# Reworking Distributions.jl

**URL:** <https://discourse.julialang.org/t/reworking-distributions-jl/32339>\
**Category:** Statistics\
**Tags:** proposal\
**Created:** [December 16, 2019, 4:08pm UTC](https://discourse.julialang.org/t/reworking-distributions-jl/32339 "2019-12-16T16:08:04Z")\
**Posts on this page:** 1\
**Showing post:** 9

<div class="post-metadata">

**Author:** ![tkf](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/tkf/32/17635_2.png) [@tkf](https://discourse.julialang.org/u/tkf)\
**Post date:** [December 16, 2019, 9:45pm UTC](https://discourse.julialang.org/t/reworking-distributions-jl/32339/9 "2019-12-16T21:45:36Z")

</div>

For specifying “parameter space” like `P`, I believe _lens_ is the best API, as I mentioned before in [DifferentialEquations, Control Systems, and Linearization - #7 by tkf](https://discourse.julialang.org/t/differentialequations-control-systems-and-linearization/31178/7). Unlike namedtuple, it can describe nested fields quite easily (it actually can describe _arbitrary_ locations). Maybe it’s useful for implementing mixture models. It also is easy to do change of parameter and adding constraints.

---

_[View the full topic](https://discourse.julialang.org/t/reworking-distributions-jl/32339)._