Shifted CholeskyQR for sparse matrices

TL;DR

This paper introduces a new shifted parameter for the Shifted CholeskyQR3 algorithm tailored for sparse matrices, enhancing its accuracy and applicability through detailed error analysis and numerical experiments.

Contribution

It proposes a novel shifted parameter based on sparsity metrics, with rigorous error analysis and validation, improving the algorithm's performance on sparse, ill-conditioned matrices.

Findings

Improved accuracy of SCholeskyQR3 with the new shift in sparse cases

Enhanced applicability of the algorithm to ill-conditioned matrices

Numerical experiments confirm the efficiency and accuracy improvements

Abstract

In this work, we focus on Shifted CholeskyQR (SCholeskyQR) for sparse matrices. We provide a new shifted item $s$ for Shifted CholeskyQR3 (SCholeskyQR3) based on the number of non-zero elements (nnze) and the element with the largest absolute value of the input sparse $X \in \mathbb{R}^{m\times n}$ with $m \ge n$. We do rounding error analysis of SCholeskyQR3 with such an $s$ and show that SCholeskyQR3 is accurate in this case. Therefore, an alternative choice of $s$ can be taken for SCholeskyQR3 with the comparison between our new $s$ and the $s$ shown in the previous work when the input $X$ is sparse, improving the applicability and residual of the algorithm for the ill-conditioned cases. Numerical experiments demonstrate the advantage of SCholeskyQR3 with our alternative choice of $s$ in both applicablity and accuracy over the case with the original $s$, together with the same level of efficiency. This work is also the first to build connections between sparsity and numerical algorithms with detailed rounding error analysis to the best of our knowledge.