Merge pull request #365 from SheffieldML/devel

patch 1.0.5

AUTHORS.txt

@@ -1 +1 @@
-[GPy Authors](https://github.com/SheffieldML/GPy/graphs/contributors)
+GPy Authors: https://github.com/SheffieldML/GPy/graphs/contributors

GPy/__version__.py

@@ -1 +1 @@
-__version__ = "1.0.3"
+__version__ = "1.0.5"

GPy/core/gp.py

@@ -532,21 +532,23 @@ def input_sensitivity(self, summarize=True):
     def get_most_significant_input_dimensions(self, which_indices=None):
         return self.kern.get_most_significant_input_dimensions(which_indices)
 
-    def optimize(self, optimizer=None, start=None, **kwargs):
+    def optimize(self, optimizer=None, start=None, messages=False, max_iters=1000, ipython_notebook=True, clear_after_finish=False, **kwargs):
         """
         Optimize the model using self.log_likelihood and self.log_likelihood_gradient, as well as self.priors.
         kwargs are passed to the optimizer. They can be:
 
-        :param max_f_eval: maximum number of function evaluations
-        :type max_f_eval: int
+        :param max_iters: maximum number of function evaluations
+        :type max_iters: int
         :messages: whether to display during optimisation
         :type messages: bool
         :param optimizer: which optimizer to use (defaults to self.preferred optimizer), a range of optimisers can be found in :module:`~GPy.inference.optimization`, they include 'scg', 'lbfgs', 'tnc'.
         :type optimizer: string
+        :param bool ipython_notebook: whether to use ipython notebook widgets or not.
+        :param bool clear_after_finish: if in ipython notebook, we can clear the widgets after optimization.
         """
         self.inference_method.on_optimization_start()
         try:
-            super(GP, self).optimize(optimizer, start, **kwargs)
+            super(GP, self).optimize(optimizer, start, messages, max_iters, ipython_notebook, clear_after_finish, **kwargs)
         except KeyboardInterrupt:
             print("KeyboardInterrupt caught, calling on_optimization_end() to round things up")
             self.inference_method.on_optimization_end()

GPy/kern/src/stationary.py

@@ -315,10 +315,10 @@ def __init__(self, input_dim, variance=1., lengthscale=None, ARD=False, active_d
         super(Exponential, self).__init__(input_dim, variance, lengthscale, ARD, active_dims, name)
 
     def K_of_r(self, r):
-        return self.variance * np.exp(-0.5 * r)
+        return self.variance * np.exp(-r)
 
     def dK_dr(self, r):
-        return -0.5*self.K_of_r(r)
+        return -self.K_of_r(r)
 
 #    def sde(self): 
 #        """ 

GPy/models/state_space_main.py

@@ -3237,6 +3237,7 @@ def lti_sde_to_descrete(F,L,Qc,dt,compute_derivatives=False,
             AB = np.dot(AB, np.vstack((np.zeros((n,n)),np.eye(n))))
 
             Q_noise_1 = linalg.solve(AB[n:,:].T,AB[:n,:].T)
+            Q_noise_2  = P_inf - A.dot(P_inf).dot(A.T)
             # The covariance matrix Q by matrix fraction decomposition <-
 
             if compute_derivatives:
@@ -3276,8 +3277,9 @@ def lti_sde_to_descrete(F,L,Qc,dt,compute_derivatives=False,
             else:
               dA = None
               dQ = None
-              Q_noise = Q_noise_1
-
+              Q_noise = Q_noise_2
+	      # Innacuracies have been observed when Q_noise_1 was used.
+
             #Q_noise = Q_noise_1
 
             # Return
@@ -3484,4 +3486,4 @@ def balance_ss_model(F,L,Qc,H,Pinf,P0,dF=None,dQc=None,dPinf=None,dP0=None):
 
     # (F,L,Qc,H,Pinf,P0,dF,dQc,dPinf,dP0)
 
-    return bF, bL, bQc, bH, bPinf, bP0, bdF, bdQc, bdPinf, bdP0, T
+    return bF, bL, bQc, bH, bPinf, bP0, bdF, bdQc, bdPinf, bdP0, T