Moved kron into slinalg, Fixed nlinalg and slinalg code and doc.

doc/library/tensor/index.txt

@@ -25,3 +25,5 @@ They are grouped into the following sections:
     utils
     extra_ops
     io
+    slinalg
+    nlinalg

doc/library/tensor/nlinalg.txt

-..  ../../../../theano/sandbox/linalg/ops.py
+..  ../../../../theano/sandbox/nlinalg.py
 
 .. _libdoc_linalg:
 
 ===================================================================
-:mod:`tensor.slinalg` --  Linear Algebra Ops Using Numpy
+:mod:`tensor.nlinalg` --  Linear Algebra Ops Using Numpy
 ===================================================================
 
 .. module:: tensor.nlinalg
@@ -14,7 +14,5 @@
 API
 ===
 
-.. automodule:: theano.tensor.nlinalg.ops
-    :members:
-.. automodule:: theano.tensor.nlinalg.kron
+.. automodule:: theano.tensor.nlinalg
     :members:

doc/library/tensor/slinalg.txt

-..  ../../../../theano/sandbox/linalg/ops.py
+..  ../../../../theano/sandbox/slinalg.py
 
 .. _libdoc_linalg:
 
@@ -14,7 +14,5 @@
 API
 ===
 
-.. automodule:: theano.tensor.slinalg.ops
-    :members:
-.. automodule:: theano.tensor.slinalg.kron
+.. automodule:: theano.tensor.slinalg
     :members:

theano/tensor/nlinalg.py

@@ -54,11 +54,7 @@ class MatrixPinv(Op):
         return Apply(self, [x], [x.type()])
 
     def perform(self, node, (x,), (z, )):
-        try:
-            z[0] = numpy.linalg.pinv(x).astype(x.dtype)
-        except numpy.linalg.LinAlgError:
-            logger.debug('Failed to invert %s' % str(node.inputs[0]))
-            raise
+        z[0] = numpy.linalg.pinv(x).astype(x.dtype)
 
     def __str__(self):
         return "MatrixPseudoInverse"
@@ -100,11 +96,7 @@ class MatrixInverse(Op):
         return Apply(self, [x], [x.type()])
 
     def perform(self, node, (x,), (z, )):
-        try:
-            z[0] = numpy.linalg.inv(x).astype(x.dtype)
-        except numpy.linalg.LinAlgError:
-            logger.debug('Failed to invert %s' % str(node.inputs[0]))
-            raise
+        z[0] = numpy.linalg.inv(x).astype(x.dtype)
 
     def grad(self, inputs, g_outputs):
         r"""The gradient function should return
@@ -346,12 +338,7 @@ class Eig(Op):
         return Apply(self, [x], [w, v])
 
     def perform(self, node, (x,), (w, v)):
-        try:
-            w[0], v[0] = [z.astype(x.dtype) for z in self._numop(x)]
-        except numpy.linalg.LinAlgError:
-            logger.debug('Failed to find %s of %s' % (self._numop.__name__,
-                                                      node.inputs[0]))
-            raise
+        w[0], v[0] = [z.astype(x.dtype) for z in self._numop(x)]
 
     def infer_shape(self, node, shapes):
         n = shapes[0][0]
@@ -394,12 +381,7 @@ class Eigh(Eig):
         return Apply(self, [x], [w, v])
 
     def perform(self, node, (x,), (w, v)):
-        try:
-            w[0], v[0] = self._numop(x, self.UPLO)
-        except numpy.linalg.LinAlgError:
-            logger.debug('Failed to find %s of %s' % (self._numop.__name__,
-                                                      node.inputs[0]))
-            raise
+        w[0], v[0] = self._numop(x, self.UPLO)
 
     def grad(self, inputs, g_outputs):
         r"""The gradient function should return

theano/tensor/slinalg.py

@@ -355,3 +355,36 @@ class EigvalshGrad(Op):
 
 def eigvalsh(a, b, lower=True):
     return Eigvalsh(lower)(a, b)
+
+
+def kron(a, b):
+    """ Kronecker product
+
+    Same as scipy.linalg.kron(a, b).
+
+    :note: numpy.kron(a, b) != scipy.linalg.kron(a, b)!
+        They don't have the same shape and order when
+        a.ndim != b.ndim != 2.
+
+    :param a: array_like
+    :param b: array_like
+    :return: array_like with a.ndim + b.ndim - 2 dimensions.
+
+    """
+    a = tensor.as_tensor_variable(a)
+    b = tensor.as_tensor_variable(b)
+    if (a.ndim + b.ndim <= 2):
+        raise TypeError('kron: inputs dimensions must sum to 3 or more. '
+                        'You passed %d and %d.' % (a.ndim, b.ndim))
+    o = tensor.outer(a, b)
+    o = o.reshape(tensor.concatenate((a.shape, b.shape)),
+                  a.ndim + b.ndim)
+    shf = o.dimshuffle(0, 2, 1, * range(3, o.ndim))
+    if shf.ndim == 3:
+        shf = o.dimshuffle(1, 0, 2)
+        o = shf.flatten()
+    else:
+        o = shf.reshape((o.shape[0] * o.shape[2],
+                         o.shape[1] * o.shape[3]) +
+                        tuple([o.shape[i] for i in range(4, o.ndim)]))
+    return o