# Advice on speeding up multivariate normal calculations

**URL:** <https://discourse.pymc.io/t/advice-on-speeding-up-multivariate-normal-calculations/1166>\
**Category:** Questions\
**Created:** [May 3, 2018, 8:44pm UTC](https://discourse.pymc.io/t/advice-on-speeding-up-multivariate-normal-calculations/1166 "2018-05-03T20:44:51Z")\
**Posts on this page:** 1\
**Showing post:** 2

<div class="post-metadata">

**Author:** ![junpenglao](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.pymc.io/junpenglao/32/8_2.png) [@junpenglao](https://discourse.pymc.io/u/junpenglao)\
**Post date:** [May 3, 2018, 9:00pm UTC](https://discourse.pymc.io/t/advice-on-speeding-up-multivariate-normal-calculations/1166/2 "2018-05-03T21:00:23Z")

</div>

Using a for loop to manually construct the correlation matrix and a bivariate normal distribution is likely very slow. In general, to speed up the computation that involves multivariate normal, it is necessary to find a way to express the covariant matrix as using its Cholesky decomposition.

There are a few similar tricks in GP, you might find posts here helpful such as: [Multiple (uncertain) function observations of the same Gaussian process](https://discourse.pymc.io/t/multiple-uncertain-function-observations-of-the-same-gaussian-process/400)

---

_[View the full topic](https://discourse.pymc.io/t/advice-on-speeding-up-multivariate-normal-calculations/1166)._