--- theano/compile/builders.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/compile/builders.py	2019-08-22 13:25:53.896813593 -0600

@@ -15,7 +15,7 @@ from theano.gradient import Disconnected

 class OpFromGraph(gof.Op):

-    """

+    r"""

     This creates an ``Op`` from inputs and outputs lists of variables.

     The signature is similar to :func:`theano.function <theano.function>`

     and the resulting ``Op``'s perform will do the same operation as::

--- theano/compile/function.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/compile/function.py	2019-08-22 13:33:40.346557969 -0600

@@ -243,7 +243,7 @@ def function(inputs, outputs=None, mode=

     if name is None:

         # Determine possible file names

-        source_file = re.sub('\.pyc?', '.py', __file__)

+        source_file = re.sub(r'\.pyc?', '.py', __file__)

         compiled_file = source_file + 'c'

         stack = tb.extract_stack()

--- theano/configdefaults.py.orig	2019-08-23 08:49:26.259908947 -0600

+++ theano/configdefaults.py	2019-08-23 08:50:57.173227993 -0600

@@ -1614,7 +1614,7 @@ AddConfigVar("compiledir_format",

 def default_compiledirname():

     formatted = theano.config.compiledir_format % compiledir_format_dict

-    safe = re.sub("[\(\)\s,]+", "_", formatted)

+    safe = re.sub(r"[\(\)\s,]+", "_", formatted)

     return safe

--- theano/gof/op.py.orig	2019-08-22 16:26:01.760679539 -0600

+++ theano/gof/op.py	2019-08-23 10:23:51.024394207 -0600

@@ -280,13 +280,13 @@ class CLinkerOp(CLinkerObject):

             string is the name of a C variable pointing to that input.

             The type of the variable depends on the declared type of

             the input.  There is a corresponding python variable that

-            can be accessed by prepending "py_" to the name in the

+            can be accessed by prepending ```"py_"``` to the name in the

             list.

         outputs : list of strings

             Each string is the name of a C variable where the Op should

             store its output.  The type depends on the declared type of

             the output.  There is a corresponding python variable that

-            can be accessed by prepending "py_" to the name in the

+            can be accessed by prepending ```"py_"``` to the name in the

             list.  In some cases the outputs will be preallocated and

             the value of the variable may be pre-filled.  The value for

             an unallocated output is type-dependent.

@@ -343,13 +343,13 @@ class CLinkerOp(CLinkerObject):

             string is the name of a C variable pointing to that input.

             The type of the variable depends on the declared type of

             the input. There is a corresponding python variable that

-            can be accessed by prepending "py_" to the name in the

+            can be accessed by prepending ```"py_"``` to the name in the

             list.

         outputs : list of strings

             Each string is the name of a C variable correspoinding to

             one of the outputs of the Op. The type depends on the

             declared type of the output. There is a corresponding

-            python variable that can be accessed by prepending "py_" to

+            python variable that can be accessed by prepending ```"py_"``` to

             the name in the list.

         sub : dict of strings

             extra symbols defined in `CLinker` sub symbols (such as 'fail').

@@ -784,7 +784,7 @@ class PureOp(object):

         folded when all its inputs are constant. This allows it to

         choose where it puts its memory/speed trade-off. Also, it

         could make things faster as constants can't be used for inplace

-        operations (see *IncSubtensor).

+        operations (see *IncSubtensor*).

         """

         return True

--- theano/gof/opt.py.orig	2019-08-22 14:06:43.820896086 -0600

+++ theano/gof/opt.py	2019-08-23 08:50:57.175227956 -0600

@@ -112,9 +112,9 @@ class Optimizer(object):

         Add features to the fgraph that are required to apply the optimization.

         For example:

-          fgraph.attach_feature(History())

-          fgraph.attach_feature(MyFeature())

-          etc.

+        fgraph.attach_feature(History())

+        fgraph.attach_feature(MyFeature())

+        etc.

         """

         pass

--- theano/gof/unify.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/gof/unify.py	2019-08-22 13:24:58.420815328 -0600

@@ -322,7 +322,7 @@ def unify_walk(a, b, U):

 @comm_guard(OrVariable, NotVariable)  # noqa

 def unify_walk(o, n, U):

-    """

+    r"""

     OrV(list1) == NV(list2) == OrV(list1 \ list2)

     """

--- theano/gof/utils.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/gof/utils.py	2019-08-22 13:22:15.449758115 -0600

@@ -307,7 +307,7 @@ def uniq(seq):

 def difference(seq1, seq2):

-    """

+    r"""

     Returns all elements in seq1 which are not in seq2: i.e ``seq1\seq2``.

     """

--- theano/gpuarray/fft.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/gpuarray/fft.py	2019-08-22 13:36:08.241971287 -0600

@@ -283,7 +283,7 @@ cuirfft_op = CuIRFFTOp()

 def curfft(inp, norm=None):

-    """

+    r"""

     Performs the fast Fourier transform of a real-valued input on the GPU.

     The input must be a real-valued float32 variable of dimensions (m, ..., n).

@@ -317,7 +317,7 @@ def curfft(inp, norm=None):

 def cuirfft(inp, norm=None, is_odd=False):

-    """

+    r"""

     Performs the inverse fast Fourier Transform with real-valued output on the GPU.

     The input is a variable of dimensions (m, ..., n//2+1, 2) with

--- theano/sandbox/linalg/ops.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/sandbox/linalg/ops.py	2019-08-22 13:37:11.588863356 -0600

@@ -199,7 +199,7 @@ theano.compile.mode.optdb.register('Hint

 def psd(v):

-    """

+    r"""

     Apply a hint that the variable `v` is positive semi-definite, i.e.

     it is a symmetric matrix and :math:`x^T A x \ge 0` for any vector x.

--- theano/sparse/basic.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/sparse/basic.py	2019-08-22 16:27:29.828134864 -0600

@@ -4219,9 +4219,9 @@ class ConstructSparseFromList(gof.Op):

         This create a sparse matrix with the same shape as `x`. Its

         values are the rows of `values` moved. Pseudo-code::

-            output = csc_matrix.zeros_like(x, dtype=values.dtype)

-            for in_idx, out_idx in enumerate(ilist):

-                output[out_idx] = values[in_idx]

+        output = csc_matrix.zeros_like(x, dtype=values.dtype)

+        for in_idx, out_idx in enumerate(ilist):

+            output[out_idx] = values[in_idx]

         """

         x_ = theano.tensor.as_tensor_variable(x)

--- theano/tensor/fft.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/fft.py	2019-08-22 13:36:23.986695912 -0600

@@ -117,7 +117,7 @@ irfft_op = IRFFTOp()

 def rfft(inp, norm=None):

-    """

+    r"""

     Performs the fast Fourier transform of a real-valued input.

     The input must be a real-valued variable of dimensions (m, ..., n).

@@ -151,7 +151,7 @@ def rfft(inp, norm=None):

 def irfft(inp, norm=None, is_odd=False):

-    """

+    r"""

     Performs the inverse fast Fourier Transform with real-valued output.

     The input is a variable of dimensions (m, ..., n//2+1, 2)

--- theano/tensor/nlinalg.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/nlinalg.py	2019-08-22 13:34:43.660450608 -0600

@@ -75,7 +75,7 @@ pinv = MatrixPinv()

 class MatrixInverse(Op):

-    """Computes the inverse of a matrix :math:`A`.

+    r"""Computes the inverse of a matrix :math:`A`.

     Given a square matrix :math:`A`, ``matrix_inverse`` returns a square

     matrix :math:`A_{inv}` such that the dot product :math:`A \cdot A_{inv}`

@@ -149,7 +149,7 @@ matrix_inverse = MatrixInverse()

 def matrix_dot(*args):

-    """ Shorthand for product between several dots.

+    r""" Shorthand for product between several dots.

     Given :math:`N` matrices :math:`A_0, A_1, .., A_N`, ``matrix_dot`` will

     generate the matrix product between all in the given order, namely

--- theano/tensor/nnet/conv.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/nnet/conv.py	2019-08-22 13:35:27.978675488 -0600

@@ -157,7 +157,7 @@ def conv2d(input, filters, image_shape=N

 class ConvOp(OpenMPOp):

-    """

+    r"""

     This Op serves a dual purpose: it can implement a vanilla 2D convolution

     (as taught in any signal processing class) or implement the

     convolutional layers found in Convolutional Neural Networks.

--- theano/tensor/nnet/neighbours.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/nnet/neighbours.py	2019-08-22 13:36:36.213482065 -0600

@@ -618,7 +618,7 @@ class Images2Neibs(Op):

 def images2neibs(ten4, neib_shape, neib_step=None, mode='valid'):

-    """

+    r"""

     Function :func:`images2neibs <theano.tensor.nnet.neighbours.images2neibs>`

     allows to apply a sliding window operation to a tensor containing

     images or other two-dimensional objects.

--- theano/tensor/nnet/nnet.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/nnet/nnet.py	2019-08-22 13:35:17.270862767 -0600

@@ -409,7 +409,7 @@ softmax_grad = SoftmaxGrad()

 class Softmax(gof.Op):

-    """

+    r"""

     Softmax activation function

     :math:`\\varphi(\\mathbf{x})_j =

     \\frac{e^{\mathbf{x}_j}}{\sum_{k=1}^K e^{\mathbf{x}_k}}`

@@ -600,7 +600,7 @@ softmax_op = Softmax()

 class LogSoftmax(gof.Op):

-    """

+    r"""

     LogSoftmax activation function

     :math:`\\varphi(\\mathbf{x})_j =

     \\e^{(\mathbf{x}_j - log{\sum_{k=1}^K e^{\mathbf{x}_k})}}

@@ -1412,7 +1412,7 @@ crossentropy_categorical_1hot_grad = Cro

 class CrossentropyCategorical1Hot(gof.Op):

-    """

+    r"""

     Compute the cross entropy between a coding distribution and

     a true distribution of the form [0, 0, ... 0, 1, 0, ..., 0].

@@ -2051,7 +2051,7 @@ def sigmoid_binary_crossentropy(output,

 def categorical_crossentropy(coding_dist, true_dist):

-    """

+    r"""

     Return the cross-entropy between an approximating distribution and a true

     distribution.

--- theano/tensor/opt.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/opt.py	2019-08-22 13:34:16.709921972 -0600

@@ -4582,7 +4582,7 @@ register_canonicalize(gof.OpRemove(T.ten

 class Canonizer(gof.LocalOptimizer):

-    """

+    r"""

     Simplification tool. The variable is a local_optimizer. It is best used

     with a TopoOptimizer in in_to_out order.

@@ -4650,7 +4650,7 @@ class Canonizer(gof.LocalOptimizer):

         return [self.main, self.inverse, self.reciprocal]

     def get_num_denum(self, input):

-        """

+        r"""

         This extract two lists, num and denum, such that the input is:

         self.inverse(self.main(\*num), self.main(\*denum)). It returns

         the two lists in a (num, denum) pair.

@@ -4751,7 +4751,7 @@ class Canonizer(gof.LocalOptimizer):

         return num, denum

     def merge_num_denum(self, num, denum):

-        """

+        r"""

         Utility function which takes two lists, num and denum, and

         returns something which is equivalent to inverse(main(\*num),

         main(\*denum)), but depends on the length of num and the length

--- theano/tensor/slinalg.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/slinalg.py	2019-08-22 13:36:53.541179005 -0600

@@ -266,7 +266,7 @@ class Solve(Op):

             return [(rows, cols)]

     def L_op(self, inputs, outputs, output_gradients):

-        """

+        r"""

         Reverse-mode gradient updates for matrix solve operation c = A \\\ b.

         Symbolic expression for updates taken from [#]_.

--- theano/tensor/tests/mlp_test.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ theano/tensor/tests/mlp_test.py	2019-08-22 16:30:23.307092085 -0600

@@ -93,7 +93,7 @@ class LogisticRegression(object):

         self.params = [self.W]

     def negative_log_likelihood(self, y):

-        """Return the mean of the negative log-likelihood of the prediction

+        r"""Return the mean of the negative log-likelihood of the prediction

         of this model under a given target distribution.

         .. math::

--- versioneer.py.orig	2019-01-15 14:13:57.000000000 -0700

+++ versioneer.py	2019-08-22 16:19:43.175333211 -0600

@@ -418,7 +418,7 @@ def run_command(commands, args, cwd=None

     return stdout, p.returncode

-LONG_VERSION_PY['git'] = '''

+LONG_VERSION_PY['git'] = r'''

 # This file helps to compute a version number in source trees obtained from

 # git-archive tarball (such as those provided by githubs download-from-tag

 # feature). Distribution tarballs (built by setup.py sdist) and build