Home

In this wiki you will eventually find some information relating to the inner workings of SpPy and how to develop the code. The core code is written primarily in Cython, and makes use of the Eigen library.

Compressed sparse matrices are wrapped using csarray_base.pyx (csarray1d_base.pyx is the equivalent class for vectors), and this class is a template for specific storage types (int, float etc.) and storage order (column or row major). The template class is specialised using TemplateExpander.py which generates a file called csarray_sub.py. Classes in both csarray_sub and csarray1d_sub are wrapped in the Python class csarray.

The directories are organised as follows:

• sppy/ contains csarray, csarray1d_base and csarray_base
• sppy/include contains C++ header files specialising those from Eigen
• sppy/lib contains useful utility functions
• sppy/linalg contains linear algebra routines such as norm, Singular Value Decompositions etc.
• sppy/test contains the test cases for the library