This extension provides the Python bindings for the ViennaCL linear algebra and numerical computation library for GPGPU and heterogeneous systems. ViennaCL itself is a header-only C++ library, so these bindings make available to Python programmers ViennaCL’s fast OpenCL and CUDA algorithms, in a way that is idiomatic and compatible with the Python community’s most popular scientific packages, NumPy and SciPy.
PyViennaCL is divided into five submodules, of which four are designed for direct use by users:
- pycore: user-friendly classes for representing the main ViennaCL objects, such as Matrix or Vector;
- linalg: an explicit interface to a number of ViennaCL’s linear algebra routines, such as matrix solvers and eigenvalue computation;
- vclmath: convenience functions akin to the standard math module (you can also access it as pyviennacl.math);
- util: utility functions, such as to construct an appropriate ViennaCL object from an ndarray (Matrix or Vector), or to provide basic debug logging;
- _viennacl: a raw C++ interface to ViennaCL, with no stable API.
Nonetheless, all of PyViennaCL’s functionality is available from the top-level pyviennacl namespace. So, if you want help on the Matrix class, you can just run:
>>> import pyviennacl as p
>>> help(p.Matrix)
For help on PyViennaCL’s core functionality in general, or PyViennaCL’s high-level linear algebra functions, run:
>>> help(p.pycore)
or:
>>> help(p.linalg)