# How to create a list of random variables instead of using size parameter?

**URL:** <https://discourse.pymc.io/t/how-to-create-a-list-of-random-variables-instead-of-using-size-parameter/4560>\
**Category:** Questions\
**Created:** [February 25, 2020, 9:44pm UTC](https://discourse.pymc.io/t/how-to-create-a-list-of-random-variables-instead-of-using-size-parameter/4560 "2020-02-25T21:44:17Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![dydx](https://avatars.discourse-cdn.com/v4/letter/d/e9bcb4/32.png) [@dydx](https://discourse.pymc.io/u/dydx)\
**Post date:** [February 25, 2020, 9:44pm UTC](https://discourse.pymc.io/t/how-to-create-a-list-of-random-variables-instead-of-using-size-parameter/4560/1 "2020-02-25T21:44:17Z")

</div>

I’m specifying some parameters for fourier terms in my regression as follows:

```auto
    #Seasonality terms
    β_ssn = pm.Laplace('β_ssn', mu = 0, b = 0.04, shape = (cat_len, 2*fourier_dim))

```

What I want to do, though, is have a separate b parameter for each prior. I want to have more regularizing priors for the higher dimensional terms. How would I go about doing this?

---

<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:** [February 25, 2020, 9:47pm UTC](https://discourse.pymc.io/t/how-to-create-a-list-of-random-variables-instead-of-using-size-parameter/4560/2 "2020-02-25T21:47:52Z")

</div>

The best way to do this is passing a `np.ndarray` to `b`, it will parameterized it to have separate prior.

---

<div class="post-metadata">

**Author:** ![dydx](https://avatars.discourse-cdn.com/v4/letter/d/e9bcb4/32.png) [@dydx](https://discourse.pymc.io/u/dydx)\
**Post date:** [February 25, 2020, 10:21pm UTC](https://discourse.pymc.io/t/how-to-create-a-list-of-random-variables-instead-of-using-size-parameter/4560/3 "2020-02-25T22:21:14Z")

</div>

Awesome this worked!!