# How do I implement an upper limit log normal distribution

**URL:** <https://discourse.pymc.io/t/how-do-i-implement-an-upper-limit-log-normal-distribution/1337>\
**Category:** Questions\
**Created:** [June 7, 2018, 11:00am UTC](https://discourse.pymc.io/t/how-do-i-implement-an-upper-limit-log-normal-distribution/1337 "2018-06-07T11:00:20Z")\
**Posts on this page:** 1\
**Showing post:** 3

<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:** [June 7, 2018, 11:26am UTC](https://discourse.pymc.io/t/how-do-i-implement-an-upper-limit-log-normal-distribution/1337/3 "2018-06-07T11:26:46Z")

</div>

Your use case is more similar to [Truncated Inverse normal distribution (also known as Wald distribution)](https://discourse.pymc.io/t/truncated-inverse-normal-distribution-also-known-as-wald-distribution/1125). You can have a look at the discussion there which is quite in-depth.

In general, if you are transforming the observed variable, you need to make sure to adjust the volume changes by adding the jacobian. In your code above, pt ~ Normal(mu, sd) does not imply y ~ULLN(exp(mu), sd)

---

_[View the full topic](https://discourse.pymc.io/t/how-do-i-implement-an-upper-limit-log-normal-distribution/1337)._