lenskit / csr

Compressed sparse matrices

Home Page:https://csr.lenskit.org

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CSR Matrices for Python

codecov

This package provides an implementation of sparse matrices in compressed sparse row format for Python. Routines are implemented with Numba, and both the CSR data structure and most related matrix and vector operations can be used from Numba's nopython mode.

Right now the feature set is very limited --- don't expect this to be a drop-in replacement for SciPy's sparse matrices. Some features we expect to develop as people take an interest in the package and contribute updates, but we also hope to keep a relatively tight scope. This package aims to provide efficient support for compressed sparse row matrices, with some routines that can also treat them as CSC matrices.

About

Compressed sparse matrices

https://csr.lenskit.org

License:MIT License


Languages

Language:Python 94.2%Language:C 5.8%