Added first draft of polynomial kernel.

GPy/kern/constructors.py

@@ -69,6 +69,26 @@ def mlp(input_dim,variance=1., weight_variance=None,bias_variance=100.,ARD=False
     part = parts.mlp.MLP(input_dim,variance,weight_variance,bias_variance,ARD)
     return kern(input_dim, [part])
 
+def poly(input_dim,variance=1., weight_variance=None,bias_variance=1.,degree=2, ARD=False):
+    """
+    Construct a polynomial kernel
+
+    :param input_dim: dimensionality of the kernel, obligatory
+    :type input_dim: int
+    :param variance: the variance of the kernel
+    :type variance: float
+    :param weight_scale: the lengthscale of the kernel
+    :type weight_scale: vector of weight variances for input weights.
+    :param bias_variance: the variance of the biases.
+    :type bias_variance: float
+    :param degree: the degree of the polynomial
+    :type degree: int
+    :param ARD: Auto Relevance Determination (allows for ARD version of covariance)
+    :type ARD: Boolean
+    """
+    part = parts.poly.POLY(input_dim,variance,weight_variance,bias_variance,degree,ARD)
+    return kern(input_dim, [part])
+
 def white(input_dim,variance=1.):
     """
      Construct a white kernel.

GPy/kern/parts/__init__.py

@@ -12,6 +12,7 @@ import mlp
 import periodic_exponential
 import periodic_Matern32
 import periodic_Matern52
+import poly
 import prod_orthogonal
 import prod
 import rational_quadratic

GPy/kern/parts/kernpart.py

@@ -56,5 +56,5 @@ class Kernpart(object):
         raise NotImplementedError
     def dpsi2_dmuS(self,dL_dpsi2,Z,mu,S,target_mu,target_S):
         raise NotImplementedError
-    def dK_dX(self,X,X2,target):
+    def dK_dX(self, dL_dK, X, X2, target):
         raise NotImplementedError

GPy/kern/parts/mlp.py

@@ -111,10 +111,10 @@ class MLP(Kernpart):
         gX = np.zeros((X2.shape[0], X.shape[1], X.shape[0]))
         
         for i in range(X.shape[0]):
-            gX[:, :, i] = self._dK_dX_point(X, X2, i)
+            gX[:, :, i] = self._dK_dX_point(dL_dK, X, X2, target, i)
             
             
-    def _dK_dX_point(self, X, X2, i):
+    def _dK_dX_point(self, dL_dK, X, X2, target, i):
         """Gradient with respect to one point of X"""
         
         inner_prod = self._K_inner_prod[i, :].T
@@ -127,10 +127,11 @@ class MLP(Kernpart):
         #arg = numer/denom
         gX = np.zeros(X2.shape)
         denom3 = denom*denom*denom
+        gX = np.zeros((X2.shape[0], X2.shape[1]))
         for j in range(X2.shape[1]):
-            gX[:, j]=X2[:, j]/denom - vec2*X[i, j]*numer/denom3
+            gX[:, j] =X2[:, j]/denom - vec2*X[i, j]*numer/denom3
             gX[:, j] = four_over_tau*self.weight_variance*self.variance*gX[:, j]/np.sqrt(1-arg*arg)
-
+            target[i, :]
     
     
     def _K_computations(self, X, X2):