# Manually update part of the model without writing a custom sampler

**URL:** <https://discourse.pymc.io/t/manually-update-part-of-the-model-without-writing-a-custom-sampler/3229>\
**Category:** Questions\
**Created:** [May 9, 2019, 10:35am UTC](https://discourse.pymc.io/t/manually-update-part-of-the-model-without-writing-a-custom-sampler/3229 "2019-05-09T10:35:37Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Hamster\_on\_wheels](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.pymc.io/hamster_on_wheels/32/1823_2.png) [@Hamster\_on\_wheels](https://discourse.pymc.io/u/Hamster_on_wheels)\
**Post date:** [May 9, 2019, 10:35am UTC](https://discourse.pymc.io/t/manually-update-part-of-the-model-without-writing-a-custom-sampler/3229/1 "2019-05-09T10:35:37Z")

</div>

I want to use NUTS to sample part A of the model, and update the remaining part B of the model manually and deterministically. Part B contains categorical distributions ([https://github.com/pymc-devs/pymc3/issues/1902](https://github.com/pymc-devs/pymc3/issues/1902)).

Log-likelihood of A and log-likelihood of B share some intermediate calculations C. I plan to define custom theano operations for A, B, and C to avoid evaluating C for twice ([Custom operation for likelihood and cache value of likelihood](https://discourse.pymc.io/t/custom-operation-for-likelihood-and-cache-value-of-likelihood/3197/2)).

I will call the theano operation for log-likelihood of A as **α**.

I am thinking about the following:

1. store the parameters of B in **α**
2. update parameters of B within the perform method of **α**.

But that means **α** is not a pure function. The following two cases give different results:

1. Evaluating **α** at point x first and then at point y
2. Evaluating **α** at point y first and then at point x

Would that cause any trouble if I use NUTS?

Within each step of NUTS, how many times would NUTS call the perform method for the same point? Can I still avoid evaluating C for the same point twice?  
Does NUTS use the same instance of **α** across all chains?

Thanks again.

---

<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:** [May 9, 2019, 5:02pm UTC](https://discourse.pymc.io/t/manually-update-part-of-the-model-without-writing-a-custom-sampler/3229/2 "2019-05-09T17:02:54Z")

</div>

It shouldnt be a problem if you write the update as a step\_methods, as in the compound step the update is sequential and always ordered in the same way. You can have a look at this notebook that I implemented the laplace approximation so you can do approximation+sampling:

> <https://github.com/junpenglao/Planet_Sakaar_Data_Science/blob/master/Ports/Laplace%2520approximation%2520in%2520pymc3.ipynb>

another example is to do EM: [https://github.com/junpenglao/Planet\_Sakaar\_Data\_Science/blob/master/Ports/Inferencing%20Linear%20Mixed%20Model%20with%20EM.ipynb](https://github.com/junpenglao/Planet_Sakaar_Data_Science/blob/master/Ports/Inferencing%20Linear%20Mixed%20Model%20with%20EM.ipynb)

---

<div class="post-metadata">

**Author:** ![Hamster\_on\_wheels](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.pymc.io/hamster_on_wheels/32/1823_2.png) [@Hamster\_on\_wheels](https://discourse.pymc.io/u/Hamster_on_wheels)\
**Post date:** [May 9, 2019, 5:13pm UTC](https://discourse.pymc.io/t/manually-update-part-of-the-model-without-writing-a-custom-sampler/3229/3 "2019-05-09T17:13:39Z")

</div>

Thank you very much!

I can see the notebook for EM.

I found the notebook for laplace approximation: [https://github.com/junpenglao/Planet\_Sakaar\_Data\_Science/blob/master/Ports/Laplace%20approximation%20in%20pymc3.ipynb](https://github.com/junpenglao/Planet_Sakaar_Data_Science/blob/master/Ports/Laplace%20approximation%20in%20pymc3.ipynb)