Let boolean indexing work with GpuArrayVariables.

theano/tensor/tests/test_subtensor.py

@@ -365,25 +365,37 @@ class T_subtensor(unittest.TestCase, utt.TestOptimizationMixin):
     def test_boolean(self):
         numpy_n = np.arange(6, dtype=self.dtype).reshape((2, 3))
         n = self.shared(numpy_n)
+
+        # indexing with a comparison (should translate to a boolean mask)
         assert_array_equal(numpy_n[numpy_n > 2], n[n > 2].eval())
         assert_array_equal(numpy_n[[0], numpy_n[0] > 2], n[[0], n[0] > 2].eval())
         assert_array_equal(numpy_n[[1], numpy_n[0] > 2], n[[1], n[0] > 2].eval())
 
+        # indexing with a mask for some dimensions
         mask = np.array([True, False])
         assert_array_equal(numpy_n[mask], n[mask].eval())
 
+        # indexing with a mask for the second dimension
         mask = np.array([True, False, True])
         assert_array_equal(numpy_n[0, mask], n[0, mask].eval())
         assert_array_equal(numpy_n[:, mask], n[:, mask].eval())
         assert_array_equal(numpy_n[:, mask], n[:, self.shared(mask)].eval())
 
+        # indexing with a boolean array
         mask = np.array([[True, False, True], [False, False, True]])
         assert_array_equal(numpy_n[mask], n[mask].eval())
         assert_array_equal(numpy_n[mask], n[self.shared(mask)].eval())
 
+        # for the GpuArray tests, self.shared will create GpuArray variables,
+        # but we also need to test the combination of GPU and normal TensorVariables
+        n_cpu = theano.shared(numpy_n)
+        assert_array_equal(numpy_n[mask], n[theano.shared(mask)].eval())
+        assert_array_equal(numpy_n[mask], n[theano.tensor.constant(mask)].eval())
+        assert_array_equal(numpy_n[mask], n_cpu[self.shared(mask)].eval())
+
+        # indexing with ellipsis
         numpy_n4 = np.arange(48, dtype=self.dtype).reshape((2, 3, 4, 2))
         n4 = self.shared(numpy_n4)
-        # with ellipsis
         assert_array_equal(numpy_n4[numpy_n > 2, ...], n4[n > 2, ...].eval())
         assert_array_equal(numpy_n4[numpy_n > 2, ..., 1], n4[n > 2, ..., 1].eval())
 

theano/tensor/var.py

@@ -467,8 +467,8 @@ class _tensor_py_operators(object):
         # Convert boolean arrays to calls to mask.nonzero()
         tmp_args = []
         for arg in args:
-            if (isinstance(arg, (np.ndarray, TensorVariable, TensorConstant,
-                                 theano.tensor.sharedvar.TensorSharedVariable)) and
+            if (isinstance(arg, (np.ndarray, theano.tensor.Variable)) and
+                    hasattr(arg, 'dtype') and hasattr(arg, 'nonzero') and
                     arg.dtype == 'bool'):
                 tmp_args += arg.nonzero()
             else: