Proposal Summary
To improve sparse structures in SciPy so they support array semantics, to deprecate SciPy’s sparse metrics, and to assist with sparse array adoption in downstream ecosystem packages.
Scientific Python
Sparse data refers to datasets where a high percentage of the values are zero or empty, which occurs where relationships across dimensions don’t exist or are neglected. This project will introduce sparse array data structures that provide array semantics to enable straightforward interaction with sparse data. This includes the design and implementation of 1D and 2D sparse array data structures, as well as expanding on the existing scipy.sparse package.