Added the infer_shape and grad method in the example Op implementation.

doc/cifarSC2011/extending_theano.txt

@@ -127,17 +127,41 @@ Op example
             z = output_storage[0]
             z[0] = x * 2
 
+        def infer_shape(self, node, i0_shapes):
+            return i0_shapes
+
+        def grad(self, inputs, output_grads):
+            return [output_grads[0]*2]
+
 Test it!
 
->>> x = theano.tensor.matrix()
->>> f = theano.function([x],DoubleOp()(x))
->>> import numpy
->>> inp = numpy.random.rand(5,5)
->>> out = f(inp)
->>> assert numpy.allclose(inp*2, out)
->>> print inp
->>> print out
+.. code-block:: python
+
+    x = theano.tensor.matrix()
+    f = theano.function([x],DoubleOp()(x))
+    import numpy
+    inp = numpy.random.rand(5,5)
+    out = f(inp)
+    assert numpy.allclose(inp*2, out)
+    print inp
+    print out
+
+Testing the gradient
+--------------------
+
+The function :ref:`verify_grad
+<validating_grad>`
+verify the gradient of an Op or Theano graph. It compare the
+analytic(symbolicaly computed) gradient and the numeric
+gradient(computed through the Finite Difference Method).
+
+To verify the grad method of the DoubleOp, you can use this:
+
+.. code-block:: python
 
+   import numpy
+   import theano.tests
+   theano.tests.unittest_tools.verify_grad(DoubleOp(), [numpy.random.rand(5,7,2)])
 
 Exercises 8
 -----------

doc/extending/unittest.txt

@@ -374,6 +374,8 @@ A few tools have been developed to help automate the development of
 unitests for Theano Ops.
 
 
+.. _validating_grad:
+
 Validating the Gradient
 -----------------------