English

Efficiently updating a covariance matrix and its LDL decomposition

Numerical Analysis 2020-02-21 v1 Numerical Analysis Computation

Abstract

Equations are presented which efficiently update or downdate the covariance matrix of a large number of mm-dimensional observations. Updates and downdates to the covariance matrix, as well as mixed updates/downdates, are shown to be rank-kk modifications, where kk is the number of new observations added plus the number of old observations removed. As a result, the update and downdate equations decrease the required number of multiplications for a modification to Θ((k+1)m2)\Theta((k+1)m^2) instead of Θ((n+k+1)m2)\Theta((n+k+1)m^2) or Θ((nk+1)m2)\Theta((n-k+1)m^2), where nn is the number of initial observations. Having the rank-kk formulas for the updates also allows a number of other known identities to be applied, providing a way of applying updates and downdates directly to the inverse and decompositions of the covariance matrix. To illustrate, we provide an efficient algorithm for applying the rank-kk update to the LDL decomposition of a covariance matrix.

Keywords

Cite

@article{arxiv.2002.08831,
  title  = {Efficiently updating a covariance matrix and its LDL decomposition},
  author = {Don March and Vandy Tombs},
  journal= {arXiv preprint arXiv:2002.08831},
  year   = {2020}
}
R2 v1 2026-06-23T13:48:18.464Z