# Case study on spatial models for areal data - Poisson CAR/IAR

**URL:** <https://discourse.mc-stan.org/t/case-study-on-spatial-models-for-areal-data-poisson-car-iar/481>\
**Category:** Modeling\
**Created:** [May 9, 2017, 8:45pm UTC](https://discourse.mc-stan.org/t/case-study-on-spatial-models-for-areal-data-poisson-car-iar/481 "2017-05-09T20:45:03Z")\
**Posts on this page:** 1\
**Showing post:** 69

<div class="post-metadata">

**Author:** ![anon75146577](https://avatars.discourse-cdn.com/v4/letter/a/f07891/32.png) [@anon75146577](https://discourse.mc-stan.org/u/anon75146577)\
**Post date:** [August 1, 2017, 6:24pm UTC](https://discourse.mc-stan.org/t/case-study-on-spatial-models-for-areal-data-poisson-car-iar/481/69 "2017-08-01T18:24:02Z")

</div>

> [@Bob\_Carpenter](#):
>
> How can we do a computation for every possible value of theta?

The same way you build a proposal kernel q( theta’ | theta\_current), for which you need to build a program that for any possible theta\_current you get a proposal distribution.

I feel the way I phrased it originally was _really_ weird. But the output from that step could be used a independence proposal kernel in a Metropolis-within-Gibbs sampler for the full conditional step x’ ~ q(x’ | y, theta)

---

_[View the full topic](https://discourse.mc-stan.org/t/case-study-on-spatial-models-for-areal-data-poisson-car-iar/481)._