# Exponential GLM

**URL:** <https://discourse.mc-stan.org/t/exponential-glm/29129>\
**Category:** rstanarm\
**Created:** [October 13, 2022, 3:31pm UTC](https://discourse.mc-stan.org/t/exponential-glm/29129 "2022-10-13T15:31:34Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![wati](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/wati/32/11265_2.png) [@wati](https://discourse.mc-stan.org/u/wati)\
**Post date:** [October 13, 2022, 3:31pm UTC](https://discourse.mc-stan.org/t/exponential-glm/29129/1 "2022-10-13T15:31:34Z")

</div>

Hi,

lets assume I have some data where my response follows an exponential distriubtion (also inlcuding zeros), what is a proper way of modelling it via rstanarm?

Unfortunately, the family objects used for `stan_glm()` doesn’t contain `Exponential()` (is there an obvious reason why not?).  
Or is it possible to use `Gamma()` instead but with a fixed `shape` parameter to 1 and still allowing also response values of zero?

Thanks!