# Most Efficient Way to Compute a Quadratic Matrix Form

**URL:** <https://discourse.julialang.org/t/most-efficient-way-to-compute-a-quadratic-matrix-form/66606>\
**Category:** Performance\
**Tags:** performance, linearalgebra\
**Created:** [August 18, 2021, 3:04pm UTC](https://discourse.julialang.org/t/most-efficient-way-to-compute-a-quadratic-matrix-form/66606 "2021-08-18T15:04:45Z")\
**Posts on this page:** 1\
**Showing post:** 10

<div class="post-metadata">

**Author:** ![RoyiAvital](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/royiavital/32/571_2.png) [@RoyiAvital](https://discourse.julialang.org/u/RoyiAvital)\
**Post date:** [August 18, 2021, 7:38pm UTC](https://discourse.julialang.org/t/most-efficient-way-to-compute-a-quadratic-matrix-form/66606/10 "2021-08-18T19:38:47Z")

</div>

I think your direction is very well understood, if the rank is low enough one could apply factorization to speed things up (Cholesky for P being Positive Definite, LDL in case it is only symmetric would be classic choice while [`KernelMatrices.jl`](https://bitbucket.org/cgeoga/kernelmatrices.jl) may offer more advanced methods [See [`KernelMatrices.jl` - A Software Package for Working with Hierarchical Matrices](https://discourse.julialang.org/t/21696)]).

What about the case the rank is high to make those method not feasible.  
Is there any function which can take advantage of the structure of the problem?

---

_[View the full topic](https://discourse.julialang.org/t/most-efficient-way-to-compute-a-quadratic-matrix-form/66606)._