Skip to content

P
pytensor
提交 7d286d04 authored 作者: Frédéric Bastien's avatar Frédéric Bastien
浏览文件
操作

Merge pull request #2121 from carriepl/doc_for_c_op

Tutorial for C op
上级 c7e88a00 3597e074
显示空白字符变更
内嵌 并排
正在显示 包含 667 行增加 和 0 行删除
doc/tutorial/extending_theano_c.txt 0 → 100644
浏览文件 @ 7d286d04
.. _extending_theano_c:
============================
Extending Theano with a C Op
============================
This tutorial covers how to extend Theano with an op that offers a C
implementation. It does not cover ops that run on a GPU but it does introduce
many elements and concepts which are relevant for GPU ops. This tutorial is
aimed at individuals who already know how to extend Theano (see tutorial
:ref:`extending_theano`) by adding a new op with a Python implementation
and will only cover the additional knowledge required to also produce ops
with C implementations.
Providing a Theano op with a C implementation requires to interact with
Python's C-API and Numpy's C-API. Thus, the first step of this tutorial is to
introduce both and highlight their features which are most relevant to the
task of implementing a C op. This tutorial then introduces the most important
methods that the op needs to implement in order to provide a usable C
implementation. Finally, it shows how to combine these elements to write a
simple C op for performing the simple task of multiplying every element in a
vector by a scalar.
Python C-API
============
Python provides a C-API to allows the manipulation of python objects from C
code. In this API, all variables that represent Python objects are of type
``PyObject *``. All objects have a pointer to their type object and a reference
count field (that is shared with the python side). Most python methods have
an equivalent C function that can be called on the ``PyObject *`` pointer.
As such, manipulating a PyObject instance is often straight-forward but it
is important to properly manage its reference count. Failing to do so can
lead to undesired behavior in the C code.
Reference counting
------------------
Reference counting is a mechanism for keeping track, for an object, of
the number of references to it held by other entities. This mechanism is often
used for purposes of garbage collecting because it allows to easily see if
an object is still being used by other entities. When the reference count
for an object drops to 0, it means it is not used by anyone any longer and can
be safely deleted.
PyObjects implement reference counting and the Python C-API defines a number
of macros to help manage those reference counts. The definition of these
macros can be found here : `Python C-API Reference Counting
<https://docs.python.org/2/c-api/refcounting.html>`_. Listed below are the
two macros most often used in Theano C ops.
.. method:: void Py_XINCREF(PyObject *o)
Increments the reference count of object ``o``. Without effect if the
object is NULL.
.. method:: void Py_XDECREF(PyObject *o)
Decrements the reference count of object ``o``. If the reference count
reaches 0, it will trigger a call of the object's deallocation function.
Without effect if the object is NULL.
The general principle, in the reference counting paradigm, is that the owner
of a reference to an object is responsible for disposing properly of it.
This can be done by decrementing the reference count once the reference is no
longer used or by transfering ownership; passing on the reference to a new
owner which becomes responsible for it.
Some functions return "borrowed references"; this means that they return a
reference to an object **without** transfering ownership of the reference to the
caller of the function. This means that if you call a function which returns a
borrowed reference, you do not have the burden of properly disposing of that
reference. You should **not** call Py_XDECREF() on a borrowed reference.
Correctly managing the reference counts is important as failing to do so can
lead to issues ranging from memory leaks to segmentation faults.
NumPy C-API
===========
The NumPy library provides a C-API to allow users to create, access and
manipulate NumPy arrays from within their own C routines. NumPy's ndarrays
are used extensively inside Theano and so extending Theano with a C op will
require interaction with the NumPy C-API.
This sections covers the API's elements that are often required to write code
for a Theano C op. The full documentation for the API can be found here :
`NumPy C-API <http://docs.scipy.org/doc/numpy/reference/c-api.html>`_.
NumPy data types
----------------
To allow portability between platforms, the NumPy C-API defines its own data
types which should be used whenever you are manipulating a NumPy array's
internal data. The data types most commonly used to implement C ops are the
following : ``npy_int{8,16,32,64}``, ``npy_uint{8,16,32,64}`` and
``npy_float{32,64}``.
You should use these data types when manipulating a NumPy array's internal
data instead of C primitives because the size of the memory representation
for C primitives can vary between platforms. For instance, a C ``long`` can be
represented in memory with 4 bytes but it can also be represented with 8.
On the other hand, the in-memory size of NumPy data types remains constant
across platforms. Using them will make your code simpler and more portable.
The full list of defined data types can be found here :
`NumPy C-API data types
<http://docs.scipy.org/doc/numpy/reference/c-api.dtype.html#c-type-names>`_.
NumPy ndarrays
--------------
In the NumPy C-API, NumPy arrays are represented as instances of the
PyArrayObject class which is a descendant of the PyObject class. This means
that, as for any other Python object that you manipulate from C code, you
need to appropriatedly manage the reference counts of PyArrayObject instances.
Unlike in a standard multidimensionnal C array, a NumPy array's internal data
representation does not have to occupy a continuous region in memory. In fact,
it can be C-contiguous, F-contiguous or non-contiguous. C-contiguous means
that the data is not only contiguous in memory but also that it is organized
such that the index of the latest dimension changes the fastest. If the
following array
.. code-block:: python
x = [[1, 2, 3],
[4, 5, 6]]
is C-contiguous, it means that, in memory, the six values contained in the
array ``x`` are stored in the order ``[1, 2, 3, 4, 5, 6]`` (the first value is
``x[0,0]``, the second value is ``x[0,1]``, the third value is ``x[0,2]``, the,
fourth value is ``x[1,0]``, etc). F-contiguous (or Fortran Contiguous) also
means that the data is contiguous but that it is organized such that the index
of the latest dimension changes the slowest. If the array ``x`` is
F-contiguous, it means that, in memory, the values appear in the order
``[1, 4, 2, 5, 3, 6]`` (the first value is ``x[0,0]``, the second value is
``x[1,0]``, the third value is ``x[0,1]``, etc).
Finally, the internal data can be non-contiguous. In this case, it occupies
a non-contiguous region in memory but it is still stored in an organized
fashion : the distance between the element ``x[i,j]`` and the element
``x[i+1,j]`` of the array is constant over all valid values of ``i`` and
``j``, just as the distance between the element ``x[i,j]`` and the element
``x[i,j+1]`` of the array is constant over all valid values of ``i`` and ``j``.
This distance between consecutive elements of an array over a given dimension,
is called the stride of that dimension.
Accessing NumPy ndarrays' data and properties
---------------------------------------------
The following macros serve to access various attributes of NumPy ndarrays.
.. method:: void* PyArray_DATA(PyArrayObject* arr)
Returns a pointer to the first element of the array's data. The returned
pointer must be cast to a pointer of the proper Numpy C-API data type
before use.
.. method:: int PyArray_NDIM(PyArrayObject* arr)
Returns the number of dimensions in the the array pointed by ``arr``
.. method:: npy_intp* PyArray_DIMS(PyArrayObject* arr)
Returns a pointer on the first element of ``arr``'s internal array
describing its dimensions. This internal array contains as many elements
as the array ``arr`` has dimensions.
The macro ``PyArray_SHAPE()`` is a synonym of ``PyArray_DIMS()`` : it has
the same effect and is used in an identical way.
.. method:: npy_intp* PyArray_STRIDES(PyArrayObject* arr)
Returns a pointer on the first element of ``arr``'s internal array
describing the stride for each of its dimension. This array has as many
elements as the number of dimensions in ``arr``. In this array, the
strides are expressed in number of bytes.
.. method:: PyArray_Descr* PyArray_DESCR(PyArrayObject* arr)
Returns a reference to the object representing the dtype of the array.
The macro ``PyArray_DTYPE()`` is a synonym of the ``PyArray_DESCR()`` : it
has the same effect and is used in an identical way.
:note:
This is a borrowed reference so you do not need to decrement its
reference count once you are done with it.
.. method:: int PyArray_TYPE(PyArrayObject* arr)
Returns the typenumber for the elements of the array. Like the dtype, the
typenumber is a descriptor for the type of the data in the array. However,
the two are not synonyms and, as such, cannot be used in place of the
other.
.. method:: npy_intp PyArray_SIZE(PyArrayObject* arr)
Returns to total number of elements in the array
.. method:: bool PyArray_CHKFLAGS(PyArrayObject* arr, flags)
Returns true if the array has the specified flags. The variable flag
should either be a NumPy array flag or an integer obtained by applying
bitwise or to an ensemble of flags.
The flags that can be used in with this macro are :
NPY_ARRAY_C_CONTIGUOUS, NPY_ARRAY_F_CONTIGUOUS, NPY_ARRAY_OWNDATA,
NPY_ARRAY_ALIGNED, NPY_ARRAY_WRITEABLE, NPY_ARRAY_UPDATEIFCOPY.
Creating NumPy ndarrays
-----------------------
The following functions allow the creation and copy of NumPy arrays :
.. method:: PyObject* PyArray_EMPTY(int nd, npy_intp* dims, typenum dtype,
int fortran)
Constructs a new ndarray with the number of dimensions specified by
``nd``, shape specified by ``dims`` and data type specified by ``dtype``.
If ``fortran`` is equal to 0, the data is organized in a C-contiguous
layout, otherwise it is organized in a F-contiguous layout. The array
elements are not initialized in any way.
The function ``PyArray_Empty()`` performs the same function as the macro
``PyArray_EMPTY()`` but the data type is given as a pointer to a
``PyArray_Descr`` object instead of a ``typenum``.
.. method:: PyObject* PyArray_ZEROS(int nd, npy_intp* dims, typenum dtype,
int fortran)
Constructs a new ndarray with the number of dimensions specified by
``nd``, shape specified by ``dims`` and data type specified by ``dtype``.
If ``fortran`` is equal to 0, the data is organized in a C-contiguous
layout, otherwise it is organized in a F-contiguous layout. Every element
in the array is initialized to 0.
The function ``PyArray_Zeros()`` performs the same function as the macro
``PyArray_ZEROS()`` but the data type is given as a pointer to a
``PyArray_Descr`` object instead of a ``typenum``.
.. method:: PyArrayObject* PyArray_GETCONTIGUOUS(PyObject* op)
Returns a C-contiguous and well-behaved copy of the array op. If op is
already C-contiguous and well-behaved, this function simply returns a
new reference to op.
Methods the C Op needs to define
================================
There is a key difference between an op defining a Python implementation for
its computation and defining a C implementation. In the case of a Python
implementation, the op defines a function ``perform()`` which executes the
required Python code to realize the op. In the case of a C implementation,
however, the op does **not** define a function that will execute the C code; it
instead defines functions that will **return** the C code to the caller.
This is because calling C code from Python code comes with a significant
overhead. If every op was responsible for executing its own C code, every
time a Theano function was called, this overhead would occur as many times
as the number of ops with C implementations in the function's computational
graph.
To maximize performance, Theano instead requires the C ops to simply return
the code needed for their execution and takes upon itself the task of
organizing, linking and compiling the code from the various ops. Through this,
Theano is able to minimize the number of times C code is called from Python
code.
The following is a very simple example to illustrate how it's possible to
obtain performance gains with this process. Suppose you need to execute,
from Python code, 10 different ops, each one having a C implementation. If
each op was responsible for executing its own C code, the overhead of
calling C code from Python code would occur 10 times. Consider now the case
where the ops instead return the C code for their execution. You could get
the C code from each op and then define your own C module that would call
the C code from each op in succession. In this case, the overhead would only
occur once; when calling your custom module itself.
Moreover, the fact that Theano itself takes care of compiling the C code,
instead of the individual ops, allows Theano to easily cache the compiled C
code. This allows for faster compilation times.
See :ref:`cop` for the full documentation of the various methods of the
class Op that are related to the C implementation. Of particular interest are:
* The methods :meth:`Op.c_libraries` and :meth:`Op.c_lib_dirs` to allow
your op to use external libraries.
* The method :meth:`Op.c_code_cleanup` to specify how the op should
clean up what it has allocated during its execution.
* The methods :meth:`Op.c_init_code` and :meth:`Op.c_init_code_apply`
to specify code that should be executed once when the module is
initialized, before anything else is executed.
* The methods :meth:`Op.c_compile_args` and
:meth:`Op.c_no_compile_args` to specify requirements regarding how
the op's C code should be compiled.
This section describes the methods :meth:`Op.c_code`,
:meth:`Op.c_support_code`, :meth:`Op.c_support_code_apply` and
:meth:`Op.c_code_cache_version` because they are the ones that are most
commonly used.
.. method:: c_code(node, name, input_names, output_names, sub)
This method returns a string containing the C code to perform the
computation required by this op.
The ``node`` argument is an :ref:`apply` node representing an
application of the current Op on a list of inputs, producing a list of
outputs.
``input_names`` is a sequence of strings which contains as many strings
as the op has inputs. Each string contains the name of the C variable
to which the corresponding input has been assigned. For example, the name
of the C variable representing the first input of the op is given by
``input_names[0]``. You should therefore use this name in your
C code to interact with that variable. ``output_names`` is used
identically to ``input_names``, but for the op's outputs.
Finally, ``sub`` is a dictionary of extras parameters to the c_code
method. Among other things, it contains ``sub['fail']`` which is a string
of C code that you should include in your C code (after ensuring that a
Python exception is set) if it needs to raise an exception. Ex:
.. code-block:: python
c_code = """
PyErr_Format(PyExc_ValueError, "X does not have the right value");
%(fail)s;
""" % {'fail' : sub['fail']}
to raise a ValueError Python exception with the specified message.
The function ``PyErr_Format()`` supports string formatting so it is
possible to tailor the error message to the specifics of the error
that occured. If ``PyErr_Format()`` is called with more than two
arguments, the subsequent arguments are used to format the error message
with the same behavior as the function `PyString_FromFormat()
<https://docs.python.org/2/c-api/string.html#c.PyString_FromFormat>`_. The
``%`` characters in the format characters need to be escaped since the C
code itself is defined in a string which undergoes string formatting.
.. code-block:: python
c_code = """
PyErr_Format(PyExc_ValueError,
"X==%%i but it should be greater than 0", X);
%(fail)s;
""" % {'fail' : sub['fail']}
:note:
Your C code should not return the output of the computation but
rather put the results in the C variables whose names are contained in
the ``output_names``.
.. method:: c_support_code()
Returns a string containing some support C code for this op. This code
will be included at the global scope level and can be used to define
functions and structs that will be used by every apply of this op.
.. method:: c_support_code_apply(node, name)
Returns a string containing some support C code for this op. This code
will be included at the global scope level and can be used to define
functions and structs that will be used by this op. The difference between
this method and ``c_support_code()`` is that the C code specified in
``c_support_code_apply()`` should be specific to each apply of the Op,
while ``c_support_code()`` is for support code that is not specific to
each apply.
.. method:: c_code_cache_version()
Returns a tuple of integers representing the version of the C code in this
op. Ex : (1, 4, 0) for version 1.4.0
This tuple is used by Theano to cache the compiled C code for this op. As
such, the return value **MUST BE CHANGED** every time the C code is altered
or else Theano will disregard the change in the code and simply load a
previous version of the op from the cache. If you want to avoid caching of
the C code of this op, return an empty tuple or do not implement this
method.
:note:
Theano can handle tuples of any hashable objects as return values
for this function but, for greater readability and easier management,
this function should return a tuple of integers as previously
described.
Simple C Op example
===================
In this section, we put together the concepts that were covered in this
tutorial to generate an op which multiplies every element in a vector
by a scalar and returns the resulting vector. This is intended to be a simple
example so the methods ``c_support_code()`` and ``c_support_code_apply()`` are
not used because they are not required.
In the C code below notice how the reference count on the output variable is
managed. Also take note of how the new variables required for the op's
computation are declared in a new scope to avoid cross-initialization errors.
Also, in the C code, it is very important to properly validate the inputs
and outputs storage. Theano guarantees that the inputs exist and have the
right number of dimensions but it does not guarantee their exact shape. For
instance, if an op computes the sum of two vectors, it needs to validate that
its two inputs have the same shape. In our case, we do not need to validate
the exact shapes of the inputs because we don't have a need that they match
in any way.
For the outputs, things are a little bit more subtle. Theano does not
guarantee that they have been allocated but it does guarantee that, if they
have been allocated, they have the right number of dimension. Again, Theano
offers no guarantee on the exact shapes. This means that, in our example, we
need to validate that the output storage has been allocated and has the same
shape as our vector input. If it is not the case, we allocate a new output
storage with the right shape and number of dimensions.
.. code-block:: python
import numpy
import theano
from theano import gof
import theano.tensor as T
class VectorTimesScalar(gof.Op):
__props__ = ()
def make_node(self, x, y):
# Validate the inputs' type
if x.type.ndim != 1:
raise TypeError('x must be a 1-d vector')
if y.type.ndim != 0:
raise TypeError('y must be a scalar')
# Create an output variable of the same type as x
output_var = x.type()
return gof.Apply(self, [x, y], [output_var])
def c_code_cache_version(self):
return (1, 0)
def c_code(self, node, name, inp, out, sub):
x, y = inp
z, = out
# Extract the dtypes of the inputs and outputs storage to
# be able to declare pointers for those dtypes in the C
# code.
dtype_x = node.inputs[0].dtype
dtype_y = node.inputs[1].dtype
dtype_z = node.outputs[0].dtype
itemsize_x = numpy.dtype(dtype_x).itemsize
itemsize_z = numpy.dtype(dtype_z).itemsize
fail = sub['fail']
c_code = """
// Validate that the output storage exists and has the same
// dimension as x.
if (NULL == %(z)s ||
PyArray_DIMS(%(x)s)[0] != PyArray_DIMS(%(z)s)[0])
{
/* Reference received to invalid output variable.
Decrease received reference's ref count and allocate new
output variable */
Py_XDECREF(%(z)s);
%(z)s = (PyArrayObject*)PyArray_EMPTY(1,
PyArray_DIMS(%(x)s),
PyArray_TYPE(%(x)s),
0);
if (!%(z)s) {
%(fail)s;
}
}
// Perform the vector multiplication by a scalar
{
/* The declaration of the following variables is done in a new
scope to prevent cross initialization errors */
npy_%(dtype_x)s* x_data_ptr =
(npy_%(dtype_x)s*)PyArray_DATA(%(x)s);
npy_%(dtype_z)s* z_data_ptr =
(npy_%(dtype_z)s*)PyArray_DATA(%(z)s);
npy_%(dtype_y)s y_value =
((npy_%(dtype_y)s*)PyArray_DATA(%(y)s))[0];
int x_stride = PyArray_STRIDES(%(x)s)[0] / %(itemsize_x)s;
int z_stride = PyArray_STRIDES(%(z)s)[0] / %(itemsize_z)s;
int x_dim = PyArray_DIMS(%(x)s)[0];
for(int i=0; i < x_dim; i++)
{
z_data_ptr[i * z_stride] = (x_data_ptr[i * x_stride] *
y_value);
}
}
"""
return c_code % locals()
More complex C Op example
=========================
This section introduces a new example, slightly more complex than the previous
one, with an op to perform an element-wise multiplication between the elements
of two vectors. This new example differs from the previous one in its use
of the methods ``c_support_code()`` and ``c_support_code_apply()`` (it does
not `need` to use them but it does so to explain their use) and its capacity
to support inputs of different dtypes.
Recall the method ``c_support_code()`` is meant to produce code that will
be used for every apply of the op. This means that the C code in this
method must be valid in every setting your op supports. If the op is meant
to supports inputs of various dtypes, the C code in this method should be
generic enough to work with every supported dtype. If the op operates on
inputs that can be vectors or matrices, the C code in this method should
be able to accomodate both kinds of inputs.
In our example, the method ``c_support_code()`` is used to declare a C
function to validate that two vectors have the same shape. Because our
op only supports vectors as inputs, this function is allowed to rely
on its inputs being vectors. However, our op should support multiple
dtypes so this function cannot rely on a specific dtype in its inputs.
The method ``c_support_code_apply()``, on the other hand, is allowed
to depend on the inputs to the op because it is apply-specific. Therefore, we
use it to define a function to perform the multiplication between two vectors.
Variables or functions defined in the method ``c_support_code_apply()`` will
be included at the global scale for every apply of the Op. Because of this,
the names of those variables and functions should include the name of the op,
like in the example. Otherwise, using the op twice in the same graph will give
rise to conflicts as some elements will be declared more than once.
The last interesting difference occurs in the ``c_code()`` method. Because the
dtype of the output is variable and not guaranteed to be the same as any of
the inputs (because of the upcast in the method ``make_node()``), the typenum
of the output has to be obtained in the Python code and then included in the
C code.
.. code-block:: python
class VectorTimesVector(gof.Op):
__props__ = ()
def make_node(self, x, y):
# Validate the inputs' type
if x.type.ndim != 1:
raise TypeError('x must be a 1-d vector')
if y.type.ndim != 1:
raise TypeError('y must be a 1-d vector')
# Create an output variable of the same type as x
output_var = theano.tensor.TensorType(
dtype=theano.scalar.upcast(x.dtype, y.dtype),
broadcastable=[False])()
return gof.Apply(self, [x, y], [output_var])
def c_code_cache_version(self):
return (1, 0, 2)
def c_support_code(self):
c_support_code = """
bool vector_same_shape(PyArrayObject* arr1,
PyArrayObject* arr2)
{
return (PyArray_DIMS(arr1)[0] == PyArray_DIMS(arr2)[0]);
}
"""
return c_support_code
def c_support_code_apply(self, node, name):
dtype_x = node.inputs[0].dtype
dtype_y = node.inputs[1].dtype
dtype_z = node.outputs[0].dtype
c_support_code = """
void vector_elemwise_mult_%(name)s(npy_%(dtype_x)s* x_ptr,
int x_str, npy_%(dtype_y)s* y_ptr, int y_str,
npy_%(dtype_z)s* z_ptr, int z_str, int nbElements)
{
for (int i=0; i < nbElements; i++){
z_ptr[i * z_str] = x_ptr[i * x_str] * y_ptr[i * y_str];
}
}
"""
return c_support_code % locals()
def c_code(self, node, name, inp, out, sub):
x, y = inp
z, = out
dtype_x = node.inputs[0].dtype
dtype_y = node.inputs[1].dtype
dtype_z = node.outputs[0].dtype
itemsize_x = numpy.dtype(dtype_x).itemsize
itemsize_y = numpy.dtype(dtype_y).itemsize
itemsize_z = numpy.dtype(dtype_z).itemsize
typenum_z = numpy.dtype(dtype_z).num
fail = sub['fail']
c_code = """
// Validate that the inputs have the same shape
if ( !vector_same_shape(%(x)s, %(y)s))
{
PyErr_Format(PyExc_ValueError, "Shape mismatch : "
"x.shape[0] and y.shape[0] should match but "
"x.shape[0] == %%i and y.shape[0] == %%i",
PyArray_DIMS(%(x)s)[0], PyArray_DIMS(%(y)s)[0]);
%(fail)s;
}
// Validate that the output storage exists and has the same
// dimension as x.
if (NULL == %(z)s || !(vector_same_shape(%(x)s, %(z)s)))
{
/* Reference received to invalid output variable.
Decrease received reference's ref count and allocate new
output variable */
Py_XDECREF(%(z)s);
%(z)s = (PyArrayObject*)PyArray_EMPTY(1,
PyArray_DIMS(%(x)s),
%(typenum_z)s,
0);
if (!%(z)s) {
%(fail)s;
}
}
// Perform the vector elemwise multiplication
vector_elemwise_mult_%(name)s(
(npy_%(dtype_x)s*)PyArray_DATA(%(x)s),
PyArray_STRIDES(%(x)s)[0] / %(itemsize_x)s,
(npy_%(dtype_y)s*)PyArray_DATA(%(y)s),
PyArray_STRIDES(%(y)s)[0] / %(itemsize_y)s,
(npy_%(dtype_z)s*)PyArray_DATA(%(z)s),
PyArray_STRIDES(%(z)s)[0] / %(itemsize_z)s,
PyArray_DIMS(%(x)s)[0]);
"""
return c_code % locals()
doc/tutorial/index.txt
浏览文件 @ 7d286d04
...@@ -43,5 +43,6 @@ you out. ...@@ -43,5 +43,6 @@ you out.
debug_faq debug_faq
profiling profiling
extending_theano extending_theano
extending_theano_c
python-memory-management python-memory-management
multi_cores multi_cores
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论