Add method SparseVariable.toarray()

doc/library/sparse/index.txt

@@ -123,6 +123,7 @@ List of Implemented Operations
       Both grad are implemented. Structured by default.
     - :class:`SparseFromDense <theano.sparse.basic.SparseFromDense>` and ``csr_from_dense``, ``csc_from_dense``.
       The grad implemented is structured.
+    - Theano SparseVariable object have a method ``toarray()`` that is the same as ``dense_from_sparse``.
 
 - Construction of Sparses and their Properties
     - :class:`CSM <theano.sparse.basic.CSM>` and ``CSC``, ``CSR`` to construct a matrix.

theano/sparse/basic.py

@@ -300,6 +300,8 @@ class _sparse_py_operators:
     # def _as_TensorVariable(self):
     #    return dense_from_sparse(self)
 
+    def toarray(self):
+        return dense_from_sparse(self)
     shape = property(lambda self: tensor.shape(dense_from_sparse(self)))
     # don't worry!
     # the plan is that the ShapeFeature in tensor.opt will do shape propagation

theano/sparse/tests/test_basic.py

@@ -708,18 +708,25 @@ class T_conversion(unittest.TestCase):
             self.assertTrue(str(val.dtype) == 'float64')
             self.assertTrue(val.format == 'csr')
 
-    if 1:
-        def test2(self):
-            #call dense_from_sparse
-            for t in _mtypes:
-                s = t(scipy.sparse.identity(5))
-                d = dense_from_sparse(s)
-                # s should be copied into the graph as a constant
-                s[0, 0] = 3.0  # changes s, but not the copy
-                val = eval_outputs([d])
-                return
-                self.assertTrue(str(val.dtype) == s.dtype)
-                self.assertTrue(numpy.all(val[0] == [1, 0, 0, 0, 0]))
+    def test_dense_from_sparse(self):
+        #call dense_from_sparse
+        for t in _mtypes:
+            s = t(scipy.sparse.identity(5))
+            s = as_sparse_variable(s)
+            d = dense_from_sparse(s)
+            val = eval_outputs([d])
+            self.assertTrue(str(val.dtype) == s.dtype)
+            self.assertTrue(numpy.all(val[0] == [1, 0, 0, 0, 0]))
+
+    def test_todense(self):
+        #call sparse_var.todense()
+        for t in _mtypes:
+            s = t(scipy.sparse.identity(5))
+            s = as_sparse_variable(s)
+            d = s.toarray()
+            val = eval_outputs([d])
+            self.assertTrue(str(val.dtype) == s.dtype)
+            self.assertTrue(numpy.all(val[0] == [1, 0, 0, 0, 0]))
 
     @staticmethod
     def check_format_ndim(format, ndim):