Change T.shape for T._shape, since the latter is an instance of the Shape Op.

theano/tensor/opt.py

@@ -159,34 +159,35 @@ register_canonicalize(local_dimshuffle_lift)
 # Shape lifters #
 #################
 
-@gof.local_optimizer([T.shape, None])
+@gof.local_optimizer([T._shape, None])
 def local_shape_lift_elemwise(node):
     """
     shape(elemwise_op(..., x, ...)) -> shape(x)
 
     Where x contains the maximal shape information.
     """
-    if not opt.check_chain(node, T.shape, T.Elemwise):
+    if not opt.check_chain(node, T._shape, T.Elemwise):
         return False
-    
+
     output = node.inputs[0]
     parent = output.owner
 
     for input in parent.inputs:
         if input.type.broadcastable == output.type.broadcastable:
-            return T.shape.make_node(input).outputs
+            return T._shape(input),
 
     return False
 
 register_canonicalize(local_shape_lift_elemwise, 'shape_lift')
+register_specialize(local_shape_lift_elemwise, 'shape_lift')
 
 
-@gof.local_optimizer([T.shape, None])
+@gof.local_optimizer([T._shape, None])
 def local_shape_lift_sum(node):
     """
     shape(sum{n}(x)) -> [shape(x)[0], ..., shape(x)[n-1], shape(x)[n+1], ...]
     """
-    if not opt.check_chain(node, T.shape, T.Sum):
+    if not opt.check_chain(node, T._shape, T.Sum):
         return False
 
     input = node.inputs[0].owner.inputs[0]
@@ -195,7 +196,7 @@ def local_shape_lift_sum(node):
         axis = range(input.type.ndim)
 
 
-    ish = T.shape(input)
+    ish = T._shape(input)
     return T.make_lvector.make_node(*(ish[i] for i in xrange(input.type.ndim) if i not in axis)).outputs
 #    return T.vertical_stack.make_node(ish[:axis], ish[axis+1:]).outputs