testing infer_shape: op SoftmaxGrad

theano/tensor/nnet/tests/test_nnet.py

@@ -19,7 +19,7 @@ from theano.tensor.nnet import (categorical_crossentropy,
                                 crossentropy_softmax_argmax_1hot_with_bias,
                                 sigmoid, softplus,
                                 Softmax, softmax, SoftmaxWithBias, softmax_grad,
-                                softmax_with_bias,
+                                softmax_with_bias, SoftmaxGrad,
                                 Prepend_scalar_constant_to_each_row,
                                 Prepend_scalar_to_each_row)
 from theano.tensor import dmatrix, dvector
@@ -125,12 +125,17 @@ class T_SoftmaxWithBias(utt.InferShapeTester):
                             [admat_val, advec_val], SoftmaxWithBias)
 
 
-class T_SoftmaxGrad(unittest.TestCase):
+class T_SoftmaxGrad(utt.InferShapeTester):
     def test_infer_shape(self):
-        a=T.constant(numpy.random.rand(3,4))
-        b=T.constant(numpy.random.rand(3,4))
-        f=theano.function([],softmax_grad(a,b).shape)
-        assert numpy.all(f()==[3,4])
+
+        admat = dmatrix()
+        bdmat = dmatrix()
+        admat_val = numpy.random.rand(3, 4)
+        bdmat_val = numpy.random.rand(3, 4)
+        self._compile_and_check([admat, bdmat], [SoftmaxGrad()(admat, bdmat)],
+                            [admat_val, bdmat_val], SoftmaxGrad)
+
+
 
 class T_CrossentropySoftmax1Hot(unittest.TestCase):
     def setUp(self):
@@ -1103,7 +1108,7 @@ class Test_softmax_opt:
 
 if __name__ == '__main__':
 
-    t = T_SoftmaxWithBias('setUp')
+    t = T_SoftmaxGrad('setUp')
     t.setUp()
     t.test_infer_shape()