Hierarchical Orthogonal Factorization: Sparse Square matrices

TL;DR

This paper introduces spaQR, a fast sparse QR factorization algorithm that uses low-rank approximations to efficiently solve large sparse linear systems with controlled approximation error.

Contribution

The paper presents a novel sparsified QR algorithm that combines nested dissection, low-rank approximations, and orthogonal factorizations for improved efficiency.

Findings

Factorization time scales as O(N log N)

Solve time scales as O(N)

Effective for benchmark unsymmetric problems

Abstract

In this work, we develop a new fast algorithm, spaQR -- sparsified QR, for solving large, sparse linear systems. The key to our approach is using low-rank approximations to sparsify the separators in a Nested Dissection based Householder QR factorization. First, a modified version of Nested Dissection is used to identify interiors/separators and reorder the matrix. Then, classical Householder QR is used to factorize the interiors, going from the leaves to the root to the elimination tree. After every level of interior factorization, we sparsify the remaining separators by using low-rank approximations. This operation reduces the size of the separators without introducing any fill-in in the matrix. However, it introduces a small approximation error which can be controlled by the user. The resulting approximate factorization is stored as a sequence of sparse orthogonal and sparse upper-triangular factors. Hence, it can be applied efficiently to solve linear systems. Additionally, we further improve the algorithm by using a block diagonal scaling. Then, we show a systematic analysis of the approximation error and effectiveness of the algorithm in solving linear systems. Finally, we perform numerical tests on benchmark unsymmetric problems to evaluate the performance of the algorithm. The factorization time scales as $\mathcal{O}(N \log N)$ and the solve time scales as $\mathcal{O}(N)$.