A fast Mixed Model B-splines algorithm

TL;DR

This paper introduces a fast algorithm for B-splines in mixed models that maintains local support and achieves linear computation time, significantly improving efficiency over existing methods.

Contribution

The paper presents a novel algorithm that preserves B-splines' local support in mixed models, enabling linear scaling of computation time.

Findings

Computation time scales linearly with the number of B-splines.

The algorithm maintains the local support property of B-splines.

It outperforms existing methods with cubic scaling.

Abstract

A fast algorithm for B-splines in mixed models is presented. B-splines have local support and are computational attractive, because the corresponding matrices are sparse. A key element of the new algorithm is that the local character of B-splines is preserved, while in other existing methods this local character is lost. The computation time for the fast algorithm is linear in the number of B-splines, while computation time scales cubically for existing transformations.