https://github.com/lenskit/csr

Compressed sparse matrices
https://github.com/lenskit/csr

numba sparse-matrix

Last synced: 3 months ago
JSON representation

Compressed sparse matrices

• Host: GitHub
• URL: https://github.com/lenskit/csr
• Owner: lenskit
• License: mit
• Created: 2021-01-30T22:48:16.000Z (almost 5 years ago)
• Default Branch: main
• Last Pushed: 2024-06-12T14:37:54.000Z (over 1 year ago)
• Last Synced: 2025-09-24T20:44:38.098Z (4 months ago)
• Topics: numba, sparse-matrix
• Language: Python
• Homepage: https://csr.lenskit.org
• Size: 3.97 MB
• Stars: 15
• Watchers: 3
• Forks: 2
• Open Issues: 9
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# CSR Matrices for Python

[![codecov](https://codecov.io/gh/lenskit/csr/branch/main/graph/badge.svg?token=6UQ4O8FCKK)](https://codecov.io/gh/lenskit/csr)

> [!NOTE]

> Development of this package is suspended as of June 2024, as LensKit 2024.x will

> not use it. Small maintenance releases to support LensKit 0.14 may be maintained.

>

> If someone is interested in taking this package over for other purposes, please

> contact @mdekstrand.

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.