From fc59ef4baf8044eb9496ef9b6d5919f8cadd9d57 Mon Sep 17 00:00:00 2001
From: Alan Saul <alan.daniel.saul@gmail.com>
Date: Mon, 28 Oct 2013 15:42:25 +0000
Subject: [PATCH] Tidying up and fixed objective being vector

---
 GPy/likelihoods/laplace.py                        | 8 ++++----
 GPy/likelihoods/noise_models/exponential_noise.py | 7 ++++---
 GPy/likelihoods/noise_models/gamma_noise.py       | 6 ++++--
 3 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/GPy/likelihoods/laplace.py b/GPy/likelihoods/laplace.py
index 8a11b146..7e570e52 100644
--- a/GPy/likelihoods/laplace.py
+++ b/GPy/likelihoods/laplace.py
@@ -340,8 +340,8 @@ class Laplace(likelihood):
                 Ki_f = old_Ki_f + step_size*dKi_f
                 f = np.dot(K, Ki_f)
                 # This is nasty, need to set something within an optimization though
-                self.Ki_f = Ki_f.copy()
-                self.f = f.copy()
+                self.tmp_Ki_f = Ki_f.copy()
+                self.tmp_f = f.copy()
                 return -obj(Ki_f, f)
 
             i_o = partial_func(inner_obj, old_Ki_f=old_Ki_f, dKi_f=dKi_f, K=K)
@@ -349,8 +349,8 @@ class Laplace(likelihood):
             #The tolerance and maxiter matter for speed! Seems to be best to keep them low and make more full
             #steps than get this exact then make a step, if B was bigger it might be the other way around though
             new_obj = sp.optimize.minimize_scalar(i_o, method='brent', tol=1e-4, options={'maxiter':5}).fun
-            f = self.f.copy()
-            Ki_f = self.Ki_f.copy()
+            f = self.tmp_f.copy()
+            Ki_f = self.tmp_Ki_f.copy()
 
             #Optimize without linesearch
             #f_old = f.copy()
diff --git a/GPy/likelihoods/noise_models/exponential_noise.py b/GPy/likelihoods/noise_models/exponential_noise.py
index 8e916353..e637cc02 100644
--- a/GPy/likelihoods/noise_models/exponential_noise.py
+++ b/GPy/likelihoods/noise_models/exponential_noise.py
@@ -40,7 +40,8 @@ class Exponential(NoiseDistribution):
         :rtype: float
         """
         assert np.atleast_1d(link_f).shape == np.atleast_1d(y).shape
-        return np.exp(np.sum(np.log(link_f*np.exp(-y*link_f))))
+        log_objective = link_f*np.exp(-y*link_f)
+        return np.exp(np.sum(np.log(log_objective)))
         #return np.exp(np.sum(-y/link_f - np.log(link_f) ))
 
     def logpdf_link(self, link_f, y, extra_data=None):
@@ -60,9 +61,9 @@ class Exponential(NoiseDistribution):
 
         """
         assert np.atleast_1d(link_f).shape == np.atleast_1d(y).shape
-        logpdf_link = np.sum(np.log(link_f) - y*link_f)
+        log_objective = np.log(link_f) - y*link_f
         #logpdf_link = np.sum(-np.log(link_f) - y/link_f)
-        return logpdf_link
+        return np.sum(log_objective)
 
     def dlogpdf_dlink(self, link_f, y, extra_data=None):
         """
diff --git a/GPy/likelihoods/noise_models/gamma_noise.py b/GPy/likelihoods/noise_models/gamma_noise.py
index 2e4e7d15..2be3106a 100644
--- a/GPy/likelihoods/noise_models/gamma_noise.py
+++ b/GPy/likelihoods/noise_models/gamma_noise.py
@@ -44,7 +44,8 @@ class Gamma(NoiseDistribution):
         assert np.atleast_1d(link_f).shape == np.atleast_1d(y).shape
         #return stats.gamma.pdf(obs,a = self.gp_link.transf(gp)/self.variance,scale=self.variance)
         alpha = link_f*self.beta
-        return (y**(alpha - 1.) * np.exp(-self.beta*y) * self.beta**alpha)/ special.gamma(alpha)
+        objective = (y**(alpha - 1.) * np.exp(-self.beta*y) * self.beta**alpha)/ special.gamma(alpha)
+        return np.exp(np.sum(np.log(objective)))
 
     def logpdf_link(self, link_f, y, extra_data=None):
         """
@@ -67,7 +68,8 @@ class Gamma(NoiseDistribution):
         #alpha = self.gp_link.transf(gp)*self.beta
         #return (1. - alpha)*np.log(obs) + self.beta*obs - alpha * np.log(self.beta) + np.log(special.gamma(alpha))
         alpha = link_f*self.beta
-        return alpha*np.log(self.beta) - np.log(special.gamma(alpha)) + (alpha - 1)*np.log(y) - self.beta*y
+        log_objective = alpha*np.log(self.beta) - np.log(special.gamma(alpha)) + (alpha - 1)*np.log(y) - self.beta*y
+        return np.sum(log_objective)
 
     def dlogpdf_dlink(self, link_f, y, extra_data=None):
         """