Extending PyTensor with a C Op

This tutorial covers how to extend PyTensor with an Op that offers a C implementation. This tutorial is aimed at individuals who already know how to extend PyTensor (see tutorial Creating a new Op: Python implementation) 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 an PyTensor 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. Listed below are the two macros most often used in PyTensor C Ops.

void Py_XINCREF(PyObject *o)

Increments the reference count of object o. Without effect if the object is NULL.

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 transferring 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 transferring 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 PyTensor and so extending PyTensor 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 an PyTensor C Op. The full documentation for the API can be found here : NumPy C-API.

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.

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 appropriately manage the reference counts of PyArrayObject instances.

Unlike in a standard multidimensional 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

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 ndarray’s data and properties

The following macros serve to access various attributes of NumPy ndarrays.

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.

int PyArray_NDIM(PyArrayObject* arr)

Returns the number of dimensions in the the array pointed by arr

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.

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.

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.

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.

npy_intp PyArray_SIZE(PyArrayObject* arr)

Returns to total number of elements in the array

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 :

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.

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.

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 an PyTensor 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, PyTensor 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, PyTensor 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 PyTensor itself takes care of compiling the C code, instead of the individual Ops, allows PyTensor to easily cache the compiled C code. This allows for faster compilation times.

The following are some of the various methods of the class COp that are related to the C implementation:

• The methods CLinkerObject.c_libraries() and CLinkerObject.c_lib_dirs() to allow your Op to use external libraries.

• The method CLinkerOp.c_code_cleanup() to specify how the Op should clean up what it has allocated during its execution.

• The methods COp.c_init_code() and CLinkerOp.c_init_code_apply() to specify code that should be executed once when the module is initialized, before anything else is executed.

• The methods CLinkerObject.c_compile_args() and CLinkerObject.c_no_compile_args() to specify requirements regarding how the Op’s C code should be compiled.

This section describes the methods CLinkerOp.c_code(), CLinkerObject.c_support_code(), Op.c_support_code_apply() and CLinkerObject.c_code_cache_version() because they are the ones that are most commonly used.

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 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:

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 occurred. 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(). The % characters in the format characters need to be escaped since the C code itself is defined in a string which undergoes string formatting.

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.

c_support_code(**kwargs)

Returns a string or a list of strings 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.

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.

Both c_support_code and c_support_code_apply are necessary because an PyTensor Op can be used more than once in a given PyTensor function. For example, an Op that adds two matrices could be used at some point in the PyTensor function to add matrices of integers and, at another point, to add matrices of doubles. Because the dtype of the inputs and outputs can change between different applies of the Op, any support code that relies on a certain dtype is specific to a given Apply of the Op and should therefore be defined in c_support_code_apply.

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 PyTensor 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 PyTensor 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:

PyTensor 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.

Also, do not use the built-in hash; it will produce different values between Python sessions and confound the caching process.

Important restrictions when implementing a COp

There are some important restrictions to remember when implementing an COp. Unless your COp correctly defines a view_map attribute, the perform and c_code must not produce outputs whose memory is aliased to any input (technically, if changing the output could change the input object in some sense, they are aliased). Unless your COp correctly defines a destroy_map attribute, perform and c_code must not modify any of the inputs.

TODO: EXPLAIN DESTROYMAP and VIEWMAP BETTER AND GIVE EXAMPLE.

When developing a COp, you should run computations in DebugMode, by using argument mode='DebugMode' to pytensor.function. DebugMode is slow, but it can catch many common violations of the Op contract.

TODO: Like what? How? Talk about Python vs. C too.

DebugMode is no silver bullet though. For example, if you modify an Op self.* during any of make_node, perform, or c_code, you are probably doing something wrong but DebugMode will not detect this.

TODO: jpt: I don’t understand the following sentence.

Ops and Types should usually be considered immutable – you should definitely not make a change that would have an impact on __eq__, __hash__, or the mathematical value that would be computed by perform or c_code.

Simple COp 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. PyTensor 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. PyTensor 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, PyTensor 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.

import numpy
import pytensor

from pytensor.link.c.op import COp
from pytensor.graph.basic import Apply


class VectorTimesScalar(COp):
    __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 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()

The c_code method accepts variable names as arguments (name, inp, out, sub) and returns a C code fragment that computes the expression output. In case of error, the %(fail)s statement cleans up and returns properly.

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 accommodate 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.

class VectorTimesVector(COp):
    __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 = pytensor.tensor.type.TensorType(
                        dtype=pytensor.scalar.upcast(x.dtype, y.dtype),
                        shape=(None,))()

        return Apply(self, [x, y], [output_var])

    def c_code_cache_version(self):
        return (1, 0, 2)

    def c_support_code(self, **kwargs):
        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()

Using GDB to debug COp’s C code

When debugging C code, it can be useful to use GDB for code compiled by PyTensor.

For this, you must enable this PyTensor: cmodule__remove_gxx_opt=True.

Then you must start Python inside GDB and in it start your Python process:

$gdb python
(gdb)r pytest pytensor/

Quick guide to GDB.

Final Note

This tutorial focuses on providing C implementations to COp`s that manipulate PyTensor tensors. For more information about other PyTensor types, you can refer to the section :ref:`Alternate PyTensor Types.