diff --git a/GPy/examples/BGPLVM_demo.py b/GPy/examples/BGPLVM_demo.py
index 056891aa..760bf4b4 100644
--- a/GPy/examples/BGPLVM_demo.py
+++ b/GPy/examples/BGPLVM_demo.py
@@ -17,7 +17,7 @@ K = k.K(X)
 Y = np.random.multivariate_normal(np.zeros(N),K,D).T
 
 # k = GPy.kern.rbf(Q) + GPy.kern.bias(Q) + GPy.kern.white(Q, 0.00001)
-k = GPy.kern.linear(Q, ARD = False)  + GPy.kern.white(Q, 0.00001)
+k = GPy.kern.linear(Q, ARD = True)  + GPy.kern.white(Q, 0.00001)
 m = GPy.models.Bayesian_GPLVM(Y, Q, kernel = k,  M=M)
 m.constrain_positive('(rbf|bias|noise|white|S)')
 # m.constrain_fixed('S', 1)
diff --git a/GPy/inference/SGD.py b/GPy/inference/SGD.py
index 425b2b95..2f2d1ef0 100644
--- a/GPy/inference/SGD.py
+++ b/GPy/inference/SGD.py
@@ -205,11 +205,10 @@ class opt_SGD(Optimizer):
                     self.x_opt -= step + momentum_term
 
                 if self.messages == 2:
-                    if count == last_printed_count + 20 or count == 0:
-                        status = "evaluating {feature: 5d}/{tot: 5d} \t f: {f: 2.3f} \t non-missing: {nm: 4d}\r".format(feature = count, tot = len(features), f = f, nm = Nj)
-                        sys.stdout.write(status)
-                        sys.stdout.flush()
-                        last_printed_count = count
+                    status = "evaluating {feature: 5d}/{tot: 5d} \t f: {f: 2.3f} \t non-missing: {nm: 4d}\r".format(feature = count, tot = len(features), f = f, nm = Nj)
+                    sys.stdout.write(status)
+                    sys.stdout.flush()
+                    last_printed_count = count
 
                 NLL.append(f)
 
diff --git a/GPy/kern/linear.py b/GPy/kern/linear.py
index a83580e1..5272e4f6 100644
--- a/GPy/kern/linear.py
+++ b/GPy/kern/linear.py
@@ -95,7 +95,11 @@ class linear(kernpart):
 
     def dpsi0_dtheta(self,partial,Z,mu,S,target):
         self._psi_computations(Z,mu,S)
-        target += (partial[:, None] * (np.sum(self.mu2_S,0))).sum()
+        tmp = (partial[:, None] * (np.sum(self.mu2_S,0)))
+        if self.ARD:
+            target += tmp.sum(0)
+        else:
+            target += tmp.sum()
 
     def dpsi0_dmuS(self,partial, Z,mu,S,target_mu,target_S):
         target_mu += partial[:, None] * (2.0*mu*self.variances) * mu.shape[0]
@@ -130,7 +134,12 @@ class linear(kernpart):
 
     def dpsi2_dtheta(self,partial,Z,mu,S,target):
         self._psi_computations(Z,mu,S)
-        target += (partial[:,:,:,None]*(2.*self.ZZ*self.mu2_S[:,None,None,:]*self.variances)).sum()
+        tmp = (partial[:,:,:,None]*(2.*self.ZZ*self.mu2_S[:,None,None,:]*self.variances))
+        if self.ARD:
+            target += tmp.sum(0).sum(0).sum(0)
+        else:
+            target += tmp.sum()
+
 
     def dpsi2_dmuS(self,partial,Z,mu,S,target_mu,target_S):
         """Think N,M,M,Q """