a/GPy/core/gp.py b/GPy/core/gp.py
index 5172d9e7..8d52a984 100644
--- a/GPy/core/gp.py
+++ b/GPy/core/gp.py
@@ -98,7 +98,12 @@ class GP(GPBase):
 
         Note, we use the chain rule: dL_dtheta = dL_dK * d_K_dtheta
         """
-        return np.hstack((self.kern.dK_dtheta(dL_dK=self.dL_dK, X=self.X), self.likelihood._gradients(partial=np.diag(self.dL_dK))))
+        #return np.hstack((self.kern.dK_dtheta(dL_dK=self.dL_dK, X=self.X), self.likelihood._gradients(partial=np.diag(self.dL_dK))))
+        if not isinstance(self.likelihood,EP):
+            tmp = np.hstack((self.kern.dK_dtheta(dL_dK=self.dL_dK, X=self.X), self.likelihood._gradients(partial=np.diag(self.dL_dK))))
+        else:
+            tmp = np.hstack((self.kern.dK_dtheta(dL_dK=self.dL_dK, X=self.X), self.likelihood._gradients(partial=np.diag(self.dL_dK))))
+        return tmp
 
     def _raw_predict(self, _Xnew, which_parts='all', full_cov=False,stop=False):
         """
diff --git a/GPy/likelihoods/ep.py b/GPy/likelihoods/ep.py
index 50df599a..aaa03938 100644
--- a/GPy/likelihoods/ep.py
+++ b/GPy/likelihoods/ep.py
@@ -77,7 +77,8 @@ class EP(likelihood):
         self.noise_model._set_params(p)
 
     def _gradients(self,partial):
-        return np.zeros(0) # TODO: the EP likelihood might want to take some parameters...
+        #return np.zeros(0) # TODO: the EP likelihood might want to take some parameters...
+        return self.noise_model._gradients(partial)
 
     def _compute_GP_variables(self):
         #Variables to be called from GP
diff --git a/GPy/likelihoods/noise_models/gaussian_noise.py b/GPy/likelihoods/noise_models/gaussian_noise.py
index 389363a3..40db423c 100644
--- a/GPy/likelihoods/noise_models/gaussian_noise.py
+++ b/GPy/likelihoods/noise_models/gaussian_noise.py
@@ -28,6 +28,10 @@ class Gaussian(NoiseDistribution):
     def _set_params(self,p):
         self.variance = p
 
+    def _gradients(self,partial):
+        return np.zeros(1)
+        #return np.sum(partial)
+
     def _preprocess_values(self,Y):
         """
         Check if the values of the observations correspond to the values
@@ -57,13 +61,14 @@ class Gaussian(NoiseDistribution):
         return 1./(1./self.variance + 1./sigma**2)
 
     def _mass(self,gp,obs):
-        return std_norm_pdf( (self.gp_link.transf(gp)-obs)/np.sqrt(self.variance) )
+        #return std_norm_pdf( (self.gp_link.transf(gp)-obs)/np.sqrt(self.variance) )
+        return stats.norm.pdf(obs,self.gp_link.transf(gp),np.sqrt(self.variance)) #FIXME
 
     def _nlog_mass(self,gp,obs):
-        return .5*((self.gp_link.transf(gp)-obs)**2/np.sqrt(self.variance) + np.log(2*np.pi*self.variance))
+        return .5*((self.gp_link.transf(gp)-obs)**2/self.variance + np.log(2.*np.pi*self.variance))
 
     def _dnlog_mass_dgp(self,gp,obs):
-        return (self.gp_link.transf(gp)-obs)/np.sqrt(self.variance) * self.gp_link.dtransf_df(gp)
+        return (self.gp_link.transf(gp)-obs)/self.variance * self.gp_link.dtransf_df(gp)
 
     def _d2nlog_mass_dgp2(self,gp,obs):
         return ((self.gp_link.transf(gp)-obs)*self.gp_link.d2transf_df2(gp) + self.gp_link.dtransf_df(gp)**2)/self.variance
diff --git a/GPy/likelihoods/noise_models/noise_distributions.py b/GPy/likelihoods/noise_models/noise_distributions.py
index 45f92950..bc4d89d6 100644
--- a/GPy/likelihoods/noise_models/noise_distributions.py
+++ b/GPy/likelihoods/noise_models/noise_distributions.py
@@ -43,6 +43,9 @@ class NoiseDistribution(object):
     def _set_params(self,p):
         pass
 
+    def _gradients(self,partial):
+        return np.zeros(0)
+
     def _preprocess_values(self,Y):
         """
         In case it is needed, this function assess the output values or makes any pertinent transformation on them.