The scikit-sparse package (previously known as scikits.sparse) is a companion to the scipy.sparse library for sparse matrix manipulation in Python. All scikit-sparse routines expect and return scipy.sparse matrices (usually in CSC format). The intent of scikit-sparse is to wrap GPL’ed code such as SuiteSparse, which cannot be included in SciPy proper.

Currently our coverage is rather… sparse, with only a wrapper for the CHOLMOD routines for sparse Cholesky decomposition, but we hope that this will expand over time. Contributions of new wrappers are very welcome, especially if you can follow the style of the existing interfaces.


The current release may be downloaded from the Python Package index at

Or from the homepage at

Or the latest development version may be found in our Git repository:

$ git clone git://github.com/scikit-sparse/scikit-sparse.git


Installing scikit-sparse requires:

Test versions are: * Python: 3.7, 3.6 * NumPy: 1.15, 1.14, 1.13 * SciPy: 1.1, 1.0, 0.19 * SuiteSparse: 5.2 (Other versions may work but are untested.)

On Debian/Ubuntu systems, the following command should suffice:

$ sudo apt-get install python-scipy libsuitesparse-dev

On Arch Linux, run:

$ sudo pacman -S suitesparse


As usual,

$ pip install --user scikit-sparse

or with conda

$ conda install -c conda-forge scikit-sparse


Post your suggestions and questions directly to our bug tracker.