# Non centered parameterization on variance parameter

**URL:** <https://discourse.mc-stan.org/t/non-centered-parameterization-on-variance-parameter/584>\
**Category:** Modeling\
**Created:** [May 17, 2017, 7:38pm UTC](https://discourse.mc-stan.org/t/non-centered-parameterization-on-variance-parameter/584 "2017-05-17T19:38:08Z")\
**Posts on this page:** 1\
**Showing post:** 3

<div class="post-metadata">

**Author:** ![Bob\_Carpenter](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/bob_carpenter/32/9230_2.png) [@Bob\_Carpenter](https://discourse.mc-stan.org/u/Bob_Carpenter)\
**Post date:** [May 18, 2017, 10:27pm UTC](https://discourse.mc-stan.org/t/non-centered-parameterization-on-variance-parameter/584/3 "2017-05-18T22:27:35Z")

</div>

I should’ve also said that you want a transform that will leave the distribution over quantities of interest unchanged, but let Stan sample from a distribution that looks more like an isotropic Gaussian than it would in the original parameterization. I don’t know how to do that for variance parameters—it would depend on the prior. You can do lognormal the same way you do normal, but I don’t know how to reparameterize something like a gamma.

---

_[View the full topic](https://discourse.mc-stan.org/t/non-centered-parameterization-on-variance-parameter/584)._